VintageBigBlue.org

 

Re: PEC makes tracking worse


Jul 8, 2017

 


----------------------------

#59574 Jul 8, 2017

I'm using a new G-11 purchased last January with Gemini 2. Early on I tried using PEC but I found that it made tracking worse. Recently, I decided to give it another try. I trained the PEC carefully this time. First, I made sure that I had a very accurate polar alignment. Next I picked a star near the meridian and the equator and recorded the PEC training about four times. The results were no different from when I tried it earlier. Using PHD2 the RMS error for RA with PEC ran from 30 to 40 with some pretty good size fluxes. Without PEC the RA RMS is in the low twenties. So, where am I going wrong? Isn't PEC supposed to make guiding more accurate?

Rob



----------------------------

#59575 Jul 8, 2017

I have had the same experience when using PEC with guiding. However, without guiding, I do notice that PEC makes tracking better. For example, when I do planetary photography, enabling PEC (but not guiding) helps keep the planet stay firmly in place, while without PEC it tends to drift around.I never have figured out why PEC and PHD seem to "fight"...

On Sat, Jul 8, 2017, 12:05 AM sactowriter@... [Losmandy_users] Losmandy_users@yahoogroups.com> wrote:

.I'm using a new G-11 purchased last January with Gemini 2. Early on I tried using PEC but I found that it made tracking worse. Recently, I decided to give it another try. I trained the PEC carefully this time. First, I made sure that I had a very accurate polar alignment. Next I picked a star near the meridian and the equator and recorded the PEC training about four times. The results were no different from when I tried it earlier. Using PHD2 the RMS error for RA with PEC ran from 30 to 40 with some pretty good size fluxes. Without PEC the RA RMS is in the low twenties. So, where am I going wrong? Isn't PEC supposed to make guiding more accurate?

Rob



----------------------------

#59577 Jul 8, 2017

I don't use PHD, but as I recall it has some ability to learn and adjust its parameters based on guiding performance.��

Perhaps the previous parameters (before PEC) need to be reset, so that PHD can re-learn these parameters once PEC is trained? Just a guess.

Regards,

�� -Paul



----------------------------

#59578 Jul 8, 2017

Hi gang,

That's an interesting finding. ..that autoguiding is making PE worse... after autoguiding, you should be getting near zero net PE on your image.

Are you really saying that the PE corrections (steps or however that is measured) are getting 2X as large after PE is turned on in the Gemini2?. That can happen if the PE recording is inverted in sign .(see below).

Next question...did you use the Gemini2 PE recording method with averaging, or another method with external software?. If external software, there is a chance the corrections are inverse in sign (i.e. negative value instead of positive).. Of you used external software, try the internal PE recording built into the Gemini. . .

Next question: are you using Ascom pulse guiding, or ST4 port cable guiding? .

If ST4 method, I thought I'd read that external pulses can interfere with PE correction pulses...but that applied to .the Orion Atlas/Synta EQG mount and in that case supposes the PE corrections are through Ascom too. . .You might change autoguider methods (Ascom to ST4 or vice versa) to see if there is a performance difference. .

All the best,.Michael.



----------------------------

#59580 Jul 8, 2017

Michael,

Are you really saying that the PE corrections (steps or however that is measured) are getting 2X as large after PE is turned on in the Gemini2?�� That can happen if the PE recording is inverted in sign ��(see below).

Yes, I took the measurements last night over several minutes.��

Next question...did you use the Gemini2 PE recording method with averaging, or another method with external software?�� If external software, there is a chance the corrections are inverse in sign (i.e. negative value instead of positive).�� Of you used external software, try the internal PE recording built into the Gemini. �� ��

I am using the hand controller to start PEC training. I don't see anywhere where it gives you any control over averaging. I assumed it did so automatically. I have trained the PEC about five times, repeating the process by hitting button -- train PEC

Next question: are you using Ascom pulse guiding, or ST4 port cable guiding? ��

I am using ASCOM Gemini.net. I found it is much more responsive than the ST4 cable guiding.

If ST4 method, I thought I'd read that external pulses can interfere with PE correction pulses...but that applied to ��the Orion Atlas/Synta EQG mount and in that case supposes the PE corrections are through Ascom too. �� ��You might change autoguider methods (Ascom to ST4 or vice versa) to see if there is a performance difference. ��

I'm not anxious to run another cable. It's really nice using the ethernet.��

Rob



----------------------------

#59582 Jul 8, 2017

I've tried to work with different parameters on PHD2 with no success. If anyone has some suggestions here, I am all ears. When I called Losmandy about the problem back in February, they told me not to use PEC. The RA tracking without PEC normally has and RMS on PHD2 in the low twenties, but this is significantly higher than the DEC which can get down to 8 to 10 RMS guided and typically runs in the teens.��

Rob



----------------------------

#59583 Jul 8, 2017

This remains a puzzle .(to me).If the internal Gemini2 PEC code is adding, instead of subtracting, .the PE corrections...then this might explain your observation of needing double corrections after turning on PEC.. It's unbelievable that the software system could have this engineering error, but... if it does, the Gemini2 can upload new software. .

Alternatively, Ray Gralak has a PEC program that is external, and its correction data can be inverted.. .and uploaded to Gemini for PEC subtraction...you might try that. .

The internal PEC training in Gemini1 has the capability to add/average PE training runs.. I don't have a Gemini2 but it would make sense to review its manual for this.

Best,Michael.







----------------------------

#59585 Jul 8, 2017

Rob

If you haven't already, I recommend running Guiding Assistant in PHD2. ��It is a better way to set parameters than trying on your own.

Bryan

---In Losmandy_users@yahoogroups.com, sactowriter@...> wrote :

I've tried to work with different parameters on PHD2 with no success. If anyone has some suggestions here, I am all ears. When I called Losmandy about the problem back in February, they told me not to use PEC. The RA tracking without PEC normally has and RMS on PHD2 in the low twenties, but this is significantly higher than the DEC which can get down to 8 to 10 RMS guided and typically runs in the teens.��

Rob



----------------------------

#59591 Jul 8, 2017

How often do you issue autoguiding corrections when programming PEC (how long is the exposure and how long the delay between successive corrections)? ��

If possible, make the interval 1 second or less. Any more and you can run into a situation where PEC corrections might be slow enough to allow the autoguider to compete with them, resulting in double the desired correction.

Regards,

�� ��-Paul



----------------------------

#59592 Jul 8, 2017

I can't speak for pec but for auto guiding you don't want corrections that fast. It will see-saw because you'd be chasing the seeing. More like 2-5 seconds between corrections for auto guiding

Brian

Sent from my iPad

On Jul 8, 2017, at 8:44 PM, yh@... [Losmandy_users] Losmandy_users@yahoogroups.com> wrote:



��How often do you issue autoguiding corrections when programming PEC (how long is the exposure and how long the delay between successive corrections)? ��

If possible, make the interval 1 second or less. Any more and you can run into a situation where PEC corrections might be slow enough to allow the autoguider to compete with them, resulting in double the desired correction.

Regards,

�� ��-Paul



----------------------------

#59593 Jul 8, 2017

Rob

What units are these? ��Typical excellent performance is an arc-second or less RMS. ��Perhaps your numbers are percentages of an arc-sec?

Bryan



---In Losmandy_users@yahoogroups.com, sactowriter@...> wrote :

I've tried to work with different parameters on PHD2 with no success. If anyone has some suggestions here, I am all ears. When I called Losmandy about the problem back in February, they told me not to use PEC. The RA tracking without PEC normally has and RMS on PHD2 in the low twenties, but this is significantly higher than the DEC which can get down to 8 to 10 RMS guided and typically runs in the teens.��

Rob



----------------------------

#59594 Jul 8, 2017

I could be wrong, but I think that Rob is not using autoguiding for PEC training. ��He is using the manual Losmandy/Gemini method. ��With that method, you issue manual corrections with the HC as often as needed to keep the star centered in the eyepiece.

Rob: correct if needed.

Bryan



----------------------------

#59595 Jul 9, 2017

I tried it tonight. Using the PHD2 guide assistant. Same result, the RA tracking is definitely worse with PEC engaged. I'm am wondering if I am doing something wrong with the PEC training?

Rob



----------------------------

#59596 Jul 9, 2017

No, I am definitely using auto guiding for PEC training, but thanks for the thoughts.

Rob



----------------------------

#59597 Jul 9, 2017

Tonight after using the PHD2 guiding assistant, RA was running from 40 to 50 RMS in PHD2 with PEC and in the twenties with PEC not engaged. There is a definite conflict going on. The gyrations with PEC are long and deep without PEC the gyrations are fare less pronounced. It gives me a head ache :).��

Rob



----------------------------

#59598 Jul 9, 2017

I'm sorry I forgot to put the decimal in. When I say RMS of 40, I mean and RMS in arc seconds of .40. My Dec typically shows an RMS in arc seconds of .1 to .2. With the PEC disengaged right now the RMS of the RA in arc seconds is .25 to .31. Whereas with the PEC engage in ran between .4 and .5.

Rob



----------------------------

#59600 Jul 9, 2017

Rob, so how frequently does your autoguider correct when you program PEC?

As I said, if you leave a long interval during PEC (e.g., as Brian suggested 5 seconds!) it'll likely produce a PEC curve that might be delayed by as much as 5 seconds with its corrections. This results in a very likely scenario that during imaging, your autoguider will detect and correct the error before PEC has a chance to correct it, resulting in double corrections.��

Try using an autoguider interval as short as you can. Find a bright star to train PEC on, so you can lower exposure to 0.5 second or so.

Regards,

�� �� -Paul



----------------------------

#59601 Jul 9, 2017

Another thing to do is to use one of the free or commercial tools to chart unguided periodic error before and after PEC. You should be able to tell if there's something odd going on with the PEC curve.

Regards,

�� ��-Paul



----------------------------

#59602 Jul 9, 2017

Rob

FYI

The number to the left is the RMS in pixels, the number to the right is the RMS in arc-seconds. ��The latter is better to use, as it is independent of guide scope focal length and guider sensor.

The former is converted to the latter using the image scale, calculated from tabs in the Brain.

Note that you can also switch between pixels and arc-seconds for the Y-axis units on the graph.

Bryan

---In Losmandy_users@yahoogroups.com, sactowriter@...> wrote :

Sorry to sound a little confused but let me clarify my answer a little more. A .15 RMS seems to be .85 arc seconds. where as a .26 RMS appears to be !.55 arc seconds. With the PEC engaged it appears that error was well over 2 arc seconds reaching 3 at times. I was giving you the number on the left instead of the number in parenthesis.

Rob







----------------------------

#59603 Jul 9, 2017

Rob ��Imho this may be more productive on the PHD forums. You should be posting your guide logs there so people helping you can really determine what.s going on, rather than just talk in generalities. ��Also imo .a .85 arcsec guiding is pretty good. I.m not sure what you are aiming for, but that should be sufficient for most astro requirements. ��You might also consider using predictive pec guiding algorithm in PHD2. ��It may be frustrating to not know entirely why it.s working or not working, but focusing on the results (i.e., your .85 arcsec guided pe) could be a lot more rewarding than fiddling with this stuff, which is endless. ��My two cents ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Sunday, July 9, 2017 1:00 AMTo: Losmandy_users@yahoogroups.comSubject: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59604 Jul 9, 2017

Brian,

Thanks for your thoughts. .85arc is my error in DEC. The RA is what I'm trying to focus on and why I have been trying to get PEC to work. It has been has been running at between 1.5 to 2.5 arc sec.

I think that I have made a major discovery. The focal length in PHD2 was significantly off. 300 versus 735 which is reality. I'm thinking that this might be the cause of the conflict as it leads to erroneous calibrations by PHD2. I'm amazed that my tracking has been as good as it has with this kind of error. I have no idea how the 300 even got there, maybe it is a default that I over looked.

As I go through the algorithms, Brian, I don't see one called predictive PEC. Where would I find that.

Thanks everyone for being so informative and patient with me. You guys are great. I'll let you know how changing to the correct focal length affects everything.

One more question, is there a way to reset the PEC training in Gemini 2? I think this error in guiding has been around for some time and might have corrupted my the training.

Thanks again,

Rob



----------------------------

#59605 Jul 9, 2017

Brian,

For clarification, my combined (DEC and RA) guiding error is running over 2 arc sec because of the gyrations in RA.

Rob



----------------------------

#59606 Jul 9, 2017

The Gemini units (certainly -1, likely -2) normally boot up with PEC off.. You have to turn it on deliberately every time you boot up the Gemini. .

However, there is a checkbox on Gemini.net for having Gemini.net turn on/enable the Gemini PEC on bootup.. Look to see if that box is checked if you are using Ascom/Gemini.net.

You don't need to reset the PEC data, just turn off PEC or leave it off...

All the best,Michael.



----------------------------

#59608 Jul 9, 2017

Rob

Two comments

1. ��The increase in focal length on your guide scope will decrease the RMS linearly (735/300), because the image scale (arc-sec/pixel) will be reduced.

P = (8 * S) / (FL)Where:P is the image scale per pixel in arc secondsS is the size of the pixel in microns (set in Cameras tab under Brain)FL is the focal length in inches (set in Guiding tab under Brain)arc-sec RMS = pixel RMS * P



2. ��If you don't see PPEC as an RA algorithm, then you have an older build. ��PPEC for RA was not implemented until PHD2 2.6.3 dev2. ��PPEC is not yet implemented as Dec algorithm.

Before getting into PPEC, I'd suggest staying with whatever algorithm you know and use now. ��I'd consider installing the latest build, 2.6.3 dev6 to ensure all the bug fixes are in place.

openphdguiding.org/development-snapshots/



I think but am not positive that each PEC training run overwrite the previous curve. ��Just re-run PEC training with your corrected focal length

Bryan



---In Losmandy_users@yahoogroups.com, sactowriter@...> wrote :

Brian,

Thanks for your thoughts. .85arc is my error in DEC. The RA is what I'm trying to focus on and why I have been trying to get PEC to work. It has been has been running at between 1.5 to 2.5 arc sec.

I think that I have made a major discovery. The focal length in PHD2 was significantly off. 300 versus 735 which is reality. I'm thinking that this might be the cause of the conflict as it leads to erroneous calibrations by PHD2. I'm amazed that my tracking has been as good as it has with this kind of error. I have no idea how the 300 even got there, maybe it is a default that I over looked.

As I go through the algorithms, Brian, I don't see one called predictive PEC. Where would I find that.

Thanks everyone for being so informative and patient with me. You guys are great. I'll let you know how changing to the correct focal length affects everything.

One more question, is there a way to reset the PEC training in Gemini 2? I think this error in guiding has been around for some time and might have corrupted my the training.

Thanks again,

Rob



----------------------------

#59613 Jul 9, 2017

Rob, ��The incorrect focal length for the guide scope should not impact calibration ��� that is part of why you do calibration. As long as there were 9 or more (20 is better) steps in the each direction of the E/W/N calibration you should be OK. .South often shortcuts if it is happy with the North motion and backlash it taken out. So if using the wrong focal length has any impact at all (aside from causing PhD to overstate the errors) it will because PhD2 is trying to be smart and use that focal length to set the calibration step duration. ��If necessary you should adjust the calibration step size so that you get a reasonable number of steps (9 to 20 being typical) ��� that lets PhD do a more accurate calibration. ��The incorrect focal length caused PhD to over-report the error. In reality your errors are less than . of what you are reporting in the emails. ��As to PEC, unless I really needed it to get satisfactory result in my imaging I would not mess with it. My philosophy is that every additional feature is just something else to get confused by. .Or to cause an error. It is very common for telescopes to have more RA error than DEC (assuming good polar alignment) ��� the RA gear is constantly turning, so any little imperfection in the gear train will show up. ��Clear Skies, ��Mark C.







----------------------------

#59614 Jul 9, 2017

You could probably easily get it to under 1 arcsec without any PE correction. ��I update my suggestion . try to solve one first, then the other. I suggest you try to get the best results from your PHD2 on guiding and then see if you can improve with PEC. ��The guys at PHD2 are really good and reading your logs and giving you feedback. ��PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest dev build and read up on how to do it. It is quite extraordinary imo. �� �� ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Sunday, July 9, 2017 1:50 PMTo: Losmandy_users@yahoogroups.comSubject: RE: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59615 Jul 9, 2017

Rob,



Have you compared periodic error with autoguiding off to when it is on? That would very likely tell what's going on.



Until you do that, here are some possibilities:



1) The PEC curve is bad or inverted.

2) The guiding rate is different than at which PEC was programmed. You cannot change the autoguider rate on a Gemini-based mount without changing the PEC curve to match the guide rate.

3) The PEC curve is causing extra drift. I have mentioned several times that I believe that programming PEC will introduce extra drift in the Gemini-based system.

4) There are bugs in PHD2's autoguider algorithms -- a strong possibility if none of the above is the culprit.



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Sunday, July 9, 2017 1:50 PM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> Brian,

>

>

>

>

> For clarification, my combined (DEC and RA) guiding error is running over 2 arc sec because of the gyrations

> in RA.

>

>

>

>

> Rob

>

>

>



----------------------------

#59616 Jul 10, 2017

Hi Brian,

> PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest dev

> build and read up on how to do it. It is quite extraordinary imo.



Looking at the PHD2 forum group messages, Predictive PEC does not always work so well. That's not surprising though. Having analyzed literally thousands of PE logs from different mounts over the last 15 years I can tell you that in many cases you cannot always accurately predict PEC from just a few points. I think that the PHD2 documentation even says that the algorithm is not a substitute for PEC because PHD2 has to learn each time the scope slews. And if periodic error is properly corrected, any algorithm will work better, because PEC done correctly will preemptively correct periodic error instead of waiting for a background software application to try to figure it out.



BTW, PHD2 only works with some imaging suites because of its non-standard interface. I've been developing an enhanced version of PHD2 which adds more features and will allow it to integrate better with other applications, as well as an alternate mechanism that works with any autoguider program that uses an ASCOM telescope driver for autoguiding. I will make that version available (for free!) when it is available.



Best regards,



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Sunday, July 9, 2017 8:37 PM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> You could probably easily get it to under 1 arcsec without any PE correction.

>

>

>

> I update my suggestion . try to solve one first, then the other. I suggest you try to get the best results from your

> PHD2 on guiding and then see if you can improve with PEC.

>

>

>

> The guys at PHD2 are really good and reading your logs and giving you feedback.

>

>

>

> PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest dev

> build and read up on how to do it. It is quite extraordinary imo.

>

>

>

>

>

>

>

> Thanks

>

>

>

> Brian

>

>

>

>

>

> Brian Valente

>

> Brianvalentephotography.com

>

>

>

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Sunday, July 9, 2017 1:50 PM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

>

>

> Brian,

>

>

>

> For clarification, my combined (DEC and RA) guiding error is running over 2 arc sec because of the gyrations

> in RA.

>

>

>

> Rob

>

>







----------------------------

#59617 Jul 10, 2017

Ray, I can only speak from my personal experience. ��I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an arcsec using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling, so I lose a few frames while it figures it out. ��My point was to try to solve one problem at a time rather than both simultaneously. �� ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Monday, July 10, 2017 5:12 AMTo: Losmandy_users@yahoogroups.comSubject: RE: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59618 Jul 10, 2017

Hi Brian,

> I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an arcsec

> using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling, so I lose

> a few frames while it figures it out.



Yes, but depending on how fast the autoguider cycle rate is autoguiding alone reduce PE to less than one arc-sec.



If PEC is programmed correctly you will get better results almost immediately.

> My point was to try to solve one problem at a time rather than both simultaneously.



I agree, but I think one should start at the lowest level... fix the mount's native tracking characteristics as much as possible so that the software application layer doesn't need to be so responsive.



Best regards,



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 7:07 AM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> Ray, I can only speak from my personal experience.

>

>

>

> I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an arcsec

> using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling, so I lose

> a few frames while it figures it out.

>

>

>

> My point was to try to solve one problem at a time rather than both simultaneously.

>

>

>

>

>

> Thanks

>

>

>

> Brian

>

>

>

>

>

> Brian Valente

>

> Brianvalentephotography.com

>

>

>

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 5:12 AM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

>

>

> Hi Brian,

>

> > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest dev

> > build and read up on how to do it. It is quite extraordinary imo.

>

> Looking at the PHD2 forum group messages, Predictive PEC does not always work so well. That's not

> surprising though. Having analyzed literally thousands of PE logs from different mounts over the last 15

> years I can tell you that in many cases you cannot always accurately predict PEC from just a few points. I

> think that the PHD2 documentation even says that the algorithm is not a substitute for PEC because PHD2

> has to learn each time the scope slews. And if periodic error is properly corrected, any algorithm will work

> better, because PEC done correctly will preemptively correct periodic error instead of waiting for a background

> software application to try to figure it out.

>

> BTW, PHD2 only works with some imaging suites because of its non-standard interface. I've been developing

> an enhanced version of PHD2 which adds more features and will allow it to integrate better with other

> applications, as well as an alternate mechanism that works with any autoguider program that uses an

> ASCOM telescope driver for autoguiding. I will make that version available (for free!) when it is available.

>

> Best regards,

>

> -Ray Gralak

> Author of APCC (Astro-Physics Command Center): www.astro-

> physics.com/index.htm?products/accessories/software/apcc/apcc

> Author of PEMPro: www.ccdware.com

> Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> Author of PulseGuide: www.pulseguide.com

> Author of Sigma: www.gralak.com/sigma

>

> > -----Original Message-----

> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > Sent: Sunday, July 9, 2017 8:37 PM

> > To: Losmandy_users@yahoogroups.com

> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> >

> >

> >

> > You could probably easily get it to under 1 arcsec without any PE correction.

> >

> >

> >

> > I update my suggestion . try to solve one first, then the other. I suggest you try to get the best results from

> your

> > PHD2 on guiding and then see if you can improve with PEC.

> >

> >

> >

> > The guys at PHD2 are really good and reading your logs and giving you feedback.

> >

> >

> >

> > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest dev

> > build and read up on how to do it. It is quite extraordinary imo.

> >

> >

> >

> >

> >

> >

> >

> > Thanks

> >

> >

> >

> > Brian

> >

> >

> >

> >

> >

> > Brian Valente

> >

> > Brianvalentephotography.com

> >

> >

> >

> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > Sent: Sunday, July 9, 2017 1:50 PM

> > To: Losmandy_users@yahoogroups.com

> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> >

> >

> >

> >

> >

> > Brian,

> >

> >

> >

> > For clarification, my combined (DEC and RA) guiding error is running over 2 arc sec because of the

> gyrations

> > in RA.

> >

> >

> >

> > Rob

> >

> >

>

>







----------------------------

#59619 Jul 10, 2017

Well that.s where I guess you and I disagree ��But whatever the approach, I think you and I would agree to solve it one step at a time �� �� ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Monday, July 10, 2017 7:58 AMTo: Losmandy_users@yahoogroups.comSubject: RE: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59620 Jul 10, 2017

Hi dear fellows...

There are bigger advantages to getting you RA drive optimized.

That is, some elements of the bearing rotations are not integer fractions of the worm period of ~399 seconds. .

That means the some RA drive errors repeat and cannot be compensated by PEC.

Of course, if the errors are long period, autoguiding will help, but those corrections come only after the guide camera has seen its target star move appreciably...which can mean your image stars also have shifted. .

Best to take the native PE data over 3 or more 8 minute cycles, examine it for its periods, and try to eliminate the root causes.. Once you get the native PE at the limits of seeing... You likely won't need the PEC at all. .

(A separate comment... If you have a power outage, I think Gemini remembers in SRAM the last worm position.. So restart and then Park.. Your PEC curve should still be in synch.. Even if you loosen your clutches and push to CWD, the worm has not rotated. .)

All the best,Michael









On Mon, Jul 10, 2017 at 9:01 AM 'Brian Valente' bvalente@... [Losmandy_users] Losmandy_users@yahoogroups.com> wrote:

.Well that.s where I guess you and I disagree.But whatever the approach, I think you and I would agree to solve it one step at a time



----------------------------

#59621 Jul 10, 2017

Ray,That's the really puzzling part. When I ran PHD2 guiding assistant (which turns off RA and DEC corrections to��measure drift)����with the PEC on, the RMS numbers where better even than my guided numbers with PEC off, especially for RA. I thought, "Oh boy, now it's going to be great." The only problem is that over the four minute period (which I understand to be the G11 time for one worm rotation) the star image drifted too far off of center to be useful for imaging unguided. Then I turned on guiding, and well, the numbers turned into crap, doubling the RMS numbers for guiding without PEC and almost tripling the unguided PEC numbers. So I was left with a giant head ache.

On a positive note: Last night I guided on M20, which is in a far southern latitude and is a typically difficult area to guide, using the PPEC algorithm. The results were good with the combined RMS arc sec of less than 1 the whole night, except for dithering between frames of course. The RA was still much worse than the DEC, but not as bad. With the correct focal length in PHD2 all of the numbers really changed in ways that I hadn't expected. For example the RMS pixel number was much higher than with the old erroneous focal length numbers running for both DEC and RA in the mid .20s to .30s. but the arc second numbers were, as stated, lower.

I would love to get the most out of my mount, which I think includes using PEC, but where do I go from here? I do think that it appears to be a��PHD2 problem and that I should probably switch forums.

Rob



----------------------------

#59622 Jul 10, 2017

Rob what.s your TVC setting on the Gemini? �� �� ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Monday, July 10, 2017 10:48 AMTo: Losmandy_users@yahoogroups.comSubject: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59623 Jul 10, 2017

Rob,��

>��I would love to get the most out of my mount, which I think includes using PEC, but where do I go from here?

Both Ray and I had suggested this: you should try to measure your mount periodic error with and without PEC, unguided, over a few worm cycles. That will help you and others diagnose what might be wrong. When you are judging using guided RMS numbers, there is no way to extract what is causing increase or decrease in error.

Regards,

�� -Paul



----------------------------

#59624 Jul 10, 2017

Also if you are experiencing total RMS of less than 1 arcsec, I.d say you are in the top 5% of gemini setups. ��Not sure how much more you.ll be able to squeak out of it, but curious on how it will progress for you. �� ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Monday, July 10, 2017 10:48 AMTo: Losmandy_users@yahoogroups.comSubject: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59625 Jul 10, 2017

Also (sorry last thought) if your focal length is 735mm I.m not sure you.ll see any difference in your results at less than 1 arcsec. You are talking sub-pixel accuracy now. ��Probably bigger issues would be along the lines of seeing limitations, flexure, etc. ��Just thinking out loud here ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Monday, July 10, 2017 10:48 AMTo: Losmandy_users@yahoogroups.comSubject: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59626 Jul 10, 2017

Rob

A couple software suggestions to help you.

1. ��PEMPro Logviewer. ��This will read PHD2 files and calculate PE,��as well as PEMPro runs. ��This is a free app.�� ��

2. ��Tools for measuring PE. ��.PEMPro (Ray Gralak - author). ��Free trial.PECPrep��(part of EQMOD Project). Free.Metaguide��Free

I only have experience with PEMPro. ��I list the others for completeness, but can't help with using them. ��8-(

For support on any of these, use the appropriate forum.

P.S. While both are measured in arc-seconds, do not confuse PE and guiding RMS.��



Bryan

---In Losmandy_users@yahoogroups.com, yh@...> wrote :

Rob,��

>��I would love to get the most out of my mount, which I think includes using PEC, but where do I go from here?

Both Ray and I had suggested this: you should try to measure your mount periodic error with and without PEC, unguided, over a few worm cycles. That will help you and others diagnose what might be wrong. When you are judging using guided RMS numbers, there is no way to extract what is causing increase or decrease in error.

Regards,

�� -Paul







----------------------------

#59627 Jul 10, 2017

The PHD2 manual does state the sometimes PEC needs to be turned off while guiding.



----------------------------

#59628 Jul 10, 2017

I have to agree with much of what Michael says here.

> There are bigger advantages to getting you RA drive optimized.

>

> That is, some elements of the bearing rotations are not integer fractions of the worm period of ~399 seconds.

>

> That means the some RA drive errors repeat and cannot be compensated by PEC.



If there are residual frequencies not integer multiples of the worm frequency then those might benefit from a learning autoguiding algorithm (or something better :-). However, if those non-repeating frequencies exist then an autoguider can't usually be used to correct them. You need a program that can pull just the frequencies that can be fixed and apply those (my PEMPro software does this). Or, you can try to fix the problem mechanically.



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 9:57 AM

> To: Losmandy_users@yahoogroups.com

> Subject: Re: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> Hi dear fellows...

>

> There are bigger advantages to getting you RA drive optimized.

>

> That is, some elements of the bearing rotations are not integer fractions of the worm period of ~399 seconds.

>

> That means the some RA drive errors repeat and cannot be compensated by PEC.

>

> Of course, if the errors are long period, autoguiding will help, but those corrections come only after the guide

> camera has seen its target star move appreciably...which can mean your image stars also have shifted.

>

> Best to take the native PE data over 3 or more 8 minute cycles, examine it for its periods, and try to eliminate

> the root causes. Once you get the native PE at the limits of seeing... You likely won't need the PEC at all.

>

> (A separate comment... If you have a power outage, I think Gemini remembers in SRAM the last worm

> position. So restart and then Park. Your PEC curve should still be in synch. Even if you loosen your

> clutches and push to CWD, the worm has not rotated. )

>

> All the best,

> Michael

>

>

>

>

> On Mon, Jul 10, 2017 at 9:01 AM 'Brian Valente' bvalente@... [Losmandy_users]

> Losmandy_users@yahoogroups.com> wrote:

>

>

>

>

>

>

>

> .Well that.s where I guess you and I disagree

>

>

>

> .But whatever the approach, I think you and I would agree to solve it one step at a time

>

>

>

>

>

>

>

>

>

>

>

> .Thanks

>

>

>

> .Brian

>

>

>

>

>

> .Brian Valente

>

> .Brianvalentephotography.com

>

>

>

>

>

>

>

> .From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> .Sent: Monday, July 10, 2017 7:58 AM

>

>

>

>

>

>

> .To: Losmandy_users@yahoogroups.com

> .Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

>

>

>

>

>

>

> .Hi Brian,

>

> .> I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an

> arcsec

> .> using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling,

> so I lose

> .> a few frames while it figures it out.

>

> .Yes, but depending on how fast the autoguider cycle rate is autoguiding alone reduce PE to less than

> one arc-sec.

>

> .If PEC is programmed correctly you will get better results almost immediately.

>

> .> My point was to try to solve one problem at a time rather than both simultaneously.

>

> .I agree, but I think one should start at the lowest level... fix the mount's native tracking characteristics

> as much as possible so that the software application layer doesn't need to be so responsive.

>

> .Best regards,

>

> .-Ray Gralak

> .Author of APCC (Astro-Physics Command Center): www.astro-

> physics.com/index.htm?products/accessories/software/apcc/apcc

> .Author of PEMPro: www.ccdware.com

> .Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> .Author of PulseGuide: www.pulseguide.com

> .Author of Sigma: www.gralak.com/sigma

>

> .> -----Original Message-----

> .> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> .> Sent: Monday, July 10, 2017 7:07 AM

> .> To: Losmandy_users@yahoogroups.com

> .> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> .>

> .>

> .>

> .> Ray, I can only speak from my personal experience.

> .>

> .>

> .>

> .> I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an

> arcsec

> .> using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling,

> so I lose

> .> a few frames while it figures it out.

> .>

> .>

> .>

> .> My point was to try to solve one problem at a time rather than both simultaneously.

> .>

> .>

> .>

> .>

> .>

> .> Thanks

> .>

> .>

> .>

> .> Brian

> .>

> .>

> .>

> .>

> .>

> .> Brian Valente

> .>

> .> Brianvalentephotography.com

> .>

> .>

> .>

> .> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> .> Sent: Monday, July 10, 2017 5:12 AM

> .> To: Losmandy_users@yahoogroups.com

> .> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> .>

> .>

> .>

> .>

> .>

> .> Hi Brian,

> .>

> .> > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the

> latest dev

> .> > build and read up on how to do it. It is quite extraordinary imo.

> .>

> .> Looking at the PHD2 forum group messages, Predictive PEC does not always work so well. That's

> not

> .> surprising though. Having analyzed literally thousands of PE logs from different mounts over the last

> 15

> .> years I can tell you that in many cases you cannot always accurately predict PEC from just a few

> points. I

> .> think that the PHD2 documentation even says that the algorithm is not a substitute for PEC because

> PHD2

> .> has to learn each time the scope slews. And if periodic error is properly corrected, any algorithm will

> work

> .> better, because PEC done correctly will preemptively correct periodic error instead of waiting for a

> background

> .> software application to try to figure it out.

> .>

> .> BTW, PHD2 only works with some imaging suites because of its non-standard interface. I've been

> developing

> .> an enhanced version of PHD2 which adds more features and will allow it to integrate better with

> other

> .> applications, as well as an alternate mechanism that works with any autoguider program that uses

> an

> .> ASCOM telescope driver for autoguiding. I will make that version available (for free!) when it is

> available.

> .>

> .> Best regards,

> .>

> .> -Ray Gralak

> .> Author of APCC (Astro-Physics Command Center): www.astro-

> .> physics.com/index.htm?products/accessories/software/apcc/apcc

> .> Author of PEMPro: www.ccdware.com

> .> Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> .> Author of PulseGuide: www.pulseguide.com

> .> Author of Sigma: www.gralak.com/sigma

> .>

> .> > -----Original Message-----

> .> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> .> > Sent: Sunday, July 9, 2017 8:37 PM

> .> > To: Losmandy_users@yahoogroups.com

> .> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> .> >

> .> >

> .> >

> .> > You could probably easily get it to under 1 arcsec without any PE correction.

> .> >

> .> >

> .> >

> .> > I update my suggestion . try to solve one first, then the other. I suggest you try to get the best

> results from

> .> your

> .> > PHD2 on guiding and then see if you can improve with PEC.

> .> >

> .> >

> .> >

> .> > The guys at PHD2 are really good and reading your logs and giving you feedback.

> .> >

> .> >

> .> >

> .> > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the

> latest dev

> .> > build and read up on how to do it. It is quite extraordinary imo.

> .> >

> .> >

> .> >

> .> >

> .> >

> .> >

> .> >

> .> > Thanks

> .> >

> .> >

> .> >

> .> > Brian

> .> >

> .> >

> .> >

> .> >

> .> >

> .> > Brian Valente

> .> >

> .> > Brianvalentephotography.com

> .> >

> .> >

> .> >

> .> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> .> > Sent: Sunday, July 9, 2017 1:50 PM

> .> > To: Losmandy_users@yahoogroups.com

> .> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> .> >

> .> >

> .> >

> .> >

> .> >

> .> > Brian,

> .> >

> .> >

> .> >

> .> > For clarification, my combined (DEC and RA) guiding error is running over 2 arc sec because of

> the

> .> gyrations

> .> > in RA.

> .> >

> .> >

> .> >

> .> > Rob

> .> >

> .> >

> .>

> .>

>

>

>

>

>

>

>

>

>

> --

>

> Michael Herman

> mobile: 408 421-1239

> email: mherman346@...

>

>

>







----------------------------

#59629 Jul 10, 2017

Hi Brian,

>> Ray wrote:

>> I agree, but I think one should start at the lowest level... fix the mount's native tracking characteristics as much

> >as possible so that the software application layer doesn't need to be so responsive.

> Brian wrote:

> Well that.s where I guess you and I disagree



Ok, but then I guess you feel you don't need to polar align because the autoguiding algorithm

will figure the tracking error and match it with autoguider bumps. Right? :-)



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 8:00 AM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> Well that.s where I guess you and I disagree

>

>

>

> But whatever the approach, I think you and I would agree to solve it one step at a time

>

>

>

>

>

>

>

> Thanks

>

>

>

> Brian

>

>

>

>

>

> Brian Valente

>

> Brianvalentephotography.com

>

>

>

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 7:58 AM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

>

>

> Hi Brian,

>

> > I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an

> arcsec

> > using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling, so I

> lose

> > a few frames while it figures it out.

>

> Yes, but depending on how fast the autoguider cycle rate is autoguiding alone reduce PE to less than one

> arc-sec.

>

> If PEC is programmed correctly you will get better results almost immediately.

>

> > My point was to try to solve one problem at a time rather than both simultaneously.

>

> I agree, but I think one should start at the lowest level... fix the mount's native tracking characteristics as much

> as possible so that the software application layer doesn't need to be so responsive.

>

> Best regards,

>

> -Ray Gralak

> Author of APCC (Astro-Physics Command Center): www.astro-

> physics.com/index.htm?products/accessories/software/apcc/apcc

> Author of PEMPro: www.ccdware.com

> Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> Author of PulseGuide: www.pulseguide.com

> Author of Sigma: www.gralak.com/sigma

>

> > -----Original Message-----

> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > Sent: Monday, July 10, 2017 7:07 AM

> > To: Losmandy_users@yahoogroups.com

> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> >

> >

> >

> > Ray, I can only speak from my personal experience.

> >

> >

> >

> > I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an

> arcsec

> > using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling, so I

> lose

> > a few frames while it figures it out.

> >

> >

> >

> > My point was to try to solve one problem at a time rather than both simultaneously.

> >

> >

> >

> >

> >

> > Thanks

> >

> >

> >

> > Brian

> >

> >

> >

> >

> >

> > Brian Valente

> >

> > Brianvalentephotography.com

> >

> >

> >

> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > Sent: Monday, July 10, 2017 5:12 AM

> > To: Losmandy_users@yahoogroups.com

> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> >

> >

> >

> >

> >

> > Hi Brian,

> >

> > > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest dev

> > > build and read up on how to do it. It is quite extraordinary imo.

> >

> > Looking at the PHD2 forum group messages, Predictive PEC does not always work so well. That's not

> > surprising though. Having analyzed literally thousands of PE logs from different mounts over the last 15

> > years I can tell you that in many cases you cannot always accurately predict PEC from just a few points. I

> > think that the PHD2 documentation even says that the algorithm is not a substitute for PEC because PHD2

> > has to learn each time the scope slews. And if periodic error is properly corrected, any algorithm will work

> > better, because PEC done correctly will preemptively correct periodic error instead of waiting for a

> background

> > software application to try to figure it out.

> >

> > BTW, PHD2 only works with some imaging suites because of its non-standard interface. I've been

> developing

> > an enhanced version of PHD2 which adds more features and will allow it to integrate better with other

> > applications, as well as an alternate mechanism that works with any autoguider program that uses an

> > ASCOM telescope driver for autoguiding. I will make that version available (for free!) when it is available.

> >

> > Best regards,

> >

> > -Ray Gralak

> > Author of APCC (Astro-Physics Command Center): www.astro-

> > physics.com/index.htm?products/accessories/software/apcc/apcc

> > Author of PEMPro: www.ccdware.com

> > Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> > Author of PulseGuide: www.pulseguide.com

> > Author of Sigma: www.gralak.com/sigma

> >

> > > -----Original Message-----

> > > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > > Sent: Sunday, July 9, 2017 8:37 PM

> > > To: Losmandy_users@yahoogroups.com

> > > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> > >

> > >

> > >

> > > You could probably easily get it to under 1 arcsec without any PE correction.

> > >

> > >

> > >

> > > I update my suggestion . try to solve one first, then the other. I suggest you try to get the best results from

> > your

> > > PHD2 on guiding and then see if you can improve with PEC.

> > >

> > >

> > >

> > > The guys at PHD2 are really good and reading your logs and giving you feedback.

> > >

> > >

> > >

> > > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest dev

> > > build and read up on how to do it. It is quite extraordinary imo.

> > >

> > >

> > >

> > >

> > >

> > >

> > >

> > > Thanks

> > >

> > >

> > >

> > > Brian

> > >

> > >

> > >

> > >

> > >

> > > Brian Valente

> > >

> > > Brianvalentephotography.com

> > >

> > >

> > >

> > > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > > Sent: Sunday, July 9, 2017 1:50 PM

> > > To: Losmandy_users@yahoogroups.com

> > > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> > >

> > >

> > >

> > >

> > >

> > > Brian,

> > >

> > >

> > >

> > > For clarification, my combined (DEC and RA) guiding error is running over 2 arc sec because of the

> > gyrations

> > > in RA.

> > >

> > >

> > >

> > > Rob

> > >

> > >

> >

> >

>

>







----------------------------

#59630 Jul 10, 2017

The PHD2 manual does state the sometimes PEC needs to be turned off while guiding.



Yes, there are a couple mounts where PEC doesn't work well.



However, there are *many* more mounts that have working PEC.



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 6:28 PM

> To: Losmandy_users@yahoogroups.com

> Subject: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> The PHD2 manual does state the sometimes PEC needs to be turned off while guiding.

>

>



----------------------------

#59631 Jul 10, 2017

Ray ��I appreciate what you are saying. ��I also appreciate that you sell products that correct PE and want to push that. I appreciate that because I own PEMPro. ��And while I appreciate your point, I find for many people including me it.s much easier to solve issues with autoguiding first, because it.s easier, there are better tools, and imho the support and user community is better suited to solving these problems. ��I.m not sure why you feel the need to continue poking at this. I already said I agree with most of what you said, I just think it.s easier to do AG first. . �� �� ��Thanks ��Brian �� ��Brian ValenteBrianvalentephotography.com

��From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com] Sent: Monday, July 10, 2017 7:15 PMTo: Losmandy_users@yahoogroups.comSubject: RE: [Losmandy_users] Re: PEC makes tracking worse



----------------------------

#59632 Jul 11, 2017

Just a heads up.. Ren. the writer of the firmware for the Gemini-2does not read the Losmandy Users thread... I asked him if he wasfollowing this thread.. Here is his response:



"No, I don't read the Losmandy user groupcurrently and Idon't get it per email. I think that PEC and autoguiding have tohave problemsbecause these are two control systems working on the same target.I supposethat PEC makes guiding somewhat unpredictable to autoguiding.Predictive filters like Kalman can't work onthis base.OTAH, with really finely tuned P, I and D parameters thecombination can bebrought to a good fit, but this tuning would require in-depthknowledge, lotsof testing and may depend on the circumstances ... a task for aperfectionist8)"



Tom Hilton



----------------------------

#59633 Jul 11, 2017

Yes this makes sense 'I think that PEC and autoguiding have to have problems because these are two control systems working on the same target. I suppose that PEC makes guiding somewhat unpredictable to autoguiding.'



They're two control systems, sampled at different intervals, actuating at different time intervals, and working independently.



Sent from my iPhone

> On Jul 11, 2017, at 5:13 AM, 'tomh@...' tomh@... [Losmandy_users] Losmandy_users@yahoogroups.com> wrote:

>

> I think that PEC and autoguiding have to have problems because these are two control systems working on the same target. I suppose that PEC makes guiding somewhat unpredictable to autoguiding.



----------------------------

#59634 Jul 11, 2017

Hi Brian,

> I also appreciate that you sell products that correct PE and want to push that. I appreciate that because I own

> PEMPro.



I purposely put tag lines in my posts so that there is no appearance that I am trying to hide what software I have produced.



However, I have a day job. I develop astronomy programs in my free time because I *love* to develop software, and I have a special interest in helping people get better results in their astro-photography. I reinvest both profit and my own money back into in astronomy equipment, better software development tools, and computer upgrades, all of which I use to test and make improvements to my software. I also develop free software.



So, I make no apologies if my software does something that can help a user get better results. But it's not about the money. It's about helping people achieve a goal. If I didn't *love* to develop software I would not do it because the little monetary return is not worth the huge fraction of my free time and energy that I devote.



Best regards,



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 10:38 PM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> Ray

>

>

>

> I appreciate what you are saying.

>

>

>

> I also appreciate that you sell products that correct PE and want to push that. I appreciate that because I own

> PEMPro.

>

>

>

> And while I appreciate your point, I find for many people including me it.s much easier to solve issues with

> autoguiding first, because it.s easier, there are better tools, and imho the support and user community is

> better suited to solving these problems.

>

>

>

> I.m not sure why you feel the need to continue poking at this. I already said I agree with most of what you

> said, I just think it.s easier to do AG first.

>

>

>

>

>

>

>

> Thanks

>

>

>

> Brian

>

>

>

>

>

> Brian Valente

>

> Brianvalentephotography.com

>

>

>

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 10, 2017 7:15 PM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

>

>

> Hi Brian,

>

> >> Ray wrote:

> >> I agree, but I think one should start at the lowest level... fix the mount's native tracking characteristics as

> much

> > >as possible so that the software application layer doesn't need to be so responsive.

> > Brian wrote:

> > Well that.s where I guess you and I disagree

>

> Ok, but then I guess you feel you don't need to polar align because the autoguiding algorithm

> will figure the tracking error and match it with autoguider bumps. Right? :-)

>

> -Ray Gralak

> Author of APCC (Astro-Physics Command Center): www.astro-

> physics.com/index.htm?products/accessories/software/apcc/apcc

> Author of PEMPro: www.ccdware.com

> Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> Author of PulseGuide: www.pulseguide.com

> Author of Sigma: www.gralak.com/sigma

>

> > -----Original Message-----

> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > Sent: Monday, July 10, 2017 8:00 AM

> > To: Losmandy_users@yahoogroups.com

> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> >

> >

> >

> > Well that.s where I guess you and I disagree

> >

> >

> >

> > But whatever the approach, I think you and I would agree to solve it one step at a time

> >

> >

> >

> >

> >

> >

> >

> > Thanks

> >

> >

> >

> > Brian

> >

> >

> >

> >

> >

> > Brian Valente

> >

> > Brianvalentephotography.com

> >

> >

> >

> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > Sent: Monday, July 10, 2017 7:58 AM

> > To: Losmandy_users@yahoogroups.com

> > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> >

> >

> >

> >

> >

> > Hi Brian,

> >

> > > I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an

> > arcsec

> > > using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling, so I

> > lose

> > > a few frames while it figures it out.

> >

> > Yes, but depending on how fast the autoguider cycle rate is autoguiding alone reduce PE to less than one

> > arc-sec.

> >

> > If PEC is programmed correctly you will get better results almost immediately.

> >

> > > My point was to try to solve one problem at a time rather than both simultaneously.

> >

> > I agree, but I think one should start at the lowest level... fix the mount's native tracking characteristics as

> much

> > as possible so that the software application layer doesn't need to be so responsive.

> >

> > Best regards,

> >

> > -Ray Gralak

> > Author of APCC (Astro-Physics Command Center): www.astro-

> > physics.com/index.htm?products/accessories/software/apcc/apcc

> > Author of PEMPro: www.ccdware.com

> > Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> > Author of PulseGuide: www.pulseguide.com

> > Author of Sigma: www.gralak.com/sigma

> >

> > > -----Original Message-----

> > > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > > Sent: Monday, July 10, 2017 7:07 AM

> > > To: Losmandy_users@yahoogroups.com

> > > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> > >

> > >

> > >

> > > Ray, I can only speak from my personal experience.

> > >

> > >

> > >

> > > I agree with your point that PPEC is not a substitute, although it is a good start. I reduced pe about an

> > arcsec

> > > using ppec. There is a downside in that it takes a few worm cycles for every new target to get rolling, so I

> > lose

> > > a few frames while it figures it out.

> > >

> > >

> > >

> > > My point was to try to solve one problem at a time rather than both simultaneously.

> > >

> > >

> > >

> > >

> > >

> > > Thanks

> > >

> > >

> > >

> > > Brian

> > >

> > >

> > >

> > >

> > >

> > > Brian Valente

> > >

> > > Brianvalentephotography.com

> > >

> > >

> > >

> > > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > > Sent: Monday, July 10, 2017 5:12 AM

> > > To: Losmandy_users@yahoogroups.com

> > > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> > >

> > >

> > >

> > >

> > >

> > > Hi Brian,

> > >

> > > > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest

> dev

> > > > build and read up on how to do it. It is quite extraordinary imo.

> > >

> > > Looking at the PHD2 forum group messages, Predictive PEC does not always work so well. That's not

> > > surprising though. Having analyzed literally thousands of PE logs from different mounts over the last 15

> > > years I can tell you that in many cases you cannot always accurately predict PEC from just a few points. I

> > > think that the PHD2 documentation even says that the algorithm is not a substitute for PEC because

> PHD2

> > > has to learn each time the scope slews. And if periodic error is properly corrected, any algorithm will work

> > > better, because PEC done correctly will preemptively correct periodic error instead of waiting for a

> > background

> > > software application to try to figure it out.

> > >

> > > BTW, PHD2 only works with some imaging suites because of its non-standard interface. I've been

> > developing

> > > an enhanced version of PHD2 which adds more features and will allow it to integrate better with other

> > > applications, as well as an alternate mechanism that works with any autoguider program that uses an

> > > ASCOM telescope driver for autoguiding. I will make that version available (for free!) when it is available.

> > >

> > > Best regards,

> > >

> > > -Ray Gralak

> > > Author of APCC (Astro-Physics Command Center): www.astro-

> > > physics.com/index.htm?products/accessories/software/apcc/apcc

> > > Author of PEMPro: www.ccdware.com

> > > Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

> > > Author of PulseGuide: www.pulseguide.com

> > > Author of Sigma: www.gralak.com/sigma

> > >

> > > > -----Original Message-----

> > > > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > > > Sent: Sunday, July 9, 2017 8:37 PM

> > > > To: Losmandy_users@yahoogroups.com

> > > > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> > > >

> > > >

> > > >

> > > > You could probably easily get it to under 1 arcsec without any PE correction.

> > > >

> > > >

> > > >

> > > > I update my suggestion . try to solve one first, then the other. I suggest you try to get the best results

> from

> > > your

> > > > PHD2 on guiding and then see if you can improve with PEC.

> > > >

> > > >

> > > >

> > > > The guys at PHD2 are really good and reading your logs and giving you feedback.

> > > >

> > > >

> > > >

> > > > PPEC is an RA guide algorithm introduced somewhere around 2.6.3 . make sure you have the latest

> dev

> > > > build and read up on how to do it. It is quite extraordinary imo.

> > > >

> > > >

> > > >

> > > >

> > > >

> > > >

> > > >

> > > > Thanks

> > > >

> > > >

> > > >

> > > > Brian

> > > >

> > > >

> > > >

> > > >

> > > >

> > > > Brian Valente

> > > >

> > > > Brianvalentephotography.com

> > > >

> > > >

> > > >

> > > > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> > > > Sent: Sunday, July 9, 2017 1:50 PM

> > > > To: Losmandy_users@yahoogroups.com

> > > > Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

> > > >

> > > >

> > > >

> > > >

> > > >

> > > > Brian,

> > > >

> > > >

> > > >

> > > > For clarification, my combined (DEC and RA) guiding error is running over 2 arc sec because of the

> > > gyrations

> > > > in RA.

> > > >

> > > >

> > > >

> > > > Rob

> > > >

> > > >

> > >

> > >

> >

> >

>

>







----------------------------

#59636 Jul 11, 2017

Hi

I believe this statementThere's one point that I think is overlooked. When properly done, PEC corrects the periodic error *before* the autoguider sees the tracking error caused by periodic error.��

would be true for an ideal system. ��However one needs to see the applied commands to the RA and DEC to see if the timing is as would be assumed in that statement. ��IE... Pec make a correction, guiding system samples the correction, calculates the correction it needs to follow the star (and if it new what PECs future corrections would be could correct accordingly), applies the correction before the next PEC sampled correction. ��Auto guiding isn't learning PEC predicted set of corrections, it's following a star.(which depending on the upper atmospheric conditions) may be counter to the last PEC and Autoguiding moves. ��Luckily these two systems aren't driving a car.

But please correct me. I'm willing to learn more details about this system. I'm just speaking from a distant engineering experience.

Cheers��

----- Original Message -----From: 'Ray Gralak (Groups)' groups3@... [Losmandy_users] Losmandy_users@yahoogroups.com>To: Losmandy users Losmandy_users@yahoogroups.com>Sent: Tue, 11 Jul 2017 13:13:42 -0000 (UTC)Subject: RE: [Losmandy_users] PEC makes tracking worse





��Hi Tom,

> No, I don't read the Losmandy user group currently and I don't get it per email. I think that PEC and

> autoguiding have to have problems because these are two control systems working on the same target. I

> suppose that PEC makes guiding somewhat unpredictable to autoguiding.

There's one point that I think is overloHioked. When properly done, PEC corrects the periodic error *before* the autoguider sees the tracking error caused by periodic error.

What is left should be mostly residual drift, which should requires much less significant autoguider movements.

-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma

> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Tuesday, July 11, 2017 5:13 AM

> To: Losmandy_users@yahoogroups.com

> Subject: Re: [Losmandy_users] PEC makes tracking worse

>

>

>

> Just a heads up. Ren�� the writer of the firmware for the Gemini-2 does not read the Losmandy Users thread.

> I asked him if he was following this thread. Here is his response:

>

> "

>

> No, I don't read the Losmandy user group currently and I don't get it per email. I think that PEC and

> autoguiding have to have problems because these are two control systems working on the same target. I

> suppose that PEC makes guiding somewhat unpredictable to autoguiding.

>

> Predictive filters like Kalman can't work on this base. OTAH, with really finely tuned P, I and D parameters the

> combination can be brought to a good fit, but this tuning would require in-depth knowledge, lots of testing

> and may depend on the circumstances ... a task for a perfectionist 8)"

>

>

>

> Tom Hilton

>

>

>

>



----------------------------

#59637 Jul 11, 2017

would be true for an ideal system. However one needs to see the applied commands to the RA and DEC to

> see if the timing is as would be assumed in that statement. IE... Pec make a correction, guiding system

> samples the correction, calculates the correction it needs to follow the star (and if it new what PECs future



Let's say an that the autoguider potentially makes a correction every 5 seconds. If an average correction is a 100 msec move, that leaves 4.9 seconds during which PEC is actively correcting periodic error without interference from the autoguider.



If the PEC curve is accurate most of the periodic error would be removed during that 4.9 seconds, leaving just residual drift. When the autoguider exposure completes the image will be the average of the drift + scintillation, not the drift + scintillation + periodic error. The only interference would come when there is an autoguider move. The periodic error is still removed for the rest of the period between moves.



And even if moves are made once a second, that would still allow PEC to work 90% of the time, assuming there is a 100 msec move each second.



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Tuesday, July 11, 2017 6:54 AM

> To: Losmandy users

> Subject: RE: [Losmandy_users] PEC makes tracking worse

>

>

>

> Hi

> I believe this statement

>

> There's one point that I think is overlooked. When properly done, PEC corrects the periodic error *before* the

> autoguider sees the tracking error caused by periodic error.

>

> would be true for an ideal system. However one needs to see the applied commands to the RA and DEC to

> see if the timing is as would be assumed in that statement. IE... Pec make a correction, guiding system

> samples the correction, calculates the correction it needs to follow the star (and if it new what PECs future

> corrections would be could correct accordingly), applies the correction before the next PEC sampled

> correction. Auto guiding isn't learning PEC predicted set of corrections, it's following a star.(which depending

> on the upper atmospheric conditions) may be counter to the last PEC and Autoguiding moves. Luckily these

> two systems aren't driving a car.

> But please correct me. I'm willing to learn more details about this system. I'm just speaking from a distant

> engineering experience.

> Cheers

> ----- Original Message -----

> From: 'Ray Gralak (Groups)' groups3@... [Losmandy_users] Losmandy_users@yahoogroups.com>

> To: Losmandy users Losmandy_users@yahoogroups.com>

> Sent: Tue, 11 Jul 2017 13:13:42 -0000 (UTC)

> Subject: RE: [Losmandy_users] PEC makes tracking worse

>

>

> Hi Tom,

>

>

> > No, I don't read the Losmandy user group currently and I don't get it per email. I think that PEC and

>

> > autoguiding have to have problems because these are two control systems working on the same target. I

>

> > suppose that PEC makes guiding somewhat unpredictable to autoguiding.

>

>

> There's one point that I think is overloHioked. When properly done, PEC corrects the periodic error *before*

> the autoguider sees the tracking error caused by periodic error.

>

>

> What is left should be mostly residual drift, which should requires much less significant autoguider

> movements.

>

>

> -Ray Gralak

>

> Author of APCC (Astro-Physics Command Center): www.astro-

> physics.com/index.htm?products/accessories/software/apcc/apcc

>

> Author of PEMPro: www.ccdware.com

>

> Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

>

> Author of PulseGuide: www.pulseguide.com

>

> Author of Sigma: www.gralak.com/sigma

>

>

> > -----Original Message-----

>

> > From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

>

> > Sent: Tuesday, July 11, 2017 5:13 AM

>

> > To: Losmandy_users@yahoogroups.com

>

> > Subject: Re: [Losmandy_users] PEC makes tracking worse

>

> >

>

> >

>

> >

>

> > Just a heads up. Ren. the writer of the firmware for the Gemini-2 does not read the Losmandy Users

> thread.

>

> > I asked him if he was following this thread. Here is his response:

>

> >

>

> > "

>

> >

>

> > No, I don't read the Losmandy user group currently and I don't get it per email. I think that PEC and

>

> > autoguiding have to have problems because these are two control systems working on the same target. I

>

> > suppose that PEC makes guiding somewhat unpredictable to autoguiding.

>

> >

>

> > Predictive filters like Kalman can't work on this base. OTAH, with really finely tuned P, I and D parameters

> the

>

> > combination can be brought to a good fit, but this tuning would require in-depth knowledge, lots of testing

>

> > and may depend on the circumstances ... a task for a perfectionist 8)"

>

> >

>

> >

>

> >

>

> > Tom Hilton

>

> >

>

> >

>

> >

>

> >

>

>







----------------------------

#59638 Jul 11, 2017

If the PEC curve is accurate most of the periodic error would be removed��



Yes, the point is that the autoguider will not see much of the periodic error if it's being corrected by PEC, and therefore there will be very little conflict between the two. Properly programmed PEC makes autoguiding easier, with fewer and smaller corrections.

Regards,

�� �� -Paul



----------------------------

#59639 Jul 11, 2017

You and everyone else helped me. Last night I took photos ngc 6924. Admittedly, it is much higher toward the zenith than m20. I used pempro to make a pec curve and the predictive algorithm in phd2. When at 4:30 I got up to put my equipment away, I was amazed to find that the combined RMS average was .48 arc seconds.��

The only problem was that the shoot terminated an hour early after Meridian flip. I am sure this happened because a cable got tangled--always something more to perfect.

Thanks, everyone, for your time and kind responses.��

Rob

Sent from my iPhone

On Jul 11, 2017, at 6:13 AM, 'Ray Gralak (Groups)' groups3@... [Losmandy_users] Losmandy_users@yahoogroups.com> wrote:



��Hi Tom,



> No, I don't read the Losmandy user group currently and I don't get it per email. I think that PEC and

> autoguiding have to have problems because these are two control systems working on the same target. I

> suppose that PEC makes guiding somewhat unpredictable to autoguiding.



There's one point that I think is overlooked. When properly done, PEC corrects the periodic error *before* the autoguider sees the tracking error caused by periodic error.



What is left should be mostly residual drift, which should requires much less significant autoguider movements.



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Tuesday, July 11, 2017 5:13 AM

> To: Losmandy_users@yahoogroups.com

> Subject: Re: [Losmandy_users] PEC makes tracking worse

>

>

>

> Just a heads up. Ren. the writer of the firmware for the Gemini-2 does not read the Losmandy Users thread.

> I asked him if he was following this thread. Here is his response:

>

> "

>

> No, I don't read the Losmandy user group currently and I don't get it per email. I think that PEC and

> autoguiding have to have problems because these are two control systems working on the same target. I

> suppose that PEC makes guiding somewhat unpredictable to autoguiding.

>

> Predictive filters like Kalman can't work on this base. OTAH, with really finely tuned P, I and D parameters the

> combination can be brought to a good fit, but this tuning would require in-depth knowledge, lots of testing

> and may depend on the circumstances ... a task for a perfectionist 8)"

>

>

>

> Tom Hilton

>

>

>

>



----------------------------

#59640 Jul 13, 2017

I also had been trying to get PEC to work on my new G-11 mount the past few nights but after I trained PEC and I turn PEC on my PE gets much worse.�� Does PEC even work on this mount?�� �� Neilson



----------------------------

#59641 Jul 13, 2017

>.Does PEC even work on this mount?



----------------------------

#59642 Jul 13, 2017

I've been able to get PEC and PHD2 to work together very well. I used Pempro to create a more accurate PEC curve. After uploading the curve. I used guiding assistant in PHD2 with PEC running. It should be noted that I am using the PPEC algorithm in PHD2. I've had excellent tracing results with this combination on my G-11.��

Rob



----------------------------

#59643 Jul 13, 2017

>..It should be noted that I am using the PPEC algorithm in PHD2.



----------------------------

#59644 Jul 13, 2017

Paul

My understanding is that they work differently but complementary

PEC is more of a curve that is loaded into the mount, and automatically makes those corrections, whereas PPEC algorithm uses observed changes over time to predict periodic errors, and then corrects those.��

If you guess that PPEC doesn't work for the first few worm cycles, you are correct! it is training on the fly.

If used together, the observed errors already have PEC factored in, and PEC corrections don't change once programmed into the mount, so they should be very complementary



Brian



----------------------------

#59645 Jul 14, 2017

What I suspect I'm seeing with PHD + PEC is that PEC and PHD both "see" the same discrepancy at the

> same time, and both end up sending the same corrective signal to the mount, thus causing overshoot. Since

> the PEC recording was made by PHD itself, the corrective pulses recorded by PEC happen just *after* the RA

> discrepancy is noted. With PHD running as well, the same RA discrepancy happens, and PHD notices and

> signals a corrective pulse, without knowing that PEC is going to issue a corrective pulse too! The result is a

> "doubled" correction that causes overshoot and an overall poorer tracking experience.



I think this is an example of the real problem: a poor PEC curve.



There are a number of problems with recording PEC via an autoguider program. I first noticed these problems over 15 years ago when programming PEC via autoguider.



1) Stellar scintillation can cause errors to be programmed into the PEC curve. For example, if there is particularly bad seeing during one autoguider cycle, a "bad" move would get programmed in that part of the PEC cycle.



So, later during normal autoguiding, the autoguider program might appear to be fighting PEC because large corrections have to be made in every worm cycle where there were "bad" moves programmed into the PEC.



2) The autoguider corrections are made "after the fact", so the PEC corrections are delayed. That is, when the autoguider makes a correction, the correction is for past periodic error, not for periodic error at the moment of the correction.



3) If the autoguider program is not setup correctly, the corrections it issues may under-correct periodic error.



4) The autoguider corrections cause the PEC curve to be made up of discrete moves at regular intervals instead of one continuous curve.



A "good" PEC curve would have a continuous curve throughout the worm cycle, so that PEC is continuously corrected. This is important because when an autoguider correction comes in the correction only "interferes" with a small portion of the PEC curve, thus the PEC can operate effectively most of the time.



An autoguider programs PEC corrections as discrete "jumps" that are played back when PEC is enabled. So, it is possible during regular autoguiding that some autoguider moves will overlap the discrete moves preprogrammed into the PEC table, thus rendering the PEC curve ineffective for that autoguider cycle because, unlike if there was a continuous PEC curve, there would be no more corrections in the mount's PEC until the next discrete move in PEC. In fact, it is possible that if one uses the same autoguider interval programming PEC as when doing regular autoguiding that an overlap could occur almost every cycle, which would mostly render the PEC curve ineffective!



So, I believe that using an autoguider to program PEC is almost always a compromise.



Best regards,



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Thursday, July 13, 2017 6:05 PM

> To: Losmandy_users@yahoogroups.com

> Subject: Re: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> >> It should be noted that I am using the PPEC algorithm in PHD2.

>

>

> That may be the "secret sauce". If PHD is able to effectively "predict" the RA movements initiated by PEC (as

> opposed to those introduced by actual tracking errors) then I could see why they would play nicely together.

> What I suspect I'm seeing with PHD + PEC is that PEC and PHD both "see" the same discrepancy at the

> same time, and both end up sending the same corrective signal to the mount, thus causing overshoot. Since

> the PEC recording was made by PHD itself, the corrective pulses recorded by PEC happen just *after* the RA

> discrepancy is noted. With PHD running as well, the same RA discrepancy happens, and PHD notices and

> signals a corrective pulse, without knowing that PEC is going to issue a corrective pulse too! The result is a

> "doubled" correction that causes overshoot and an overall poorer tracking experience.

>

>

> PPEC, after 8x worms, would theoretically begin to notice that the guide pulses aren't actually needed

> (because PEC ends up taking care of them) and so it scales back its corrective movements resulting in

> smooth guiding.

>

>

> ...just a theory.

>

> On Thu, Jul 13, 2017 at 4:58 PM sactowriter@... [Losmandy_users]

> Losmandy_users@yahoogroups.com> wrote:

>

>

>

>

>

> .I've been able to get PEC and PHD2 to work together very well. I used Pempro to create a more

> accurate PEC curve. After uploading the curve. I used guiding assistant in PHD2 with PEC running. It should

> be noted that I am using the PPEC algorithm in PHD2. I've had excellent tracing results with this combination

> on my G-11.

>

> .Rob

>

>







----------------------------

#59646 Jul 14, 2017

There are difference between programming PEC using PemPro and programming it directly into Gemini using an autoguider.

When using PemPro and similar software, PEC curve is first computed based on the collected data, and then many frequent, tiny corrections are programmed into the mount to apply this computed curve. The programmed PEC is smooth, and changes in step with the actual error occurring at the worm, often many times a second.

This is why PemPro curve has very little chance of interfering with autoguider input since the error is corrected very, very often, and in tiny increments, as the error is occurring. Autoguider will not see any large error that it will need to correct, since it's being corrected as it occurs.



----------------------------

#59647 Jul 14, 2017

Hi Ray,

> I think this is an example of the real problem: a poor PEC curve.��

Funny.. I was posting almost exactly the same explanation, but you beat me to it :)

Regards,

�� �� -Paul



----------------------------

#59650 Jul 14, 2017

>So, I believe that using an autoguider to program PEC is almost always a compromise.



totally agree. when I think about PEC i always think of pempro or similar. Programming via autoguider and THEN using autoguiding seems like asking for trouble (and PPEC basically replaces this kind of approach imho).

Brian



----------------------------

#59651 Jul 14, 2017

Hi Brian,

> >>So, I believe that using an autoguider to program PEC is almost always a compromise.

>

>

>

> totally agree. when I think about PEC i always think of pempro or similar. Programming via autoguider and

> THEN using autoguiding seems like asking for trouble (and PPEC basically replaces this kind of approach

> imho).



Except:



1) that PPEC has to relearn each time the scope is moved



And,



2) Just like autoguiding, it does not provide the continuous correction that a proper PEC curve can.



The best solution for now is probably to use PEC + PPEC.



BTW, not that there is anything inherently bad about PHD2, but having to use PHD2 to get PPEC is a limitation that I plan to solve.



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Friday, July 14, 2017 7:57 AM

> To: Losmandy_users@yahoogroups.com

> Subject: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> >>So, I believe that using an autoguider to program PEC is almost always a compromise.

>

>

>

> totally agree. when I think about PEC i always think of pempro or similar. Programming via autoguider and

> THEN using autoguiding seems like asking for trouble (and PPEC basically replaces this kind of approach

> imho).

>

>

> Brian

>

>



----------------------------

#59652 Jul 14, 2017

>The best solution for now is probably to use PEC + PPEC.

agreed.��

my only point was if your plan is to use autoguider to program PEC, you'd be better off not doing that and just using PPEC.��

But the best results are a god PEC + autoguide





Brian



----------------------------

#59655 Jul 14, 2017

Neilson

groups.yahoo.com/neo/groups/Losmandy_users/conversations/topics/59574



Bryan

---In Losmandy_users@yahoogroups.com, neilsonshepard@...> wrote :

I am not able to get my PEC trained from the mount itself to work. Unfortunately my free trial of Pempro was wasted on a CEM60 that was defective.�� Pempro is an excellent software for training your PEC but I cannot afford to buy it.So is there any other way to train my PEC that works. I wanted to use PEC while guiding and unguided but I can't get it to work. ��I guide my mount with PHD2 and I select train PEC using the hand controller. When it's finished I have PEC turned on and check my PE using PHD2 unguided but the PE is more than doubled.��

Neilson



----------------------------

#59660 Jul 14, 2017

my only point was if your plan is to use autoguider to program PEC, you'd be better off not doing that and just

> using PPEC.



The one caveat, and my original point, is that PPEC doesn't always work well according to issues reported on the PHD2 forum.



Given what I have seen in PE curves over the last 15 years it is not surprising that the algorithm can make mistakes.



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Friday, July 14, 2017 8:11 AM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> >>The best solution for now is probably to use PEC + PPEC.

>

>

> agreed.

>

>

> my only point was if your plan is to use autoguider to program PEC, you'd be better off not doing that and just

> using PPEC.

>

>

> But the best results are a god PEC + autoguide

>

>

>

>

>

>

> Brian

>

>

>

>







----------------------------

#59667 Jul 17, 2017

>The one caveat, and my original point, is that PPEC doesn't always work well according to issues reported on the PHD2 forum.��

perhaps in some cases, but in actual use I've found it to be quite good and an improvement over standard algorithm. it's not just predictive PE but folds in elements of hysteresis too, and of course you can control how much of each.

Again, not a substitute for a good quality PE correction plus guiding, but a solid step forward.��





Brian



----------------------------

#59671 Jul 17, 2017

Hi Brian,

> Again, not a substitute for a good quality PE correction plus guiding, but a solid step forward.



I'm not so sure it's a big improvement over simply autoguiding if no PEC is used. If it calculates incorrectly you can get worse guiding, which is exactly what has been reported. I think that the only reason it can work well with PEC is that that with PEC there would only be linear drift for the predictive algorithm to remove, and that.s a pretty easy calculation.



BTW, I read the PPEC author's paper and I think he makes a number of poor assumptions. I'm not surprised he had to add hysteresis because using guider pulses to determine periodic error is often non-deterministic. :-)



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Monday, July 17, 2017 8:01 AM

> To: Losmandy_users@yahoogroups.com

> Subject: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> >>The one caveat, and my original point, is that PPEC doesn't always work well according to issues reported

> on the PHD2 forum.

>

>

> perhaps in some cases, but in actual use I've found it to be quite good and an improvement over standard

> algorithm. it's not just predictive PE but folds in elements of hysteresis too, and of course you can control how

> much of each.

>

>

> Again, not a substitute for a good quality PE correction plus guiding, but a solid step forward.

>

>

>

> Brian

>

>

>



----------------------------

#59672 Jul 18, 2017

Hi Ray

>>I'm not so sure it'sa big improvement over simply autoguiding if no PEC is used.



All I can say is my results speak for themselves. PPEC is an excellent guide algorithm and a marked improvement over hysteresis for me, and a number of people i've seen on the openPHD forum.

I have not seen any "worse guiding" reports after the beta where they were shaking out the bugs. I have seen plenty of people trying to figure out how to configure it, and in most (all?) cases they resolved it and had much better results.

If you are using it yourself I'd love to see some of your results and why you think it's worse. I'm more interested in actual guiding results than the author's paper.

Personally I think it's great, the guys do a great job supporting it on their own time. I don't think you need to continuously cast dispersions on this algorithm, particularly if you aren't using it or participating in it.

It's open source, so I would hope any input you do have you would add back to the open source community by reporting bugs, making suggestions, etc.��As far as I can tell you aren't participating in that, and that's fine. I support you through purchase of your software and it's great stuff.��



It's mystifying to me why it appears you only disparage progress which isn't very constructive imho.��



Brian



----------------------------

#59673 Jul 18, 2017

Hi Brian,

> I have not seen any "worse guiding" reports after the beta where they were shaking out the bugs. I have seen

> plenty of people trying to figure out how to configure it, and in most (all?) cases they resolved it and had

> much better results.



Easy to say, but kind of hard to prove. With exactly the same settings autoguiding results can change based on changes to seeing conditions. The time of year and time of night can matter.

> It's open source, so I would hope any input you do have you would add back to the open source community

> by reporting bugs, making suggestions, etc. As far as I can tell you aren't participating in that, and that's fine. I

> support you through purchase of your software and it's great stuff.



I have contributed code to the PHD2 project. I also offered to test and provide feedback for the PPEC project at it's infancy.

> >>I'm not so sure it's a big improvement over simply autoguiding if no PEC is used.

>



> It's mystifying to me why it appears you only disparage progress which isn't very constructive imho.



I don't think this comment was warranted. The topic is about "PEC making tracking worse", which I think in an earlier post I explained in detail possible reasons why that might happen. The topic wasn't about PPEC, but I think I explained why PPEC will not provide the best auto-guiding possible. The reasons are simple facts, just like for example, "mount X" tracking better than "mount Y".



To recap:



1. First, PEC and auto-guiding don't fight each other. PPEC acts like "auto-guiding" to the mount.



2. A good PEC curve provides continuous tracking correction for periodic error, while any form of autoguiding, including PPEC, only provides stair-stepped corrections.



3. With a good PEC curve, autoguiding with PPEC immediately becomes better because there is nothing for PPEC to "learn" except for the residual drift.



4. Autoguider moves are non-deterministic. This means that a move commanded for X milliseconds doesn't always (in fact not usually) moves for X milliseconds. In fact, I have found huge time discrepancies for some mounts. So, why does this matter? Because the PPEC algorithm is accumulating the supposed times for those moves, but they can be often very inaccurate! This means that the amplitude and phase of a periodic movement may be miscalculated by the algorithm.



5. Thus, because of the previous item, PPEC inherently won't always work effectively.



So, why should you believe me? Because I have lots of experience with the tracking characteristics and periodic error correction of many mounts, going back over 15 years



Anyway, I'm glad you are happy with the autoguiding performance that you have, but I don't think it is the best possible.



Best regards,



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Tuesday, July 18, 2017 7:44 AM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> Hi Ray

>

> >>I'm not so sure it's a big improvement over simply autoguiding if no PEC is used.

>

>

>

> All I can say is my results speak for themselves. PPEC is an excellent guide algorithm and a marked

> improvement over hysteresis for me, and a number of people i've seen on the openPHD forum.

>

> I have not seen any "worse guiding" reports after the beta where they were shaking out the bugs. I have seen

> plenty of people trying to figure out how to configure it, and in most (all?) cases they resolved it and had

> much better results.

>

> If you are using it yourself I'd love to see some of your results and why you think it's worse. I'm more

> interested in actual guiding results than the author's paper.

>

> Personally I think it's great, the guys do a great job supporting it on their own time. I don't think you need to

> continuously cast di spersions on this algorithm, particularly if you aren't using it or participating in it.

>

> It's open source, so I would hope any input you do have you would add back to the open source community

> by reporting bugs, making suggestions, etc. As far as I can tell you aren't participating in that, and that's fine. I

> support you through purchase of your software and it's great stuff.

>

>

> It's mystifying to me why it appears you only disparage progress which isn't very constructive imho.

>

>

> Brian

>

>

>







----------------------------

#59685 Jul 20, 2017

well Ray, it seems this thread is just you and me, so i'll leave it at this:

1. PEC plus autoguiding (and imho likely PPEC) gives the best results. you made this comment, i made it before, I think we are in agreement

2. trying to use an autoguider for PEC training is not a great idea, probably worse results than if you just autoguided alone. I think we are in agreement on that too

3. no one including me is trying to disparage PEC. quite the contrary, it's an important and good tool. It's also true i have seen many threads with people struggling to get good results with PEC: inverted correction, not enough cycles, using autoguider etc. these are all solveable, that's why we have forums like this right?��

4. no one including me is suggesting PPEC is a replacement for good PEC. I think we agree on that.

5. you can say whatever you want about PPEC, but again the results (not theoretical, i'm talking actual results) are consistently better than standard hysteresis algorithm for many people who have used it correctly. I you have specific data you want to post about PPEC as an algorithm being worse, love to see it. Otherwise while i think we all understand autoguiding is at best approximate. PPEC approximates it better than other algorithms.

I was on the PPEC beta, it was an open and public beta. I don't recall seeing you in any of those threads, so if you are working behind the scenes, I guess that's great? not sure why though.��



----------------------------

#59688 Jul 20, 2017

Hi Brian,

> 1. PEC plus autoguiding (and imho likely PPEC) gives the best results. you made this comment, i made it

> before, I think we are in agreement



But are we? I thought you previously disagreed with me that correcting tracking should start from a lower level (i.e. mount mechanicals, then moving up, PEC) instead of trying to correct errors after the fact at the application level. I could find that post if you don't recall? Or, did I just misunderstand what you meant?

> 2. trying to use an autoguider for PEC training is not a great idea, probably worse results than if you just

> autoguided alone. I think we are in agreement on that too



I don't think you are following through with your own logic... since PPEC uses autoguiding to learn, that is exactly why it may not be able to always correctly analyze amplitude and phase of periodic error.

> 5. you can say whatever you want about PPEC, but again the results (not theoretical, i'm talking actual

> results) are consistently better than standard hysteresis algorithm for many people who have used it correctly.



Instead of just saying that you *think* algorithm X is better than Y, how about some concrete reasons why? As I said previously, atmospheric seeing can change from moment to moment. Settings that worked well 5 minutes before might not work well later. It's hard to make a conclusion under random skies.



And even if there looks like improvement in a few cases, that doesn't prove that there will always be improvement with all mounts. There is inherent error in the algorithm caused by the indeterminate nature of small mount moves. This is not a guess on my part. I have done lots of analysis across many mount types. I've even put logic in software to try to account for those deviations. Apparently some aren't even aware that this can happen!

> I you have specific data you want to post about PPEC as an algorithm being worse, love to see it. Otherwise

> while i think we all understand autoguiding is at best approximate. PPEC approximates it better than other

> algorithms.



Even if that were true, it's a moot point. It seems to me that a smart user would just program PEC and be done with it. Then the user wouldn't have to wait those few minutes for the algorithm to try to converge. Nor would they have to face possible bad algorithm calculations or unanticipated move distances.

> I was on the PPEC beta, it was an open and public beta. I don't recall seeing you in any of those threads, so if

> you are working behind the scenes, I guess that's great? not sure why though.



I said I offered to help. You will have to ask the PHD2 list owner why he would not let my post go through. I've asked, but have not received a *good* answer! But that's OK, I'm developing an alternative solution that is already working better (yes that should answer your question about if I have tried PPEC). I am working on submitting my technique for patent approval before anyone can copy the idea. In the long term, it may be a moot point though because I think that within 10 years even inexpensive mounts will likely have encoders to correct PE.



Best regards,



-Ray Gralak

Author of APCC (Astro-Physics Command Center): www.astro-physics.com/index.htm?products/accessories/software/apcc/apcc

Author of PEMPro: www.ccdware.com

Author of Astro-Physics V2 ASCOM Driver: www.gralak.com/apdriver

Author of PulseGuide: www.pulseguide.com

Author of Sigma: www.gralak.com/sigma



> -----Original Message-----

> From: Losmandy_users@yahoogroups.com [mailto:Losmandy_users@yahoogroups.com]

> Sent: Thursday, July 20, 2017 11:47 AM

> To: Losmandy_users@yahoogroups.com

> Subject: RE: [Losmandy_users] Re: PEC makes tracking worse

>

>

>

> well Ray, it seems this thread is just you and me, so i'll leave it at this:

>

> 1. PEC plus autoguiding (and imho likely PPEC) gives the best results. you made this comment, i made it

> before, I think we are in agreement

>

> 2. trying to use an autoguider for PEC training is not a great idea, probably worse results than if you just

> autoguided alone. I think we are in agreement on that too

>

> 3. no one including me is trying to disparage PEC. quite the contrary, it's an important and good tool. It's also

> true i have seen many threads with people struggling to get good results with PEC: inverted correction, not

> enough cycles, using autoguider etc. these are all solveable, that's why we have forums like this right?

>

> 4. no one including me is suggesting PPEC is a replacement for good PEC. I think we agree on that.

>

> 5. you can say whatever you want about PPEC, but again the results (not theoretical, i'm talking actual

> results) are consistently better than standard hysteresis algorithm for many people who have used it correctly.

> I you have specific data you want to post about PPEC as an algorithm being worse, love to see it. Otherwise

> while i think we all understand autoguiding is at best approximate. PPEC approximates it better than other

> algorithms.

>

> I was on the PPEC beta, it was an open and public beta. I don't recall seeing you in any of those threads, so if

> you are working behind the scenes, I guess that's great? not sure why though.

>

> B

>

>

>







----------------------------

#59689 Jul 20, 2017

Any chance you could move this to a private discussion?

Bryam



----------------------------

#59692 Jul 21, 2017

Bryan,

Why on Earth should they do that? Do you know all of this stuff already, have you tested these ideas and assertions? If you have then please contribute your results and experiences. If you have not please let the rest of us learn and consider the topic issues. These discussions are what makes boards like this have value for many of us, if you do not want to follow along don't click on the topic.��

Chip��



Contact Us
This Site's Privacy Policy
Google's privacy policies

S
e
n
i
o
r
T
u
b
e
.
o
r
g