Re: quick question on FITS Liberator


Jul 12, 2005

 


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

#19312 Jul 12, 2005

When I open fits files generated by MaxIm I find that the file is

displayed upside down in PS (as opposed to the way it is displayed in

MaxIm). Is this to be expected or have I done something wrong either

in MaxIm or with FITS Liberator?



Mike



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

#19315 Jul 12, 2005

It is normal- FITS Liberator assumes a different origin for the data. Just

do a 'flip vertical' in PS to get the image back to the correct orientation.



Chris



*****************************************

Chris L Peterson

Cloudbait Observatory

www.cloudbait.com



----- Original Message -----

From: "zeidler02" mzeidler@...>

To: MaxImDL@yahoogroups.com>

Sent: Tuesday, July 12, 2005 11:14 AM

Subject: [MaxImDL] quick question on FITS Liberator





> When I open fits files generated by MaxIm I find that the file is

> displayed upside down in PS (as opposed to the way it is displayed in

> MaxIm). Is this to be expected or have I done something wrong either

> in MaxIm or with FITS Liberator?

>

> Mike



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

#19318 Jul 12, 2005

Isn't the origin for FITS data part of the standard?





> It is normal- FITS Liberator assumes a different origin for the data. Just

> do a 'flip vertical' in PS to get the image back to the correct orientation.

>

> Chris

>

> *****************************************

> Chris L Peterson

> Cloudbait Observatory

> www.cloudbait.com

>

>

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

> From: "zeidler02" mzeidler@...>

> To: MaxImDL@yahoogroups.com>

> Sent: Tuesday, July 12, 2005 11:14 AM

> Subject: [MaxImDL] quick question on FITS Liberator

>

>

> > When I open fits files generated by MaxIm I find that the file is

> > displayed upside down in PS (as opposed to the way it is displayed in

> > MaxIm). Is this to be expected or have I done something wrong either

> > in MaxIm or with FITS Liberator?

> >

> > Mike

>

>

>

>

> Yahoo! Groups Links

>

>

>

>

>

>

>



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

#19319 Jul 12, 2005

Adam Block wrote: > Isn't the origin for FITS data part of the standard?



Nope!



Doug



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



Doug George

dgeorge@...



Diffraction Limited

Makers of Cyanogen Imaging Products

www.cyanogen.com



25 Conover Street

Ottawa, Ontario,

Canada, K2G 4C3



Phone: (613) 225-2732

Fax: (613) 225-9688



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



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

#19320 Jul 12, 2005

That's the good thing about standards. There are so many of them (and

variations). Take Micro$oft for example, ...



Adam Block wrote: > Isn't the origin for FITS data part of the standard?

>

>

>

>

>>It is normal- FITS Liberator assumes a different origin for the data. Just

>>do a 'flip vertical' in PS to get the image back to the correct orientation.

>>

>>Chris

>>

>>*****************************************

>>Chris L Peterson

>>Cloudbait Observatory

>>www.cloudbait.com

>>

>>

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

>>From: "zeidler02" mzeidler@...>

>>To: MaxImDL@yahoogroups.com>

>>Sent: Tuesday, July 12, 2005 11:14 AM

>>Subject: [MaxImDL] quick question on FITS Liberator

>>

>>

>>

>>>When I open fits files generated by MaxIm I find that the file is

>>>displayed upside down in PS (as opposed to the way it is displayed in

>>>MaxIm). Is this to be expected or have I done something wrong either

>>>in MaxIm or with FITS Liberator?

>>>

>>>Mike

>>

>>

>>

>>

>>Yahoo! Groups Links

>>

>>

>>

>>

>>

>>

>>

>

>

>

>

>

> Yahoo! Groups Links

>

>

>

>

>

>

>



--





Carpe Noctem



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

#19323 Jul 12, 2005

Hmmm, I guess not. But there is a commonly accepted way of interpreting the

data. I recall a note from the FITS Liberator authors that they considered

the image in PS to be flipped vertically, but that they weren't planning on

doing anything about it.



Normally, you would have WCS coordinates in your data so the display

orientation shouldn't matter. Among amateur images, I see all kinds of

orientations, too. Many people using diagonals or AO7s don't remember to

reverse their images before publishing them. So really, it doesn't much

matter where the origin of a FITS array is interpreted to be.



Chris



*****************************************

Chris L Peterson

Cloudbait Observatory

www.cloudbait.com



----- Original Message -----

From: "Douglas B. George" dg@...>

To: MaxImDL@yahoogroups.com>

Sent: Tuesday, July 12, 2005 2:04 PM

Subject: Re: [MaxImDL] quick question on FITS Liberator





> Adam Block wrote:

>> Isn't the origin for FITS data part of the standard?

>

> Nope!

>

> Doug







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

#19324 Jul 12, 2005

Chris Peterson wrote: > Yeah, but somebody got it wrong.



No. There is no definition of this in the FITS standard. Some applications do

it one way, others do it the other way. Very annoying.



FITS is a rather uneven standard. They've adopted a very comprehensive set of

"World Coordinate Systems" keywords for recording astrometric solutions, yet it

lacks a standard keyword for exposure time! Not to mention CCD temperature,

temperature setpoint, filters, exposure type, and all kinds of other useful

things. So you have to make those up for yourselves as so-called "Additional

Keywords."



Some things have become more-or-less "defacto standards," largely because

someone coded it a certain way when they wrote a piece of IRAF.



To make sure that our software interoperates correctly, SBIG, Bisque and

Diffraction put our heads together and created our own standard for Additional

Keywords. That standard is available on the SBIG website.



Doug



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



Doug George

dgeorge@...



Diffraction Limited

Makers of Cyanogen Imaging Products

www.cyanogen.com



25 Conover Street

Ottawa, Ontario,

Canada, K2G 4C3



Phone: (613) 225-2732

Fax: (613) 225-9688



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



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

#19330 Jul 12, 2005

I had several discussions about this probably over a year ago on this

list and the MIRA list because Michael Newberry stands by the claim

that the origin of the data is at the bottom left and not the top left

and more than one person has told me that is what WCS specifies but I

have yet to get any of them to point me to a paragraph in the spec. So

the FITS Liberator and MIRA display the image the same way which is

flipped from MaxImDL. After several of us complained we finally got

Michael to add a user preference in MIRA to flip the image upon opening

but I'm sure it's secretly dubbed the "stupid amateur option" since

several claim that professionals always interpret the data the other

way.



That's true about flipping the images. I was always guilty of that

myself when using an AO7. The funny thing is I never had anyone tell

me an image was backwards unless it was the HorseHead. Plus no one ever

seemed to mind the severe rotation angles to get a good guide star. :)



Gil Jones

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

> Subject: Re: [MaxImDL] quick question on FITS Liberator

> From: "Chris Peterson" clp@...>

> Date: Tue, July 12, 2005 1:28 pm

> To: MaxImDL@yahoogroups.com>

>

> Hmmm, I guess not. But there is a commonly accepted way of interpreting the

> data. I recall a note from the FITS Liberator authors that they considered

> the image in PS to be flipped vertically, but that they weren't planning on

> doing anything about it.

>

> Normally, you would have WCS coordinates in your data so the display

> orientation shouldn't matter. Among amateur images, I see all kinds of

> orientations, too. Many people using diagonals or AO7s don't remember to

> reverse their images before publishing them. So really, it doesn't much

> matter where the origin of a FITS array is interpreted to be.

>

> Chris

>

> *****************************************

> Chris L Peterson

> Cloudbait Observatory

> www.cloudbait.com

>

>

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

> From: "Douglas B. George" dg@...>

> To: MaxImDL@yahoogroups.com>

> Sent: Tuesday, July 12, 2005 2:04 PM

> Subject: Re: [MaxImDL] quick question on FITS Liberator

>

>

> > Adam Block wrote:

> >> Isn't the origin for FITS data part of the standard?

> >

> > Nope!

> >

> > Doug

>

>

>

>

> Yahoo! Groups Links

>

>

>



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

#19334 Jul 12, 2005

Gil Jones wrote: > I had several discussions about this probably over a year ago on this

> list and the MIRA list because Michael Newberry stands by the claim

> that the origin of the data is at the bottom left and not the top left

> and more than one person has told me that is what WCS specifies but I

> have yet to get any of them to point me to a paragraph in the spec. So

> the FITS Liberator and MIRA display the image the same way which is

> flipped from MaxImDL. After several of us complained we finally got

> Michael to add a user preference in MIRA to flip the image upon opening

> but I'm sure it's secretly dubbed the "stupid amateur option" since

> several claim that professionals always interpret the data the other

> way.



were still working out the details of WCS When WCS is present, it doesn't

matter which way around things are flipped. WCS automatically takes account of

any flips in X or Y. There's a sign convention in there somewhere, but it's

rather buried. Besides, it's quite a stretch to go to WCS for the definition,

because FITS predates WCS by two decades. Heck, they were still working out the

details of WCS long after MaxIm DL was released.



Certainly, some professionals do read the arrays in the opposite direction. The

POSS FITS images are written that way. But mostly no one seems to really care

about it.



This discrepancy arose because (a) the standard didn't address it, and (b) many

of the software systems were built in isolation; a lot of them predate the

widespread adoption of the internet. FITS was created in the late 1970's and

adopted by the IAU in 1982. As a result there are various mainframe-oriented

anachronisims in it. The headers follow the format for FORTRAN punch cards*.

And space for the header is allocated in blocks that match the size required for

reel-to-reel magnetic tapes.



(*For those too young to remember, people used to program computers by punching

little squares out of stacks of cards using a noisy machine called a keypunch.

You then dropped your stack of cards into a reader, hoping none of them got

shredded, and then waited patiently for the printout. If that sounds really

awful... trust me, it is!!!)



These shortcomings are all in all extremely minor and easily dealt with. The

extensible and flexible architecture of FITS gives you the freedom to establish

and document your own conventions, while maintaining a high level of

interoperability. That's what makes it so indispensible.



Just as an aside, the FITS header data is so incredibly useful that we've taken

to tucking a copy of it into TIFF headers as well. If you check, you'll notice

that TIFF images work just like FITS images in MaxIm DL V4.



Doug



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



Doug George

dgeorge@...



Diffraction Limited

Makers of Cyanogen Imaging Products

www.cyanogen.com



25 Conover Street

Ottawa, Ontario,

Canada, K2G 4C3



Phone: (613) 225-2732

Fax: (613) 225-9688



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







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

#19335 Jul 12, 2005

Ah, but it seemed nearly miraculous back then!



Chris



*****************************************

Chris L Peterson

Cloudbait Observatory

www.cloudbait.com



----- Original Message -----

From: "Douglas B. George" dg@...>

To: MaxImDL@yahoogroups.com>

Sent: Tuesday, July 12, 2005 4:31 PM

Subject: Re: [MaxImDL] quick question on FITS Liberator





> (*For those too young to remember, people used to program computers by

> punching

> little squares out of stacks of cards using a noisy machine called a

> keypunch.

> You then dropped your stack of cards into a reader, hoping none of them

> got

> shredded, and then waited patiently for the printout. If that sounds

> really

> awful... trust me, it is!!!)



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

#19341 Jul 13, 2005

Thanks Chris,

Thats what I wanted to know.



Mike

--- In MaxImDL@yahoogroups.com, "Chris Peterson" clp@a...> wrote:

> Yeah, but somebody got it wrong. I seem to recall there being some

issue of

> how the origin is interpreted in PS. FITS Liberator uses the CFITSIO

library

> for actually reading the file, so there is no doubt that the app is

> completely compliant with the FITS standard, but that doesn't

preclude a bug

> like the wrong origin creeping in where the FITS array gets

converted to a

> PS array.

>

> When I used the word "normal" I didn't mean there wasn't a bug, only

that

> Mike wasn't doing anything wrong with either the plugin or Maxim.

>

> Chris

>

> *****************************************

> Chris L Peterson

> Cloudbait Observatory

> www.cloudbait.com

>

>

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

> From: "Adam Block" ablock@o...>

> To: MaxImDL@yahoogroups.com>

> Sent: Tuesday, July 12, 2005 1:59 PM

> Subject: Re: [MaxImDL] quick question on FITS Liberator

>

> > Isn't the origin for FITS data part of the standard?



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

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