×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Conversion ti fits from CR2 changes geometry?

  • Posts: 643
  • Thank you received: 62
Hi!

I know we have had this up before... but I don't know what came of it.

I use a Canon 1000D DSLR. I'm playing with using native format from the camera. And I've noticed that the geometry of the native CR2-files is different from the Fits-files that Ekos converts to.
My native CR2-files are 3906 x 2602. The fits files are 3893 x 2595.

Is there a sensible reason for this? Is it possible to have the fits files preserve the geometry (there are for instance processes in PixInsight that likes even numbers and dislikes uneven...)?

Magnus
5 years 2 months ago #33085

Please Log in or Create an account to join the conversation.

  • Posts: 984
  • Thank you received: 160
Hi, Magnus!

The issue comes up regularly. Ekos converts CR2 into FITS internally when a photo is taken. An option to do it later on has not been implemented. In this thread

indilib.org/forum/general/3323-question-...-cr2-conversion.html

Copello suggests a hack to do it anyway. I tried and it worked perfectly for me. I wish a conversion option would be implemented though.

CS
Last edit: 5 years 2 months ago by Alfred.
5 years 2 months ago #33088

Please Log in or Create an account to join the conversation.

Can you post logs? So take one Native, then one FITS. Let see where the discrepancy comes. I don't have this issue on 600D here.
5 years 2 months ago #33091

Please Log in or Create an account to join the conversation.

  • Posts: 643
  • Thank you received: 62
Hi!

Here's a log. OK?

Magnus
5 years 2 months ago #33153
Attachments:

Please Log in or Create an account to join the conversation.

Looks like it is getting subframed. Clear your camera configuration (the trash bin icon next to the camera name in Ekos camera module). Then enter the correct dimensions again and see how it goes.
5 years 2 months ago #33158

Please Log in or Create an account to join the conversation.

  • Posts: 643
  • Thank you received: 62
Hi!

I guess part of the problem might be to understand what the correct dimensions are. I get some information here, for instance:

www.digicamdb.com/specs/canon_eos-1000d/

Which is what I typed in before. But the CR2-file has a different size, according to PixInsight. Should I use those numbers, the geometry of the CR2-file as given by PI, or what should I use?

Magnus
5 years 2 months ago #33159

Please Log in or Create an account to join the conversation.

I would try the CR2 dimensions and see how it goes from there. As long as you keep the format RAW I don't think you'll have troubles. Try and see the results.
5 years 2 months ago #33162

Please Log in or Create an account to join the conversation.

  • Posts: 111
  • Thank you received: 40
I think one should use the value LibRaw infers after unpacking (and debayering) the data. According to your posted logs it should be:
3906 x 2602. This value should also be shown by PixInsight.

I have a similar issue with my Nikon D50. The Linux image viewer geeqie shows for RAW the dimension: 3008 x 2000,
also this value is presented at www.digicamdb.com/specs/nikon_d50/ whereas PixInsight shows
3039 x 2014.

I just debugged the related struct field of libraw (to look inside). And indeed also there is the value: 3039 x 2014
$1 = {raw_height = 2014, raw_width = 3040, height = 2014, width = 3039, top_margin = 0, left_margin = 0, iheight = 2014, iwidth = 3039, raw_pitch = 6080, pixel_aspect = 1, 
  flip = 0, mask = {{0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}}, raw_crop = {cleft = 65535, 
    ctop = 65535, cwidth = 0, cheight = 0}}

See also this post for an additional issue on dimensions: indilib.org/forum/ccds-dslrs/3792-canon-...rame-size.html#28826
The following user(s) said Thank You: Magnus Larsson
5 years 2 months ago #33164

Please Log in or Create an account to join the conversation.

  • Posts: 984
  • Thank you received: 160
With resolution set to the 5D Mk2's nominal 5616x3744 the results are:

FITS: 5616x3744 (FitsView)
CR2: 5634x3755 (UFRaw) or 5626x3745 (RawTherapy).

After setting the resolution to 5634x3755 in Ekos the results are like this:
FITS: 5634x3753 (FitsView)
CR2: 5634x3755 (UFRaw) or 5626x3745 (RawTherapy).
Last edit: 5 years 2 months ago by Alfred.
5 years 2 months ago #33169
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 984
  • Thank you received: 160
Is there a certain reason for not implementing a simple file requester that enables the user to select an already taken CR2 file and have it converted by Ekos? The whole conversion code is already there so why not use it? Copello's hack did work perfectly and the resulting FITS files had the desired resolution. One only has to select an alternative input source (file instead of camera), no?
5 years 2 months ago #33170

Please Log in or Create an account to join the conversation.


Thanks for logs. So Ekos/INDI are indeed sticking to the dimensions given whenever it is possible to do so.

Libraw reports the CR2 resolution as 5634x3753.

So when you take 5616x3744, INDI can subframe this to the desired value. However, for 5634x3755, INDI cannot add two extra pixels to make 3753 become 3755, so it remains at the maximum reported height by Libraw which is 3755.
5 years 2 months ago #33182

Please Log in or Create an account to join the conversation.

  • Posts: 984
  • Thank you received: 160
When I set frame size to 5634x3753 the results are

FITS: 5634x3753
CR2: 5634x3755 (UFRaw), 5626x3745 (RawTherapy)
5 years 2 months ago #33184
Attachments:

Please Log in or Create an account to join the conversation.

Time to create page: 0.208 seconds