×

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

Bi-monthly release with minor bug fixes and improvements

FITS Viewer DeBayering Oddness

  • Posts: 388
  • Thank you received: 17
I am in the process of taking flats, normally I don't use the auto debayering in the Fits Viewer, but in this case I did. I took the flats using Kstars 3.5.0 on my Rpi, using the white tee shirt method, the main camera is a fairly new ZWO ASI533. The "issue" shows up in the Fits Viewer on the Rpi and also if I load the file into the Fits Viewer on my laptop (same Kstars version) It is not file dependent, they all behave the same.
With auto debayering on I load a file. It looks flat, but I need to see the statistics to get the correct exposure. I turn on the Statistics tab and it looks like this:



the Red channel show a maximum of 42140, while Green and Blue show a maximum of 65532, the Blue mean is much lower than the Green or the Red.

I then look at the Histogram:



If I scroll over the right side peak, it shows Intensity levels as well as RGB frequency levels, however Blue frequency is zero. I am assuming the peak to the left are the Blue numbers, but scrolling over it shows nothing.

If I view the exact same image in ASISTudio viewer, the histogram there looks like this:



The mean values are roughly the same for the R, G numbers in the Fits Viewer, but here you can see Blue is also the same mean.

In any fits image I open the Red channel never has a maximum equal to the Blue and the Green (65532) for this camera. I think it must have something to do with the number of bits in the color image, as a processed jpg image shows all three colors at 255 maximum.

Or I don't understand how to use the Fits Viewer with Debayering correctly!

thanks for any feedback.
3 years 2 months ago #66585
Attachments:

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

  • Posts: 910
  • Thank you received: 86
My guess is that ASIStudio viewer ignores the white balance (WB) values in ASI camera drivers.
Ekos should do this too but it doesn't. This is actually an issue which must be fixed (in Ekos).
ZWO ASI default WB_R=52 WB_B=95 produces twice as much blue in Ekos (and not only in Ekos).
When set to WB_R=50 WB_B=50, the histogram looks as it should (in Ekos).
ZWO software also ignores the offset value (they use some fixed number depending on the camera).
-- Max S
ZWO AM5. RST-135. AZ-GTI. HEQ5. iOptron SkyTracker.
TPO RC6. FRA400. Rokinon 135 and other lenses.
ZWO ASI2600MC. D5500 modified with UVIR clip-in filter.
ZWO ASI120MM Mini x 2. ZWO 30F4 guider. Orion 50mm guider.
ZWO EAF x 2.
Last edit: 3 years 2 months ago by maxthebuilder.
3 years 2 months ago #66592

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

  • Posts: 1009
  • Thank you received: 133
Maybe add that (IMO) you should mostly ignore the maximum value. It's usually from hot/warm pixels anyhow (unless you're really overexposed).

ASIStudio has it's own default settings, as Max says, and tweaks the white balance. That is a definite no-go for proper astro processing as the ZWO driver/library changes the raw values in the FITS file if your WB isn't 50/50. So (in EKOS) verify you're at 50/50, then look at the mean/median in the channels. Those of your example do look fine, this is normal.
Maybe also switch the saturation display on (red dot button) and verify that the maxima are just single pixels and not extended areas.
3 years 2 months ago #66595

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

  • Posts: 388
  • Thank you received: 17
You are both way ahead of me. First, you say "in EKOS verify you're at 50/50," could you explain, I do not know how to set the White Balance to 50/50. Is there a setting in the driver for the ZWO camera, (there is nothing in the FITS header)? I am looking at the files on my Windows machine and not on the Rpi at the moment, so I don't seem to have access to the .indi directory to check the drivers.

Second, Are you saying that ASIStudio does not do a proper job of de-Bayering images from their own camera? I am just beginning processing with a color camera (previously used an SBIG mono with filters), so I have to depend on some program to do the de-Bayering. I have had issues with DSS, it reads the flats in as grayscale and I have not figured out how to adjust that or if it matters. I don't have PI, I have used DSS and PhotoShop so far.

thanks,
Ron
3 years 1 month ago #66621

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

  • Posts: 388
  • Thank you received: 17
OK, I did a little more hunting and I see there are settings in the Indi Control Panel for the camera for WB, etc. That explains that, thanks. I did change the saturation setting as it did have a small spike. But mostly interested in the mean value.

I have been imaging with the ASI 533 with (what I believe is) unity gain at 100, but I have not messed with offset. It is off topic, but I would like to be pointed in the right direction for what settings I should be using for this camera.

thanks,
Ron
3 years 1 month ago #66622

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

  • Posts: 910
  • Thank you received: 86
I am not familiar with ASIStudio at all. Maybe it's doing auto WB and balances the channels for display purposes.

About the offset.
ZWO claim that they now have it fixed to some value. Maybe it's only for newer cameras, maybe it's only for their software. I don't know.
My ASI294MC and ASI071MC do respond to the offset changes in INDI/Ekos.

Offset is easy to understand - it moves/offsets the histogram left or right.
The higher the gain the larger the offset should be to avoid clipping the black (left side of the histogram).
ZWO's default of 8 can be too low for higher gains (100 and up) - that's for my ASI294.
Some people keep it at 30 to be safe.
I use offset 10 for unity gain of 120 and 20 for gain 200.
-- Max S
ZWO AM5. RST-135. AZ-GTI. HEQ5. iOptron SkyTracker.
TPO RC6. FRA400. Rokinon 135 and other lenses.
ZWO ASI2600MC. D5500 modified with UVIR clip-in filter.
ZWO ASI120MM Mini x 2. ZWO 30F4 guider. Orion 50mm guider.
ZWO EAF x 2.
Last edit: 3 years 1 month ago by maxthebuilder.
3 years 1 month ago #66625

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

  • Posts: 1009
  • Thank you received: 133
OK, good you found it :)
The 'Saturation setting' likely refers to the FITS Viewer, yes? This is only for visualization, it doen's change things. It's only to check where you have too high intensity.

I can't give you proper numbers, as I don't have that camera. Maybe someone can chime in. Else take darks, look at the histogram, and verify that the peak is clearly detached from the left edge (at zero; IIRC the ZWO driver will actually report 16 as lowest value). Then your camera electronics ("ADC" = Analog-to-Digital-Converter) will not clip your data.

(I just tried this with the FITS viewer histogram, but that one seems unusable - so better wait for some answers from people that have the camera :( )
3 years 1 month ago #66630

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

  • Posts: 91
  • Thank you received: 18

Replied by Dean on topic FITS Viewer DeBayering Oddness

funny I was working on the same things with Ekos fits viewer (EFV) and ASIFitsView (AFV) last night to
sort out offset on my ASI533MCPro. Like you, I see funky results in the EFV histogram and stats
reporting. Minimum is always zero in Ekos stats but AFV seems to be reporting the lowest non-zero values for
minimums and highest non-zero for maximums. For darks, those maximums can be quite high due to a single hot
pixel which is near impossible to see on a histogram (freq. of 1). I was shooting at 120 gain, 10 offset
but wanted to drop the gain closer to the 100 gain setting where you get a discontinuity drop in read noise. So now
I am trying 110 gain and may go to 105 just to ensure that drop is firm: I cant explain it, but I just don't trust
the exact 100 gain value to transition the drop.
AFV plots and stats for 30 sec exp, 110 gain, 20 and 40 offset, -10C. Pretty substantial shift away from zero for each offset 20 increment.
For bias frames, very similar minimum and average values but the maximums are much lower, probably because there
are no "hot" pixels ?

The following user(s) said Thank You: Peter Sütterlin
Last edit: 3 years 1 month ago by Dean. Reason: added content
3 years 1 month ago #66638
Attachments:

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

  • Posts: 388
  • Thank you received: 17
Thanks all,
DerPit - thanks for the explanation of offset, that makes sense. I have read on CloudyNights (CN) that for the ASI533 people use offset numbers like 10-21 at gain of 100 to 120 (close to unity). Now I understand why.

I am having as tough a time trying to find the right tools to 'understand' my calibration images and then use them in processing. I have FITS Liberator, AStroImageJ, DSS, ASI, EKOS and PhotoShop (non fits files) and I am not sure any of them agree!
3 years 1 month ago #66666

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

Time to create page: 5.182 seconds