Bad debayering with Skyris 132C?

Post Reply
skaven
Posts: 5
Joined: Sun Feb 25, 2018 6:02 am

Bad debayering with Skyris 132C?

Post by skaven » Sun Feb 25, 2018 6:11 am

Hi, I am a licensed user trying to use SharpCap with my Skyris 132C camera. SharpCap seems to have no trouble at all controlling the camera, but the video that gets recorded seems to not be debayered properly (or something). The color seems right (so I believe the RGB translation is at least correct), but the recorded video has a "screen door" effect, resulting in awful moire patterns when scaling the image. Stacking a capture with AutoStakkert and then attempting wavelet processing in Registax doesn't work because of the "grid" pattern in the final image, which indicates this is not just an artifact of the capture software -- the grid pattern is being recorded in the final video frames.

I have tried setting the capture mode to RGB24, RGB32, MONO16, and MONO8, and all of them have the same effect. I have also tried with and without flat frames -- same effect.

I've looked around in the SharpCap settings and I can't seem to find anything to configure how debayering works. And it may not even be a debayering problem...

It's probably worth noting that I get the exact same behavior from FireCapture, but not with iCap (which is the "recommended" software from Celestron and The Imaging Source). So I'm pretty sure that whatever is going on in SharpCap and is a software setting

I'm running the latest version of SharpCap 3.1 on Windows 10 64-bit. The Skyris 132C is plugged into a USB3 port. I can't think of anything exceptional about my system configuration.

Any ideas about what this may be?

User avatar
admin
Site Admin
Posts: 1761
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Bad debayering with Skyris 132C?

Post by admin » Sun Feb 25, 2018 1:38 pm

Hi,

yes, sounds like something is not quite right with the debayering - can I ask some questions?

1) Is the image shown in SharpCap itself the right colours? Does it show the screen door effect when zooming in/out?
2) What file format are you trying to capture to?
3) Do you have an option for RAW8 or RAW16 in the colour space list? They should show for a colour cam rather than the MONO8/MONO16 options
4) Do you have a short capture that you could share via dropbox etc?

thanks,

Robin

skaven
Posts: 5
Joined: Sun Feb 25, 2018 6:02 am

Re: Bad debayering with Skyris 132C?

Post by skaven » Sun Feb 25, 2018 4:07 pm

If I select RGB24 or RGB32, the color seems fine in the preview window. The screen door effect is definitely visible in the preview window even when zoomed to 100%.

I'm capturing to AVI (the default for RGB)

There is no RAW option in the color space drop down, and MONO options appear despite this being a color camera.

I'll capture a short flat field and post a link later today.

User avatar
admin
Site Admin
Posts: 1761
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Bad debayering with Skyris 132C?

Post by admin » Sun Feb 25, 2018 4:37 pm

Hi,

ok, looks like the code is failing to spot that you have a colour camera properly, which is why you get MONO8/16 not RAW8/16. I can fix that easily for the next version - then you can capture in RAW to SER output and you will find that processing programs like AutoStakkert should automatically debayer correctly.

Why you see a grid pattern in RGB24/32 is a bit of a mystery - in those modes all the colour handling and debayering is left to the Celestron SDK - SharpCap is just copying data around and showing it on the screen or saving it to file. I have sometimes seen this sort of thing when the colour balance is very badly adjusted, so it may be worth playing with the colour balance controls.

cheers,

Robin

skaven
Posts: 5
Joined: Sun Feb 25, 2018 6:02 am

Re: Bad debayering with Skyris 132C?

Post by skaven » Sun Feb 25, 2018 4:53 pm

iCap works fine, so perhaps there is some additional post processing they do after pulling frames from the SDK?

That would explain why FireCapture has the same problem.

skaven
Posts: 5
Joined: Sun Feb 25, 2018 6:02 am

Re: Bad debayering with Skyris 132C?

Post by skaven » Sun Feb 25, 2018 7:38 pm

Here's a dropbox folder with some data that might be helpful for debugging: https://www.dropbox.com/sh/ptr3j822l1v5 ... mTkka?dl=0
  • iCap RGB32 screenshot.PNG - iCap preview in RGB32. Note histogram. Camera has a translucent cover over it for a "flat" image
  • iCap RGB32.avi - a 10-frame capture using above settings, using iCap.
  • SharpCap RGB32 NoFlat.PNG - SharpCap preview in RGB32. Note histogram (good color reproduction, flat gray). I also note that the screen door pattern is gone...I think?
  • SharpCap RGB32 WithFlat.PNG - After capturing and using a flat frame. There's the moire pattern! So perhaps that's the problem -- flats are not being captured properly.
  • 17_34_46.png - the flat file that created the moire pattern in the image above
  • 18_19_37_AS_p2_g5_ap177.png - AutoStakkert-created stack (best 2%) from a 500-frame capture of the moon with the flat file in use
  • registax.PNG - When trying to sharpen the above stacked image with Registax, the screen door effect is apparent, and wavelets fall apart
  • firecapture.PNG - It looks like FireCapture has a similar problem as SharpCap -- it seems to be treating my camera as monochrome, note the moire pattern in the preview window. It might be trying to capture in RAW mode, and so the preview is the raw, un-debayered image? Not sure. I've never actually been able to get FireCapture to capture a usable video from my Skyris -- it always comes out with the screen door effect.
While gathering this data I see that I did not do any captures without a flat file. And there's a pretty darn good chance that the problem is with the flat file, since it gets captured in "monochrome", which if SharpCap isn't properly recognizing my camera as color, would very likely be creating a flat file that is not properly debayered.

So it sounds like the best/easiest first thing to try would be to get a version of SharpCap that properly recognizes the camera as color, then try again both with and without flats, to see if I can reproduce the problem.

User avatar
admin
Site Admin
Posts: 1761
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Bad debayering with Skyris 132C?

Post by admin » Mon Feb 26, 2018 6:55 am

Yes, I will try to publish a new version this evening that sould recognize the camera as colour and we can see what happens then. The right approach then would be to select either RAW8 or RAW16 from the options as that means that SharpCap will be getting the unprocessed pixel data from the camera and run its own debayering algorithm on it. Also, save in '.SER' format by preference as that includes the bayer pattern as part of the file header so proessing software like Registax or AutoStakkert will automatically select the right mode.

cheers,

Robin

skaven
Posts: 5
Joined: Sun Feb 25, 2018 6:02 am

Re: Bad debayering with Skyris 132C?

Post by skaven » Tue Feb 27, 2018 2:22 am

I see you have posted a new version. I'll give it a try as soon as I can and let you know.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest