Problem saving snapshot with ASI294PRO

[turfpit]

One of the advantages of being retired

On CN, I found a few references to colour issues with the ASI294 Pro.

There is something going on, Robin will have a better idea now we have some pointers.

Dave

[admin]

Ok, that's weird - didn't know about that.

I don't have any code to debayer that matrix pattern. If the PNG contains truly raw output from that grid then the answer may be to do a 2x2 bin in PIPP and then debayer the result.

I do have a prototype 294 here from another manufacturer, but I'm still waiting for an SDK update to support it, so I don't have anything to test on.

I would suggest testing with a CS/C lens (meteor lens) and normal images to work out how to handle the output files from the 294 - much easier to tell if you have the colour handling correct!

Robin

[LaCasaCorp]

i can take a couple of terrestrial images and upload them here, would that be ok of you?

[robrj]

From what I've read, it still functions like an RGGB. The pixels in the CN thread are actually sub-pixels that can be arranged via software in the camera for a special HDR mode. But the way that ZWO implements it, it is still RGGB. When I open a fits file in Fitswork, it loads like RGGB. If you read later in the thread, Astrojedi verifies that his understanding of it was mistaken after talking with Sam at ZWO.

[robrj]

I hope you don't mind but I downloaded your image and ran it through Fitswork. One thing I noticed about your image that's different than mine is that the png file has a debayer pattern in it (like it was a fits file). In Fitswork, I opened it and the selected "Processing->Bayer Interpolation->RGB to Bayer Pattern". Then I selected "Processing->Bayer Interpolation->Demosaic Camera Raw Image" and used a RGGB pattern.

That gave me an image without the bayer pattern in it. So I saved it in .TIF format and opened it with Startools. I adjusted use the DEV module and the Color Module to correct a blue tint and cropped it and I got this (I'm storing it in the staging area of my Astrobin account so no one can see it unless I link to it. I can delete it when you say):



It looks right to me so something seems odd in how it's being saved. When I saved a fits file and opened it with Fitswork, my image comes up already debayered.

[LaCasaCorp]

looks like with Fitswork the image is debayered correctly (at least, looks a lot more like what i see on the sharpcap live screen... i never used Fitsworks, nice tool!

In any case, i will take a daytime image, just to provide all the data

F.

[admin]

I guess what is still confusing me is that it displays OK in SharpCap but doesn't debayer afterwards - SharpCap uses the same debayering algorithms as PIPP etc, expecting a standard bayer grid. I could understand if RGB mode worked OK in SharpCap and RAW looked funny as in RGB mode the debayer happens in the ZWO SDK and they would correctly deal with the weird pixel layout. I'm sure we will get to the bottom of this in the end though.

cheers,

Robin

[turfpit]

That last image looks more like one would expect - thanks for the Fitswork pointer robrj, it looks a useful tool.

Any money on "a new SDK is belatedly released which exposes a new camera control"?

Dave

[LaCasaCorp]

Here i am.... not the best day for a daylight picture (exceptionally gray and rainy) but i will be traveling the next 2 days.... at least i can give you something to check

http://jmp.sh/hAT3oKi

[robrj]

I guess what is still confusing me is that it displays OK in SharpCap but doesn't debayer afterwards - SharpCap uses the same debayering algorithms as PIPP etc, expecting a standard bayer grid. I could understand if RGB mode worked OK in SharpCap and RAW looked funny as in RGB mode the debayer happens in the ZWO SDK and they would correctly deal with the weird pixel layout. I'm sure we will get to the bottom of this in the end though.

cheers,

Robin

It's confusing to me as well. My ASI294MC Pro saves .png files just fine. They look like they do on the screen. None of them have that debayer pattern in them.