Bayer Matrix for ZWO ASI224MC

bellavia

Does anyone know the Bayer Matrix for ZWO ASI224MC?

Is it R-G-G-B?

I am getting terrible colors when I do a de-bayering of the 16-bit FITS file, using Nebulosity (latest version, 4.17).

I've tried changing the diagonal terms of the color balance matrix, but nothing works

There are two attachments:

The default is using R-G-B = 1.0 1.0 1.0, and the nebula looks yellow

I tried this, R-G-B = 0.6 0.8 1.6, and the nebula is starting to look normal at the expense of all the stars and background being much too blue.

Any help is appreciated.

Steve

P.S. Sorry if this issue has been addressed, but I did a search on the forum and surprisingly, nothing came up on this topic.

Sam@ZWO

it's RGGB, but different app call it different

I think the yellow one is correct

you need to adjust the white balance to make the color better

bellavia

Sam wrote:
it's RGGB, but different app call it different

I think the yellow one is correct

you need to adjust the white balance to make the color better

This is an image of a blue cable on a brown desk. When I debayer it with RGB , the cable is green.

When I use CMYG, it gets much closer.

But I cannot use CMYG for the nebula, as it creates an odd aritfact.

bellavia

bellavia wrote:
Sam wrote:it's RGGB, but different app call it different

I think the yellow one is correct

you need to adjust the white balance to make the color better

This is an image of a blue cable on a brown desk. When I debayer it with RGB , the cable is green.

When I use CMYG, it gets much closer.

But I cannot use CMYG for the nebula, as it creates an odd aritfact.

This is interesting: I opened the same file in APT's previewer.

The cable is still green, but the desk is closer to brown (yellow?)

Sam@ZWO

this is because there is no IR-CUT filter

you can add one IR-CUT filter so the final image is not so red

bellavia

Sam wrote:
this is because there is no IR-CUT filter

you can add one IR-CUT filter so the final image is not so red

But the image is NOT too RED. it is too GREEN, which is the opposite of RED.

I think there is something wrong with the method the FITS header is being written OR the matrix is NOT R-G-G-B.

I was using this camera as a test to see if I would purchase a lager chip, cooled ZWO color camera.

But after this experience, I will not. It is not working out at all, and there is very little help or technical support of these products.

bellavia

bellavia wrote:
Sam wrote:this is because there is no IR-CUT filter

you can add one IR-CUT filter so the final image is not so red

But the image is NOT too RED. it is too GREEN, which is the opposite of RED.

I think there is something wrong with the method the FITS header is being written OR the matrix is NOT R-G-G-B.

I was using this camera as a test to see if I would purchase a lager chip, cooled ZWO color camera.

But after this experience, I will not. It is not working out at all, and there is very little help or technical support of these products.

Hi again,

It took much effort, but I resolved the issue.

You need to apply an offset, as the software does not "know" which pixel to start with for the R-G-G-B mosaic.

For Nebulosity, the offset is 1,1 for X,Y

Then, after that, you can tweak further. This is not final (as I am still waiting for my photo color card to arrive), but does present fairly good colors. I will post again if I improve upon this but for now, the color balance matrix should be as follows:

0.87 0.10 0.00

0.10 0.89 0.00

0.00 0.25 1.30

These are the rows and columns for R-G-B

the columns are the input, and the rows are output

For example, row 1 above is applying 87% of the red to the red channel, 10% of green to the red channel and 0% of blue to the red channel. (This does not have to add up to 100%, nor does the determinant of this matrix need to be 1.0)

Attached is the summary of what I have done so far.

As I said, if I get it better, I'll let you know, but this is a good place to start

Sam@ZWO

white balance is always need to adjust

and different software use different way

that's good news there is some improvement

bellavia

Sam wrote:
white balance is always need to adjust

and different software use different way

that's good news there is some improvement

I finally got my color card, and the latest settings to get better colors are extremely simple (actually embarrassing, considering what I had in my last post, as it is just the diagonal terms, and nice round numbers).

This seems to work well enough to call it a day:

Matrix offset x,y: 1,1

Color Control:

R G B

R 0.80 0.00 0.00

G 0.00 0.80 0.00

B 0.00 0.00 1.25

Sam@ZWO

thanks for sharing

jcoldrey

Great info Steve - much appreciated.

I can recall identical white balance challenges using Nebulosity with a QHY12 a couple of years ago.

"Out of the box" white balance for these sensors is a tad whacky.

Thanks for refreshing me on the process, as I now need to do similar for my ZWO ASI 094.

I absolutely love the ASI 094 ... but nailing the white balance will make processing much easier.

Cheers,

Jeff