OSC Flats - colour differs too much warning
Stacking program, astrometric (plate) solver, fits file viewer
Brought to you by:
han59
Menu
▾
▴
Hi Han,
first of all - ASTAP is an excellent, a small and easy to use program i like and use very much. Great job! I prefer Astap for stacking instead of pixinsight.
Now to my problem - i use an OSC Camera 294MC Pro. For Flats i use an Wanderer Astro Flat panel, where i can change the intensity.
But even if i use 0.1s / 4s / 5s flats - see attachments. Your warning is always coming when stacking
█ █ █ █ █ █ Warning flat pixel colour values differ too much. Use white light for OSC flats!!. Will compensate accordingly." █ █ █ █ █ █
Perhaps the limits are too high? Is it really necessary to compensate / change the flat colour values? And what exactly are you doing with the values in the background ;-) ?
I always get a little excited, when reading this error-message.
Greetings
Hi Meister,
This warning comes if the intensities of red, green or blue pixels of the flat are unequal. The cause could very blue twilight or an aging electroluminance panel loosing the red.(so very blue)
Normally the exposure duration time does not have an influence.
The histogram you attached looks normal. That would be weird. What does the viewer popup menu " Show statistics report? Can you attach one original flat to your reply?
Cheers, Han
Hi Han,
thank you for your fast reply. Attached you find the 100ms OSC Flat of my ZWO ASI 294MC.
Here you find the statistics for the 100ms OSC Flat.
Greetings
Rolf
Hi Rolf,
The statistics doesn't tell much about the image.when it is still raw. But when you demosaic it without autolevels you can see green is strong 30856. Red is 17651 and blue is weakest at 15086. That indicates you either:
1) Have a green filter in front
2) Your camera is more sensitive in green then in red and blue. (weird)
3) Camera settings are wrong. (Possible?)
Is doesn't do great harm. The red and blue signals are still strong enough. What do you see in your light/astronomical images? Does any of above observations make sense?
See screenshot below.
Mit freundlichem Gruss, Han
Red:
x̄ : 17650.98 (sigma-clip iterations=3)
x̃ : 17698.50
Mo : 18030
σ : 379.8 (sigma-clip iterations=3)
σ_2: 594.4
mad: 289.5
m : 0
M : 18554 (4 x)
Flux: 7.841E+09 (Counts 11680440, Average 671/px)
BG : 16967
Hot pixels: 0.0%
Hot pixels RMS: 8578.3
≥64E3 : 0
Green:
x̄ : 30855.39 (sigma-clip iterations=3)
x̃ : 30902.50
Mo : 31090
σ : 438.9 (sigma-clip iterations=3)
σ_2: 896.2
mad: 330.8
m : 0
M : 32142 (1 x)
Flux: 1.621E+11 (Counts 11680440, Average 13878/px)
BG : 16967
Hot pixels: -0.2%
Hot pixels RMS: 11694.8
≥64E3 : 0
Blue:
x̄ : 15086.76 (sigma-clip iterations=2)
x̃ : 15097.00
Mo : 15146
σ : 176.4 (sigma-clip iterations=2)
σ_2: 217.0
mad: 129
m : 13599
M : 15842 (4 x)
Flux: -2.198E+10 (Counts 11680440, Average -1881/px)
BG : 16967
Hot pixels: -0.1%
Hot pixels RMS: 658.9
≥64E3 : 0
Vignetting [Mo corners/Mo]: -10%
Bit depth data: ?? Image is not saturated.
Greyscale levels: 1614
Rectangle: 1, 1, 4144, 2822
Filename: MasterFlat_Stack20_100.0ms_Bin1_294MC_L_gain120_20250309-200944_-0.3C.fit
Noise in electrons can be set with the popup menu of the status bar.
Legend:
x̄ = mean background | x̃ = median background | Mo = mode or most common pixel value or peak histogram, so the best estimate for the background mean value | σ = standard deviation background using mean and sigma clipping| σ_2 = standard deviation background using values below Mo only | mad = median absolute deviation | m = minimum value image | M = maximum value image | Flux = total flux inside shape above BG | BG = median background outside the shape | Hot pixels = percentage of pixels with a value above 3 sigma Gaussian noise | Hot pixels RMS = the root mean square value of the hotpixels | ≥64E3 = number of values equal or above 64000
Hi Han,
thank you for the info.
I don't get it exactly what to do with the OSC raw .fit -> What i do is to "De-Mosaic Bayer matrix" with the raw an set range to "off / manual" - but i dont get the R G B colored graph in the Histogram view. Where can i switch off the auto-leveling?
The second point is perhaps the issue - i read the following in the ZWO forum and the answer direct from the manufacturer was
So this seems right - this camera is green-heavy.
Do i have a problem with this flat while stacking or is it correct.
Should i "auto level" the flats before stacking, for better performance?
When stacking a bunch of images, i set the RGB Stacking to "Auto levels" and everything is normal.
Attached you find the auto-level disabled in my raw fits viewer - now i get the same result with greenish tint.
Greetings and thank you
Last edit: Meister Eder 2025-05-21
1) If your luminance panel is producing a reasonable white light then the conclusion the camera is more green sensitive reads plausible. But I have the impression it is more sensitive to green then other ASI294 images I have seen. But I have no hard data.
2) Do not modify the flats. The software will correct automatically.
3) ) So ignore the error message(s). ASTAP will correct and auto balance.
Auto balance of the ned result of stacking can be switched off. I will upload an other screen shot showing how with next message.
I just released a new version of ASTAP you could downloaf. The red, green and blue or OSC indication label of the selected files was not shown in the previous ASTAP version. This has been fixed now.
I will modify the statistics option to work on raw image directly. In the previous screen shot you can see the the Bayer pattern show a checker pattern due to the weaker pixels values for the blue and red sensitive pixels.
Han
See attached screenshot for the setting auto levels=off. This is only for the image displayed. It doesn't do anything for the flat correction. That is a separate routine.
Again. The program should correct all automatic. The end result should be good.