50 replies to this topic

[kevinbreen]

So far this season has been dogged with poor seeing as a result of jet stream, low pressure weather systems, Covid-19, Brexit, iffy pints and late night kebabs.

In yet another desperate attempt to eke even a modicum of improvement out of my imaging sessions, let me ask -

What are your (yes, you! you firecapturing ASI224MC Martian hunter) preferred wred and wblue settings for Mars?

I use 52 and 92.
I also use an IR cut filter. My Mars images are a dull brown with little contrast. Martian dust issues aside, maybe there are firecapture settings that are better than mine?

[KiwiRay]

52 for red, 95 for blue.  This shouldn't matter in your final outcome, as colours can always be balanced in post-processing.  There doesn't seem to be a lot of dust activity right now, so that's not your problem either.  I think you know the reason...it's definitely the iffy pints and late night kebabs.  The Gods of Seeing don't look kindly on poor food and beverage choices.

[kevinbreen]

52 for red, 95 for blue. This shouldn't matter in your final outcome, as colours can always be balanced in post-processing. There doesn't seem to be a lot of dust activity right now, so that's not your problem either. I think you know the reason...it's definitely the iffy pints and late night kebabs. The Gods of Seeing don't look kindly on poor food and beverage choices.

Cheers Ray.
I put those references in to see who actually reads posts. My local country pub is still shut here since March, the poor chap. Kebabs? It would take a few pints to make me brave one of those!

OK, thanks, I'll chalk it up to poor seeing.

[Tulloch]

52 for red, 95 for blue.  This shouldn't matter in your final outcome, as colours can always be balanced in post-processing. 

While these are the "official" values from ZWO, I believe the numbers are closer to Wred = 62, Wblue = 99 with the blue channel multiplied by 1.1 in Registax (to correct for atmospheric distortions).

 

These are my own settings calculated from colour card measurements from the sun, if you have a spare couple of hours you can read the series of measurements here  .

https://www.cloudyni...acy/?p=10067895

 

What I never put in that posting (but which clearly shows the effect of the Wred and Wblue values) is the raw data from the grey patches on the colour chart. These should have equal values for R, G and B channels, and I used these as a metric to get the colour balance correct. The table below shows the "true" values (ie the colour checker chart specs), together with the RGB for the ASI224MC with Wred=52 Wblue=95, the ASI224MC with Wred=62 Wblue=99, and the Canon 700D taken in the same way for reference.

 

While the absolute numbers for the ASI224MC do not match the colour checker chart (due to the sRGB gamma correction not being applied), the RGB values for Wred=62 and Wblue=99 are more or less equal to each other for each patch, similar to those for the DSLR (although it is coming out a little too blue). This is why I think that Wred=62 and Wblue=99 are the correct values to use.

 

However, as Ray says you can always balance the colours in post, which is easy for Jupiter, harder for Saturn, while for Mars, Uranus and Neptune if you don't follow a "standard", you are just guessing (IMHO) .

 

Andrew

Attached Thumbnails

• 

Edited by Tulloch, 03 August 2020 - 10:47 PM.

[RickD_99]

While these are the "official" values from ZWO, I believe the numbers are closer to Wred = 62, Wblue = 99 with the blue channel multiplied by 1.1 in Registax (to correct for atmospheric distortions).

Andrew any idea if those values would be appropriate for the 290MC as well?

[RedLionNJ]

Hi Kev,

 

I've historically used 54/96 for my ASI224MC.  Then in Registax, I color-equalize (if AS!3 didn't do it already - sometimes it does, sometimes it doesn't) and bump the Red value up a notch or two.  And I set the saturation to +2.

 

This brings a color cast very similar to that observed through the eyepiece.  Color comparison is really the only reason I ever put an eyepiece into my scope.

 

Best,

 

Grant

[Tulloch]

Andrew any idea if those values would be appropriate for the 290MC as well?

Hi RickD, different sensors will have different sensitivities, so no, I don't think they will be the same for the 290MC. Having said that however, there is a relatively easy way to work out what values that might be appropriate to give you better colour balance straight out of the camera. 

 

Take for instance the raw frame of Jupiter that Darryl posted recently (see his image below compared to mine taken at a similar time and elevation) in the thread linked below for the brand new ASI462MC. He used the "standard" Wred=52 Wblue=95 and it is quite obvious that the raw frame has a distinct blue hue compared to mine. Registax's "Auto balance RGB" tool is excellent for colour matching the channels on Jupiter, and based upon this image it suggested that the Red channel needed to be multiplied by 1.24, Green by 0.94 and Blue by 0.85. Since you cannot change Wgreen (it's all relative anyway), this led to the conclusion that the "correct" settings for the 462 chip should be closer to Wred=69, Wblue=86.

https://www.cloudyni...etc/?p=10376048

 

"Why bother the change these values if you can simply change them in post?", I hear you ask. Simply because the "Auto balance" tool in Registax doesn't work as well for Saturn (Saturn does not have an equal mix of colours), and it's terrible for Mars, Uranus and Neptune because they are (primarily) single colour planets. OK, that's an over-simplification, but you get what I mean.

 

The point is that there should be no good reason why the colour shift required for Jupiter should be any different to the shift required for the other planets (neglecting elevation differences, which I've discussed in this post). So once you've found a set that works for Jupiter, it should be the same for the other planets if they were imaged at a similar elevation. One caveat to this is that it appears that the wavelet sharpening levels affect the colour balance also, different colour channels are affected differently in a non-linear way by the wavelet settings used, I'm still looking into this effect.

 

Anyway, for your ASI290MC, my advice would be to take the best raw footage of Jupiter you have, stack it in AS!3, sharpen it in Registax, apply the Auto-balance RGB tool and adjust your Wred and Wblue values accordingly. Note that even for the Wblue setting of 99 for my ASI224MC, the blue channel is still too low, which means I need to adjust it in Registax by a factor of 1.1 to get the histogram to look "right".

https://www.cloudyni...etc/?p=10376649

 

 

Hi Kev,

 

I've historically used 54/96 for my ASI224MC.  Then in Registax, I color-equalize (if AS!3 didn't do it already - sometimes it does, sometimes it doesn't) and bump the Red value up a notch or two.  And I set the saturation to +2.

 

This brings a color cast very similar to that observed through the eyepiece.  Color comparison is really the only reason I ever put an eyepiece into my scope.

 

Best,

 

Grant

Thanks Grant, for whatever reason (probably inexperience) I've never been able to see the colours of Jupiter clearly through the eyepiece, at least not well enough for a colour cast determination. It's always far too bright for me, I can see the clouds and GRS fine, but I find trying to turn that into a colour match impossible. 

 

Do you use an ND filter (or something else) to determine colours?

 

Andrew

Attached Thumbnails

• 
• 

Edited by Tulloch, 05 August 2020 - 04:54 PM.

[GSBass]

Only one night experience with 462 but for me the colors of jupiter seemed quite accurate, I post processed with excessive red really to bring out detail but that was just a creative choice, in comparison to view through eyepiece I felt the difference was more of saturation, ie the camera image is just more pale, mine is qhy instead of zwo but I doubt that’s the difference being discussed, [btw I used the ir cut filter]

Attached Thumbnails

• 

Edited by GSBass, 05 August 2020 - 05:27 PM.

[Tulloch]

Only one night experience with 462 but for me the colors of jupiter seemed quite accurate, I post processed with excessive red really to bring out detail but that was just a creative choice, in comparison to view through eyepiece I felt the difference was more of saturation, ie the camera image is just more pale, mine is qhy instead of zwo but I doubt that’s the difference being discussed, [btw I used the ir cut filter]

Thanks, which values for Wred and Wblue did you use during capture?

[GSBass]

I’ll go see if it’s listed in the data but it was default.... I seriously didn’t know enough to mess with them

[GSBass]

Here it is, these were defaults for qhy in sharpcap

 

[QHY5III462C]
Debayer Preview=On
Output Format=SER file (*.ser)
Binning=1x1
Capture Area=1920x1080
Colour Space=RAW8
Pan=0
Tilt=0
Force Still Mode=Off
Enable Live Broadcast=Off
USB Traffic=0
Offset=51
Amp Noise Reduction=On
Frame Rate Limit=Maximum
Gain=180
Exposure=7.76912632177017
Timestamp Frames=Off
White Bal (B)=114
White Bal (G)=69
White Bal ®=64
Contrast=0
Brightness=0
Gamma=1
Banding Threshold=35
Banding Suppression=0
Apply Flat=None
Subtract Dark=None
#Black Point
Display Black Point=0
#MidTone Point
Display MidTone Point=0.5
#White Point
Display White Point=1
Notes=
TimeStamp=2020-08-02T04:40:59.2744440Z
SharpCapVersion=3.2.6396.0

[GSBass]

Posting this one too because I’m not positive which belonged to that image, lessened learned, need to keep file name when saving after process

5III462C]
Debayer Preview=On
Output Format=SER file (*.ser)
Binning=1x1
Capture Area=1920x1080
Colour Space=RAW8
Pan=0
Tilt=0
Force Still Mode=Off
Enable Live Broadcast=Off
USB Traffic=0
Offset=51
Amp Noise Reduction=On
Frame Rate Limit=Maximum
Gain=342
Exposure=1.94228154165886
Timestamp Frames=Off
White Bal (B)=114
White Bal (G)=69
White Bal ®=64
Contrast=0
Brightness=0
Gamma=1
Banding Threshold=35
Banding Suppression=0
Apply Flat=None
Subtract Dark=None
#Black Point
Display Black Point=0
#MidTone Point
Display MidTone Point=0.5
#White Point
Display White Point=1
Notes=
TimeStamp=2020-08-02T04:29:30.3057914Z
SharpCapVersion=3.2.6396.0

[GSBass]

Btw not sure why blue is so high, I did go back and look at raw, I’m correct about natural color but there definitely is a blue tinge preprocessing, did not realize at time

[GSBass]

Do y’all think dark frame is needed when taking 2000 frame or so avi for stacking, I didn’t bother my first time out although I know it’s needed for longer exposure deepsky stuff

[RickD_99]

Anyway, for your ASI290MC, my advice would be to take the best raw footage of Jupiter you have, stack it in AS!3, sharpen it in Registax, apply the Auto-balance RGB tool and adjust your Wred and Wblue values accordingly. Note that even for the Wblue setting of 99 for my ASI224MC, the blue channel is still too low, which means I need to adjust it in Registax by a factor of 1.1 to get the histogram to look "right".

https://www.cloudyni...etc/?p=10376649

 

Andrew

 

OK so I followed your suggestion above and picked out the best footage from my Jupiter captures on July 14. After processing as above I ended up with the following values in the Registax 6 auto balance tool (normalized values are in parentheses):

 

Red: 0.97 (0.98)

Green: 0.99 (1.0)

Blue: 1.15 (1.16)

 

So since this footage was captured with Wred at 52 and Wblue at 95 that would mean that for my 290MC I would need to set Wred to 51 and Wblue to 110 for proper white balance within Firecapture. Since the Wblue setting in Firecapture maxes out at 100 I would need to set the colorweight for blue at 1.1 in the Registax RGB align histogram box correct (assuming my Firecapture settings are Wred=51 and Wblue=100)?

[Tulloch]

Anyway, for your ASI290MC, my advice would be to take the best raw footage of Jupiter you have, stack it in AS!3, sharpen it in Registax, apply the Auto-balance RGB tool and adjust your Wred and Wblue values accordingly. Note that even for the Wblue setting of 99 for my ASI224MC, the blue channel is still too low, which means I need to adjust it in Registax by a factor of 1.1 to get the histogram to look "right".

https://www.cloudyni...etc/?p=10376649

 

Andrew

 

OK so I followed your suggestion above and picked out the best footage from my Jupiter captures on July 14. After processing as above I ended up with the following values in the Registax 6 auto balance tool (normalized values are in parentheses):

 

Red: 0.97 (0.98)

Green: 0.99 (1.0)

Blue: 1.15 (1.16)

 

So since this footage was captured with Wred at 52 and Wblue at 95 that would mean that for my 290MC I would need to set Wred to 51 and Wblue to 110 for proper white balance within Firecapture. Since the Wblue setting in Firecapture maxes out at 100 I would need to set the colorweight for blue at 1.1 in the Registax RGB align histogram box correct (assuming my Firecapture settings are Wred=51 and Wblue=100)?

Hi Rick, sounds about right to me . Let me know how it turns out.

 

Andrew

[GSBass]

Is there a reason manufactures set the blue so high as default?

[RedLionNJ]

Hi Andrew - you use enough visual magnification and Jupiter may look blurry, but it's not overly-bright. Colours (esp the EZ, belts, festoons and larger spots) are fairly clear.  I've never been too sure on the colors of the polar areas.

 

Grant

[Tulloch]

Is there a reason manufactures set the blue so high as default?

Remember that there are twice as many green pixels as either red/blue pixels in the Bayer matrix (see image below), and that the Wred and Wblue values are relative to the green (set to 1). This is OK, as the human eye has a peak efficiency near at around 550nm, or in the green part of the spectrum.

https://www.gigahert...pectr-sens-eye/

https://www.kth.se/s...Fotoslides4.pdf

 

The cameras we use have their own sensitivity to wavelengths which don't match our own colour vision, so corrections need to be made. For example, the spectral response of the ASI290MC is shown below, where you can see the sensitivity to blue light is much lower than the other colours. The camera makers cannot change the response from the sensor, but we can introduce corrections for it in software.

 

 

Hope this helps,

 

Andrew

Attached Thumbnails

• 

Edited by Tulloch, 06 August 2020 - 04:33 PM.

[GSBass]

Remember that there are twice as many green pixels as either red/blue pixels in the Bayer matrix (see image below), and that the Wred and Wblue values are relative to the green (set to 1). This is OK, as the human eye has a peak efficiency near at around 550nm, or in the green part of the spectrum.

https://www.gigahert...pectr-sens-eye/

https://www.kth.se/s...Fotoslides4.pdf

 

The cameras we use have their own sensitivity to wavelengths which don't match our own colour vision, so corrections need to be made. For example, the spectral response of the ASI290MC is shown below, where you can see the sensitivity to blue light is much lower than the other colours. The camera makers cannot change the response from the sensor, but we can introduce corrections for it in software.

 

 

Hope this helps,

 

Andrew

Thankyou very much , that helped me understand alot 

[DMach]

Andrew any idea if those values would be appropriate for the 290MC as well?

After playing around with various settings, I tend to use Red = 66 and Blue = 99. But colour balance does have a degree of personal preference to it (I tend to like a slightly warm colour balance) so I won't comment as to whether these are "appropriate".

 

Blue does tend to need all the help it can get with the 290MC.

[GSBass]

Here is freq response for the 462 just for comparison to one above

Attached Thumbnails

• 

[John Verderame]

I'm hoping to get one last shot at Jupiter and Mars this week, and just picked up a Baader Contrast Booster filter to try out on them with a ZWO ASI462MC.  Any idea what filter setting I should use for that filter in FC?

[Tulloch]

I'm hoping to get one last shot at Jupiter and Mars this week, and just picked up a Baader Contrast Booster filter to try out on them with a ZWO ASI462MC.  Any idea what filter setting I should use for that filter in FC?

Neither - I wouldn't use the filter with a camera. You can boost whatever wavelengths you like afterwards in Photoshop/Gimp, the Contrast Booster is designed for visual observing.

 

Andrew

[John Verderame]

Thanks, Andrew, but I'm gonna have to beg to differ.  Baader's own specs state, "Together with the Baader UV / IR cut filter Baader Contrast Booster is additionally an optimal filter for CCD images from light-polluted big city and metropolitan area (see second filter curve)."  Now the filter itself supposedly has the IR cut built in.

 

Anyhow, though I don't have much problem with LP at my location, still I've found that colored filters like the 23A and 80A do enhance details on both Jupiter and Mars, so I thought I'd give it a try (then I'll know for sure!).  Also, from my reading, it's better to use the filter rather than try to boost wavelengths in post-processing.  The idea with the filter is to sharpen the original image, not just enhance colors, and the filters definitely sharpen the original image.