Calibrating correct RGB values using white paper.

[RobH]

Hello all.

I decided to set the correct RGB levels on my SXV M25C today, and, rather than using any old G2V star, I decided to use the closest one....the sun

Standard G2V calibration can be a bit tricky....for one thing, you need to see some stars, which doesn't seem to be happening much at the moment, and you also have atmospheric absorption to deal with, meaning that the RGB values will change with altitude.

As I'd had the camera set up in order to get the guide cam parfocal in the daytime, it seemed sensible to set a true white for the camera too.

In order to cut down the light levels reaching the chip (at full aperture the whitepoint was completely clipped), I covered the objective with tinfoil, and made a small hole in the centre.

I then put a white sheet of paper on a chair in direct sunlight and pointed the scope at it.

I then took and image of this, and converted it to colour with a value of 100% for each colour channel and saved as a TIFF.

This was then opened in photoshop, and the Info window opened too, and the colour balance checked with this.

The intial reading showed that there was quite a bit too much green, and blue was a bit low.

I then repeated the process (using the same image for consistency) and adjusted the percentages of each colour until the info window in PS told me that I had equal colour values in each channel.

This took just 4 runs.

I now have the correct values to get an accurate colour balance with this camera/IR filter combination and will repeat the process with my Atik 16HR and H18, which are mono cameras with filters used.

Here's a pic of the setup I used (the scope is my old Helios ST120).

Cheers

Rob

[roundycat]

An interesting setup. You could probably use a Baader astro solar film as a neutral density filter, that would leave you at full aperture in case there are any odd effects from shooting at f200.

Not sure I understand the bit about 100%. Does this mean full exposure as in 65,000 ADU or something else. Sounds like a good idea for all those DSLR users that pratt about with custom white balance when the proper white is there all the time.

Do you think you would get the same sort of result under light cloud?

Dennis

[RobH]

Hello Dennis.

By 100%, I mean that in Maxim, the 'convert RGB' command in the colour menu gives you an option to change the balance of the RGB when debayering an OSC camera.

The settings I came up with for the M25C and Baader IR/UV blocker are....

Red. 100%

Green. 85%

Blue. 106%

When I run this with my mono cams, I will simply use the info pallete in PS and adjust the relative levels of RGB accordingly until they are of equal value.

The I can work out the correct relative exposures for the filters.

I don't see any reason that this won't work in overcast conditions.....light clouds are pretty neutral in colour most of the time.

I would imagine also that shooting at F2 million or so , with the old 120, is probably better than shooting at full aperture, as at F5 it's half way between a prism and a lens!!

Cheers

Rob

[blinky]

Good thinking Rob -I've been trying to figure out how to do this for ages, never dawned on me to use the sun/white paper Doh!

[steppenwolf]

This is an interesting way of doing it, Rob.

With my own M25C I simply installed my IDAS LP Filter (this also has an for IR filter) and took a daytime image of the inside of my conservatory with a bird feeding station outside (different coloured feeders) using a 50mm prime lens. I then made a series of RGB conversions in MaximDL until I got a realistic colour image on the monitor that I use for all my post processing.

You are going to love that M25C, Rob!

[roundycat]

Hi Rob,

now that you mention the OSC camera the 100% thing makes a bit more sense. I am not sure how MaxIm does it and I'm also not sure how you can set those values using PS. Maybe by using layer opacity, colourising the layers and screening them. Is there another way to re-set colour balance in PS without 'bending' the curve?

Dennis

[RobH]

Hi Dennis.....in PS, I'd guess that I'll use the levels control and adjust the output levels of each channel until they all read the same in the info window.

Obviously, I don't want to boost the levels of any channel or I'll clip the whitepoint, so I'll pick the channel with the lowest level, and use that as my reference, dropping the levels on the other 2 channels to match that one.

Then I can read off the output levels of each colour channel, convert that number to a percentage, and apply that to my imaging time per filter.

Cheers

Rob

[roundycat]

Rob, that much I understand. You can vary the output levels using Curves, just move the ends up or down, it does the same thing as Levels. The thing is, with Curves, you can see what it is doing to the graph. If you drop the output the contrast goes down as the output is dropped pro rata from 0 to 255.

Are the two things really the same, ie, MaxIm's colour ratio and PS's Levels?

Dennis

[RobH]

Are the two things really the same, ie, MaxIm's colour ratio and PS's Levels?

Dennis

No idea Dennis.....I guess a bit of experimentation is needed

I was going to do a bit of demolition work in the garden today, but playing with astro kit sounds a lot more fun, so I may well do that instead

Rob.

[ollypenrice]

I wonder if the use of an achromatic lens will throw things out slightly? Maybe not, since you are using an unfocussed light source but it might be the case that there will be a slightly different transmission spectrum from the cheap coatings. Probably vanishingly small, I suppose. Also the reflection spectrum may be affected by the paper. Getting silly!!!

I really must get Maxim...

Olly

[RobH]

I wonder if the use of an achromatic lens will throw things out slightly? Maybe not, since you are using an unfocussed light source but it might be the case that there will be a slightly different transmission spectrum from the cheap coatings. Probably vanishingly small, I suppose. Also the reflection spectrum may be affected by the paper. Getting silly!!!

I really must get Maxim...

Olly

Probably true Olly, but, as you say, I think that any of these effects will be extremely small, so I'm not concerned about them.

I would imagine that any inconsistencies will be considerably less than the effects of atmospheric absorption/ pollution etc when trying to calibrate on a G2V star.

Cheers

Rob

[antimorris]

I had some astro equipment work (changing the setup for the next planned imaging target/filter) and saw your thread. I think I might add this to my plan for astro work today as well. I tend to use the B-V color calibration method rather than G2V as it is easier and quicker for me, but I am kinda interested to see what values I get using the good ol' sun like you did.

Hmm - just started raining ... again ... sigh - might have to delay the whole plan for a few hours. Guess I should have gone out earlier rather than typing on the forum

[roundycat]

Another thought just ocurred to me; why not use a Macbeth Colour Checker. OK, they cost the earth but you can get good figures for all colours. I scanned mine years ago, back in the days of film and we were all getting used to Photoshop 4. It comes with the correct RGB values for each swatch written on the back. Or somewhere.

Dennis

[RobH]

Well ....didn't get the demolition work done, or get the other cameras out...,.too nice lounging in the garden in the sun

Dennis....the colour checker sounds like it would be very accurate, but, TBH, I reckon the sun on white paper will be good enough, especially when you consider that the images will be maily viewed on uncalibrated monitors of differing quality.

I took an image of the view out over the harbour with the colour values I've set, and it looks pretty well spot on, so that's good enough for me, and a cheap way to do it

Cheers

Rob

[Shibby]

Good plan Rob but I'd recommend getting a grey card as these are guaranteed to reflect equal amounts of R, G and B. I have one of these: Colour Confidence Total Balance Grey Card

[blinky]

OK, tonight I did more or less as has been suggested but I used DeepSkyStacker and set the values for RGB in the 'colours adjustment' section in the FITS files tab. My values were - Red Scale 1.0430 and Blue 1.4200

I stacked an image of the NA Nebula and whats strange is that the background colour seems pretty much in balance and the levels graph agrees but in the cores of stars - the ones that are burnt out, I get a green blob! Anybody know what I have done wrong?