One shot colour CCD advice.

[ollypenrice]

Hi All,

Atik have kindly agreed to lend me a one shot colour version of the Atik 4000 which we already use here. I aim to write a review for a magazine comparing the two in action. While I am familiar with using mono, I would be glad to receive any tips on getting the best out of colour, so if you are in the know I'd love to hear from you. Eg dithered guiding. (Reckon I'm already quite good at that, unintentionally!) Processing? (Eg I gather the synthetic lum idea is a good one.) Ideas please, if you have a mo.

Cheers,

Olly

[MartinB]

My OSC work has been a bit limited sadly but I've done a bit. Obviously you will notice your sensitivity taking a hit, will be pretty much the same as doing filtered exposures with the mono I guess. The bayer matrix isn't IR blocking as far as I'm aware so you will still need an IR filter. The pros and cons for dithering are just the same as for mono.

I have heard that it is best to calibrate before debayering. If using Maxim and other software I guess, you can choose a weighting for individual channels when debayering to give you the truest colour. you can do this after debayering if you want. Some cameras need an adjustment to the offset to get the colour right, best to practice with daylight exposures.

There are a good few debayering algorithms. The most popular for quality is "variable number of gradients". This is the algorithm Maxim uses for it's "high quality" setting.

Process just as you would your combined RGB. You can prepare a luminence with the same data, sharpen it up to taste and blend in using luminosity and adjust the blend % to suit. The effectiveness depends on the target and you do loose some saturation. The best bet is to play around and see what you like best. So no different to LRGB I suppose.

Anyway, I'm sure SteveL, Steppenwolf and others will be along shortly with much more expert OSC advice.

[steppenwolf]

Martin has covered most of the main items but I can confirm that you should calibrate (BIAS, FLATS and perhaps DARKS as well if necessary although I never use them with my own OSC) before you de-Bayer. With regard to the offset mentioned by Martin, this is critical to correct de-Bayering and you may need to experiment - there are two offsets (X and Y) and these are each either 'on' or 'off'.

With regard to the weighting of the individual colour channels (RGB), I suggest that you take a daytime image of a colourful object and adjust the weightings to get a realistic daytime image BUT you must use an IR filter for both this colour calibration and for the actual imaging itself.

After calibration, de-Bayering and stacking, you can process the image just like a mono and RGB filters image. I do sometimes use a false luminance channel for some images but this is not a pre-requisite, I only use them if I need to boost the saturation strongly as layering in a false lum. channel (made from the original RGB image)to make an LRGB image holds the detail and by 'painting' the saturation boosted RGB data on top of the false lum. and blurring it, the artefacts of the saturation increase are lost.

[ollypenrice]

Strewth, I'm flounderiing already! (But I have been on three hours a night sleep for about fifteen nights so maybe it will all make more sense in the morning!) Many many thanks.

Olly