Comparing cameras to our eyes

[raymo]

From what I have read about astro imaging cameras, it seems that
different makes and models, [modded and unmodded] have varying
sensitivity to the colours of the spectrum. If that is so, which, if any,
cameras produce an image which would closely approximate what
the human eye would see if the object was bright enough for us to
see it's colours fully?
raymo

[cometcatcher (Kevin)]

It would be 99% the OIII regions Ray. Our eyes are virtually blind to Ha, (Ha is near infrared) which is the red we see in photos but we have peak sensitive to green. The bluish / green areas of photos would be what our eyes see. That's why OIII filters work so well visually for nebula.

[raymo]

Thanks Kevin.
raymo

[rustigsmed (Russell)]

wouldn't it be nearly equivalent to an unmodded DSLR camera? otherwise they wouldn't be very good at taking photos that appear 'real'?
you would therefore still see some red and blue but with the 'green' being most evident?

[jsmoraes (Jorge)]

which, if any,
cameras produce an image which would closely approximate what
the human eye would see if the object was bright enough for us to
see it's colours fully?

If I understand your question, you want know which camera model in the market is more similar to ours eyes. Mainly in color response.

Undoubtly: none of modified cameras. They perceive more infrared information than ours eyes.

Now, about the most similar camera ... is very hard to say. They are digital, our eyes is analog. The photos produced always have some internal graphic process, as contrast, tone, saturation, and many others.

I don't know if there is a curve of color response with default set to compare with the curve of response of ours eyes.

And with ISO - or gain and gama of CCD - and time of expositon, they change that response. Ours eyes haven't these ressources.

But your question makes sense. There will be a camera more similar to our eyes.

[cometcatcher (Kevin)]

A box brownie.

[LightningNZ (Cam)]

Thanks to a dynamic iris, memory and bunch of other weird things like dark adaptation, the human eye is capable of an amazing 30 stops of dynamic range (1,000,000,000:1 contrast ratio of black to white levels), though "only" 1,000,000:1 is available without chemical adjustments. Without iris adjustments as well we have a static contrast ratio of somewhere between 100:1 and 1000:1. (see http://wolfcrow.com/blog/notes-by-dr...the-human-eye/ and http://en.wikipedia.org/wiki/Human_eye).

So for a single "exposure" we actually have less "grey levels" than even a cheap digital camera, and far less than a deep-well CCD.

Colour response changes things again. DSLRs with their Bayer Matrix of RGGB (Red, Green, Green, Blue) pixels are designed to give a very "human" colour response, but an individual pixel will likely respond differently to a human "cone" (and don't forget your low-light "rods" are essentially grey-scale cyan).

As to exposure times the human eye is potentially able to "expose" for up to 6 seconds (sorry no link for this), with sub-sampling and memory helping to "smooth over" what we actually perceive.

Hope this helps. Just a few bits and pieces I've picked up over the years of working with sensors and studying psychology and biology.
-Cam

[Shiraz (Ray)]

There would not be much point in building a camera with anything like the eye's response. To see an image with realistic colour, it must have the same colour balance as the real world. Then your eye imposes it's own colour distortion - which your brain interprets - and you perceive colour as it would be if the image was a real scene.

If you have a camera with the eye's response already built in, the images will not have the same colour balance as the real world and will look green as Kevin says - you don't want to put in the eye response twice (once while getting the image and again when you look at it).

If you want a camera that gives you astro images that are something close to what you would see if your eyes were more sensitive, take Russell's advice - one of the DSLRs should do - their "daytime" colour setting should provide images that are fairly true to reality if viewed on a reasonably well calibrated screen. Of course, your brain will probably still try to impose some form of colour distortion to keep the scene "balanced", but there is not much you can do about that - you won't even be aware of it. And there is also the atmosphere that will add sky glow (colour varies) and attenuation (mainly of blue) - these effects will limit the colour accuracy of DSLR images of dim targets under polluted skies. So, find a dark sky site on a night with high transparency and use a DSLR on a relatively bright target - that should give you at least some idea of "the actual colour", which is what I suspect you are seeking.

If you use an astro camera on dim targets, colour calibration must be carried out in processing. Software such as Pixinsight includes modules that remove background colour cast from skyglow and balance the colour channels to produce star colours that are consistent with a "typical" population of stars, or galaxy colours that are consistent with typical galaxies. The end result before stretching should be fairly close to "the actual colour", although there could be some extension of the red spectrum that your eyes could not see well (if at all), but that may be mapped into the visible region.