On dynamic range, monitors and the human eye

[Jessun]

This is a spin off from an ongoing 'what camera thread', and deals with the concept of dynamic range.

The question I have is basically: How important is dynamic range for the appreciation of an image?

Some simple basics dictate that two same f-ratio telescopes will expose a given surface unit at the exact same rate. Since the introduction of CCD's replacing film, we have a choice of pixel size to make on top of everything else.

If we aim at the same resolution - here meaning arc seconds per pixels for the image - we can have two scopes of different size imaging at the same f-ratio, and by putting a CCD with larger pixels in the big telescope, we can arrive at the exact same arc sec per pixels. Both telescopes will also saturate their respective CCD at a very similar moment. In essence, you can achieve the precise same image with two widely different telescope/CCD combos, and achieve them just as fast.

But that is not all. The big telescope produces the image on a CCD with big pixels. These pixels have a greater full well capacity, so whilst it 'blows out' at the same rate as the small setup, the range is perhaps 100.000 instead of 25.000 during a given time frame.

Also, if you aim to catch say only 5.000 photons with one pixel, the bigger setup will achieve this quicker. So it can be said to be more sensitive. Particularly useful of course in the dark that we typically aim at, or so it would seem.

So here the question arises: How important is this dynamic range for how well we perceive the actual image, on a computer screen, or perhaps in print?

The monitor for instance is perhaps not perfect, and the human eye is very complicated since it involves not just optics and chemicals, but also a brain that is all but perfect as an interpreter of inputs.

We also tend to look at images in the day time, in certain brightness, when if I understand it correctly our eyes' sensitivity is at its lowest, and finally to complicate matters more, before we look at the image we used software to fill in the gaps in the range as we stretch our data.

For scientific work I appreciate that big is better - the data is perhaps scrutinized far more accurately than what we do as we produce 'pretty pictures'. I also limit my thoughts here to the consumer market.

I would like to hear from you all, what you think. Is dynamic range a real deal breaker for an image or plain data sheet boosting.

/Jesper

[lensman57]

This is a spin off from an ongoing 'what camera thread', and deals with the concept of dynamic range.

The question I have is basically: How important is dynamic range for the appreciation of an image?

Some simple basics dictate that two same f-ratio telescopes will expose a given surface unit at the exact same rate. Since the introduction of CCD's replacing film, we have a choice of pixel size to make on top of everything else.

If we aim at the same resolution - here meaning arc seconds per pixels for the image - we can have two scopes of different size imaging at the same f-ratio, and by putting a CCD with larger pixels in the big telescope, we can arrive at the exact same arc sec per pixels. Both telescopes will also saturate their respective CCD at a very similar moment. In essence, you can achieve the precise same image with two widely different telescope/CCD combos, and achieve them just as fast.

But that is not all. The big telescope produces the image on a CCD with big pixels. These pixels have a greater full well capacity, so whilst it 'blows out' at the same rate as the small setup, the range is perhaps 100.000 instead of 25.000 during a given time frame.

Also, if you aim to catch say only 5.000 photons with one pixel, the bigger setup will achieve this quicker. So it can be said to be more sensitive. Particularly useful of course in the dark that we typically aim at, or so it would seem.

So here the question arises: How important is this dynamic range for how well we perceive the actual image, on a computer screen, or perhaps in print?

The monitor for instance is perhaps not perfect, and the human eye is very complicated since it involves not just optics and chemicals, but also a brain that is all but perfect as an interpreter of inputs.

We also tend to look at images in the day time, in certain brightness, when if I understand it correctly our eyes' sensitivity is at its lowest, and finally to complicate matters more, before we look at the image we used software to fill in the gaps in the range as we stretch our data.

For scientific work I appreciate that big is better - the data is perhaps scrutinized far more accurately than what we do as we produce 'pretty pictures'. I also limit my thoughts here to the consumer market.

I would like to hear from you all, what you think. Is dynamic range a real deal breaker for an image or plain data sheet boosting.

/Jesper

Hi Jesper,

I can only express my own opinion as a newbie to astro imaging but also as an old timer in classic photographic ways, mostly fiilm and paper, what we used to call the wet stuff. It took me over 3 years of study of the medium ( technical and not arty forty stuff ) in the late 70s and early 80s , mostly self taught, untill I felt confident to call myself a photographer, I eventually setteled on landscape and still life trying to emulate Adams, Weston and many other greats, too many for my tired mind to remebmer now. I learned the zone system of exposure, I learned how to compress the dynamic range of a subject of 14 stops of exposure to match a specific film emulsion - developer combination or how to expand a flat scene so I could expand the values and extract max dynamic range so the values were printable on a suitable paper of 8 stops ( zones ) of exposure and still preserve detail. In the final analysis what we did was to compress a zonal value of lets say 12 to 16 stops on a paper that could at the very best only cope with 7 to 8 zones of exposure without loosing important aestethic detail. I very much doubt that these aestethic values have changed much, even with the domination of digital processing these days. Preservation of a subjects dynamic range is as important now as it was during the old days of photography as it preserves detail and therefore data that could be analysed at a later stage. Sorry to be long winded but I think that I am correct in my analysis.

Regards,

A.G

[JamesF]

I think I'm in agreement. Though it may be impractical or financially impossible beyond a certain point, I think it's worth collecting as much information as is reasonably possible at the time. Today you might only be concerned about rendering on an LCD monitor say, but who knows what you might want to do with the image in five years time, or if the technology changes? Or if processing techniques change.

James