Re: Nikon D200 and Canon 350D/Rebel XT Noise Tests

I just noticed it doesn't open DNG files... bummer.


JPS@xxxxxxx wrote:

Paul Furman <paul-@xxxxxxxxxxxxx> wrote:

Nothing on [the camera] page does anything to your loaded NEFs except the
rectangle at the bottom. The rest is for controlling some other devices
from the computer.

Ah, remote control, thanks.

This is interesting for analyzing the data but how do I get Iris to apply a proper gamma & take the linear aspect out? I see in the view menu: 'Logarithm, Equalization and Modified Equalization' but are those really desirable? Similar to Photoshop's Image > Adjust > Equalize, which I believe sort of flattens images by making a flat histogram.

If the image is already color, "Gamma" appears in the "View" menu. If
for some reason you want to see the checkered, greyscale RAW with gamma,
you have to convert 16-bit to 48-bit in the "Digital camera" menu, then
gamma will appear, but you have to adjust all three channels separately,
unfortunately.

Hmm, I getting something normal looking with gamma around 3.

OK I see now that the 'median' and 'gradient' look a little better once doing the CFA conversion. But still doesn't look normal and I don't understand what exactly those do.

None of them will look normal, because they do not "demosaic" in the
traditional sense; they only interpolate the three color planes
separately. No attempt is made to find the luminance of each pixel.

Hmm, so it gets a tad more detail & less noise for astronomy type work but is just not suited to color photography? Again I wonder what the use is then.

Any idea what proper RGB WB values should be?

That depends on the camera and the light, and is often subjective. A
good starting value for most cameras for daylight would be R=1.9, G=1.0,
B=1.4.

Yeah that got me in the ballpark.


From there just save as a tif & when opening in photoshop, sRGB seems

as good as any color space to assign. Then I do a levels adjustment to set the brightness of the middle gamma level & make it look like a normal exposure.

There must be some way to assign a color space though. It has to be something.

No; it's meaningless in this context. This is a RAW viewer, not a RAW
converter. The program knows *nothing* about the color characteristics
of the CFA; all it knows is the CFA pattern of the camera you select.
The red, green and blue on your monitor are not the red, green, and blue
of the CFA, exactly. You don't use this software to get pretty color
pictures (although it can make some very nice B&W images).

The astronomers do though right? I can twiddle WB till the colors are about right & tag it sRGB since that's the color space of my monitor but I guess what I need to do is apply a curve to it, to remove the linear nature of the capture. Adjusting gamma approximates this but very crudely compared to what I imagine the correct procedure would be. What's more the gamma adjustment in Iris has no impact on the TIF I save opened in PS, all the data is still crammed over in the far left of the histogram. I can adjust levels in PS but it's posterizing severely & the whole point is to expand into the desired range before chopping 4095 down to 255, right? How do I do that? What I see in photoshop is everything crammed into 0-40 (out of 0-255) in levels.

When I was messing with DCRAW, I came on the advice to use the results only as b&w (luminance) and use a regular conversion for the color (chrominance). This can work OK but I suspect the mismatching causes some blur and that may lose the advantage. The reason is DCRAW (at least older versions) does a less aggressive demosaicing which leaves some crazy looking color 'noise' artifacts, but supposedly it produces a sharper luminance image although it might show moire quicker.

You can, of course, adjust hue and saturation in PS to get a nice
looking color image, especially if you downsample the color image first
so that the interpolation is nullified.

Again, about blackpoint - your Nikon probably has a blackpoint of 0, but

Yes when I adjust threshold for 0-10 or 0-5 there is a little bit showing.

If the blackpoint were higher than that, you'd still see something in a
dark image, as a negative; there is image below the blackpoint in a
Canon, or any camera that doesn't have a zero blackpoint!

What I see is not negative. If I set the range 0-5 I see a few black flecks in the darkest part of the image.
.

Relevant Pages

  • Re: Nikon D200 and Canon 350D/Rebel XT Noise Tests
    ... rectangle at the bottom. ... Similar to Photoshop's Image> Adjust> Equalize, ... That depends on the camera and the light, ... and you have not subtracted the blackpoint, there will be a color cast ...
    (rec.photo.digital.slr-systems)
  • Re: fuji finepix 810
    ... Brian wrote: ... A very fine camera, though keep in mind that it's different from ... - Not sure what you mean by photos being a little blurred, ... there's a slider on the left that allows you to adjust ...
    (rec.photo.digital)
  • Re: Why use raw?
    ... all-digital P&S camera with an EVF where you can adjust the diopter of the ... I use reading glasses but have to use ... The +/-5 diopter adjust for the EVF ...
    (rec.photo.digital)
  • Re: Gamma & linearity - how to use them
    ... Those settings amount to the same as the often mentioned "gamma 2.2" ... Linearity also appears to change saturation to some ... If you want to use custom profiles, ... Then adjust "exposure" to just below where highlights ...
    (rec.photo.digital)
  • Re: Focus calibration
    ... But the regular punter won't get a free repair unless the target shot shows resolution is down by 30 per cent, or exposure is more than 1/3rd stop out, or the sensor is out of horizon level by more than 1 degree... ... On my camera the right half of the frame is softer focus than the left getting worse the further right you look. ... It's probably within their spec but I don't feel I should have to turn the camera to an awkward position to take sharp pictures. ... If it's outside these QC figures, they adjust the camera. ...
    (rec.photo.digital.slr-systems)