[A200/Tamron 18-250, Promaster MC UV filter, ISO400 1/60s F8.0 90mm effective -0.3eV > dcraw H0W A200 updated ICC, AHD 3-color +0.0eV > Gimp rotate, crop, brightness +40, contrast +47, usm 1.0-0.60-10]

Without NI but the same processing as before.
Just a hint of grain in the sky...if you're a stickler for low-noise, that will annoy you to death :) but admittedly there's a difference between some if any grain and a buttery-smooth sky. So check out the NI version of this with the same processing.

It's debatable as to whether the NI version is better but in any case this is serious "playing around in the margins". The two versions are almost exactly the same when viewed at full-image which ultimately is how it's going to be evaluated. The big issues to me are whether the contrast is high enough, do the colors still look somewhat muted and pastel, or is the shot "rich & wet" enough.

With a HDR scene is impossible to get that "wet" effect without blowing-out some of the shadows & highlights. So the overall-exposure determines how much of the shadows & highlights will be lost in boosting the contrast. So it becomes a simple tradeoff between exposure and contrast, with noise not really a big factor here even at ISO400. The bigger factors are the DR and CS of the sensor at ISO400, because the addition of contrast is an attempt to restore the vitality and richness of the original scene, which naturally is muted somewhat by the gamut-issues of the camera at the ISO at which the shot was taken. The bottom-line is that the camera is irrelevant here, this shot should not be taken at ISO400. Technically no "landscape" shot should be taken above ISO100 or 200 at most. So it has to be evaluated in the context of having been shot at an ISO that is much higher than optimal for the scene. Then of course a comparison between cameras &-/ lenses can be made ;)

Did you get the shot that you want...are you happy with it...would you be happier if you had used a different camera or lens or shooting-technique...or a slight change in PP...note that you can change those 4 things (and there are infinite variation in all of them) and still not be happy with the results. Ken Rockwell, for example, seems to be happiest shooting film. Despite all his talk about and purchase & shooting of digital gear. Would this scene just simply look better if it were shot with film, to the point where really no affordable digital camera could compare? It might, but if I tried to shoot this with ISO100-400 film, it would have to either be with an IS lens or a much-faster lens than F8 or both. Or a tripod. I probably can't use a tripod on the bridge that I shot this, even this early in the morning. But the question remains: would film-results be so-much better here as to be worth the effort? Well, even if they are, what happens when the result is scanned to digital format and then displayed on Flickr? If the film-shot contains the entire visible gamut, for argument, and the Flickr result is limited by display-gamut, then what?

How can we really know how much device-gamut really affects the results if we have never seen the original scene? Or even if we *have* seen it, it's of unrealistically-high DR for a digital camera to shoot and a digital device to display? How can a HDR scene ever look good when displayed on a small-gamut device? The use of better gear and/or better technique can only improve the image when the potential-improvement is supported by the display & the rendering software. And clearly the biggest advantage that the display has relative to the camera is in SNR, really the fact that it is always at "the optimum ISO". Technically there is some ISO at which the cameras' gamut is smaller than the displays, but who cares because by that point the SNR of the camera (or DR or CS) will be the overwhelming problem. So we are bound on one side by display-gamut and on the other side by SNR, really by ISO. And clearly of the two the ISO is the bigger problem here.

But what if I had shot this at ISO100 off a tripod?

How much better would it look?

It's easy to tell when the SNR is not high enough. But what about the DR or CS? Again: can I *really* tell the difference between the 22.5 bits of CS that the A200 has at ISO100 vs the 18 bits that the G9 has at ISO80, when I look at the results on my "calibrated" display? Or is the difference that I see mostly due to the 37dB vs 30dB of SNR of the two cameras at their lowest ISOs? Or even the 11.11eV of DR of the A200 vs the 9.79eV of the G9 at their lowest ISOs?

What if my monitor is so crappy that it only supports 12 bits of CS and 6 bits of DR, and as long as the SNR is "good enough" and the camera-gamut is "large enough", there really is no visible-difference? Well then obviously buying a much-more expensive, much-cleaner camera and/or a faster lens makes sense because it would allo

[Canon 5D/Tamron 28-300VC, B+W MC UV filter, ISO800 1/125s F8 28mm effective -0.5eV > dcraw 5D ICC H0W AHD 3-color +0.0eV > Gimp contrast +5, usm 1.0-0.70-05]

So I finally got my batch-file modified to use the proper ICC profile, ran off a few 5D shots and came up with this as an "example". Please forgive the induced keystoning :) I shot this before I even knew what "keystoning" was LOL

And this is "color-corrected"?!?

Wow if this color is accurate then I will seriously have to rethink a few things. I can't believe that this is accurate. I can see the color-checker card itself being out of gamut for this camera. Wow, it's too bad that I can't just take a jaunt up to Manhattan tomorrow to take a comparison-shot with my A200...maybe I'll just ride my motorcycle up there after I get the wheel-bearings replaced tomorrow ;) then ride back all night ;) stay up there all week and hang out in the Village...too bad I can't park my bike anywhere up there and trust that it will be there when I get back ;)

Hell I'd have to be a nut to try to ride my motorcycle from DC to Manhattan in one day. That's what buses are for. I can barely stand to *drive* that far on the Turnpike much less ride a bike on it for 5 hours LOL 6 hours for stops, 7 hours for traffic...another hour or two for rain...each way...plus parking...shit...

...then the effing tunnels and bridges going into Manhattan...

Anyway this is a prototypical sucky-color 5D shot. I took it just south of Chelsea Piers around 18th St on the West Side highway. It's gray, ok, fine. It's not *that* gray, the sky is orangish crap with gray, basically just a gray shot with barely any color in it. It is like someone took a black & white photograph and then added some color with colorizing software. The very-best color that I could get out of this camera was to add some reddish metallic gloss to it with a tone-curve in Bibble4. Next to a D300 it simply sucked ass for color. Hell it even made my 400D look good. The ICC profiling fixed the black-point which is good, and hid luminance-noise in the shadows, but otherwise made no change here.

I appreciate the lack of noise but that's a hell of a compromise to make.
And before you talk about the camera jpegs and the in-camera tone-curve, the embedded camera jpeg looks just like this but the overall exposure is a little higher.

I just couldn't see going back to this...it's great for handheld lowlight shooting around twilight especially at the price, but for most of the shots that I want to take it is just too color-blind. Though after looking at the next 5D shot in my stream you'll think this is downright purty.