RustyBug wrote:
Seeing how this has turned into an algorithm writing debate ... YUP, that's just what the average user wants to incorporate into his workflow.
My point as a wedding photographer was one that several others have mentioned or hinted at. I come back with x thousand images from a wedding, cull them down to 1000 or so then have to edit them. Using ACR I can do that at a rate of about 5-10 seconds per image, it's usually just WB adjust, simple crop, brightness tweak and possibly a quick dodge/burn. Now with a fashion shoot your final picks are sent to the retouchers. Not a big deal if you have moire, that's part of their job to fix and the final pick is in the tens or less, not the thousands. Ditto landscape/street shooting, you are only giving the final treatment to a few images per shoot not hundreds. When you need to process up hundreds of images you are far more likely to accept the slight loss of contrast/sharpness that the AA filter on the 800 will bring, especially as the images are not anyway being processed in a style where it becomes crucial in print, rather than having to face the nightmare of fixing moire locally on hundreds of images just to be able to post a web gallery!
theSuede wrote:
*The DFT (FFT) of the image only shows spikes if the moire is present on a perfectly flat surface covered in a perfectly repetitive (and uniform) pattern. Take a wavy veil over a bride - Where's the perfect repetition that would give a spike?
Try an FFT on the image in that was linked earlier, on photo.net.
Beni wrote:
I'm a wedding photographer using 5Dc's, I get moire every single wedding from the bridal veil. Not terrible but certainly there. As a wedding photographer I wouldn't touch the 800E with a bargepole, the advantages are outweighed for this profession by the downsides. We aren't printing as large as fashion shoots are neither are we lighting as hard, etc.
Complete agreement. If I were a wedding photographer, there is no way I would get the D800E. Mind you, I would have thought that the D700 would be a better choice that the 5Dc given it's superior low light performance, and I think better than the D800 for that matter.
On the other hand I wonder what a D800 file downrezzed to about 18 megapixels would look like in comparison to the D700 at high iso. I'd always known my 5Dc to be far superior to my 1Ds3 at iso 1600 till I downrezzed to match resolution when suddenly not only was the noise better but even at the same pixel count, so was the detail.
Setting up a downrez in ACR or LR is a matter of applying a preset, it's seconds and the results are remarkeable. If you want a one solution does all then at least at the computer end, you can have 36 megapixels of portrait detail and 12-18 megapixels of high iso reception shooting detail, all from one camera and without having to do anything that batch apply a preset. You would still need to capture in full rez which is a pain of course but I would assume that for high iso the D800 will give a very good 12 megapixels worth of clean detail for those scared D700 shooters.
Theoretically of course, we don't have any raws to play with yet...
Beni wrote:
My point as a wedding photographer was one that several others have mentioned or hinted at. I come back with x thousand images from a wedding, cull them down to 1000 or so then have to edit them. Using ACR I can do that at a rate of about 5-10 seconds per image, it's usually just WB adjust, simple crop, brightness tweak and possibly a quick dodge/burn. Now with a fashion shoot your final picks are sent to the retouchers. Not a big deal if you have moire, that's part of their job to fix and the final pick is in the tens or less, not the thousands. Ditto landscape/street shooting, you are only giving the final treatment to a few images per shoot not hundreds. When you need to process up hundreds of images you are far more likely to accept the slight loss of contrast/sharpness that the AA filter on the 800 will bring, especially as the images are not anyway being processed in a style where it becomes crucial in print, rather than having to face the nightmare of fixing moire locally on hundreds of images just to be able to post a web gallery!...Show more →
makes sense i've only shot weddings for fun and never actually even been to a wedding where the bride wore a veil over her face (only been to 20 or so), so i've never had any moire to deal with in such circumstances despite my cameras having pretty weak AA filters. i suspect this camera is targeted mostly at landscape photographers. i really don't generally feel that maximum sharpness is necessary in photos of people (maybe just me). that said cityscapes are where i see moire the most.
sebboh wrote:
i suspect this camera is targeted mostly at landscape photographers. i really don't generally feel that maximum sharpness is necessary in photos of people (maybe just me). that said cityscapes are where i see moire the most.
+1 I think the D800E makes sense for the natural landscape photographer. I think the conundrum comes in because a lot of folks don't shoot just this subject matter and cannot afford to buy the D800E as a very specific, dedicated tool. For that niche it makes sense.
Tariq Gibran wrote:
+1 I think the D800E makes sense for the natural landscape photographer. I think the conundrum comes in because a lot of folks don't shoot just this subject matter and cannot afford to buy the D800E as a very specific, dedicated tool. For that niche it makes sense.
On diglloyd's blog he suggests that situations causing moire will be less for the D800e than it would be the Leica M9 (for example) becuase of the much smaller pixel pitch.
To quote, "If I thought that moiré were a serious general issue, I would not buy a D800E. But I do not think that— with 36 megapixels, the chances for moiré are even more reduced over 18 megapixels."
It that a reasonable assertion or will just different subjects/content provoke it?
Lotusm50 wrote:
On diglloyd's blog he suggests that situations causing moire will be less for the D800e than it would be the Leica M9 (for example) becuase of the much smaller pixel pitch.
To quote, "If I thought that moiré were a serious general issue, I would not buy a D800E. But I do not think that— with 36 megapixels, the chances for moiré are even more reduced over 18 megapixels."
It that a reasonable assertion or will just different subjects/content provoke it?
Different subjects/ content wil matter but I don't know the definitive answer about how much it's going to be a real issue. We need to see more tests between the D800 and D800E. Just going by the full size Nikon sample photo's, I tend to like the rendering of the D800 better. The shots from the D800E look sort of thin, artificial/ digital to me overall and the D800 shots look to have enough real detail that I suspect very good sharpening technique would satisfy me...but I want to see more from both.
Lotusm50 wrote:
On diglloyd's blog he suggests that situations causing moire will be less for the D800e than it would be the Leica M9 (for example) becuase of the much smaller pixel pitch.
To quote, "If I thought that moiré were a serious general issue, I would not buy a D800E. But I do not think that— with 36 megapixels, the chances for moiré are even more reduced over 18 megapixels."
It that a reasonable assertion or will just different subjects/content provoke it?
It is very reasonable. And in any given situation where the 18MP M9 would give "a certain amount" of moire, the 36MP D800 with a similar field-of-view lens will give slightly less strong moire - at a point slightly further away from the camera, OR in a pattern slightly smaller.
Beni wrote:
On the other hand I wonder what a D800 file downrezzed to about 18 megapixels would look like in comparison to the D700 at high iso. I'd always known my 5Dc to be far superior to my 1Ds3 at iso 1600 till I downrezzed to match resolution when suddenly not only was the noise better but even at the same pixel count, so was the detail.
.../snip/...
Theoretically of course, we don't have any raws to play with yet...
There are some raw's :-)
From the samples I have I would say that the D800 has almost exactly the same strength of noise at ISO12800, when downsampled "plain & stupid" to 12.3MP size (D700). What makes a HUGE difference is that the noise - even if it is just as strong, according to a "measurement" - is very much finer in grain. The colour noise is also less blotchy.
So given some practice, or just an optimized workflow, I'd say the D800 is better than the D700 at high ISOs.
Looks good to me. As we get better raw-converters (the interpolation engine I used is definitely NOT good for noisy images!) expect better results too.
The "normal" 800 is capable of a lot better detail than what you've seen from the published full-size images.
I will probably never be able to understand what the "big-uns" are thinking when they publish what should be showcases of optimal captures with new products.
Are they trying to avoid being accused of "cheating"? If so, they're trying to hard!
Looks good to me. As we get better raw-converters (the interpolation engine I used is definitely NOT good for noisy images!) expect better results too.
Honestly, that looks thoroughly unimpressive to me. Take this as a counter example - M9, ISO10,000 (ISO2500 pushed two stops in post), some noise reduction in LR applied and resized to D700 size.
Here's the original without the +2 stop exposure, resize and NR: Original (ISO2500)
I would expect the D800 image to be dramatically better both because of the increased pixel count and the more modern sensor. The M9 is not exactly considered to be a high performer when it comes to high ISO. Perhaps the reduced D800 image is a bit better, although I'm not even sure of that - but certainly not dramatically better.
Looks good to me. As we get better raw-converters (the interpolation engine I used is definitely NOT good for noisy images!) expect better results too.
theSuede wrote:
Looks good to me. As we get better raw-converters (the interpolation engine I used is definitely NOT good for noisy images!) expect better results too.
To go off on a slight tangent, I agree completely about improvements in raw processors. I've been reprocessing a lot of my high ISO files using the latest version of ACR's camera process and the change is dramatic.
The ACR processing for both sets of images is almost identical--I adjusted the sharpening in the 2010 versions, and pulled a nasty highlight in the first image, but otherwise left them as they were when I first processed them. For the first image, the post-ACR processing is 99% identical--I literally dropped the new version into the smart objects I was working with, re-cloned out a ghost, tweaked the opacity of my sharpening a bit and called it done. The second one was completely reworked post-ACR.
... YUP, that's just what the average user wants to incorporate into his workflow. 
