I'm not really keen on writing this article anymore - when I had the D700 for the first few days I couldn't help but being disappointed with the D700 and it's "normal" sensor. After the first couple hundred exposures with the D700, I could certainly say that the S5pro is SUCH a CLEAR winner. Fuji's SR sensor is a strike of genius, period. It seems that no other sensor than Fuji's can provide the smooth and "natural" highlight shoulder/rolloff for photographers thanks to the "two sensors per pixel" speciality - if you haven't used a Fuji with the SR sensor, you may not understand what a fundamental difference that really is, but if you remember the "old days of film" and how highlights slowly turned into overexposure (heck, even the most basic photos made with a compact camera and film from the supermarket had it!), you know what I am talking about (just look at that old holiday pictures you made on film with a compact camera, like I did:-). Compared to that, digital clipping is harsh, a brick-wall, ugly - and Fuji's sensor addresses exactly this problem.
But as the weeks passed by and I used the D700 more and more, I understood that the dynamic range difference is not that dramatic. After I began using UniWB and controlled my exposures more carefully with the histograms I can say that the fundamental difference between the two cameras is where the dynamic range difference comes from: the S5pro has it's extra bits in the highlights, the D700 has it in the shadows (and a good part of it is maybe because of it's huge pixels - after all, the D700 has "only" 12 megapixels on a full-frame sized sensor).
I've prepared some example photos. I don't include them in this post but uploaded them to an unpublic Picasa Web Album instead (just follow the link). I added captions to explain what you're seeing.
The question of course is: which concept is better? And I am absolutely convinced that the Fuji approach is the better solution. Why?
For someone who uses only JPEG and expects pleasing and natural images out of the camera, it provides a "highlight safety" that is simply essential. Nikon addresses this problem with it's Active D-Lighting technology, but I must admit I haven't played with it much (it doesn't make much sense if you shoot raw).
Knowing that the D700 often clips the highlights in scenes with high contrast, the only solution for me is bracketing over three frames: normal, -1 and -2. Just to be on the safe side with the highlights. Which means more work at home: three files per photo, and then checking which exposure is "perfect", highlight-preservation wise, at home. :-/
So furthermore and more importantly: exposure-wise, the S5pro is really a "fire and forget" camera no matter if you're using raw data or JPEG - take a walk on a sunny day and try to capture the average scene with a bright sky and some foreground stuff. Nikon D700: dial in -1EV exposure compensation and you're good to go (if you shoot raw!). Fuji S5pro: just shoot. It's just depressing to experience that. I forgot how comfortable and worry-free it is to shoot the S5pro. It simply requires "less fiddling around" with the camera and that is a luxury I've learned to value very much...
So I guess you are keeping S5Pro then. ^ ^
ReplyDeleteI read the beginning of this post earlier, and now I got back to it because it was still "unread" in my Reader. After the "negative" beginning I expected something different for the rest of the post. My surprise was that it was the same "rant" all the way :D
ReplyDeleteSeriously, nice to see tests like this. There is one thing to say of what I saw: the shadow details (levels) were identical to my eye, except for the noise. D700 does better noise-wise, but this is expected. I noticed a slight difference in the exposure time, should it have been manually set for identical exposure? Unless it was also a AE test, but that would be unfair because of the differences between exposure metering and the exact framing of the target. However, the time was 1/800 vs. 1/900, so maybe it's not a big deal.
For astro photography I'd definitely choose less-noisy body, as the lowest level details need to be preserved to the last bit. (Another unrelated note: the full-frame may not be ideal for hobbyist astro-optics). However, if I need a second-hand daylight camera, I'd be thrilled to try S5Pro! I can immediately think of many objects and situations where the highlight preservation would be extremely handy. For example storm/lightning photography, scenes with bright clouds, close-ups on kids in (usually partial) sunlight...
The exposure difference is due to the camera's matrix metering "fine tuning", or something. In the viewfinder, I saw 1/250s for the first photo set (without compensation), and 1/500s for the second set (which I then corrected manually by -1 stop to 1/1000s).
ReplyDeleteAs for the noise... I have the same impression Sami, and this is exactly why I consider the D700 to be more "usable" in the shadows, compared to the S5pro being more "usable" in the highlights.
Hello I come from D300s and D90. Before I had a Fuji s5 pro. I have always been disappointed with the colors and the DR of the Nikons. I can get the colors from my Nikons nearly the same as from the Fuji but it takes a lot of time. The colors of the Fuji are not always correct but the color shift is very predictable and easy to correct with a standard preset. I tried the D700 because I expected this to be a big upgrade from my D300s, but the results are exactly the same. Only at high iso's the D700 shows less noise. Last week I bought back a Fuji S5 pro and I am crazy happy with it. In general de JPEGS of the fuji are better than the raws from the Nikon. I put the fuji on maximum DR and just give a little bit more punch in postprocessing. Fantastic! Files are just between 3 and 5 mb (D700 up to 25 mb!. What a time saver is the Fuji! Fuji colors are so much more rich than Nikon! What a sensation to have no blown out highlights anymore. No more oversaturated "digital" greens from the nikons. Fuji is a clear winner when you are talking about IQ>
ReplyDelete