Blasphemy

[sojournerphoto]

A 5D1 is capable of resolving around 11.5 stops of dynamic range in the right conditions. Obviously getting the best from it requires some careful technique. Getting exposure right is critical to avoid burning highlights beyond recovery. The 1Ds3 is similar and I would expect the 5D2 to be pretty much the same, maybe edging towards 12 stops. How you deal with this in printing is another matter.

Bit depth in raw is directly related to dynamic range as raw capture is linear data (unlike jpgs which have a gamma curve encoded in the colour space so dynamic range is not determined directly by bit depth). However, although 14 bits is capable of covering 14 stops of range, the noise floor limits the useable dynamic range to somewhat less than the bit depth would otherwise allow. But to knock an old tale on the head - raw bit depth does not describe the number/size of steps on a same size staircase - it gives you many smaller stairs and allows you to take several smaller steps in place of the first step on a courser staircase, which is to say it expands the dynamic range (as noted noise floor considerations aside!).

Whatever, nice camera.

Mike

 

[Harry Lime]

I think one company Fuji or someone else had a Fovian (sp) sensor that had two 'layers' of sensors that increased the tonal range and as far as I know it hasn't been a success, so maybe they have already won.

That was the Fujifilm S5 PRO. It used a custom 6+6MP CCD sensor. The receptors were arranged in a honeycomb pattern. In addition to the main receptor there was an additional smaller one that was dedicated to capturing the upper zones. The camera combined these two data sets in what amounted to an internal HDR merge. The S5 PRO got a solid 10 stops, which is about 2-3 stops more than anyone else. But it was only 6MP, when everyone else had moved on to 8 and 10MP. The AA filter was also pretty aggressive, so it was a little soft, but wedding shooters loved this camera. If you had to shoot a bride and groom in a white dress and tux, it was the only game in town...




Personally I think we need about 12 stops of range and at least 16bit color.

 

[Harry Lime]

Bill, I have a 5D. I have a substantial investment in Nikon glass. I have an adapter. If I were to contemplate another 5D (as I haven't played with the new one yet, obviously), the ergonomics are the issue. I hate looking in a menu for something that should be on a dedicated switch. And that's what we pay for in F6s D3s and 1ds's and so on: direct access to important functions. .

Ergonomics. That's the exact reason why I switched from Canon to Nikon. The dial on the back of the D-series drove me nuts. I even hated it back when I shot the EOS 1-V.
The D700 feels right to me (as does the M8, but it's too expensive...)

 

[Harry Lime]

A 5D1 is capable of resolving around 11.5 stops of dynamic range in the right conditions. Obviously getting the best from it requires some careful technique. Getting exposure right is critical to avoid burning highlights beyond recovery. The 1Ds3 is similar and I would expect the 5D2 to be pretty much the same, maybe edging towards 12 stops. How you deal with this in printing is another matter.
Mike

I hear this claim repeated quite often, but I have yet to see it in practice. I know that there are certain testing programs that will calculate the DR to this figure, but it flies in the face of anything I have experienced in real life. Also sites like DPreview do not rate the useable DR of these cameras beyond 8-9 stops. You certainly can expose for the highlights and then dig out the shadows (because of the lack of noise), but there still is something lacking.

I've been involved with digital manipulation for a long time, but I have yet to see a DSLR (except for the S5 PRO) that yields DR that is marginally better than slide film.

I'm shooting with a Nikon D700 right now and have developed the RAW files in everything from Lightroom and Aperture to Adobe Camera RAW. There is no way that these files have 10 or more stops of range. I see about 3-4 stops above gray and maybe 4-5 below (I tend to reverse the balance when I shoot). The images are very low in noise and the color is extremely accurate, if you WB properly, but in terms of DR film negative blows them away.

 

[kshapero]

I can't believe you guys. I had a Sony Mavica from 1998, it cost $1000. I still see no improvement in dynamic range. Plus, the color isn't very good on any of them unless you use the sterile, plastic approach of PS. I'm not saying that digital is bad. I've had 5 d-cams and still have three. I still use them, but they suck the soul out of my photos.

With you on that.

 

[pphuang]

That was the Fujifilm S5 PRO. It used a custom 6+6MP CCD sensor. The receptors were arranged in a honeycomb pattern. In addition to the main receptor there was an additional smaller one that was dedicated to capturing the upper zones. The camera combined these two data sets in what amounted to an internal HDR merge. The S5 PRO got a solid 10 stops, which is about 2-3 stops more than anyone else.

Actually, 11.8 at iso 100-200. http://www.dpreview.com/reviews/fujifilms5pro/page18.asp

But it was only 6MP, when everyone else had moved on to 8 and 10MP. The AA filter was also pretty aggressive, so it was a little soft, but wedding shooters loved this camera. If you had to shoot a bride and groom in a white dress and tux, it was the only game in town...

The camera has an interpolation algorithm that uses the info from all the pixels to give a resolution comparable to 10 MP. And the AA filter can be removed... http://www.maxmax.com/hot_rod_visible.htm

Which gives you a pretty amazing camera, despite its limitations. I've spent time with many other DSLRs, including the D3, but the DR of the S5 keeps me coming back. Unfortunately, this just never caught on, and the S5 may be the last pro DSLR for Fuji. They are now very cheap, and I've been tempted to pick up another body for back-up.

I just hope that this concept is incorporated by one of the "big boys" in the future...