Quick update looking at my 16mp CMOS shots (from various models) I'm surprised that not much progress has been made with "highlight headroom" it's better than the older sensors, possibly not v the 12mp CMOS. The main improvement seems to be you can be more aggressive with the shadows on the newer sensors. Still easy to blow out the sky even a few years down the road.

We've still got some way to go here..that's my view at least APS-C wise

How do you think a new camera with a new sensor shooting raw and converting with LR 4.3/ACR 7.3 would do?

I've just done an HDR from one raw file from the Canon 6D. As Barry says, the gain is all in shadow zones, the highlight headroom is still about 2 stops as it always was for Min/Sony. But previous Canons had only 1 stop highlight headroom, so it's an improvement within their range. It now matches Sony

David

I would appreciate an explanation of this since it is something that I may be confused about. It sort of seems to me that how one wants to use the DR is up to the photographer so I don't quite understand talk of gain in the shadow zones or headlight headroom. This may just be my misunderstanding though. With color negative film, I think, there was more room at the top than the bottom so it was often better to give more exposure and with slides it was generally the opposite. One can meter and set exposure as one wants with raw digital too and if one wants more in the highlight end then expose a bit less and then adjust when the raw conversion in LR, ACR, etc. is done. If one wants more in the shadows end then expose a bit more and then adjust when the raw conversion in LR, ACR, etc. is done.

Here are some interesting articles that I have found in the last few months:

http://www.dpreview.com/articles/4241806072/sense-and-sensitivity

RAW is not Raw

http://theonlinephotographer.typepad.com/the_online_photographer/2012/10/raw-is-not-raw.html

Why ISO Isn't ISO

http://theonlinephotographer.typepad.com/the_online_photographer/2012/09/why-iso-isnt-iso.html

The problem with just exposing more, or less, is that digital raw files will at some point clip entirely (no more data present). It's best to give a normal exposure, so that there will a reasonable amount of hidden range at the highlight end that can be pulled back in without colour clipping or total clipping. Some 10-bit raw files have very limited information in the shadows (hence the outdated Canon 'expose to the right' advice) so you can't brighten up the shadows without introducing noise. Most 12-bit files are a bit more robust, but 14-bit raws tend to be far better - you can dial in +2 exposure and the shadows will brighten up perfectly and stay clean without tone breaks or noise. When this is combined with good highlight recovery headroom, you can really work the raw file to an extreme.

It goes far beyond the range ever possible with film of any type.

David

David, that matches what my understanding is, so, I think it hinges on what is a "normal exposure"? I know what you mean, but my thoughts are that one can choose the exposure for a particular photo based on what one wants to achieve and one's knowledge of the characteristics of the camera and the raw converter. Even with film that was common. I think Ansel Adam's Zone System is one of the more famous variations on that. These days most cameras have a matrix metering system with sophisticated software running on a computer inside the camera and the result is sometimes they will give a different exposure than a center-weighted meter or averaging meter. The matrix metering system probably takes into account the characteristics of the sensor and camera as a whole. Some cameras (I think the E-M5 does this and I also think I have read that some Sony cameras do this) "underexpose" a bit in contrasty situations and then internally gives a boost for jpegs and the camera profile for the raw converter does it in order to try to keep highlights from blowing out.



Since this thread was started a long time ago digital has gotten better. Earlier the comments in the thread were that film was better in this regard so it is cool to see the progress that has been made. I don't do testing, but my feeling is that my 16mp E-M5 is somewhat better in DR and better in high ISO than my 12mp A700 and 18mp Canon 60D even though it has a smaller sensor than both of them. I think DxO says the same thing.

Yes, the E-M5 has better DR and high ISO than the A700 and 60D:

http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/%28appareil1%29/793|0/%28brand%29/Olympus/%28appareil2%29/663|0/%28brand2%29/Canon/%28appareil3%29/562|0/%28brand3%29/Sony

Since my A700 and 60D were the cameras that I was using at the time I got the E-M5 that is the comparison that was meaningful to me. Of course, later Sony models with their larger and newer sensors would do better.

I just went back to the beginning of this thread that started in 2008 and skimmed through it again. Pretty interesting and worth doing to see the discussion and also how digital has evolved over the last 4+ years.