ISO 50 - Reduced DR on original DP's

SRT201 wrote:

I noticed that Sigma pointed out the reduced dynamic range of ISO 50 when it was introduced in the DP1s. I gather base ISO is actually 100. Has anybody noticed just how much DR is lost. I originally figured ISO 50 would be the obvious choice for landscapes but if much DR is lost then ISO 100 would be the one to stick with.

Since the DP1s is ISO-less, the DR isn't actually reduced. ISO 50 is similar to a ISO 100 +1.0 (exposure compensation) shot. So you get cleaner shadows but lose 1 stop highlight recovery.

SRT201 wrote:

I noticed that Sigma pointed out the reduced dynamic range of ISO 50 when it was introduced in the DP1s. I gather base ISO is actually 100. Has anybody noticed just how much DR is lost. I originally figured ISO 50 would be the obvious choice for landscapes but if much DR is lost then ISO 100 would be the one to stick with.

You have to avoid overexposure with ISO 50. Highlights get  permanently blown

Interesting.  So there's no gain on the imager and it's all just a matter of underexposure and multiplication during the development stage.

In that case then the ISO 50 with -1 compensation should be identical to ISO 100's normal exposure.  Correct?  Fairly easy to test.

Are the Quattro's ISO-less as well?

SRT201 wrote:

Interesting. So there's no gain on the imager and it's all just a matter of underexposure and multiplication during the development stage.

In that case then the ISO 50 with -1 compensation should be identical to ISO 100's normal exposure. Correct? Fairly easy to test.

That's correct in theory, but since SPP is correcting the "overexposure" of ISO50 internal, it might be slightly different than ISO 100.  Same with the ISO-less SD1, it's theoretically the same if you under expose ISO 100 by 2 stops or choose ISO 400. The RAW data would be the same. But SPP is interpreting the image in a way that we can't control, so the result is slightly different too.

Are the Quattro's ISO-less as well?

Quattro, SD15, DP1x and DP2x are different, because they have an AFE (analog front end)

SRT201 wrote:

Interesting. So there's no gain on the imager and it's all just a matter of underexposure and multiplication during the development stage.

There is gain in terms of better IQ. I'm puzzled why people obsess about DR. For the most pleasing photos, people have reduced the DR in post anyway.

Here is a photo I took years ago. In this I reduced the DR in post quite a bit to obtain something more pleasing: https://www.flickr.com/photos/prebenr/6284631384/in/album-72157627678060149/

Not saying it is a goo picture, but if one wants to have an indefinite DR then things starts to look like snapshots.

In that case then the ISO 50 with -1 compensation should be identical to ISO 100's normal exposure. Correct? Fairly easy to test.

No.

Are the Quattro's ISO-less as well?

No.

Kendall Helmstetter Gelner wrote:

SRT201 wrote:

Interesting. So there's no gain on the imager and it's all just a matter of underexposure and multiplication during the development stage.

In that case then the ISO 50 with -1 compensation should be identical to ISO 100's normal exposure. Correct? Fairly easy to test.

"In theory there is no difference between theory and practice. In practice there is."

- Yogi Berra

Are the Quattro's ISO-less as well?

All of the Sigma cameras are.

That said, when using SPP to actually process the raw files they do not behave as if they were ISO-less.

You should test and see what happens (especially in the shadows), but I agree with the statement that you lose about a stop of highlight recovery.

I thought the Quattros had AFE. If it is only in SPP, then it would be interesting to hack the RAW images of the Quattros to change ISO. Roland, any pointers?

The subject can have a brightness ratio that is pretty low for dull scenes (say 1:4) or pretty high in bright scenes (say 1:20,000).  If your sensor was capable of infinite dynamic range it would capture all the detail of the original scene.  You would then make the decision during editing as to how you wanted to render the scene in your image.

If your sensor has a dynamic range that is less than the brightness range of the scene then your camera cannot record all the detail and your choice of exposure determines what you end up  losing. In this case, you do not have the same breadth of artistic renderings available to you because your file doesn't contain all the data.

Generally (assuming you edit photos rather than using the untouched OOC files), you are better off with the first situation as it gives you far more editing options.

PrebenR wrote:

SRT201 wrote:

Interesting. So there's no gain on the imager and it's all just a matter of underexposure and multiplication during the development stage.

There is gain in terms of better IQ. I'm puzzled why people obsess about DR. For the most pleasing photos, people have reduced the DR in post anyway.

Here is a photo I took years ago. In this I reduced the DR in post quite a bit to obtain something more pleasing: https://www.flickr.com/photos/prebenr/6284631384/in/album-72157627678060149/

Not saying it is a goo picture, but if one wants to have an indefinite DR then things starts to look like snapshots.

In that case then the ISO 50 with -1 compensation should be identical to ISO 100's normal exposure. Correct? Fairly easy to test.

No.

Are the Quattro's ISO-less as well?

No.

ISO 50 is just overexposed ISO 100. It is useful for lower contrast subjects as you get a bit more exposure in the shadows. But shooting ISO 100 with +1EV exposure comp will yield the same file.

Tom Schum wrote:

I remember Carl Rytterfalk ISO 50 shots from the SD15, so smooth, so brilliant. I had to try it with my SD15. Same result.

You get a different look at ISO 50. I still have my classic DP2, need to try ISO 50 again.

I think ISO 50 is very worth the trouble, but overexposure is definitely fatal.

DMillier wrote:

The subject can have a brightness ratio that is pretty low for dull scenes (say 1:4) or pretty high in bright scenes (say 1:20,000). If your sensor was capable of infinite dynamic range it would capture all the detail of the original scene. You would then make the decision during editing as to how you wanted to render the scene in your image.

If your sensor has a dynamic range that is less than the brightness range of the scene then your camera cannot record all the detail and your choice of exposure determines what you end up losing. In this case, you do not have the same breadth of artistic renderings available to you because your file doesn't contain all the data.

Generally (assuming you edit photos rather than using the untouched OOC files), you are better off with the first situation as it gives you far more editing options.

One thing to get had to video filming, is that here you want to get a result!t as flat and little colour as possible. Then you grade in post.

My point would be that sure you would have more options in post, but would it not also kill creativity when you took photos?

I really don't know why Sigma had to artificially make that sounds so hi-tech.  The "Analog Front End" a term very common in signal processing is a gain stage that amplifies the signal from the imager (in this case).