DaSigmaGuy wrote:
> You cannot and indeed should not draw any meaningfull conclusions
> from these review images because the test is obviously
> fundimentally flawed from the start.

I agree, it is not the best test, but the same splotches are reported by every other SD14 user who tried to shoot JPEGs at ISO 1600. Have a look here for a recent example:

http://peterbernik.blogspot.com/2007/05/sigma-sd14-hi-iso-sample.html

> Firstly, all the shots appear to be in-camera jpegs but it is well
> known amongst Sigma users that that SD14 in-camera jpegs are often
> of poor quality...Serious SD14 users ONLY use RAW.

The same is true with other "serious" DSLR users, but the majority of them also shoot JPEGs. Sigma should have developed a more decent in-camera processor or should have left the JPEG support out. As it is, the dreadful JPEG quality is going to bite their rear-end, as all the major review site focus on JPEG comparisons.

> And secondly native size SD14 images are 2640 x 1760 in size, NOT
> 4,608 x 3,072 as the pics are in the comparison so its hardly fair
> to compare greatly upsized SD14 images against native sized Ricoh
> GX100 images, especially as the end results depend greatly on the
> upsizing algorithm used and the fact that upsized images always
> look a lot softer than native sized images.

Fair enough, but I've also seen SD14 ISO 1600 JPEGs that were shot in the native resolution, and the GX100 is still much better.

> As for the noise, it would have been much less that shown if the
> photographer knew what he was doing as the SD14 shot in question is
> underexposed...Correct exposure would have given very similar noise
> pattern to the GX100, perhaps even better.

Sorry to disappoint, but it won't. Here's an ISO 1600 JPEG from an SD14 review posted today at ephotozine.com:

http://forums.ort.org.il/files/30/2435594/2140760.jpg

And here's a 100% crop from that shot:

http://forums.ort.org.il/files/30/2435595/2369550.jpg

Don't you find those splotches sickening? Just compare it with a similar crop from the ISO 100 shot:

http://forums.ort.org.il/files/30/2435596/9709516.jpg

The entire review can be found here:

http://www.ephotozine.com/article/Sigma-SD14-reviewed-1

There's no way around it. For anyone who shoots JPEGs, the GX100 + IS + f/2.5 lens + filmlike grain is *way* better for low light photography than a DP1 + f/4.0 + splotchy noise will ever be. In RAW it will likely be a closer match, but I don't hold my hopes too high for the DP1. This is a camera built for landscape photography, not street shooting, and definitely not street-shooting in dim light.

I think Sigma should limit their cameras to ISO 400 (like most medium format film backs) and should forget about JPEG until they develop better in-camera processing. The Foveon sensor has fabulous low-ISO quality, and this should be their main selling point. There's no point in ruining their reputation with sub-par high-ISO quality.

Prog.