A few shots on the streets of Hermosillo

xpatUSA wrote:

richard stone wrote:

xpatUSA wrote:

xfajardo wrote:

xpatUSA wrote:

richard stone wrote:

The older DP series cameras used (I think) a sensor not too different from the one in the SD10. Excellent color and not much edge-roll-off, just sharp three color per-pixel definition.

Yes, they had the DP13 sensor - same as the SD14/15 cameras but with different micro-lenses and a built-in UV/IR blocking filter.

In theory the sdQ on lo-res should behave somewhat similarly, but, in terms of reality I'm not so sure about that.

Nobody knows for sure (meaning no Sigma references) how they do that.

You would think that they use pixel-binning 2x2 on the top layer and pre-Quattro raw conversion but that has yet to be proven to me.

Oh my, I didn't quite understood anything

Xavier una vez más:

It all depends on what is meant by "not too different" ...

For example, the DP1s and SD10 sensors are the same size (20.7 x 13.8mm) and both have three layers.

But they have a different number of pixels, different fill factors, different micro-lenses, different firmware, different UV/IR blocking filters, different raw-to-color conversion multipliers and . . different sensor part numbers and data sheets.

Are they "not too different" or are they just "different"?

P.S. For "pixel-binning", please see: http://www.foveon.com/article.php?a=71

Ted

I know you're right on the technical aspects.

I just remember, years ago, looking at the same scenes shot with the SD10 and then the SD15, comparing the detail in each image and the result was, as I saw it at the time...not much difference. If any really. I know the pixels on the SD15 are "better," compared to the SD10, but there are not that many more of them. At the time I had been reading Norman Koren's site on a regular basis, and his wish, as expressed on that site was for a Foveon sensor with many more pixels. Meaning enough to easily see a difference.

One significant difference between my SD10 images and my SD14/DP2s is detail contrast, Richard. The smaller the pixel, the more "detail contrast".

Interestingly, SD10 edges or lattices have "clarity" but lower contrast. As a consequence, an SD10 image can appear to be soft when it is just "dull" - especially in FastStone or in RawDigger. Koren explains that with a sinc formula where the size of the pixel is a factor.

http://www.normankoren.com/Tutorials/MTF2.html

Did you ever see: https://www.dpreview.com/forums/post/59664792

Fortunately, RT can make up that difference in a flash with CBDL.

Ted

Thank you for your post and information. Yes, I checked your references, after looking into CBDL (Contrast By Detail Level) in RT. I have not used that function in RT yet. I have been using RT on some of the sdQ images. I don't think the sdQ images  need more detail, or more or less grittiness, as a general rule, but I am curious about making some sdQ images look more like Merrill images. Curious, but not desperate or determined.

I recall reading the Koren section long ago, and your earlier post more recently, but I did not see a direct use for that information in the pictures I was producing with the SD10 at the time. I tend not to crop the SD10 images much. Or examine them too much at 100-200%. What I have noticed in those SD10 images is that the printer programs seem to remove (or somehow fix) the jaggies. I also recall the great unhappiness of some Merrill users who wanted Sigma to stay on the gritty Merrill path, forever and always.

From what I could see in the CBDL-related threads, and as I recall from some initial threads regarding the Merrill sensor and cameras, the detail and sharpness of the Merrill sensors can present "grittiness" issues in portraits and skin in general. If there is such a thing in optics and cameras and images as (somehow defined) grittiness. Removing or reducing grittiness is probably desirable in portraits. For other images "grittiness" may be, or contribute to, the unique Merrill "look" that "makes" those images. I tend to see grittiness in the nature of an artifact.

I will try RT and CBDL on some of the SD10 shots. Combined with the better lenses I have now (The 30mm 1.4 is really very sharp.) it should make the SD10 more like a newer digital camera. Alternatively I could look into the results on sdQ images shot on the low resolution setting.

I would imagine RT and CBDL could be helpful on the older DP cameras too.

Richard