Thinking about the Sigma fp

xpatUSA wrote:

Very good comparison set, Frank.

At 200%

In the red cloth, an excellent illustration of:

a) Merrill v. Bayer and Merrill v. Quattro

b) Horrible moire from the fp.

c) The opinion that a big Bayer image downsized can equal the Foveon.

The Quattro colors are quite different to the other three - surprisingly so!

Thank you Ted. I can make it more confusing. Playing around with the samples from Imaging Resource, I came across this observation:

Hope you can read the camera names. Please look at full resolution.

All modern 24MP cameras have problems with that red cloth. The Panasonic S1, the Fuji X-Pro3 (despite of the X-Trans sensor, but it is also the only APS-C in this comparison), the Nikon Z6 and the Leica SL. Interestingly, the Leica SL is clearly better in this respect. Leica Magic? But... don't ask why, I also checked older cameras, the Sony A900 and the Nikon D700. Both about 12 years old. The D700 with only 12 MP. And wow, not as good as the SD1 Merrill but way better than the modern cameras. Both have an AA filter. So, I checked out the Nikon D800 and D800E. The closest I can get to a camera in both variants, with and without an AA filter. But unfortunately with 36 MP, so not quite the same as the 24 MP. Anyway:

D800 with AA filter. Not bad.

D800E , no AA filter. Also not bad. Slightly worse??

Ok, so, it is the AA filter or better, the absence of an AA filter? Doesn't seem so.

Any ideas? It was easy to understand for me why the Merrill beats the Quattro in this respect and why both beat the Bayer sensors by a large margin. But what to say about the old Sony A900?

One thing come to my mind: can it be that this red cloth is loosing its color (the black one) over time? So, maybe, today it really has much less contrast and my whole research was in vain...? OK, I checked out the latest and highest MP camera I could find, the Fuji GFX100.

Here, I don't believe that we loose the contrast due to Bayer color aliasing and interpolation. And it seems indeed as if the red cloth lost the contrast. I mean, as good as I can judge from this picture. Still no perfect proof, we need to see the real cloth....

The pictures are from:

Sony A900 and Nikon D700 : 2008

Nikon D800, D800E : 2012

Sigma SD1 Merrill : 2012

Sigma dp2 Quattro : 2014

Leica SL : 2015

Phase One XF 100 : 2016

Panasonic S1, Fuji X-Pro3, Fuji GFX100, Nikon D780, Sigma fp : 2019/2020

What does this comparison shows: it is easy to get tricked by assumptions. In this case, that the red cloth is always the same and that a loss of contrast in the image of that red cloth is due to sensor technology.

A fail? In some sense yes, but a "fail forward".