Top vs bottom comparison....

Not trying to start a holy war just looking to better understand some differences between 1:1:1 and 1:1:4 sensors.

For those that don't have it download RawDigger. (25% off right now) It will let you view X3F files layer by layer and export them as TIFFs if desired. I have adjusted exposure of the bottom layer (red) tiff to roughly match the top layer tiff. Can not match exactly as they are not the same data. I did not do anything else.

It is interesting to compare Merrill files and Quattro files top vs. bottom layer to see how the two compare and the challenges and advantages both sensor types have.

Do not compare Merrill vs. Quattro images.

The point is to compare the differences between top and bottom layer of Merrill and top and bottom layer of Quattro.

*Disclaimer* I fully realize that this is just a comparison with flaws in it. Ideally these would be done with the exact same scene, through the same lens and with the same exposure. I don't have a SD1,(or any 1:1:1 camera) so I can't do that.

First up SD3M.

Botton on left, Top on right. Bottom layer boosted 1.65 stops.

Bottom layer is lacking in contrast/DR compared to top layer.

Acuity is somewhat lower on bottom layer, lower DR/contrast

Acuity is somewhat lower on bottom layer, lower DR/contrast.

Noise is evident in shadows on bottom layer.

Now the Quattro..... Same subject... years later...

Note: Subject is smaller in this example, will be more zoomed in due to increased resolution of files and the lens is totally different. (10-20mm)

Reminder: Do not compare Merrill vs. Quattro images.

The point is to compare the differences between top and bottom layer of Merrill and top and bottom layer of Quattro.

If you want to compare Merrill and Quattro directly use the same scene to do that.

Top on left, bottom layer on right. RawDigger interpolated the bottom layer. Bottom layer boosted 0.66 stops. (1 stop less needed compared to Merrill)

In this image DR/contrast seems more similar between top and bottom layer compared to Merrill. If anything the bottom layer looks like it has more contrast.

Bottom layer actually seems to have more DR/contrast.

Starting to see interpolation errors in weave of hat.

Interpolation clearly seen at 4:1

Looking around the Quattro image there is not an obvious place where noise is more visible in bottom vs. top.

So, in a nutshell... Merrill bottom layer has less acuity than top layer, more noise (due to being boosted 1.65 stops) and less DR/CR. (Blows out red easier) You can zoom into bottom layer more (>1:1) and the layers do not show interpolation errors.

Quattro bottom layer does not show a large difference in noise compared to top layer (is not boosted as much), has comparable DR/CR (less likely to not blow out red.... maybe also why colors seem more accurate overall) but lower layers shows interpolation errors when viewed very large in isolation. (Remember, in this case the interpolation is done by RawDigger. Might be interesting to do the same thing with a Sigma generated DNG to see if their interpolation is different but then that adds a load of processing to the image)

Frankly: The amount of 'black magic' processing that goes into interpreting either sensors images is impressive. Merrill has to deal with decreasing DR/CR (tonal range) and increasing noise with each layer which means this is far more complex then just simply blending 'R,G,B.'

Quattro has to deal with less color resolution on the lower layers, but with what appears to be more similar DR/CR (tonal range) and noise levels on the layers.

Shawn (Putting on flame resistant gear right now....)