>>2942286So here's basically the effect of using my colour separation development technique at iso 400, although it's applicable to all isos except 100 where it isn't really worth the trouble because you barely gain anything.
Top is sigma default (with incorrect white balance damn you sigma just let me choose in kelvin) bottom is my lengthly export and recombine process
As you can see, you're genuinely getting the output of each layer of the sensor. When you turn down the luminance noise reduction in SPP while in B&W, it genuinely completely removes any kind of reduction. This isn't true of the regular RGB output because there's some kind of lowlevel correction done in the colour processing which isn't surprising, but what is surprising is how terrible their reduction algorithms are. Having noise reduction on anything but 0 in SPP is stupid, it does nothing but remove the high frequency information that is the point of the Merrill in the first place.
Yeah, as I was saying using my technique brings up issues as you can see, but it completely rids you of that mustard/purple puke coloured mess you always get in the shadows of non-primary colours, instead you have colour noise - which is good, because you can cure it with any regular colour noise reduction and not end up with the Sigma puke colour that's their idea of noise reduction - and the image is also a little less sharp, which is also good because it's not pretending, just telling you what it actually saw. Finally remarkably luminance noise seems also to be reduced, you don't get that veil of banding grain you'll see on any merril output iso 400+, I suspect because of the absence of a sharpening step.
One more benefit is that you don't get the green/red banding at high iso WHATSOEVER. Somehow it's completely absent, don't know why it happens in the default if this is possible.
As for your battery concern, buy an OVF like I did and turn the back screen off, I get 200+ per battery. Have not used my spares.
