Capture One Testing (2/??)

Published: 12/23/2019

Update 01/28/2020: Capture One returned my bug report and have stated they have fixed this issue. However, my 30 day trial expired so I can't confirm.

In my last blog post comparing two RAW converters: Nikon's Capture NX-D (NX-D) and Capture One 20 (C1), I was really impressed by Capture One's ability to recover highlight detail. Some general thoughts in playing with two RAW converters since then:

1) The layout of Capture One is very nice and gives you a large view of the images. NX-D in comparison is a bit cluttered.

2) I don't really like the cataloging system in comparison to the (extremely) barebones Capture NX-D.

3) I go back and forth between the colors -- my thoughts (which I'll probably explore soon) are that I like NX-D, but NX-D is adding more contrast (or C1 is taking some away...). Not entirely sure of my thoughts here yet.

4) I think I like the quality of the noise better in NX-D, and not surprisingly then, I think I like the quality of the detail more in the NX-D shots. I'll briefly look at that here, and also touch on an issue I'm finding with C1.

Similar to the last post, I'll look at a Sacred kingfisher, or kōtare, shot from New Zealand, taken in reasonably low light at 700mm f/5.6 1/400 and ISO 1,600. This isn't an especially high ISO, but given the generally dark setting, noise will be visible and therefore will have to be handled in post production, i.e., I will have to do some noise reduction in the RAW converter.

Getting somewhat at point 3) above, the colors, contrast, and WB are different between converters when the file is loaded. To just compare the actual quality of the noise and noise reduction (NR) in the file, I've chosen to keep all sliders except sharpening and noise reduction at zero. This isn't 100% fair as contrast will impact noise and detail, but for this purpose it's fine.

In each converter I adjusted the sharpening and luminance and color noise to optimize detail while removing noise. I've talked about this trade off before in this blog post. Below, I've shown a 100% crop (e.g., 1 pixel in the original capture is 1 pixel on the image) with optimized sharpening and NR. You can click the image to see it larger.

Right away there are some clear differences here I think. The first obvious one is touched on by my point 3) -- the C1 file is definitely lighter. I'm not entirely sure if I think that's a good thing. The more subtle points are of course NR and sharpening. Detail is in my opinion pretty much equal between the two. I could probably back off the sharpening a tad in C1, but I think it's pretty even. The bigger difference to me is the noise. To my eyes, C1 has "finer" looking noise. Not necessarily "grainier", but almost, especially in darker areas like the bill and background. In contrast, I like the quality of the NX-D file more. It just looks cleaner and smoother without killing detail. In a weird way, the NX-D noise seems a bit more "confident" if that makes sense. This is very unscientific and some folks may like the C1 file more, or would have optimized the files to look different from the start even.

I think if you didn't the RAW file, and I showed you the processed image with additional noise reduction and sharpening, you would probably be happy with either.

However...

I've run across an issue with exporting 16 bit tif files in C1 where certain parts of the image, always in a square, will be significantly noisier than the rest. This isn't in the original file, but shows up in the export. I've tried a reinstall a couple times and have reached out to the C1 team, but it makes processing an optimal file pretty much impossible as is. You can see an example of these "noise squares" below.

Social Media

Capture One Testing (2/??)

Update 01/28/2020: Capture One returned my bug report and have stated they have fixed this issue. However, my 30 day trial expired so I can't confirm.