For part #1 click here.
This is part #2 of a series of posts regarding RAW image converters and their performance.
What’s being tested?
The converters being tested are:
The test image:
This part of the test looks at the converters’ ability to provide basic image corrections, in particular initial sharpening of the image. Smart RAW developers sharpen the image in stages – once during RAW processing, once after enhancing the image and once after sizing the image for print/display. The final step depends on the output medium – printed images tend to need more sharpening than ones displayed on a screen.
All the RAW converters being tested only allow for one sharpening step, with the exception of DXO, which has a specific camera/lens auto-sharpening step as well as ‘regular’ unsharp mask sharpening.
Personally, I don’t like unsharp mask sharpening, as it is too global in nature and lacks subtlety. Other photographers get on well with it. In any case, this stage of the Shootout does the following:
- DXO – default camera/lens auto sharpening and auto lens corrections
- Photolemur – it does what it does. There is no control over it. At this point we’ll see the final Photolemur image as there is no way to ask it to only do the sharpening.
- Lightroom – its default sharpening and lens corrections.
- Exposure X3 – its default sharpening and lens corrections.
- Luminar – its default sharpening and lens corrections.
- Smart Photo Editor – This tool provides a number of user-define sharpening presets. I used Subtle Sharpen as we don’t want to overdo things at this stage.
The inclusion of these processing steps affected the time taken to process the images. Development times were:
Lightroom is still the fastest, but what about the results?
There’s no doubt about it, when the images are compared below, that DXO has produced the sharpest results. This isn’t surprising as DXO has the smartest algorithm of the RAW converters – it sharpens the image according to its knowledge of the camera/lens combination and selectively sharpens areas of the image according to its needs. Unsharp mask sharpening is global by comparison.
The important thing to note is that none of the converters have over sharpened the image. At this stage in the workflow, it’s better to under sharpen the image than to overdo it. More sharpening can, and will, be applied later.
Photolemur’s final result is shown here – it doesn’t allow for any tweaking of what it does. The final image looks pretty decent to me. Interestingly, Photolemur can (as can the other tools tested here) process regular images such as TIFF and JPEGs and it’s instructive to see what it does with the DXO output image, for instance. But more on that later.
In the following table, you can see full size sections of the image and you can compare the results.
Results – the images
You can view larger sections of the results below. Click the arrows top right for a full screen view.
The image is showing a fair bit of noise, especially in the water. So the next test will be to see how the RAW converters handle it whist still keeping the image sharp and detailed…