Sharpening: CaptureOne 9 or 3rd Party Software

[James R]

 I've used Pixel Genius for years; however, I would like to incorporate CO9's sharpening but I just don't have a feel for the tool. I have not seen any tips on using this tool or posts championing its quality.  So, any advise or tips would be appreciated. 

[Bart_van_der_Wolf]

I've used Pixel Genius for years; however, I would like to incorporate CO9's sharpening but I just don't have a feel for the tool. I have not seen any tips on using this tool or posts championing its quality.  So, any advise or tips would be appreciated.

Hi James,

Capture One is first and foremost a Raw converter. Its post-processing functionality is constantly being improved further, so Sharpening may also follow. It's not bad as it is (with sharpness fall-off correction and local sharpening adjustment layers), but it can be better if real deconvolution sharpening was used for Capture sharpening, and different techniques for Creative/Output sharpening.

What the Pixel Genius software does, is more akin to Creative sharpening and Output sharpening. It offers a functionality more aimed at postprocessing. There are other/better solutions than Capture One can (currently) offer. My personal preference is Topaz Labs Detail, which works as a Plugin that can be used with almost all Photoshop plugin aware photoeditors. But it does much more than just sharpen ...

Cheers,
Bart

[Bob Rockefeller]

My understanding is that CO "knows" different sensors and applies a default sharpening intended to replace the sharpness lost in the image capture. So it's taking into account the RGB pattern of the sensor, its AA filter (or not), and probably magic sauce, to get the "right" sharpness. If you add in lens correction for profiled lenses, it will fix the loss of sharpness from the lens, too. So classic 'capture sharpening' should be covered - if the image was sharp when it was taken, CO will recover all of that sharpness.

'Creative sharpening' seems OK enough in CO, to me. It doesn't seem to be as good as Lightroom/Photoshop at salvaging somewhat out of focus images, but I think that's asking a bit much. If it wasn't sharp when it was taken, that's mostly on the photographer. To bring a little attention to eyes, or other important image points, CO seems to cover the bases.

The 'output sharpening' end of the chain is CO's weakest point, IMHO. And it's pretty weak with just a single slider for "sharpening." But Jeff Schewe (sharpening guru and Adobe alpha/beta tester) doesn't think Lightroom's output sharpening is all it should be either - so this must be a tough area to get right.

And I'm not at all sure that printing is very high on PO's development roadmap for CO. Maybe most CO users print from dedicated RIP software?

[Bart_van_der_Wolf]

My understanding is that CO "knows" different sensors and applies a default sharpening intended to replace the sharpness lost in the image capture.

Hi Bob,

While there are defaults, they can be set to anything else (and be made a new default).

I have mine set to get a good idea of the degree of focus/DOF, but I disable sharpening (with a checkbox) in the output recipe to files. I then use Focusmagic for Capture sharpening and Topaz Detail for Creative 'sharpening' and Output sharpening.

Cheers,
Bart

[James R]

Thanks for the replies.  I've use PG for years, maybe I'll check out Topaz and Focus Magic.

[ErikKaffehr]

Hi,

In general I would agree with Bart quite a bit. Personally I take a different approach and use Lightroom for my routine processing. But, sometimes I want a processing workflow with less artefacts. In that case I would switch of sharpening in Lightroom and use a tool like FocusMagic for sharpening the image. Actually, Lightroom may not be the best converter to this as I feel it may induce some unnecessary artefacts in raw conversion.

For normal size prints, like A2, I would say Lightroom works just fine and Capture One does too, I am pretty sure!

Best regards
Erik

[James R]

Hi,

In general I would agree with Bart quite a bit. Personally I take a different approach and use Lightroom for my routine processing. But, sometimes I want a processing workflow with less artefacts. In that case I would switch of sharpening in Lightroom and use a tool like FocusMagic for sharpening the image. Actually, Lightroom may not be the best converter to this as I feel it may induce some unnecessary artefacts in raw conversion.

For normal size prints, like A2, I would say Lightroom works just fine and Capture One does too, I am pretty sure!

Best regards
Erik

For website or small prints, I'm fine with the built in sharpening.  I'm more concerned with larger prints.  I just finished looking at some plug-ins.  It seems there is an emphasis on dealing with out-of-focus images, which I tend to discard.  FocalBlade2 and Focal Magic looked interesting.  I'll probably just continue using Pixel Genius.

[Paul2660]

It was Bart that clued me into the power of Focus Magic, however with all processing from or LR, I always apply some sharpening, then work the files with either Topaz In Focus  or Focus Magic. 

I have always felt that raw conversion without some sharpening leaves a less than acceptable result, as too much detail is lost in the bayer pattern conversion as 2 out 3 colors are being guessed for each pixel.  Just what seems to work best for me.  If I try to pull back sharpening after the raw conversion I feel it's too great a delta.  Just what works for me.

However with Focus Magic I find I definitively don't need as much sharpening during the raw conversion.  Focus Magic is well worth the cost. 

For web work, I have an action that scales down the image that does a good enough job, as I feel there are too many variable out on the web anyway that can contribute to how an image is viewed that I can't control.

Paul C

[alain]

Hi James,

Capture One is first and foremost a Raw converter. Its post-processing functionality is constantly being improved further, so Sharpening may also follow. It's not bad as it is (with sharpness fall-off correction and local sharpening adjustment layers), but it can be better if real deconvolution sharpening was used for Capture sharpening, and different techniques for Creative/Output sharpening.

What the Pixel Genius software does, is more akin to Creative sharpening and Output sharpening. It offers a functionality more aimed at postprocessing. There are other/better solutions than Capture One can (currently) offer. My personal preference is Topaz Labs Detail, which works as a Plugin that can be used with almost all Photoshop plugin aware photoeditors. But it does much more than just sharpen ...

Cheers,
Bart

Bart

Would running a imagemagick batch job (after raw conversion) also be possible?
Although I have the impression that imagemagick has "problems" with tiff and non sRGB color spaces.

Alain

[Bart_van_der_Wolf]

Bart

Would running a imagemagick batch job (after raw conversion) also be possible?

Hi Alain,

Yes, one can point the Basic output recipe of C1 to "Open with" an application that takes an input filename as parameter.

Although I have the impression that imagemagick has "problems" with tiff and non sRGB color spaces.

Correct, that's also my experience. I just tried it again, and C1P V9 is no exception. There must be some incompatibility between the TIFF variant written by C1, and the TIFF format expected by ImageMagick. I always have to load and save in Photoshop before the TIFF is handled well in ImageMagick. Have to look into that one day with the ImageMagick folks.

Cheers,
Bart

[alain]

Hi Alain,

Yes, one can point the Basic output recipe of C1 to "Open with" an application that takes an input filename as parameter.

Correct, that's also my experience. I just tried it again, and C1P V9 is no exception. There must be some incompatibility between the TIFF variant written by C1, and the TIFF format expected by ImageMagick. I always have to load and save in Photoshop before the TIFF is handled well in ImageMagick. Have to look into that one day with the ImageMagick folks.

Cheers,
Bart

Thanks

[James R]

I've spent sometime looking at the options and have now decided to give FocalBlade 2 a try.  Hopefully, it isn't too much of a resource hog.  Never thought about batch sharpening, but, when appropriate, I'll give it a try. 

Thanks again to all.

[Bart_van_der_Wolf]

Hi Bob,

While there are defaults, they can be set to anything else (and be made a new default).

I have mine set to get a good idea of the degree of focus/DOF, but I disable sharpening (with a checkbox) in the output recipe to files. I then use Focusmagic for Capture sharpening and Topaz Detail for Creative 'sharpening' and Output sharpening.

To add a small but hopefully useful tip:
One can hand off output from Capture One, e.g. a TIFF, to a TopazLabs plugin directly from Capture One without the need for an additional photo-editor. This is done by specifying in the output recipe that the file must be opened with "Fusion Express 2". Here is a Post in the Topaz forums with a link to a document that has some more details.

Sending the output off to e.g. Topaz Detail, will allow to use its "Deblur" control for Deconvolution Capture sharpening, and the awesome other functionality to really make detail 'pop' (or not pop, e.g. in smooth sky) in your image.

This looks really useful for e.g. web-publishing, with the improved rescaling of C1 Pro Version 9 a near perfect downsampling is possible, sharpening can be disabled in the output recipe, and sharpening and final detail work at the actual output size can be done immediately with Topaz Detail. Sending multiple files to Fusion Express 2, will open multiple instances of FE2.

Cheers,
Bart

[brandon]

To add a small but hopefully useful tip:
One can hand off output from Capture One, e.g. a TIFF, to a TopazLabs plugin directly from Capture One without the need for an additional photo-editor. This is done by specifying in the output recipe that the file must be opened with "Fusion Express 2". Here is a Post in the Topaz forums with a link to a document that has some more details.

Sending the output off to e.g. Topaz Detail, will allow to use its "Deblur" control for Deconvolution Capture sharpening, and the awesome other functionality to really make detail 'pop' (or not pop, e.g. in smooth sky) in your image.

This looks really useful for e.g. web-publishing, with the improved rescaling of C1 Pro Version 9 a near perfect downsampling is possible, sharpening can be disabled in the output recipe, and sharpening and final detail work at the actual output size can be done immediately with Topaz Detail. Sending multiple files to Fusion Express 2, will open multiple instances of FE2.

Cheers,
Bart

Hi Bart, thanks for this. Is your preference to use Topaz detail on an output TIFF/jpeg rather than use the "edit with" opyion and have the sharpened file reimported to CO (eg catalogue)? Is this because you are using it as "output sharpening" after all other adjustments have been done? Do you do earlier external sharpening eg Focus Magic via the "edit in" route to bring you back to CO?

[Bart_van_der_Wolf]

Hi Bart, thanks for this. Is your preference to use Topaz detail on an output TIFF/jpeg rather than use the "edit with" opyion and have the sharpened file reimported to CO (eg catalogue)?

Hi Brandon,

No particular preference, other than that I rarely re-import TIFFs in C1. When I produce output I'm targeting for a specific purpose, and there is no need to re-open in C1, and when I'm still preparing for output, then I'm working on the Raw conversion data itself for as long as possible.

That's the only reason I only mentioned the output recipe.

Is this because you are using it as "output sharpening" after all other adjustments have been done?

Yes, although I'd prefer if I could do the Capture Sharpening in C1 itself. Currently I cannot judge the sharpening effect on rescaled output with a preview, so I have to postpone it till I'm done with C1 processing.

Do you do earlier external sharpening eg Focus Magic via the "edit in" route to bring you back to CO?

No, I prefer to only sharpen at the final output size (sharpening and then downsampling can cause artifacts), and when I output full size, there is no real need to do it in C1 when better quality is available outside. Creative 'sharpening' is (in my book) not real sharpening , it's more detail enhancement, and that is also better done with Detail than within C1, IMHO.

Cheers,
Bart

[Bob Rockefeller]

Less than ideal capture sharpening (getting back the original sharpness lost from the AtoD conversion and the sensor's color pixel array) and creative sharpening (as Bart said, enhancements to specific areas of the image) may be a real downside of CO for me.

Every image needs capture sharpening, by definition. And any image being output to anywhere other than the web at web sizes, especially to be printed, needs output sharpening.

All these 16-bit TIFFs, if output for sharpening by another program, are a pain from a storage size and a management point of view.

I was hoping CO's default capture sharpening, created by PO specifically for each support camera and lens combination, would do what is needed.

Creative sharpening, unless done to a great extent, shouldn't need the fancier deconvolution sharpening, I would think.

CO's output sharpener appears rather crude with only an amount slider to adjust. My hope here is that whatever it does would be good enough for smaller prints; I rarely print over 11x17. The occasional larger print would likely benefit from a dedicated sharpening tool regardless of the RAW converter/adjustment program.

It is much better for me to stay in a RAW workflow all the way to output, if at all possible. Lightroom seems to do that well. But perhaps CO can't?

[Ligament]

Bumping this topic.

Has the Capture one sharpening improved with the latest version release?

Are you guys still recommending focus magic and topaz detail as before?

I'm really hoping the sharpening capabilities of capture one have finally incorporated deconvolution.

[N80]

I can't say as I just came on board with CO at 9.2. But I'll be honest, I find CO's sharpening tools to be perfectly adequate especially since I do not print larger than 13x19. With a very typical sharpening work flow my images seem perfectly sharp and not over sharp to me.

There is certainly the possibility that I am a weak critic of my own prints. There is also the possibility that I misunderstand output sharpening, or at least not knowing what is available since I have not used PS for this purpose or other plug ins, even though I do own the Nik collection.

In CO9.2 you have one slider for ouput sharpening and it is a simple number, a percentage I believe. I print almost entirely on glossy papers and I leave it at 25 or less.

I may be exposing my ignorance here but isn't output sharpening always a black box in which the quality, or lack thereof, can only be seen in the print? In other words, no soft proof for output sharpening?

[Bart_van_der_Wolf]

Bumping this topic.

Has the Capture one sharpening improved with the latest version release?

Hi,

No differences with previous versions, although the resampling quality has dramatically improved with version 9.x.

Are you guys still recommending focus magic and topaz detail as before?

I do, can't speak for others.

I'm really hoping the sharpening capabilities of capture one have finally incorporated deconvolution.

It's the highest on my personal wishlist, but it depends also on how it's implemented.

The current sharpening is not bad, and given (aliased) non-AA filtered input (like from Medium Format cameras), there is traditionally not that much one can safely do, unless the image is diffraction blurred. But with  denser sampling sensors (smaller sensel pitch), and either AA-filtered sensors or multiple step exposures, the restoration capabilities are sorely missed.

Another thing missing, IMHO, is a sharpening preview for downsampled or upsampled output/print files. The print sharpening is (at best) experimental, but without preview, so multiple actual prints must be made to see if the (post resizing ?) sharpening is enough or too much.

For a better output preview, there should also be a more seamless zooming. Although display and actual print can have vastly different resolutions, it should be possible to simply zoom on display to the exact same size as the printed output (a zoom ratio of display_PPI / output_PPI). That will at least take some of the guesswork out of the equation.

I think that the Capture One engineers would be wise to implement such a solution soon, because there will be new contenders that might otherwise steal the show. Iridient's Raw Developer is praised for its Richardson-Lucy deconvolution and some other mixed methods, but it's a Mac OS application and not yet available for Windows 10 (although announced it will be). RawTherapee has Richardson-Lucy deconvolution that works well, and I'm looking forward to what Affinity Photo will offer when its Windows version comes out  (coming month?). OnOne is also coming with a new Raw converter, and Photo Ninja's results are described as very detailed and sharp, but that's probably more due to their sharpening than to the Raw conversion as well.

Cheers,
Bart

[Ligament]

Hi,

No differences with previous versions, although the resampling quality has dramatically improved with version 9.x.

I do, can't speak for others.

It's the highest on my personal wishlist, but it depends also on how it's implemented.

The current sharpening is not bad, and given (aliased) non-AA filtered input (like from Medium Format cameras), there is traditionally not that much one can safely do, unless the image is diffraction blurred. But with  denser sampling sensors (smaller sensel pitch), and either AA-filtered sensors or multiple step exposures, the restoration capabilities are sorely missed.

Another thing missing, IMHO, is a sharpening preview for downsampled or upsampled output/print files. The print sharpening is (at best) experimental, but without preview, so multiple actual prints must be made to see if the (post resizing ?) sharpening is enough or too much.

For a better output preview, there should also be a more seamless zooming. Although display and actual print can have vastly different resolutions, it should be possible to simply zoom on display to the exact same size as the printed output (a zoom ratio of display_PPI / output_PPI). That will at least take some of the guesswork out of the equation.

I think that the Capture One engineers would be wise to implement such a solution soon, because there will be new contenders that might otherwise steal the show. Iridient's Raw Developer is praised for its Richardson-Lucy deconvolution and some other mixed methods, but it's a Mac OS application and not yet available for Windows 10 (although announced it will be). RawTherapee has Richardson-Lucy deconvolution that works well, and I'm looking forward to what Affinity Photo will offer when its Windows version comes out  (coming month?). OnOne is also coming with a new Raw converter, and Photo Ninja's results are described as very detailed and sharp, but that's probably more due to their sharpening than to the Raw conversion as well.

Cheers,
Bart

Thank you very very much.

on a tangent, I really only shoot non-AA equipped cameras (Ricoh GR, D810, Sigma Merrill series). That said, I'm curious about you comment "non-AA filtered input (like from Medium Format cameras), there is traditionally not that much one can safely do." Do you mean this mathematically (image needs to be blurred by AA or motion or defraction) to benefit from deconvolution or any form of sharpening in general?