C1 sharp versus PS sharp?

[lowep]

For large prints do you export from C1 to PS with the default sharpening option enabled, disabled or both?

When I export 16-bit TIFFs from C1 with the sharpening option disabled then process the files in PS and add sharpening as the last step in the PS processing I am not able to get as good results in areas with detail (to my eye) as when I start off by exporting from C1 with sharpening enabled. Is this because C1 uses a different (better?) algorithm for sharpening than PS?

That said I also find that in areas with little detail (such as walls and sky) the default sharpening added by C1 during export is too much for the large prints I want to make.

So at the moment I am exporting two separate TIFFS for each image that I then import into PS as separate layers and sandwich them together - one with C1 sharpening enabled for areas with a lot of detail, and a separate one with C1 sharpening disabled for areas with little detail.

Is there not a better way to do this?  

[Bart_van_der_Wolf]

C1 sharpening is prehistoric USM and well left alone IMHO.

Hi Nick,

Interesting, how did you reach that conclusion?

Cheers,
Bart

[fredjeang]

IMO, the Capture 1 sharpening is a weak point of the averall very good software. I always disable sharpening on export in C1.

Ps does the job. I'm not in favor of a uniform sharpening on an image and generally sharpen with the brush where I'd like to enhance details and fool the eyes sensation that way.

Sharpening is very much like heating a room. You don't need to have 20 grades in all the volume, all you need is 20 grades on your feet and the all body feels 20. 

[lowep]

Thanks for the input.
 

Apart from the usefulness or not of a program such as C1 to target where the sharpening is done and other user interface controls that give the user control over how much sharpening is done, is there also significant differences in the way different programs such as C1, Photoshop, Lightroom etc apply sharpening to an image or is it all the same process dressed up in different outfits?

[Bart_van_der_Wolf]

Hi Bartvanderwolf

Detailed testing.

Hi Nick,

Really? What specifically was it then that convinced you that it was nothing different than USM?

Cheers,
Bart

[lowep]

This useful LL article explains in a way even I can understand how sharpening works though it brings me no closer to understanding the obvious differences I see in the way C1 and PS does the same task.

If C1 and PS do sharpen in the same way then I should be able to replicate the sharpening I get from exporting a C1 file with sharpening enabled by sharpening a PS file exported from C1 with sharpening disabled, right? However I have not managed to do this yet.

[lowep]

Meanwhile back at the ranch, the cautionary comments about C1 sharpening in this thread have raised my concern about my current strategy of exporting from C1 with sharpening enabled as I have been doing until now for detailed areas of my images.

I do not have a desktop printer or easy access to a lab so the obvious solution to check the effects of this on a print is not easily available where I am currently located in west africa. Instead I have tried to test the limits of C1 sharpening by reducing the resolution of my final image to 100 dpi (usually I print at 300dpi) and blowing this 100dpi image up on my monitor to 100% in PS. The result is that even at 100dpi and 100% I do not see any problems caused by the sharpening in the detailed areas of my image - I only start to see unwanted artifacts when I blow up the image to 120%.

So I am thinking that despite the cautionary comments to the contrary at least C1 sharpening is probably usable in terms of my image size requirements (printing further down the track at 200-300dpi) as it does not seem to leave visible artifacts when I reduce my image to a much lower dpi and higher image zoom than I require and "looks" sharper than the best results I can get from PS sharpening. Do you agree?

[robgo2]

Capture One has several different sharpening presets from which to choose.  The default sharpening is fairly strong, but Presharpening 1 & 2 are less so.  I usually use one of them to counteract the blurring that is introduced by the camera's AA filter.

Ro

[lowep]

Hmmm... if so, any idea how you can figure out the minimum amount of sharpening required to offset the effect of the AA filter - just trial and error or is there a more scientific way to work this out.

Does this also mean that a camera with no AA filter does not require a minimum amount of sharpening to offset the way digital capture works?

[Bart_van_der_Wolf]

Hmmm... if so, any idea how you can figure out the minimum amount of sharpening required to offset the effect of the AA filter - just trial and error or is there a more scientific way to work this out.

Take an image of a sharp edge (abrupt transition from one uniform tone to a much lighter/darker uniform tone), from a distance (say 50x focal length). How many pixels does it take in the Raw conversion to go from light to dark or vice versa? Zoom in on the image of the edge and determine how large a radius setting is required to achieve an abrupt transition, instead of a gradual one over several pixels. When you start with a large amount setting it is easier to see, after which you reduce the amount to reduce halo artifacts (overshoots at the edge transition). Do note that for different apertures (=different trade-off between residual lens aberrations and diffraction) different radii and amounts may be needed.

Does this also mean that a camera with no AA filter does not require a minimum amount of sharpening to offset the way digital capture works?

The Bayer CFA demosaicing will usually introduce some blur. Sensor arrays without an AA-filter risk having the aliasing artifacts exaggerated by sharpening. Therefore one usually uses much less capture sharpenng with such sensor arrays. 

Cheers,
Bart

[lowep]

Thanks Bart, I tried to follow your suggestion as best I could and found that CO1 5.2.1's default radius settings of 80 for my sinar dng file seemed to provide quite a good compromise between having smooth transition with no clear border between a very light and very dark area (when radius was reduced to zero) and a thick black line (when radius was increased to max). With radius set at 80 the border was clearly discernible but there was no black line between the dark and light areas.

To do this test I pumped up the amount of sharpening to max to make the black line between light and dark zones as easy to see as possible, then after fiddling with the radius I reduced the amount of sharpening until there were no visible rouge pixels in the dark area, which also ended up being the same amount of sharpening as the default setting that was 180.

So looks like in this particular case the default sharpening settings (amount 180, threshold 0.8, threshold 1.0) are close enough to nirvana for preliminary output to Tifs prior to the final adjustments to sharpening done at the end of the post adjustment process just prior to final output depending on final image size, media etc.

[Bart_van_der_Wolf]

[...]
So looks like in this particular case the default sharpening settings (amount 180, threshold 0.8, threshold 1.0) are close enough to nirvana for preliminary output to Tifs prior to the final adjustments to sharpening done at the end of the post adjustment process just prior to final output depending on final image size, media etc.

Great, sounds like a plan to me. You can alway pull back on the amount a little in case the image doesn't stand another (output) sharpening at a later stage well. The radius and threshold values seem reasonable enough.

Cheers,
Bart