Sharpening software comparisons

[Coho]

I recently posted a question about sharpening software availability

and based upon the responses have posted an image which was sharpened

with 4 different software programs:

 

http://www.photo.net/photo/3584250

 

Thought you might be interested.

[peter_nelson1]

I didn't see a ton of difference between ANY of those images, but the CS2 one seemed a <B>little</B> better than the others. <P>

 

One of the problems with these comparisons is that <U>so much</U> depends on the settings. And since algorithms and UI's vary so much it is virtually impossible to say what the "equivalent" settings are for any two tools.<P>

 

In Unsharp Mask there are various tradeoffs you can make between the radius, degree of sharpening, and threshold. Some people prefer a slightly larger radius with a smaller degree of sharpening, others prefer it the other way around. Also, a lot depends on the image and the relationship between the pixel resolution and cutoff spatial frequency of the actual detail.

[ellis_vener_photography]

Are those 100% views ? That is what you should be looking at to make a valid comparison.

[rich815]

I'm wondering how a small JPEG view of these files (and not knowing what the exact operator workflow was either) helps in understanding which tool is better for sharpening? Especially for printing purposes. Sharpening for web display is not such a big issue but for printing, yes.

[sergey_oboguev]

I am not sure what this comparison is supposed to suggest.

 

Focal Blade gives you a lot of interactive control over separate sharpening of edges and smooth areas, halo suppression, sharpening suppression in shadows if grainy and highlights if artefacts appear etc., whatever may be appropriate for a particular image.

 

Comparing it to USM (even luminosity channel USM) does not make a lot of sense.

[kirk_thompson]

I believe andrew Rodney has written a couple of notes about why comparisons like this

don't really work.

 

1. Prints are the only place to make a judgment - sharpening looks different on-screen

than in a print, & looks different on CRT & LCD.

 

2. You're trying to maximize a whole system, not just choose a single sharpening product.

Cameras have different anti-aliasing screens, and need different kinds & degrees of

sharpening. Some folks do only output sharpening; others use tools (e.g. Photokit

Sharpener) that were designed on the assumption that you should do some capture

sharpening & some output sharpening.

 

3. What works on one print size often results in painfully visible artifacts on another.

 

4. Different tools for res-ing up interact with different sharpening tools & techniques.

 

IMO the point can only be that there's no one right answer, no one 'best' sharpening

technique. We're all stuck with doing our own experimenting in our own system. And we

can get good second opinions only from folks who've looked at our test prints.

[Coho]

All images were sharpened in 100% view but obviously there is a limitation to what you can do on PN. All images are sized at 500 by 500 pixels for inclusion in my posting. Yes, everyone is correct to some degree. Some of the programs have a lot more control than others. And I onlt included one image. That said, I do notice differences on the monitor (19 inch LCD, Acer AL1931 calibrated with ColorVision Spyder Pro.) While I print many of my images, I also post many on PN and I need a product that does a good job on screen. When I did this, I put all images side by side. I have two monitors and can move images between monitors in order to do comparisons. On my own monitor, I did notice differences between edge sharpness and colour nois and sharpening artifacts. Obviously, I can never know how they look on your monitors.

[digidurst]

David, thanks for sharing your results. While it IS difficult to truly judge what's what on

screen, your intentions are notable and I appreciate your efforts. Take care :)

[jtk]

I think the "unsharpened" file actually was sharpened by Nikon, by default and perhaps unavoidably. The various sharpeners do seem to have contributed.

 

Monitors and print are two different and equally worthwhile media. The monitor isn't necessarily just a casual sharing and proofing vehicle...they'll probably become more important than prints as time passes.

 

With experience one increasingly learns to interpolate from what we get from monitors to what we expect and then get in prints. Of course, if Internet viewers are using strangely profiled or poor quality monitors, all bets are off.