| Author | Thread |
|
|
06/14/2003 01:49:33 PM · #1 |
| I have Amount=50%, Radius=1.4 pixels and Threshold=0 levels. I normally just do it once for all my images, sometimes 2x. What is the best setup? |
|
|
|
06/14/2003 02:10:39 PM · #2 |
1.4 radius is pretty big if you're taling about contest-sized images.. I often use 0.4 to 0.5 radius values, and anywhere from 50 to 150% .. and no threshold.. (well, sometimes i'll use the threshold if the sharpening is adding alot of grain.. though i sometimes like the grain)
|
|
|
|
06/14/2003 02:20:05 PM · #3 |
Amount=varries per image, Radius=.4 pixels and Threshold=0 levels O can go up to 250% and its very subtle but does the job nice. I would think that a 1.4 radius would be too much for the most part.
|
|
|
|
06/14/2003 02:39:19 PM · #4 |
I've seen two theroies, one suggesting radii >1 and low sharpening amounts (<75%), the other suggesting radii <<1 and sharpening amounts >100%.
I personally like results from the latter more, though my typical value for radius is about 0.8, a bit higher than many use. The sharpening amount I use depends on the effect I want to convey, the amount of noise in the original image, and the softness of the original image.
If the original image is not tack sharp, a small radius will work poorly; then radii of 1.5 or higher can yield better results.
With regard to threshold, I nearly always use 3, 4 or 5. It's almost never a benefit, IMHO, to use a lower threshold, since you don't want to exaggerate noise, and transitions needing sharpening nearly always have differences greater than 4 or 5.
My 2¢
|
|
|
|
06/14/2003 03:00:12 PM · #5 |
| Unless you're shooting the same subjects and compositions every shot. Using some kind of standard or universal setting for USM is a tad foolish. What settings work great for a portrait aren't exactly going to give the greatest results for say a autumn leaves on a yard. |
|
|
|
06/14/2003 03:55:24 PM · #6 |
Some sites suggest using a multi-pass approach. One I've seen recently was: 20%, 100 rad., 0 thresh.; 20%, 20 rad., 0 thresh.; 30%, 4 rad., 0 thresh.; and finally 90%, 0.6 rad., 0 thresh. I tried this and seems to work great on some images. But as someone suggested there is no universal way to do it for all images. I wish there was. The closest thing to that would be Intellisharpen v3.0 from Fred Miranda Actions. I've never tried it but if it works as good as described it may be well worth the $12.50 US.
|
|
|
|
06/14/2003 08:38:16 PM · #7 |
Originally posted by orussell:
Some sites suggest using a multi-pass approach. One I've seen recently was: 20%, 100 rad., 0 thresh.; 20%, 20 rad., 0 thresh.; 30%, 4 rad., 0 thresh.; and finally 90%, 0.6 rad., 0 thresh. I tried this and seems to work great on some images. But as someone suggested there is no universal way to do it for all images. I wish there was. The closest thing to that would be Intellisharpen v3.0 from Fred Miranda Actions. I've never tried it but if it works as good as described it may be well worth the $12.50 US. |
Hi Owen,
Thanks for the multi-pass sharpening example. I tried this with a couple or three images, and found it definitely does a better job than I was able to do with a single pass. I found that the zero threshold setting is not mandatory, in fact it creates real problems with areas like sky & skin. I used threshold = 2 for all but the last pass, and threshold = 4 for the last pass, with very significantly improved results. I also experimented with the sharpening amount and radius on the various passes and concluded that again "it depends" on the image as to what works best.
I will very definitely remember this technique.
BTW, I wonder if the Fred Miranda action would be DPC legal? It would be great to try it out before plunking down the change, though for $12.50 it would be very little lost. Maybe I will give it a try...
|
|
|
|
06/14/2003 08:55:51 PM · #8 |
I usually use 0.6-0.8 radius, 66-88%, and a threshold of 5-7.
Use less sharpening and a higher threshold for smooth gradients/flesh tones to avoid creating a banding effect. Use higher sharpening and lower threshold if you have a lot of details and edges you need to bring out.
I often use the lower settings and apply it twice, but I check carefully after the first pass because it is very easy to over-sharpen the small images we are using here. You may have to apply higher settings or apply more passes for large print images. |
|
|
|
06/14/2003 09:02:22 PM · #9 |
OK, so curiosity got the best of me, LOL.
I sprung for the Fred Miranda action, and found that it yields slightly better results than I got with the multi-pass technique.
It looks like Fred is using a three-pass unsharp mask, but he's also got more going on than that.
My conclusion, based on what he is doing, is that the action would NOT be DPC legal.
|
|
|
|
06/14/2003 09:44:05 PM · #10 |
Originally posted by kirbic:
OK, so curiosity got the best of me, LOL.
I sprung for the Fred Miranda action, and found that it yields slightly better results than I got with the multi-pass technique.
It looks like Fred is using a three-pass unsharp mask, but he's also got more going on than that.
My conclusion, based on what he is doing, is that the action would NOT be DPC legal. |
Probably not....LOL
|
|
|
|
06/15/2003 05:33:40 PM · #11 |
Input are photo's without in camera sharpening.
640x480 sized; 150 to 300% at 0.4 0 treshold, preferably lab mode lightness channel.
6mp print files 200 to 400% at 2 4 treshold, ............
|
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 07/27/2025 02:56:33 AM EDT.
