JThompson From Australia, joined Mar 2009, 21 posts, RR: 0 Posted (5 years 10 months 1 week 5 days ago) and read 6688 times:
I have been reading the Masterclass thread and other's on Sharpening, where a lot of people use multiple passes at the same percentage of USM, or different percentages of USM.
My question is, does everyone that uses this method Expand before each new pass, as Fergul describes in his Tutorial, if so, by how much, Fergul says by "1", is this the general rule of thumb?,......also, is this all done on the one Duplicate Layer, or are there more than one Duplicate Layer being used?
Any help on the process of this will be very much appreciated.
NicolasRubio From Argentina, joined Sep 2005, 585 posts, RR: 3
Reply 1, posted (5 years 10 months 1 week 5 days ago) and read 6675 times:
I use an action I made up myself by following a tutorial I once read. It goes something like this:
- Duplicate layer
- Select all and copy
- Create alpha channel
- Filter -> Stylize -> Find edges
- Bump up the levels to get things "blacker"
- Select with the tool that is to the left (first one) where the trash can to delete layers is
- Invert selection
- Delete the channel
- Go to the duplicated layer, apply USM to your personal taste
- Select none
- Flatten image
Hope it helps!
Gripped 7D + Sigma 10-20mm + 17-40L + 50mm f/1.8 II + 70-200mm f/4L IS + EF 400mm f/5.6L + 580EX II
LOCsta From United States of America, joined Sep 2006, 306 posts, RR: 7
Reply 2, posted (5 years 10 months 1 week 4 days 11 hours ago) and read 6619 times:
I haven't read the tutorial you are refering too, but each pass of sharpening should be done on its own layer so you can erase any parts from that specific layer that become over sharpened after the action is completed. Expanding or contracting the selection between layers is to prevent a noticable "line" where the sharpened areas meet the non sharpened areas.
RonS From United States of America, joined Feb 2009, 763 posts, RR: 21
Reply 3, posted (5 years 10 months 1 week 4 days 5 hours ago) and read 6588 times:
Yep, that's what I do. Create the Duplicate Layer, Expand by 1 pixel. Then do my Sharpening (in my case Smart Sharpen). Basically every time now I run the eraser around the edges and fuselage to erase any jaggies, even if I don't really see any. I just always do it as routing because I hate getting a jaggie rejection. In the rare case I need more sharpening, I create another Duplicate Layer and add a little more.
All opinions expressed by me are my own opinions & do not represent the opinions in any way of my employers.
Klemmi85 From Germany, joined Mar 2009, 212 posts, RR: 0
Reply 5, posted (5 years 10 months 1 week 3 days 22 hours ago) and read 6524 times:
One more thing...
If you don't want to worry about jagged edges, try this:
Open your original image,
apply highpass filter with 1-3px (depends on picture, just try)
set highpass layer to "soft light" within the blending modes.
To get a stronger result, just duplicate the highpass layer again until you are satisfied.
I do sharpening only that way as it doesn't come up with the jagged edges. Keep in mind that you can get halos at high contrast edges if your highpass pixel setting is to high. Depending on image size, anything more than 8-10px makes the result unusable.
If there are further questions, feel free to post.
Cpd From Australia, joined Jun 2008, 4889 posts, RR: 37
Reply 9, posted (5 years 10 months 5 days 22 hours ago) and read 6336 times:
If you use Nikon and have Capture NX2, try the high-pass sharpening method in Capture NX2. It works very well. See the circled part in my screenshot:
To use it go to:
Adjust > Focus > High Pass
It will appear on the adjustments palette. Set the blending mode to overlay. When you export to Photoshop, you'll still have some scope to apply a bit extra sharpening once you've resized down to 1280px wide or whatever you are using if you need to, without making many jagged edges.