<
  News     DIFP   Benefits Profiles Account  


articles

Creating Panoramas
Monday May 10, 2004
by Ellen Anon

Ellen Anon illustrates the creation of Panoramas in Photoshop.

Working with Pressure Sensitive Tablets
Monday May 3, 2004
by Katrin Eismann

Katrin Eismann outlines the use of pressure sensitive tablets in image processing.

Book Review: Mastering Digital Photography and Imaging(Updated)
Monday April 26, 2004
by IR Staff

IR Staff reviews Peter K. Burian's latest book, "Mastering Digital Photography and Imaging."

Digital Workflow for RAW Processing: Part IV - Batch
Thursday April 1, 2004
by Jeff Schewe

The fourth in a five part series, Jeff demonstrates the application of settings from one image to a series of images within Photoshop.

PDF Presentations
Thursday March 11, 2004
by John Paul Caponigro

John Paul Caponigro discusses an under-used feature of Photoshop. Allowing you to use the automate function to create universally viewable PDF slideshows!

...see more articles(many more!)

forums



Software > Inside Photoshop > Rule of Thumb for Feathing a Selection

Please Login

AuthorSubject: Rule of Thumb for Feathing a Selection   [Go to Bottom of Page]
John Hollenberg
Posted on Thu Oct 23, 2003 11:49 PM

Is there a rule of thumb for how much to feather a selection to
blend it in with non-selected areas? Of course this will be somewhat subjective,
and vary with the type of transition, but I thought the Photoshop gurus might offer
a starting point.

My other question was going to be about a more detailed description of the feather radius, and how this affects the number of pixels feathering is spread across, but I found a great description on the Adobe Photoshop forum from which I will liberally quote:

[Begin quoted section]

"...I'm also interested in the Feather question. Can someone explain exactly what it does?..."

OK, Martin, here's "Exactly" what feathering does:

When feathering a selection, the selection "edge" (where the marching ants appear) becomes the 50% opacity locus. All pixels on that edge have an opacity of 50%. Since opacity and transparency are inverses of each other, you might prefer to think of the edge as the 50% transparency locus.

From the edge inward, the opacity increases toward 100% (transparency decreases toward zero). From the edge outward, the opacity decreases toward zero (transparency increases toward 100%).

The gradation of opacity at the edge is exponential (not linear). The feather radius which you set allows you to control the gradation. At a distance inward from the edge of one feather radius, the opacity has changed 63% of the way toward 100%, becoming 82% there. The opacity gets to 100% at 2.7 feather radii inward.

Outwardly, likewise, the opacity drops to 18% at one feather radius and to zero at 2.7 radii.

When there is no feathering (radius = 0), there is a discontinuous change in opacity from 100% to zero at the selection edge.

Since you're working with digitized rather than analogue variables, the opacity changes with distance are stepped rather than continuous and the opacities you read are not going to be precise---but they will be close enough for all reasonable intents.

Feathering does not "drop" any pixels. It keeps all of them. But it does change the opacity attribute of some of them, as detailed above

[End quoted section]

To my surprise, feathering a selection with a 3 pixel setting spreads the feather out over about 16 pixels, with 90% of the change spread out over about 8 pixels--much more than I had thought. I feathered a selection of a black square on a white background and zoomed in to 1600% and checked the spread of opacities--very enlightening.

--John


... this thread contains expert advice from Andrew Rodney, Katrin Eismann and Jeff Schewe.




sign-up now


Imagine being able to ask any one of TEN of the leading industry experts in the digital imaging field any question for just pennies a day...well stop imagining and sign-up now! member login

Login with an existing account.
E-mail:
Password:

Forgot your password?
CLICK HERE

calendar

May 2004
S M T W T F S
            1
2 3 4
5
6
7
8
9
10
11 12
13
14 15
16
17
18
19
20
21
22
23
24 25 26 27
28
29
30 31          
June 2004
S M T W T F S
    1 2 3 4 5
6 7 8 9 10 11 12
13
14
15
16
17
18
19
20 21 22 23 24 25 26
27 28 29 30      
 

Upcoming Events


The ClearFocus Studio
    5/13/04 at 12:00 PM
DIGITAL WORKFLOW, NOT WORKslow (A D-65 workshop led by IR expert Seth Resnick)
    5/16/04 at 9:00 AM to 5/22/04
Article - Tutorial
    5/17/04 at 5:00 PM
Photo Imaging World 2004
    5/21/04 at 9:00 AM to 5/23/04
Article - Tutorial
    5/21/04 at 5:00 PM

...see more upcoming events

imagingrevue Members receive special discounts on most events listed in calendar

  


customer.support@ imagingrevue.com
Privacy Policy
Terms of Service