9 Replies Latest reply on Jan 11, 2012 4:26 PM by Chris Cox

# Photoshop lens blur: radius < 4 = no effect?

Why allow me to specify a radius < 4 if values below that will have no effect?

Try it yourself: set up a Lens Blur with no other effects (brightness, DOF etc) and apply it to a duplicated layer. Now run difference between the two layers - all balck = no difference.

Why can't I get more subtle values below 4?

Same thing with Gaussian blur, though the threshold in that case is .25

• ###### 1. Re: Photoshop lens blur: radius < 4 = no effect?

It's been like that ever since I can remember.

I've always figured it's just one of those things.

For more subtle/smaller blurs, you could upsample the image and run the minimum sized blur on that, then re-downsample when you're done.

-Noel

• ###### 2. Re: Photoshop lens blur: radius < 4 = no effect?

Gaussian Blur works down to radius 0.1, with distinct differences at each radius.

Lens Blur should work down to radius 1, if it's not, then that's a bug.

• ###### 3. Re: Photoshop lens blur: radius < 4 = no effect?

C'mon, Chris, "if"?

It's not hard to try it yourself.  I just timed it:  Took me 24 seconds to reproduce the problem, and 3 of that was opening Photoshop.

-Noel

• ###### 4. Re: Photoshop lens blur: radius < 4 = no effect?

He who is neck deep in converting third party data files for a deadline doesn't always have 24 seconds or a spare machine to test on. :-)

• ###### 5. Re: Photoshop lens blur: radius < 4 = no effect?

Hey Chris, actually your assessment of Gaussian Blur (in 12.0, at any rate) is not correct.

Here's the proof:

1. new RGB 8-bit doc. Fill with 50% grey

3. duplicate layer

4. on new layer apply Gaussian Blur radius 0.1

5. apply Difference blend mode to layer

6. View histogram and review level 0 and level 1. Level 0 will be 100% of the distribution and all other levels are 0

both layers are identical after the gaussian blur is applied. I believe you will find this to be the case up to .24 as my OP asserts.

• ###### 6. Re: Photoshop lens blur: radius < 4 = no effect?

Radius 0.1 is the lower limit -- where the gaussian coefficients collapse to an impulse function (within the precision available).

Now, increase the radius to 0.2 and try again, then 0.3, etc.

• ###### 7. Re: Photoshop lens blur: radius < 4 = no effect?

Chris Cox wrote:

He who is neck deep in converting third party data files for a deadline doesn't always have 24 seconds or a spare machine to test on. :-)

Yet no one seemed too interested in hiring help for you...

Let's ask Tom to post a bug report over on:  http://feedback.photoshop.com/photoshop_family

-Noel

• ###### 8. Re: Photoshop lens blur: radius < 4 = no effect?

Does anyone (official) actually visit Photoshop Family? I thought it was a marketing gimmick. I'll be happy to post a suggestion or bug report in JIRA or anywhere else it makes sense to do so.

• ###### 9. Re: Photoshop lens blur: radius < 4 = no effect?

We not only visit feedback.photoshop.com, but answer questions every day.

We're using the feedback from that site to help guide our feature planning, and  bug fixes.