Helpful Posts Helpful Posts:  0
Results 1 to 14 of 14

Thread: Wavelet sharpening?

  1. #1

    Join Date
    Nov 2011
    Location
    Brisbane, Australia
    Posts
    1,107
    Real Name
    Tony Watts

    Wavelet sharpening?

    There was a passing remark about wavelet sharpening by Donald in a recent post. Since then, I have been doing some Google research and wavelet sharpening seems to offer something worthwhile.

    Is there a general view about this? Is it better than Photoshop's smart sharpening? I think I understand how the unsharp mask works, or doesn't work depending on the circumstances, but have no idea about "smart sharpening.

    In my professional years, the theory of wavelets passed me by but when I look at the mathematics behind them now they seem to have potential in this application.

    Is it possible to get a wavelet plug-in for Photoshop that works on a Mac?

  2. #2
    IzzieK's Avatar
    Join Date
    Dec 2013
    Location
    Chesterfield, Missouri/Melbourne, Australia
    Posts
    17,827
    Real Name
    Izzie

    Re: Wavelet sharpening?

    I can't answer your question, Tony but it will be nice to know the answer to this...but is it worth having it?

  3. #3
    Black Pearl's Avatar
    Join Date
    Jan 2011
    Location
    Whitburn, Sunderland
    Posts
    2,422
    Real Name
    Robin

    Re: Wavelet sharpening?

    New one on me too.

    I suppose the real question should be what issues are you having with sharpening your images now and what have you tried to remedy it. With the experience here I'd wager someone will be able to help get your output files correctly sharpened without the need for an esoteric route.

  4. #4

    Join Date
    Jun 2013
    Location
    North West of England
    Posts
    7,178
    Real Name
    John

    Re: Wavelet sharpening?

    Looks like it's a GIMP only plug-in and from one of the Google references, seems to work in a similar way to Unsharp mask - but I'm no expert.

  5. #5

    Join Date
    Apr 2011
    Location
    Western MA, USA
    Posts
    455
    Real Name
    Tom

    Re: Wavelet sharpening?

    Quote Originally Posted by John 2 View Post
    Looks like it's a GIMP only plug-in and from one of the Google references, seems to work in a similar way to Unsharp mask - but I'm no expert.
    I got the impression that it was more like microcontrast than unsharp masking from what I read. But I may well be wrong.

  6. #6
    Moderator Donald's Avatar
    Join Date
    Feb 2009
    Location
    Glenfarg, Scotland
    Posts
    21,402
    Real Name
    Just add 'MacKenzie'

    Re: Wavelet sharpening?

    Quote Originally Posted by tclune View Post
    I got the impression that it was more like microcontrast than unsharp masking from what I read. But I may well be wrong.
    It's neither one nor the other and it's a bit of both. The difficulty is in trying to define it in relation to what folk have knowledge of and currently use; i.e. USM, Local Contrast Enhancment (LCE). It's 'different'. But as I wrote elsewhere, I do all output sharpening with it. Before I started using DxO to process RAW files and the NIK suite of software to make my images, I used the Wavelet Sharpening tool for all phases of sharpening (Capture, Creative & Output) ... at different settings of course.

    It is a GIMP plug-in.

  7. #7
    dje's Avatar
    Join Date
    May 2011
    Location
    Brisbane Australia
    Posts
    4,636
    Real Name
    Dave Ellis

    Re: Wavelet sharpening?

    I've never come across this either so it's interesting to hear about it. It almost makes me want to play with Gimp !

    The other sharpening tool often talked about is de-convolution sharpening. Some say it's used in Smart Sharpen and ACR/Lightroom sharpening when the Detail slider is set towards the top end but it's difficult to find much definitive information on this from Adobe. It is availble in RawTherapee.

    Dave

  8. #8

    Re: Wavelet sharpening?

    Focus Magic is a deconvolver and motion blur remover. It even has a forensics mode. Reasonable price, easy to use. There's another app somewhat like it I believe, can't remember the name, but Google is your friend.

  9. #9

    Join Date
    Apr 2011
    Location
    Western MA, USA
    Posts
    455
    Real Name
    Tom

    Re: Wavelet sharpening?

    Quote Originally Posted by Richard Lundberg View Post
    Focus Magic is a deconvolver and motion blur remover. It even has a forensics mode. Reasonable price, easy to use. There's another app somewhat like it I believe, can't remember the name, but Google is your friend.
    Topaz InFocus is another deconvolution sharpener. I have both, but find Focus Magic massively easier to use. OTOH, Topaz is considerably more flexible if you can ever figure out how to control it. FWIW

  10. #10

    Re: Wavelet sharpening?

    Blurity is a deconvolver also. Its sample images show considerable ringing, possibly due to their FFT process. it costs twice as much as Focus Magic. I agree about Focus Magic ease of use. I have InFocus but have only used it once or twice. Images Plus is another app, and there is something called Sharpen Smarter or whatever that was mentioned in another forum that I can't find right away.

  11. #11

    Join Date
    Nov 2011
    Location
    Brisbane, Australia
    Posts
    1,107
    Real Name
    Tony Watts

    Re: Wavelet sharpening?

    Richard, can you explain what you mean by ringing?

    Deconvolution sharpening is something else I had not heard of before either. With the help of our friend Google, I understand the basic idea of it but I would have thought there are traps in its application. The blurring of the image due to out of focus lens, Bayer filter, etc, is somewhat analogous to a diffusion process and reversing that is like solving a diffusion equation backwards in time, which is unstable and not guaranteed a solution.

  12. #12

    Join Date
    Nov 2009
    Location
    Provence, France
    Posts
    990
    Real Name
    Remco

    Re: Wavelet sharpening?

    Ringing is an artifact after Fourier transformation caused by sharp edges in the original data. In short: Fourier transformation tries to describe the data as the sum of sine and cosine functions. Sharp edges cannot be described this way with a finite number of such functions, so we get artifacts.

    See http://en.wikipedia.org/wiki/Ringing_artifacts for a more in-depth description.

    One disadvantage of deconvolution sharpening (and I guess with wavelet sharpening) is the amount of computation involved, which makes those techniques slow relative to USM. But they can (at least in theory) increase the amount of detail in the image, which USM cannot do.

    Images can look a bit soft after deconvolution sharpening and similar, as those techniques do not increase the edge contrast (contrary to USM).

  13. #13

    Join Date
    Nov 2011
    Location
    Brisbane, Australia
    Posts
    1,107
    Real Name
    Tony Watts

    Re: Wavelet sharpening?

    Thanks for that, Remco. I understand about Fourier series/transforms - I just did not know this terminology.

    When you look at the effect of USM sharpening in detail, i.e. by pixel peeping, you see that for with small radius there is something like genuine sharpening but with larger radiuses, there are artefacts that give an illusion of sharpening but do not produce what I would call genuine sharpening.

  14. #14
    dje's Avatar
    Join Date
    May 2011
    Location
    Brisbane Australia
    Posts
    4,636
    Real Name
    Dave Ellis

    Re: Wavelet sharpening?

    Hi Tony

    I've been fiddling with the Slanted Edge method of measuring sharpness in recent times. These plots from a test image may be of interest. They show the effects of various levels of sharpening in ACR. The labels such as ACR 100/0.5/25/20 refer to the ACR settings Amount, Radius, Detail and Masking respectively. This is not Unsharp Mask but the results give a feel for what sharpening actually does to an edge. You can see how increasing Amount and Radius increase overshoot and ringing.

    I used to sharpen a lot of my landscapes with a radius of 4 but over time I've come to the view that radii around the 0.7 to 1 are better to get cleaner sharpening. Higher radii can give more "pop" but this is doing more than just sharpening edges and, as you say, can give some undesirable effects.

    Dave



    Wavelet sharpening?

    Wavelet sharpening?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •