Individual clicks/pops highlighted in whole wave view

Pc Pete, thanks for your review of Rip Vinyl.

In case you're interested (and I think you are), have a look at AFDeClick, http://www.andreas-flucke.homepage.t-online.de/index_eng.html. It's a stand-alone program, doing just one thing: declicking. No settings, no fiddling around. It can handle 24 bit/32 bit float files. It's part of my favorite declick setup: declick, decrackle with Younglove script and, if required, start AFDeClick to finish declicking/decrackling part of the process. I find the results very impressive indeed.

How the s/w works is described on the site; it has a different approach than, e.g., AA and ClickFix. It has a drawback: it is slow. Fortunately, it has a batch function, so you can leave the program to work at night.

I made a trial test of AFDeClick, to see if it indeed would process without removing much material that is not a click or pop. All automated declickers remove much more than is needed; AFDeClick's author claimes otherwise for his program.

My test material did not seem particularly difficult, or in any way special. The program removed energy from a lot of transients that were just part of the music, the same as every other declicker I've ever tried. Whether or not it did more or less than other programs seemes to depend on the settings of those other programs; more aggressive settings of course remove more "extraneous' material, while less aggressive settings remove less. The program might do a good enough job, but I failed to find any relative advantage to it.

AndyH, thanks for your review. I use AFDeClick only if the previous steps (click removal, crackle removal with Younglove script) don't give results I expect; I don't use it as a single decklicker/decrackler; perhaps that might explain the difference in our experience.

Apart from AA with ClickFix and Younglove, and quit often AFDeClick, I haven't yet found other declickers which can satisfy my needs.

Just in case there is any question, my tests were strictly objective. I did not attempt to evaluate the sound of things after AFDeClick processing.

The program removed energy from a lot of transients that were just part of the music

What would be, in your opinion, the best way to check this? How did you do the objective tests (comparing spectral frequency views, inverse mix past)?

An inverse mix-paste will provide a file of everything changed by the declicking, minus the original waveform. When compared to the unprocessed file, you can see what the changed bits were. In Spectral View, if you have enough familiarity with using it, you can easily tell -- most of the time -- whether or not a change was necessary at any given spot without actually listening to it. You will find that any automated declicking removes more invalid than valid material.

Once you have the difference file from the mix-paste, there are two ways to do the comparison. Both must be done close up if you wish to have much idea about what is on screen.

When Synchronize Cursor Across Windows is checked, you can open both files at once and switch between them (via Window or Waveform List). You will be at exactly the same place in each, so you can see what the bit in the difference file corresponded to in the unprocessed file.

An easier way to view, but more effort to set up, is to open both (waveform view, not spectral view, is fastest for this step). Select the right channel of one and copy. Switch to the other. Select the left channel, clear it, then paste. Now change to Spectral View (or zoom in somewhere, then make the change).

Now you are looking at only one channel (the right in this example) at a time but you have the same channel from both files on one screen, perfectly aligned. As before, you will see that many of the changes from the difference file were not clicks but were musical transients that would have been better off left alone. Although AFDeClick claimed not to effect these, it did remove many in my tests.

Andy, thanks for sharing.