ClickRepair Best Settings

That doesn't sound right at all. My setting needs to be at 12 before I start hearing the sound being affected negatively.

My guess is something isn't properly configured. I'd suggest for your friend to take screenshots of all his settings as well as make a recording of it applied with a setting of 1 and off for comparison's sake and post them here.

 

It’s not. Placebo effect.

 

That is almost impossible. ClickRepair does not affect the music at all, except for what it identifies as a click or pop. I have verified this by comparing the before and after files (subtracting one from the other). The only bits that are changed are the few right at a click or pop. All the rest of the bits are exactly the same as the unprocessed file. At a setting of 1, only very obvious clicks would be identified. If you set the threshold too high, then you can start falsely identifying transients in the music (fast guitar plucks, sharp horn blasts and such) as clicks and they will be processed just as a real click would be. I've never seen that happen on any recording unless the setting is at 10 to 15 or more. For most music, it doesn't happen until the setting is at least 20. Even then, I don't see how it could be described as veiled. It only affects the transients. 99.999% of the music is totally unaffected. The best way to check for this is to output only the repairs. You can listen to the unchanged original, the changed output, or the difference (only the repaired or changed bits). You should hear mostly silence, with random clicks. If you hear anything that resembles the beat or rhythm of the music, then you have set the threshold too high and you are changing the fast transients in the music in addition to the clicks and pops.

 

Good call. Rereading the post, that seems plausible.

I (perhaps wrongfully) assumed the sound being referred to was *quite* veiled. Wonder if Janet heard the difference or is merely acting as a messenger.

 

My blunt response assumed everything is set up “reasonably”.

If it isn’t, I suppose anything is possible - but I can’t see how it could sound “veiled”. But that could just be semantics.

 

I use iZotope RX2 and I do all mine with the track reversed, with that anything between 3 to 5 normally removes all clicks. In fact I recently did a track for two forum members and they were pleased with the results. iZotope also does crackle repair, for that the setting is normally 3.

 

If your settings are too high it will take out high frequency transient = veiled sound

 

In this case, though, we're talking about the minimum setting of 1 out of 20 (I think). The more I think of it, the more I'm inclined to believe Janet hasn't heard the A/B comparison and is relaying her friend's concerns. Therefore, the comments typed by @marcb wouldn't be out of line.

 

May well be I've only used iZotope so it was pure speculation, perhaps just a cautionary warning of what is possible if you over egg it

 

I’ve used CR hundreds and hundreds if times at various settings. I wouldn’t call it “veiled” when set too high.

IMO, “veiled” is a consistent lack of high end. CR doesn’t act in that way. It impacts transient dynamic peaks, not a particular frequency or frequency range - even at a very high setting. High frequencies still get through. If anything, I would say it can sound dulled/tempered dynamically or smeared.

Tom-AY-to, Tom-AH-to, I suppose.

 

We agree veiled. The vagaries of language, a seriously flawed system!!

 

World’s laziest ClickRepair user here. To make 24/96 needledrops I use the 5 setting for automatic declicking of fairly quiet LPs, 8 to 10 for moderately noisy records, 15 for seriously noisy vinyl, and 25 for near-disastrous albums that make a racket but don’t skip.

The key thing is not caring about perfection. What CR accomplishes automatically with low settings is stellar and if a few muted flaws remain, so what?

 

One should see thin red vertical lines where ticks and clicks are being removed. If one sees red blocks or anything larger than thin vertical lines, or excessive lines where the music does not originally sound that ticky, one should stop and listen to the original files and consider aborting mission with this software. Some music registers as full of ticks where there are none. Chicago - Live in Japan has horns with bright spiky edges. I did a 24/96 needle drop that came out wonderful. I tried CR on it and got thousands or millions or false positives. Luckily the 2LP set was in very clean and nice condition, I hand removed the few real ticks by hand (with other software) and was was done with it.

But having large blocks of red running through entire songs on a NM Japanese pressing was strange to see.

 

At what setting?

 

I've mentioned my preferred setting(s) earlier in this very short thread. 12 to 15 is my common settings and what I used on this LP set was likely 14.

 

I use 2 for most LPs. Works fine. But still, 14 shouldn’t be getting that kind of correction.

Millions of corrections? While I’ve run across the odd song that fools ClickRepair, I’ve never run across anything remotely like what you’re describing.

 

There are several (or many) albums that have certain tracks that cause false positives by the millions. Pbathal the world famous needle dropper has found a few. Synth parts of tunes can wreck havoc. Wish You Were Here has a part on Welcome to the Machine is one such part.

But yeah, when you see more than you can count, and solid red blocks, that would equal millions on an album side.

 

Assuming this is all true, what’s the point again?

 

Can't imagine needing that high a setting. 9 takes care of everything on the few LPs I own which aren't in NM condition.

Remember that the music is louder than the clicks and that the software is reading nothing but.

 

That's what I like about ClickRepair. It is not like a filter affecting the entire track. It only affects what it thinks are clicks and pops. A few posts back I showed the statistics of a typical side of an album. Out of the 21 minutes of the album side, ClickRepair changed 0.017 seconds of the music. No way that is veiled sounding.

 

Yeah when you see more than you can count, and solid red blocks, that would equal millions.

Knowing when to halt, or continue but of course.

 

I know how it works. It gives you a count of corrections.

 

So you know when to about mission, good.

 

I want to resurrect this thread instead of creating "yet another clickrepair thread".

Lately with click repair I have been having problems. I am getting quite a few leftovers here and there after processing.

Many of these clicks that I was still hearing were well buried in the waveform and can't just be selected and removed. At first I thought maybe my latest system revision has made the sound more revealing, but this issue is still present on various systems I own.

Anyways, I decided to do click repair at a bit higher setting (18-30) as it didn't seem to do too much transient damage. I kept running the file through CR until there was 0 repairs and definitely helped diminish the louder of the ticks and pops I was hearing. I then decided to listen through the file and mark anything that's left. I then would run those small sections through CR again at 40-60. That worked for a few recordings, but the issue started to come back again after a few rips.

I am not sure if the software is just not cutting it anymore, or another issue entirely. I tried izotope rx and it seems to be able to get rid of a few things CR missed, but at the same time missed stuff CR got. I tried the demo which does not allow saving, and I definitely don't want to spend 100+ dollars for something marginally better in some places, and worse in others. I tried simply running the file as high as I dared in CR (around 60) and it seems to have gotten rid of 99% of what's left but unfortunate at this point the transients have been affected to the point that I can hear the beat on the inverted output.

Anyhow, I'm not sure what else I can try and this is starting to get a bit frustrating. I have about 10-12 transfers sitting that need processing and cannot get to them. The only other thing I just thought of when writing this is static... I don't know if these are pops in the vinyl or just static buildup. I have always had a static issue though, and this issue just started more recently.

Appreciate any input on this. Willing to try things.

Also.. if anyone is having the issue with random pops being added into the recording after CR, I found this entirely went away when running at wavelet instead of x3.

 

All this writing and all that experimenting, and the time and effort. Yet you never tried the reverse method? Try reverse first at 16 or 18, then standard forward.