Click Repair questions - vinyl

[markshan]

I don't know about CR.
I can say that iZotopeRZ works by isolating the click artifact and then nulling it out. If you use the 'output clicks' function and zoom in on the results, you'll see. So it's not removing an entire slice of time which would then have to be replaced with something. It's isolating JUST the click artifact (spectral-ly), polarity reversing it, and mixing it back in to remove it from the source.

Older click repair tools may be doing something like slice and replace. I'm pretty sure that's what the old real time units did. It turns out that a lot of vinyl click and pop artifacts are so much more dynamic/transient than the music that even the older cruder methods can yield good results. You can really get artifact/loss free results with iZotope though.

Huh? If it were simply nulling it you would have silent spots instead of click. Any software which is removing a click is creating something to put in it's place.

CR also has the output clicks monitoring function.

 

[atrocity]

If you read the manual, you will find that CR declicks first then sums second. That is critical.

Most of what I digitize lately is 78s (sooner or later I'll get them all!) and I've semi-recently started using ClickRepair in two steps: First I do a DeClick pass at 50 (yes, it's high, but these are seriously noisy records) without summing to mono. When monitoring in real time it can sometimes sound alarming--stuff is getting removed that shouldn't be and it can sound genuinely bad when listening in "stereo". But then I do a second pass with DeClick off, DeCrackle set at 50 and Stereo>Mono checked. The "holes" that the first step created largely seem to be filled during the summing phase.

One of my very, very few complaints with ClickRepair is that summing to mono slightly breaks the "In/Out/Noise" monitoring functionality: If you select "In" to hear the unprocessed input, it's clearly already in mono. I would expect that to be the unaltered "stereo" signal with mono being heard when "Out" is selected.

 

[jimfisheye]

Huh? If it were simply nulling it you would have silent spots instead of click. Any software which is removing a click is creating something to put in it's place.

CR also has the output clicks monitoring function.

Nope. Try it and take a look yourself if you're able and willing.
IzotopeRX uses spectral technique to isolate ONLY the click artifact from the rest of the program at that spot. It does NOT crudely erase the whole slice of audio. Really truly and I'm not being dishonest here! IzotopeRX really does this! And it gets much more accurate precise results than you can get with the pencil tool. (There's a reason that pencil tool technique is kind of depreciated nowadays.)
There have been some big advances in some of this software especially in the last 10 years.

 

[markshan]

IzotopeRX uses spectral technique to isolate ONLY the click artifact from the rest of the program at that spot. It does NOT crudely erase the whole slice of audio.

Jim, you aren't making a bit of sense. That click exists in time. Even if it is 1/100th of a second, it has length. And if (as you say) iZotope "nulls" that part, that is 1/100th of a second of silence. It doesn't. It creates 1/100th of a second of alternate material to fill that void. Never mind arguing about this program or that one (I've used at least a half dozen including all of the ones discussed on this thread), the simple fact is that clicks have length. If they didn't, you wouldn't be able to hear them. Even if the click in spectral shows only info from 10k up (for example), the software is still generating alternate info in that range. Otherwise the spectrals would look like low rate MP3s with sections being shaved off the top.

 

[markshan]

Most of what I digitize lately is 78s (sooner or later I'll get them all!) and I've semi-recently started using ClickRepair in two steps: First I do a DeClick pass at 50 (yes, it's high, but these are seriously noisy records) without summing to mono. When monitoring in real time it can sometimes sound alarming--stuff is getting removed that shouldn't be and it can sound genuinely bad when listening in "stereo". But then I do a second pass with DeClick off, DeCrackle set at 50 and Stereo>Mono checked. The "holes" that the first step created largely seem to be filled during the summing phase.

One of my very, very few complaints with ClickRepair is that summing to mono slightly breaks the "In/Out/Noise" monitoring functionality: If you select "In" to hear the unprocessed input, it's clearly already in mono. I would expect that to be the unaltered "stereo" signal with mono being heard when "Out" is selected.

I trust you know to never clean 78s with anything alcohol based.

 

[atrocity]

I trust you know to never clean 78s with anything alcohol based.

Yes. I keep my Nitty Gritty full of alcohol-base LP cleaner and keep alcohol-free 78 cleaner in a spray bottle.

Though if I remember correctly, Edison Diamond Discs are an exception to that rule, not that you could EVER fit one on a Nitty Gritty!

 

[markshan]

Yes. I keep my Nitty Gritty full of alcohol-base LP cleaner and keep alcohol-free 78 cleaner in a spray bottle.

Though if I remember correctly, Edison Diamond Discs are an exception to that rule, not that you could EVER fit one on a Nitty Gritty!

Correct because they aren't made of shellac. Some later 10" 78s are vinyl and those are fine too, but it's just so much easier to just say "78s" instead of putting out qualifiers which could be misunderstood and lead to destruction. Have you ever seen alcohol on shellac? I tried it once on a beater just out of curiosity. It literally melts it.

 

[atrocity]

Have you ever seen alcohol on shellac? I tried it once on a beater just out of curiosity. It literally melts it.

No, though I'm morbidly curious about it. As fate would have it, my only truly unplayable, sacrificeable (?) 78s are all other-than-shellac. I'm not sure if they're really vinyl or styrene or some other oddball substance.

Wandering further off-topic, it's weird how many records were pressed in the 1950s that were intended to be played at 78 but with a microgroove stylus. I mean, not a LOT of them as far as I can tell, but I have at least two cheapo 10" ones (both damaged beyond playability) as well as a couple of 7" Bells. It's only Discogs notes and my iffy eyes that say the 10" ones are MG, but the Bells actually have a "For best results..." note on them suggesting the smaller stylus.

What on earth was the intended audience for something like that? *Especially* in the cheap realm?

 

[markshan]

No, though I'm morbidly curious about it. As fate would have it, my only truly unplayable, sacrificeable (?) 78s are all other-than-shellac. I'm not sure if they're really vinyl or styrene or some other oddball substance.

Wandering further off-topic, it's weird how many records were pressed in the 1950s that were intended to be played at 78 but with a microgroove stylus. I mean, not a LOT of them as far as I can tell, but I have at least two cheapo 10" ones (both damaged beyond playability) as well as a couple of 7" Bells. It's only Discogs notes and my iffy eyes that say the 10" ones are MG, but the Bells actually have a "For best results..." note on them suggesting the smaller stylus.

What on earth was the intended audience for something like that? *Especially* in the cheap realm?

Yeah, lots of weird things during periods of transition.

 

[thevrp]

You didn't get many useful replies to your original post. A bunch of suggestions on other software, and then no tips on settings in CR. But glad you did get the Andy Williams LP sounding rather good and clean in the end.

I'd like to give you some setting tips that I have found useful after nearly 8 years and thousands of LPs transferred to digital but in 16/44.1 and the 24/96kHz digitally captured vinyl.

Do a first pass in reverse (check box reverse)
Setting at level 14 (out of 1 to 100)

then listen to the treated tracks and if you had a particularly bad tick and click ridden LP, go for another pass in forward (reverse box unchecked).

14 I feel is a good and rather low setting and unlikely to do any damage to the fidelity at all. I can tell it's low because sometimes a tick or click can have made it past this low setting.

No crackle repair checked, and normal settings, "simple" checked.

There are two types of recordings that flag false positive ticks and clicks, these should not be treated with CR, or maybe but only in the worst circumstances. LPs with loud horns like Chicago, TJB, and LPs with synth like Kraftwerk or parts of Pink Floyd like "Welcome to the Machine" generally cause CR to see and repair tons of ticks that are not there. My SQ copy of WYWH really freaked out CR on that track.

Feel free to pm me if you would like to ask about other tips or settings.

Someone emailed the developer with questions about the software, Brian Davies emailed back that he is ill with no time to work further on it, and does not expect to recover. Very sad news, but he leaves a legendary piece of software for vinyl transfer process. So much better than others that I tried.

Wonderful product. I too have used it for years and found it to be the best in terms of non-intrusive. Thank you for sharing your method too. I settled on a setting of 20 but never tried the reverse setting. I will give that a go at some point.

Also, I wonder if you have used Brian's DeNoise product? I have not and am wondering how it compares to Izotope's RX 7 Advanced version. I like Izotope's version. Is there another thread that already addresses this? If so, apologies.

Finally, it is sad to hear about Mr. Davies. It looks like the old website is now defunct but another new website might be in the works, perhaps?

 

[quicksrt]

Wonderful product. I too have used it for years and found it to be the best in terms of non-intrusive. Thank you for sharing your method too. I settled on a setting of 20 but never tried the reverse setting. I will give that a go at some point.

Also, I wonder if you have used Brian's DeNoise product? I have not and am wondering how it compares to Izotope's RX 7 Advanced version. I like Izotope's version. Is there another thread that already addresses this? If so, apologies.

Finally, it is sad to hear about Mr. Davies. It looks like the old website is now defunct but another new website might be in the works, perhaps?

Thank for commenting.

I moved my setting up to 16 as a standard. I try and see if I can get away with no second pass in forward. But usually, I do get more cleanliness with that second pass.

If one if going to do only a single pass then reverse would be able to get and catch more ticks, as a reverse reading of a file makes pop and tick noises stand out against everything more.

I've not used any other pop or tick removal sw. I do most of my editing in GoldWave. All crossfades are done in Vegas which is a video editor that can handle audio just fine. Crossfades imo need to be looked at visually, and adjusted in real-time and played back in real-time without having to render. So a good video editor is more accurate for that purpose. Like a live album with applause at end of a side and applause at beginning of next side. Sometimes it blends perfectly with a good careful crossfade at that point. King Biscuit and BBC shows need the commercials removed at those points as well. It is nice when one can do the edit and have it appear there never were any commercials there.

 

[thevrp]

Thank for commenting.

I moved my setting up to 16 as a standard. I try and see if I can get away with no second pass in forward. But usually, I do get more cleanliness with that second pass.

If one if going to do only a single pass then reverse would be able to get and catch more ticks, as a reverse reading of a file makes pop and tick noises stand out against everything more.

I've not used any other pop or tick removal sw. I do most of my editing in GoldWave. All crossfades are done in Vegas which is a video editor that can handle audio just fine. Crossfades imo need to be looked at visually, and adjusted in real-time and played back in real-time without having to render. So a good video editor is more accurate for that purpose. Like a live album with applause at end of a side and applause at beginning of next side. Sometimes it blends perfectly with a good careful crossfade at that point. King Biscuit and BBC shows need the commercials removed at those points as well. It is nice when one can do the edit and have it appear there never were any commercials there.

Yes, agreed and thanks for the reply.

But I just take a very minimalistic, delicate approach to all of it, as it sounds like you do as well, and it pays big dividends in the end. But there is no substitute for working with the best possible audio vinyl source to begin with. You can get so much better results starting with a very clean, low-noise floor vinyl copy and working from there than you can by starting with a copy that is in some sort of degraded state.

 

[atrocity]

Also, I wonder if you have used Brian's DeNoise product? I have not and am wondering how it compares to Izotope's RX 7 Advanced version.

I can't make a comparison as I have not tried anything else, but DeNoise can be useful sometimes. I've found that I'm much more aware of it producing artifacts than I am with ClickRepair, though with a few really worn 78s it's done a pretty good job. More often than not I find the artifacts worse than the noise, but A) That's just me and B) Maybe I just need to learn it better.

DeNoiseLF, on the other hand, is pure gold if you just need to get rid of hum and/or rumble. I've found that I don't hear hum in normal situations, but DeNoiseLF shows that it's often really there.

 

[MidiMagic]

I use software that lets me enlarge and see the waveforms in the copy I made on the PC.

The tallest peaks are usually the clicks.

I enlarge each click that is annoying. Then I low pass just that tiny piece of sound (both channels).

 

[quicksrt]

But I just take a very minimalistic, delicate approach to all of it, as it sounds like you do as well, and it pays big dividends in the end. But there is no substitute for working with the best possible audio vinyl source to begin with. You can get so much better results starting with a very clean, low-noise floor vinyl copy and working from there than you can by starting with a copy that is in some sort of degraded state.

Yes I am pretty minimalist in my approach to digital capture of vinyl. In fact I do not use any digital EQ or any digital treatments other than CR. I just bought an analog EQ, the Loki+ model which is getting notices as one of the most transparent EQs out there. So it I think an LP needs some bass or reduced treble I'd be more inclined to do the touch up in analog before digital capture.

And yes, having really good vinyl to work with is helpful. But as long as the grooves are not worn out, and the inner grooves do not have distortion, and ticks and clicks are the only real issue with the LP, then I'm fine with moving forward with the ticky LP as long as everything else is fine with it.

I was doing an album tonight to 24/96kHz capture. It is a non-Beatles Apple LP. Not sure if you follow that group or their catalog of album on their Apple label but it's an interesting story. For 25 years the LPs were almost all OOP. Maybe one Badfinger LP got a 1978 reissue in Aust or Germany and that was it. 1993 came a big reissue campaign of the Apple albums. At that time a few got UK issues only, and some Apple albums did not get a CD reissue at all in 1993, but most that got a CD were US, UK or EU, and Japan pressings. Then in 2005 another Apple reissue campaign. Some missing albums on CD that were skipped over 1993. And some that were issued in '93 did not get another remastered reissue.

This leaves me with the option of finding a '93 CD of the two I need (there are no 2005 CDs available of these), or a vinyl LP transfer I can do myself. So tonight I broke out a couple of my obscure Apple LPs and did the digital drops of them. The sound quality was good but not great, CR removed all ticks, and required both for and rev passes to get it scrubbed clean.

Now I am tempted to go for the CD issues of both these two albums because the CDs were issued in 1993, and '95, at the very end of the '93 campaign. And they were very small runs on these two titles. And they did not get reissued again when the 2005 reissues came along. And a copy of each album on CD is (currently) priced just under $10. I think that is low for a strong cult collectible that is unlikely to see another CD issue ever. And my digital capture at 24/96 is not mind-blowing enough in fidelity. I need to know how much better this can sound. One more point I should make is that the 2005 reissues of the Apple albums improved upon the '93 transfers when the given album was issued twice on CD.

So anyway, it comes down to 1) Does the LP have a CD issue at all, 2) does the LP sound better than the CD issue, 4) is my LP copy a good enough to use, and 4) is the music actually something I am going to want to hear again, or will want on my music server?

I think I will spring for the $9.00 CD copies and compare them to my vinyl.... figuring that I can always sell off those CDs for perhaps x2 or x3 what I paid. I pretty much analyze each title case by case. I have to do this as I've spent a lot of money already, have way to much music as it is.

 

[quicksrt]

I use software that lets me enlarge and see the waveforms in the copy I made on the PC.

The tallest peaks are usually the clicks.

I enlarge each click that is annoying. Then I low pass just that tiny piece of sound (both channels).

Sounds reasonable.... until you have to work on thousands of tiny ticks. I would not be satisfied with just the biggies being removed. And I can't deal with it manually unless it is just a couple at the begging or each side.

 

[quicksrt]

I can't make a comparison as I have not tried anything else, but DeNoise can be useful sometimes. I've found that I'm much more aware of it producing artifacts than I am with ClickRepair, though with a few really worn 78s it's done a pretty good job. More often than not I find the artifacts worse than the noise, but A) That's just me and B) Maybe I just need to learn it better.

DeNoiseLF, on the other hand, is pure gold if you just need to get rid of hum and/or rumble. I've found that I don't hear hum in normal situations, but DeNoiseLF shows that it's often really there.

For hum I would check cables and phono, pre, and amp connections, and not want that filtered out. And rumble is not an issue with a good tables and nicer records. Of course 78s are a different story, I can see how they would require some additional work over LPs.

 

[kfbkfb]

CD-4 and Click Repair:
https://www.quadraphonicquad.com/fo...software-demodulator.24900/page-8#post-439134

Kirk Bayne

 

[atrocity]

For hum I would check cables and phono, pre, and amp connections, and not want that filtered out.

I think it's often baked into the recordings, given that when having CR just output the noise it's at different levels on different records. Sometimes it's barely there at all, other times it's surprisingly loud, though I never actually hear it when the music's playing.

 

[markshan]

For hum I would check cables and phono, pre, and amp connections, and not want that filtered out.

Filters have their place but I certainly subscribe to the idea that an ounce of signal is worth a pound of processing.