I've been using the Izotope RX series to remove clicks and pops (and crackles) from my 10,000+ vinyl rips for the past 6+ years. Very fine product. Used by the pros. You can do an inversion so you can just hear the parts that were removed to see whether there is any musical content. Depending on the version it runs from the low hundreds to high hundreds of dollars. Also does many other things, like invert phase, switch channels, and even removes frequency related bumps. It is typically used in a recording studio to correct imperfections in the recording process. However, for use in ripping vinyl, you do need to have separate ripping software and hardware, which I do (you can see in my signature).
I flip flop between iZotope RX and ClickRepair. Currently I like ClickRepair best. I find iZotope RX5 is too aggressive, even at the lowest settings I feel it overly smooths things.
My usual ClickRepair settings are:
Declick Strength: 1
Decrackle: Off
Reverse: On
Pitch Protection: On
Full auto is OK the vast majority of the time however if you have music with synths (potential square-ish waveforms) or noise intentionally on the track you need to be very careful with auto de-click algorithms as you can get a ton of false positives which will destroy the transient attack/decay of the instruments.
I 100% agree with astrotoy that RX is the best DAW for vinyl rips or any sort of restoration work. (I just prefer a different declicker)
Edit: With reference to the original topic, I would not be surprised at all to discover they are using ClickRepairRT inline in the SugarCube. I think the SugarCube is a very cool product.
Thanks guys. Had not heard (or remembered ) these other options. This is a very challenging problem. The ear easily distinguishes pops and clicks from music but to a computer, it looks like high frequency transients. Lossy codecs attempt to find transitions too (to adjust their window) and I remember we worked very hard to make them bullet proof but at the end, it was not so. The sugarcube buffers 5 seconds of audio and looks ahead to see if there is any harmonic structure (using wavelets). If there is, then it considers it music. In listening to it, I could not find a single instance that it misfired. I doubt that they were using what others have but i am just guessing.
They were quite interested in continuing the conversation with me so I hope to get more information about it in the future.
I don't rip vinyl but I have a way of removing an estimated 90-95% of ticks and pops by way of the record cleaning machine and steam cleaning the LP's (safely). You would be surprised at how many if them are removed even from a brand new LP.
I don't rip vinyl but I have a way of removing an estimated 90-95% of ticks and pops by way of the record cleaning machine and steam cleaning the LP's (safely). You would be surprised at how many if them are removed even from a brand new LP.
+1000..... I've never found a product that will remove pops/clicks from vinyl without touching the music. I've used everything from Cedar to Izotope. That being said, I've found the higher end turntables/carts and phono pre's, and a good KLAudio clean allow for a silent (or near silent) transfer. People mistake my vinyl transfers for tape transfers!!
+1000..... I've never found a product that will remove pops/clicks from vinyl without touching the music. I've used everything from Cedar to Izotope. That being said, I've found the higher end turntables/carts and phono pre's, and a good KLAudio clean allow for a silent (or near silent) transfer. People mistake my vinyl transfers for tape transfers!!
Bruce, I have heard four of your rips in a very good system. They were very near CD silent in terms of pops/clicks. I was extremely impressed and it has me considering replacing my labor intensive Loricraft RCM/4 step liquid process with a KLAudio ultrasonic RCM. Whatever your exact process is, the results are truly impressive.
Bruce, I have heard four of your rips in a very good system. They were very near CD silent in terms of pops/clicks. I was extremely impressed and it has me considering replacing my labor intensive Loricraft RCM/4 step liquid process with a KLAudio ultrasonic RCM. Whatever your exact process is, the results are truly impressive.
Yes I completely agree; getting the record as clean as possible before starting is a must. I built a DIY ultrasonic (Elma P60H + vinyl stack) machine last year and that was one of the best things to happen to my vinyl replay.
Anyhow, prompted by this thread, I did a some listening tests over the past few days. First result was that I still feel that the current version of ClickRepair (3.9.5) is better then the current version of RX (RX5). RX5 is too aggressive, even with the lowest setting I think it's making way way way too many fixes to things which are not audible. Second result was that I feel that both algorithms have the effect of closing in the sound a bit. The effect is very small and for a noisy record the I think trade off in reduced surface noise is worth it.
Based off the above, it seems the best way is using the manual interpolate function in RX to remove anything audible which interferes with the music. The downside is that it takes a (very) long time to do this.
Random question for Bruce... What's your opinion on recording in DSD and converting to PCM for editing using something like DSD Master or Saracon? I think the DA3000 produces better results this way. (I use DSD Master to convert)
Random question for Bruce... What's your opinion on recording in DSD and converting to PCM for editing using something like DSD Master or Saracon? I think the DA3000 produces better results this way. (I use DSD Master to convert)
Based off the above, it seems the best way is using the manual interpolate function in RX to remove anything audible which interferes with the music. The downside is that it takes a (very) long time to do this.
Click Repair is java based and (ironically for a Java app) is the fastest of them all by a large margin. It takes under a minute to process a side captured in 24/192 on my 2014 MBP. There is a RT (realtime) version of Click Repair which is designed to be stuck inline. That, the talk of wavelets and then the "negative" where only the pops are output are all features of ClickRepair which is why I strongly suspect the SugarCube guys are licensing it from Brian Davies. Also, the cost per license for an individual is $40. If the SugarCube guys negotiated a bulk licensing deal I'm sure the cost would be would be very much suitable for integration into a $2000 product.
Got it. The lowpass filter in DSD Master is set around 45Khz (much higher then Saracon) which is perfect for me as my final output is usually 24/96. I don't think you would want higher as the quantization noise is rising very rapidly at that point.
Got it. The lowpass filter in DSD Master is set around 45Khz (much higher then Saracon) which is perfect for me as my final output is usually 24/96. I don't think you would want higher as the quantization noise is rising very rapidly at that point.
+1000..... I've never found a product that will remove pops/clicks from vinyl without touching the music. I've used everything from Cedar to Izotope. That being said, I've found the higher end turntables/carts and phono pre's, and a good KLAudio clean allow for a silent (or near silent) transfer. People mistake my vinyl transfers for tape transfers!!
I largely agree with this. Even though most of the programs have an "inversion" function, I'm not convinced they don't touch the music, however small the effect may be. It is very easy to create a test scenario where you run a segment through the software, invert it, do not hear any music but then when you actually do the processing, the music is very obviously effected, sometimes dramatically. It is for this reason that I largerly eschew them as Bruce appears to. For my own part, after I transcribe a given side, I will run through it visually, place a marker on any obvious glitches then fix them up manually by either re-drawing the wave form with a pencil tool or using interpolation very sparingly, depending on the situation. I will still use software for the final touch, but at an exceptionally conservative setting that is so low I have difficultly hearing any effect on the music even though it still exists.
And as Bruce says, the right turntable, cartridge and cleaning regime / equipment will make a world of difference.
I have, however, come up with some extremely challenging situations in my time. They are very difficult and time consuming to fix but there is a lot of pride in getting it right. In a relatively recent situation I was left with a situation where there were some severe places of non-fill. Anyone who knows how horrible this sounds will tell you that fixing it transparently is almost impossible. It is pretty easy with a mono recording if one channel only is effected since you can simply copy and paste the unaffected channel onto the effected one (one reason I will stick to using a stereo cartridge on modern mono reissues - once everything is perfect I will take the "better" channel and paste it over the "worst" one so that I have "perfect" mono as an end-result). Stereo is far more difficult. So in this recent situation I actually went to an alternative source (24/192 source files), spent quite some time massaging those with level changes and some EQ until it matched as closely as my ears could tell the vinyl that I was transcribing. I then created a noise floor file based on the existing noise on the vinyl and mixed it with the massaged digital file. I then painstakingly mixed the digital bit over the stuffed analogue channel and the result was so perfect you would probably have to donate your ears to science if you were able to hear what I did and where I did it.
So in this recent situation I actually went to an alternative source (24/192 source files), spent quite some time massaging those with level changes and some EQ until it matched as closely as my ears could tell the vinyl that I was transcribing. I then created a noise floor file based on the existing noise on the vinyl and mixed it with the massaged digital file. I then painstakingly mixed the digital bit over the stuffed analogue channel and the result was so perfect you would probably have to donate your ears to science if you were able to hear what I did and where I did it.