Vinyl to CD - again

[Gordon]

AIC #74

Back to your original objective: I can't see why you're bothering to make digital copies of your LPs. I have over 2,000 LPs. I still play them and they sound excellent.

Thanks. As I said at the start this is a friend getting paranoid but has now a better understanding of what’s involved. His music taste and sound preferences are different to mine and to be honest his system [REGA/Dynavector/NAIM] sounds bass light on both vinyl and CD and also a little shrill at the top but he likes it like that!! I really don’t fully understand what that paranoia is [see earlier post] but I’ve said my piece of advice and so now it’s down to him. It has been useful to see what posters here have done to see what solutions may exist. I did all my LP conversions years ago using a Yamaha HD recorder and v good it was until it died!!

The only ones I've ever digitised have been for times when friends have heard an LP, can't get a digital copy/CD and asked for one from me. Then I play the LP on my usual set-up (SME/Lyra deck, Chord preamplifier) and pipe that into a Sony PCM1 recording at CD standard. There's another thread where you stress the importance of the front-end design of amplifiers, and this is especially true of the LP input, which is not easy to design). The Chord tackles this very well. I've experimented with higher bit lengths/rates and yes, there's an improvement (slight) but the files are rather too cumbersome for present-day domestic technology (and I have gigabit LAN, tera bytes of storage).

I have done the same for other friends that have not got the same paranoia!! I have tried auditioning different file formats and I cannot hear a sufficiently compelling difference between LP done at 96/24 and 44.1/CD or to the commercial CD to be bothered about. Others here will disagree I’m sure but that’s audio for you….

The key to surface noise control is to buy a good quality cleaner such as a Keith Monks machine. There was a glut of second hand machines a few years back when the libraries ditched their LP stock, but they're still around. Remember how quiet the BBC LPs used to sound? Well, that's what you get with one of these.

If you're going through this exercise to create an archive of your LP catalogue, it would be probably a lot less effort to buy second copies of the LPs. Most of them will be still out there.

I agree re surface noise; a clean disc is usually a quieter one except [in my experience] if it has been badly treated with a too heavy playing weight in its past. And speaking of that, the BBC used heavy playing weights as I recall, they shift the muck a bit more effectively!! My experience with second hand discs, even some from “reputable” sources has been disappointing and frankly not worth the prices charged, but sometimes rarities [back to my paranoid friend who prides himself on those he has] just aren’t available. I copied most, not all, of my 3,500 LPs to get rid of them but I kept just a few rarities back but my friend is intent on keeping all his and wants digital backup for the rarities and also for convenience on iPod. I was fed up with the faff, cleaning, destaticising, warps, swishing noises etc etc, including wow due to bad centring it simply was no pleasure any more. One Decca LP I had, part of the Solti Gotterdammerung set, was at least a semitone out in wow on one side but fine on the other!!!– It went back to them and was replaced by an identically faulty one, never did get a good one so the set was disposed of!! Touch wood, never yet had a bad CD – except the bronzed ones few a few years ago that were all replaced and still play perfectly now.

DAVE2002 #75:

Absolutely agree. Would be a standards issue, and again my guess is that it would give more accurate tracking than a radial cutter with most pickup arms even if not absolutely compliant.

At the end of the day, how much difference does a 1 degree tracking error make in either the vertical or horizontal direction? How does it translate into distortion? I suspect not much serious distortion arises from small angular errors, and also my thinking is that the distortion would be fairly benign.

Not a lot perhaps, but to a dyed in the wool audiophile it is a grave matter and one to fret about. And anyway, that is not the only angular error to consider [just think about them all, in 3 dimensions] so when they all add up there is a measurable consequence so keeping each contribution down as much as possible helps. The result is distortion and crosstalk especially with spherical stylii although those are not common any more and are actually more tolerant in some respects than ellipticals because if their rotational symmetry. With those ellipticals, designed to trace the groove better but not ideally by any means, one has to be more sure of the mounting precision [several sources of errors possible] otherwise all is lost. Which takes us on to SRA/VTA….

I recall that vertical tracking error was strange, because it wasn't just a simple question of making the cutter have a well defined angle to the vertical. Ben Bauer discovered that the material used to cut the disc relaxed a bit, and this gave an effective vertical tracking angle significantly different from the cutter angle.

I don’t see how anyone can believe that the SRA/VTA, two different but confusible things but call them what you want, are precisely and consistently set during playing an LP or that there is an ideal setting that removes all error – settings should vary at least with the brand of cutter, whose cutting angle isn’t guaranteed fixed either statically or dynamically, and obviously with any warp in the disc. I can believe that a well set up playback system can optimise the sound but it is a matter of reducing sources of problem to a minimum rather than eliminating them. A compromise as usual as few minutes with a piece of paper, a pen and a bit of school trigonometry will tell you.

Acetate elasticity has been known about for decades and only when those surround sound LPs were being touted back in the late 70s and then DMM came along did the matter get some serious attention. It’s an insoluble one and a lot of nonsense has been committed to print about it too.

[An_Inspector_Calls]

Gordon, Dave 2002
For a graph of the difference between the tracking error a 9 " and a 12 " tone arm see here:

http://www.sme.ltd.uk/content/Series-V12-1515.shtml

There's not a lot of difference between the two. Half percent distortion (second harmonic) seems about average. And the improvement offered by a 12 " tonearm hardly seems worth the candle.

Re playing weights: well, you'd think putting an old threepenny bit on the end of a tonearm to stop it jumping would cure the LP of any top response, but not a bit of it. I've got LPs played like that which sound fine (apart from being old recordings). I think the reason is that our modern cartridges are reading different depths in the groove.

Re LP cleaning: I think the reason a lot of people complained about LP noise was not down to bad handling but caused by the growth of mould in the groove. It was a usual trick to play the LP with a dampened dust-bug, and then the LP was stored in quite warm rooms - just the place for the growth of mould. A Keith Monks cleaner will fettle that. I've had LPs brought to me that were quite dreadful but after a couple of sessions with the record cleaner were quite clear of all clicks and swishes.

[Dave2002]

AIC

Thanks for that. Indeed I wonder if it would be possible to compensate for any tracking error if this is known. If the location of the arm on the disc is known, then it should be possible to determine the error. If the signal off the disc has been digitised I'll bet it's theoretically possible to remove most of the distortion due to tracking. Whether it'd be worth it I'd very much doubt, unless there was a disc which was very rare and where the quality of the recording and performance justified it. For most discs I'd expect other factors to be more important.

[Gordon]

AIC

Thanks for that. Indeed I wonder if it would be possible to compensate for any tracking error if this is known. If the location of the arm on the disc is known, then it should be possible to determine the error. If the signal off the disc has been digitised I'll bet it's theoretically possible to remove most of the distortion due to tracking. Whether it'd be worth it I'd very much doubt, unless there was a disc which was very rare and where the quality of the recording and performance justified it. For most discs I'd expect other factors to be more important.

Yes that tracing error can be calculated as in the SME link and is down to some simple geometry in which case it is removable once you know what it is. Snag is distortion isn't reversible a posteriori without a lot of faff - generating it again and adding it back out of phase so as to cancel it. Best not to cause it in the first place. Maybe some digital post would do the job but you'd need to know precisely how much and its phase. I'd suspect that that the latest versions of CEDAR etc are able to do this eg the process used by Pristine but with some human intervension.

EDIT: Here's how the tracking error varies with the overhang used in set up for a 9" arm, as AIC says the difference for a 12" isn't that much:



These agree with the SME ones - their green line, but they don't give the overhang. You can actually get plus or minus 0.5 degree for most of the disc surface from innermost to a radius 102mm if you choose the right overhang of 12.5mm [my yellow line]. This then helps with the inner groove performance but the error at the outermost groove is then 4 degrees.

Note that most set ups using a protractor with that overhang set the zero error somewhere in the middle of the disc, meaning there will be an error at the inner groove just where you want it least unless that point is at 82mm radius!! The pale blue line shows that plus or minus 1 degree can be achieved with an overhang of 15mm and gives a zero error at 110mm where the protractor should point for that desired error. At the outermost groove, 146mm where error can be tolerated better because of the longer wavelengths, the error is now 3 degrees. All in all not bad.

The curves are normalised so that there is zero error at the innermost groove defined by the disc standard, 60.3 mm. Notice that an overhang of 20.3 mm secures zero error at both the Inner and Outer grooves but with a peak error of 2.5 degrees in between [dark blue line]. I have not done the sums but I reckon that the crosstalk/distortion produced in the vertical channel [L-R] due to a largish lateral modulation will produce more than this.

[An_Inspector_Calls]

" tracing error can be calculated "

I wonder if it can? Firstly, are we guilty of merging two forms of distortion: tracing distortion introduced by the cutter, and tracking distortion introduced at playback? The distortion of both types will depend on the geometry of the cutter/stylus assembly and then be modified by the recording levels and the equalisation curve operating at the time. Too many unknowns perhaps?

Does it really make a big difference? I can't tell, but perhaps the experience of second harmonic distortion we hear when we listen to moving coil loudspeakers can give us a clue. For many years the big advantage electrostatic loudspeakers had over moving coil was in distortion level. Moving coils always suffered second harmonic distortion due to self induction. Then ATC introduced their super linear drivers which heavily reduced this effect and got their moving coils down to electrostatic speaker distortion levels. Fortunately, ATC owners were offered upgrades to the driver units which I took. The difference was stunning, esp. for string music such as cellos and violas. But the improvement applies to both CD and LP listening which suggests to me that LP tracking distortion is perhaps not such a problem.

[cloughie]

Ref the Keith Monks machines - I used to use the Lenco Clean, an arm which delivered fluid and played the record wet thus removing the crackle and pop. The down side was that thereafter the disc if played dry had even more noise, as the mud created in the grooves dried to a silty dust, so playing wet became the norm for best results. Nowadays my use of LPs is mostly for transfer to CD or Minidisc and so I use a mist spray to reduce the noise. I digitise the LP tracks using a Boss recorder with USB downloading. Time consuming but worthwhile for those LPs the record companies stubbornly decline to put out on CD.

[Dave2002]

" tracing error can be calculated "

I wonder if it can? Firstly, are we guilty of merging two forms of distortion: tracing distortion introduced by the cutter, and tracking distortion introduced at playback? The distortion of both types will depend on the geometry of the cutter/stylus assembly and then be modified by the recording levels and the equalisation curve operating at the time. Too many unknowns perhaps?

Does it really make a big difference? I can't tell, but perhaps the experience of second harmonic distortion we hear when we listen to moving coil loudspeakers can give us a clue. For many years the big advantage electrostatic loudspeakers had over moving coil was in distortion level. Moving coils always suffered second harmonic distortion due to self induction. Then ATC introduced their super linear drivers which heavily reduced this effect and got their moving coils down to electrostatic speaker distortion levels. Fortunately, ATC owners were offered upgrades to the driver units which I took. The difference was stunning, esp. for string music such as cellos and violas. But the improvement applies to both CD and LP listening which suggests to me that LP tracking distortion is perhaps not such a problem.

AIC
Your comments about loudspeakers and distortion are very interesting, and had me rushing to find out more about super linear drivers - articles such as this - http://www.tagnz.co.nz/uploads/pdf/S...hite_Paper.pdf
This article does, however, specifically refer to reduction in 3rd harmonic distortion. My knowledge, and assumption, is that generally small amounts of second harmonic distortion are not too unpleasant, and relatively insignificant. I'm not suggesting that such distortion is acceptable, but simply that in terms of sound quality it is relatively benign compared with odd order harmonic distortion. Clearly your experience with speakers has led you to believe that some manufacturers have improved the SQ considerably with their own designs in the last decade or two.

My own experience with cartridges over the years - though clearly not recent - has been that severe distortion is possible if tracking fails badly, or if the LP is damaged. On the other hand, if a cartridge is fitted properly, and in a good arm, then it should be possible to recover the signal pretty well. It was on that assumption that I thought it might be feasible to compensate for the tracking error - though it could be that there are too many other variables to make it worthwhile. Other factors, such as the corrections for RIAA curves would also interfere, and as discussed earlier these might be hard to do in the digital domain. Indeed, if one really cared, for digitising purposes, it would be possible to adjust the orientation of the cartridge in a conventional arm so that over (say) 1cm radial traverses of an LP the tracking error could be close to zero. This would be no good for regular listening, but for transcription purposes would reduce tracking error to very low levels, thus enabling the whole of an LP to be played with close to zero tracking error throughout. If anyone really wanted to do it, I expect that the cartridge could actually be adjusted dynamically by some suitable electro-mechanical device, though it would have to be a good one, and not introduce unwanted vibrations etc. Another example of rethinking things in analogue terms perhaps.

[An_Inspector_Calls]

Dave2002

Alzheimer's sets in: you're right, of course loudspeaker self induction is third harmonic, not second harmonic. ATC used to have some superb pdfs on their super linear technology. Now all I can find is this:

http://cyrille.pinton.free.fr/electroac/lectures_utiles/tech_haut-parleurs/ATC_super_linear_technical_white_paper.html

They achieve a reduction of 10-15 dB in third harmonic distortion between 100 and 3,000 Hz so the improvement isn't subtle, it's huge.

Re tracking. Most modern cartridges completely cure any serious tracking problems. All this discussion of overhang/tracking error reminds me that my father had a parallel tracking deck - sorry I can't remember the make. My father's hi fi budget was never that high so it must have been quite affordable. And it seemed to work very well. Perhaps it didn't look as pretty as an SME arm!

[Dave2002]

Dave2002

They achieve a reduction of 10-15 dB in third harmonic distortion between 100 and 3,000 Hz so the improvement isn't subtle, it's huge.

Re tracking. Most modern cartridges completely cure any serious tracking problems. All this discussion of overhang/tracking error reminds me that my father had a parallel tracking deck - sorry I can't remember the make. My father's hi fi budget was never that high so it must have been quite affordable. And it seemed to work very well. Perhaps it didn't look as pretty as an SME arm!

It is still interesting that you rate the speakers, and attribute their properties to the reduction in 3rd harmonic distortion. There could be other factors - such as they might just be better! The 10 to 15 dB improvement takes the distortion down from about 40 dB below signal level - I'll believe that's audible if you tell me so. Also, are ATC the only manufacturer using those techniques? I've not had serious new speakers for over 20 years now, and could be interested in new developments, if affordable.

The Garrard Zero 100 had parallel tracking, though apparently that didn't make the arm sound better than others. For passive models bearing friction was a significant adverse factor. Some models used powered tracking, where, for example, the pivot moves across parallel to a radius. Vibrations can be induced by the tracking motor. Some systems used a much smaller and probably lighter arm (given that the pivot was moving end externally powered), and that could affect the arm resonances and influence the cartridge response.

[Gordon]

" tracing error can be calculated "

I wonder if it can? Firstly, are we guilty of merging two forms of distortion: tracing distortion introduced by the cutter, and tracking distortion introduced at playback? The distortion of both types will depend on the geometry of the cutter/stylus assembly and then be modified by the recording levels and the equalisation curve operating at the time. Too many unknowns perhaps?

Does it really make a big difference? I can't tell, but perhaps the experience of second harmonic distortion we hear when we listen to moving coil loudspeakers can give us a clue. For many years the big advantage electrostatic loudspeakers had over moving coil was in distortion level. Moving coils always suffered second harmonic distortion due to self induction. Then ATC introduced their super linear drivers which heavily reduced this effect and got their moving coils down to electrostatic speaker distortion levels. Fortunately, ATC owners were offered upgrades to the driver units which I took. The difference was stunning, esp. for string music such as cellos and violas. But the improvement applies to both CD and LP listening which suggests to me that LP tracking distortion is perhaps not such a problem.

You are right of course, there are many sources of error/distortion in analogue recording and playback systems and not only 2nd harmonic which is prevalent but more tolerable and actually contributes to the analogue sound that people like [eg valves].

The calculation I did in #79 was only that for the angular non-tangential "tracing" error [nomenclature needs tidying up] for a stylus set in a groove on the end of a conventional arm. Distortion of the waveform because of this stylus tracing will be 2nd harmonic [its amplitude is reported on that SME link] but there is also that due to the stylus not having the same shape and trajectory as that of the cutter blade which is also 2nd harmonic. This is caused by the groove narrowing [it isn't 90 degrees between groove walls as the playback stylus sees it] as it deflects laterally thus lifting the stylus tip [same thing happens in the vertical direction] even though there may be no vertical modulation. For spherical tips this produces more distortion than the tracing error and is most prevalent at short wavelengths. A quick sum says that this can be as much as 5% for a small narrowing of a few degrees.

These figures [and the SME ones] are illustrative because, as you say, there are many other things going on as well. Provided that the equipment used is well designed and set up the distortions are tolerable at 2nd harmonic partly because we have got used to it but it's also naturally present in music anyway so a bit more isn't such a big deal - it's a colouration that we don't hear unless we can also hear the source. Intermodulation is perhaps a more important phenomenon because non linearity adds components that were not in the original. The interesting question is the engineering ie how much distortion is tolerable and how do we allocate it to the parts of a chain because on the whole, although its phase can change, it does add on a power basis. So many things to account for so this becomes a thankless and perhaps futile task and thus contributes to the "vinyl" and/or " analogue" phenomenon.

[Stunsworth]

Revox and B&O both had parallel tracking turntables (in the late 70s/80s?).

The were a few arms from the States that used an air bearing to provide frictionless movement. These used an air pump similar to those used in aquariums to provide the air pressure.

All of which reminds me that I ought to have my Roxsan Xerxes seen to - it's suffering from the dreaded sagging armboard.

[OldTechie]

There are other ways to get that vinyl sound: http://www.izotope.com/products/audio/vinyl/ - and they don't even charge you for it (except for the fact that you need Avid ProTools to use it.) Unfortunately it only does the bits you normally spend ages trying to avoid.

[MrGongGong]

There are other ways to get that vinyl sound: http://www.izotope.com/products/audio/vinyl/ - and they don't even charge you for it (except for the fact that you need Avid ProTools to use it.) Unfortunately it only does the bits you normally spend ages trying to avoid.

What is interesting is that the sounds that many of us spent the first 30 years of our lives trying to exclude from our recordings are now the very things that are added or even what music is composed of. I think that there is an interesting change in the way that people listen and what is regarded as an "error" , Many of the young people I work with don't hear scratches and crackle, static and distortion as "errors" even in straightforward mainstream "classical" music recordings. There is sometimes interesting discussion of the aesthetics of this on the CEC list....