Digital / Analogue: some observations

[iMark]

Over the last couple of weeks I have digitized old LP's for friends. This has also led me to playing some of the great records that I inherited from my parents. These were operas from the 1950's and 1960's. After a thorough clean with the Knosti Disco Antistat these records sounded quite good. But the digital transfers I made sound even better after applying some filtering in Vinyl Studio. I can thoroughly recommend that software. Incidently, I record to Apple Lossless files, which is CD quality if I want to burn CD's and stream from iTunes to the Airport Express.

Analogue or digital recordings on vinyl. It's interesting that classical recordings from the early 1980's were released on vinyl before they were released on CD. Even though it was a new technology, some of these LP's sound excellent. One of the interesting cases is Murray Perahia's set of recordings of the Mozart Piano Concertos. The recording cycle started in 1976 and halfway the project (then) CBS Masterworks switched to digital recording. On vinyl the records from both eras sound great. On CD the digital recordings sound a tiny bit better because of the lack of any tape hiss. Of course Sony has tinkered with the remasters and the current box set sound much better than the earlier transfers. Some other digital recordings on vinyl I have listened to: Ravel Piano Concertos by Pascal Rogé and the Orchestre Symphonique de Montréal and Charled Dutoit (Decca) and a Schubert (8th Sympony) and Mendelssohn (Italian Symphony) by the Philharmonia Orchestra and Giuseppe Sinopoli (DG).

But some of the best sounding records in the collection are late 1960's recordings by the Academy of St Martin-in-the-Fields on Argo. There's Mozart, Handel, Haydn. The recorded sound of the strings is astonishingly good and the pressings are excellent. If you ever come across any of the ASMF records on Argo, don't hesitate to buy them. The Argo recording engineers certainly knew what they were doing. I have transfered these records too to the computer so I can enjoy them on my iPod and we can listen to them in the car. I don't think you lose any of the 'vinyl magic' by transfering them. After some slight filtering you can rid of some irritating noise and just focus on the music.

So after a lot of listening I have come to the conclusion that what I like is well recorded quality playing and it doesn't really matter if I hear this on vinyl, CD or any other type of disc. Modern classical CDs and SACDs can sound astonishingly good too. My main gripe with modern CDs is the compression in many 'popular' recordings. That really kills the music for me.

One final thought however is on distortion. Many people claim that vinyl sounds more 'natural' and 'warmer' than digital. I'm not at all convinced by this argument. I have noticed that on quite a few records there is quite a bit of distortion when the stylus comes nearer to the centre of the record. This is very noticable in a lot of classical recording when there is a crescendo at the end of a work. It's also more notacible when there's more music on a side. Any record side that has more than 22 minutes on it seems to suffer more from distortion than other records. I suppose the distortion and the lack of surface noise in very quiet passages is one of the main reasons that classical music collectors were happy to switch from LP to CD.

 

[matt49]

Thanks for those very interesting observations.

Since the late 1950s classical music has in general been very well recorded and benignly mastered. The problem was the storage medium. I think you're absolutely right about the reasons why most lovers of classical music migrated to CD, even though many recordings on vinyl do sound excellent. CD offered a transparency that at least equalled LP but suffered from none of the infuriating problems of vinyl: the need for obsessive care in handling LPs and maintenance of the replay equipment, and the issue of groove cramming.

I plan to do some archiving of vinyl sometime soon. As you rightly say, much rock/pop music of the last 20 years suffers from heavy-handed use of compression in digital formats, whereas the vinyl recordings don't.

Matt

 

[CnoEvil]

"So after a lot of listening I have come to the conclusion that what I like is well recorded quality playing and it doesn't really matter if I hear this on vinyl, CD or any other type of disc. "

This I agree with.

 

[ID.]

Interesting. Thanks for sharing your impressions.

 

[The_Lhc]

One final thought however is on distortion. Many people claim that vinyl sounds more 'natural' and 'warmer' than digital. I'm not at all convinced by this argument. I have noticed that on quite a few records there is quite a bit of distortion when the stylus comes nearer to the centre of the record. This is very noticable in a lot of classical recording when there is a crescendo at the end of a work. It's also more notacible when there's more music on a side. Any record side that has more than 22 minutes on it seems to suffer more from distortion than other records.

This is largely unavoidable as the cartridge is usually aligned for the centre of the record and obviously the angle of the tone arm changes as the cartridge tracks across the record, so by the end of the record it is subtly out of alignment compared to the centre. There isn't much you can do except experiment with your alignment settings to see which gives the best compromise across the whole disc.

 

[iMark]

Last night I read an online article about the first digital recordings in the 1970's and why the record companies were very quick in adopting digital recording for classical music, years before there was a digital medium to release them on.

http://www.aes.org/aeshc/pdf/fine_dawn-of-digital.pdf

Even though the bitrates were quite low in the beginning, the engineers obviously thought the sound was better than using analogue tape. It's interesting to hear great analogue recordings from the 1950's, 60's and 70's and pretty bad digital recordings from today. On the other hand I know some awful recordings from half a century ago and some terrific digital recordings that were made last year. In the end it boils down to knowledgable recording engineers that know a thing or two about setting up their microphones. Whether they record digitally or analogue doesn't make a lot of difference.

A very interesting question remains why so many of us like analogue sound. And would we like analogue sound if it had never been available? After a lot of listening I have come to the conclusion that provided the sound is not compressed, CD is very good medium for classical music (and jazz). SACD is a touch better and vinyl still sounds OK but depends heavily on the quality of the pressing, the way it has been handled and the playback equipment. One of my favourite digital recordings is on hybrid SACD: Mozart Symphonies 38-41 by the Scottish Chamber Orchestra under Charles Mackerras on Denon.

Modern popular music should be released in two different masters. A compressed version for the mp3-generation and an uncompressed version for the audiophiles. Let the consumers decide which version they like best. Incidently, I begin to understand more and more why people like popular music on vinyl. A lot of effects and studio trickery depend on distortion. So a bit more distortion from the playback medium only adds to the experience while in jazz and classical recordings the listeners want as little distortion as possible.

 

[Vladimir]

I would like to contribute to your investigations with this video.

 

[iMark]

Thanks. Interesting video.

 

[CJSF]

Mmm . . . I dont desput the video and what it apears to show, however, it does not tell us what TT/arm/cartridge and therefor stylus is being used, all of which will affect the findings, not to mention the overal TT stetup?

I remember back in the 70's/80's, the Japanese were producing amps that mesured so low on the destotion/noise meter it was almost unreadable, and yet the same amps were so 'dry and flat' they were out performed in listening tests by amps with more normal figures. People came to realise that the human ear likes a bit of 'rough and tumble' to sort out.

I think Rotel were one of the first to do their testing using the human ear as a majour factor, I used to know one of their home listeners, in Rotel terms, the rest is history.

In the same period, Jardine valve amps, were producing a unit that was making everyone who heard it sit up and listen, I had the pleasure to hear it in a private evening session at one of the Heathrow shows. I seem to remember the measured destortion was somewhere in the 7% reagon? . . . it was certainly a very high figure? It is a long time ago now!!!

Such issues are a fact of our 'normal' life . . . ?

CJSF

 

[iMark]

I think that we can conclude from all these discussions that the Holy Grail of recording, faithful reproduction, is no longer at the heart of recording of reproduction for some people. Could it be that humans actually prefer distortion over faithful reproduction?

It's pretty obvious from the video that when you play the same test recording from a CD and an LP the CD will give a more faithful reproduction of that test signal than the LP. And it doesn't matter what turntable, cartridge, phono stage or quality of vinyl you use: there will always be distortion from the LP medium.

I actually prefer systems that don't colour too much. They tend to tire me much less than other systems. I compare classical recordings to what I have heard in concert halls. I tend to like recordings that to me sound similar to performances I have heard in a good concert hall. I obviously prefer digital recordings because they don't have any tape hiss or other distortion and I can hear all the details.

At the same time I like to play vinyl but I will never claim that it is a more faithful reproduction than a CD. In theory at least a CD must sound more faithful (or better) than an LP. The fact that a lot of people don't like the sound of modern CDs has absolutely nothing to do with the medium itself. I own a couple of LPs that I prefer to their CD counterpart but that only has to do with the mastering.

I wonder if recording and reproduction never would have had the stage of tape and vinyl but would have gone straight to digital we would have this discussion.

 

[andyjm]

I think that we can conclude from all these discussions that the Holy Grail of recording, faithful reproduction, is no longer at the heart of recording of reproduction for some people. Could it be that humans actually prefer distortion over faithful reproduction?

I wonder if recording and reproduction never would have had the stage of tape and vinyl but would have gone straight to digital we would have this discussion.

An interesting point.

With modern DSP techniques, it is possible to engineer a near-flat frequency response (at least in one location) in a listening room. I have done it, and it sounds pretty sad and lifeless. This has lead to discussions about a 'house curve' where the frequency response is deliberately skewed to make the system 'sound better'.

It is possible that this is because the original mix was done in a studio which itself had questionable acoustics, and you need to reproduce the dodgy acoustics in your own home to get the same experience as the mixing engineer did. A case of two wrongs making a right.

Or it could be we have just got used to a certain sound and we feel comfortable with it, even if it is nothing like the original.

An interesting parallel is the use of high frame rate to overcome motion judder found in 24fps films. Movie audiences have generally reacted badly as 'it felt like I was watching TV and not a real film'. We expect motion judder in films, and video to be much smoother. If we had grown up with 100Hz frame rates in movies, I doubt anyone would be demanding to go to 24fps.

 

[Glacialpath]

I would like to contribute to your investigations with this video.

That is an interting video. How about this one?

https://www.youtube.com/watch?v=hh-23EUP-s8

 

[iMark]

Those guys in the video make some intersesting points. So generally analogue recordings sound better on vinyl and digital recording sounds better on CD. Obviously the question remains what 'better' actually means. My suggestion is that a record has to be as faithful to the original master tape as possible, possibly with filtering out the tape hiss. This should be relatively easy to accomplish with analogue recordings if transfered properly to the digital domain. Things go bad when engineers start compressing the digital transfer. (An example of great digital transfers are the RCA Living Stereo SACDs, based on tapes from the 1950's. They sound awesome and better than original vinyls of which I have heard a few.)

Then we come to modern digital recordings. I don't think the recording process in itself is the problem. It's the mastering with compression that is driving us mad. In classical and jazz recordings modern engineers show us how good digital recordings can sound if they leave out (or use minimal) compression. That has to be the main reason that the whole discussion about the loudness wars focuses on popular music and there are no complaints about modern digital classical recordings. (Obviously some classical recordings sound better than others but that is mainly to do with the recording venue, the engineering and the microphone positioning.)

I still think that there is nothing inherently wrong with the Redbook CD format. The dozens of excellent classical recordings we have show this. (SACDs and Bluray Audios sound a little bit better on our system but not very much.) Modern stuff sounds dire and the mastering engineers have to take the blame for that. Just put the LP masters out on CD for the audiophiles.

 

[Vladimir]

Ah... can't remember the full details about the video test setup, it's all burried in a lengthy thread over at the Harbeth Forums (link in video description). I do remember it was a Thorens (TD 166 or 160, I think) with a good cartridge, not a Koetsu but something decent. I trust Alan Shaw to know how to setup a TT, especially considering his age, expertise and him being a vintage nutt.

 

[Vladimir]

I think that we can conclude from all these discussions that the Holy Grail of recording, faithful reproduction, is no longer at the heart of recording of reproduction for some people. Could it be that humans actually prefer distortion over faithful reproduction?

Carefully dosed distortion yes.

This sound file has violins playing. The loop goes pure, clipped, pure, clipped... The component clipping is the amplifier. Which one of these sounds better to you?

www.harbeth.co.uk/library/mp3files/threeviolins.mp3

 

[cheeseboy]

Modern stuff sounds dire and the mastering engineers have to take the blame for that.

Personally I'd say it's the record companies themselves. Any studio engineer, mastering or mixing would not want to put out the over compressed limited stuff, it's the record companies telling them to do it, becuase if mastering engineer A doesn't do it, it'll take them about 5 seconds to get one in that will. Jobs in the music industry are hard to come by, therefore if the person paying the wages says make it loud, that's what you do

 

[Vladimir]

I would like to contribute to your investigations with this video.

That is an interting video. How about this one?

https://www.youtube.com/watch?v=hh-23EUP-s8

At what point in the video do they start measuring/testing?

EDIT: Just watched it, apparently it's just a discussion with sharing some good points. Plenty of similar videos online but I like Alan's approach of getting down to business (science).

 

[Thompsonuxb]

Not all vinyl is recorded well - that I think is the first myth that needs busting.

And not all CD is poorly mastered and sounds rubbish.

The best CD's sound better than the best vinyl too at a lower cost point ref equipment.....lol.

Watched the videos.... Lol.

The one measuring the sound wave was funny. As one of the comments correctly identifies - the professor doing the measurements made no allowance for the TT being a mechanical device.

While the signal was oscillating on the metre the frequency/pitch remained constant/stable - chances are you would not hear that oscillation.....actually you might....I doubt it though.

 

[Vladimir]

While the signal was oscillating on the metre the frequency/pitch remained constant/stable - chances are you would not hear that oscillation.....actually you might....I doubt it though.

How about jitter in SATA cables?

 

[Thompsonuxb]

This is were it gets interesting. All the arguments over the past few weeks ref 'is it audible' comes full circle.

Could go some way in explaining people's preference for vinyl or cables.

Note the sine wave for the TT - on the test tone you have 3 peaks could this be an audible echo?

Ambience on each harmonic presented to a speaker, a note lasting a few micro seconds longer than the dry single peak on the digital signal?

Jitter in sata cables - tis a can of worms and I'm sure no one wants another cable thread.

I mean if we go there will have to re visit the burn in topic. How things or how the characteristics of capacitors for example may change after a thermal cycle, cables of different resistance etc - is it audable......?

Lol...... Hopefully when davedotco looks over this topic he'll contribute.

While the signal was oscillating on the metre the frequency/pitch remained constant/stable - chances are you would not hear that oscillation.....actually you might....I doubt it though.

How about jitter in SATA cables? 

 

[Vladimir]

This is were it gets interesting. All the arguments over the past few weeks ref 'is it audible' comes full circle.

Could go some way in explaining people's preference for vinyl or cables.

Note the sine wave for the TT - on the test tone you have 3 peaks could this be an audible echo?

Ambience on each harmonic presented to a speaker, a note lasting a few micro seconds longer than the dry single peak on the digital signal?

What you are describing is microphonic feedback that a cartridge would capture from the speakers. On that test the speakers are very small computer speakers. If all that jitter is not from the mechanical contact between the stylus and the grooves and it is purely microphonic feedback, imagine the rubbish you are hearing in your home with big hi-fi speakers.

So your theory is that if the professor had his small PC speakers turned off, the signal coming from the TT on the scope would be pure as the one coming from the CD? God forbbid he coughed or sneazed! Migh be worse than a Keith Jarrett concert! *shok*

 

[cheeseboy]

How about jitter in SATA cables?

lolz

 

[davedotco]

This is were it gets interesting. All the arguments over the past few weeks ref 'is it audible' comes full circle.

Could go some way in explaining people's preference for vinyl or cables.

Note the sine wave for the TT - on the test tone you have 3 peaks could this be an audible echo?

Ambience on each harmonic presented to a speaker, a note lasting a few micro seconds longer than the dry single peak on the digital signal?

What you are describing is microphonic feedback that a cartridge would capture from the speakers. On that test the speakers are very small computer speakers. If all that jitter is not from the mechanical contact between the stylus and the grooves and it is purely microphonic feedback, imagine the rubbish you are hearing in your home with big hi-fi speakers.

So your theory is that if the professor had his small PC speakers turned off, the signal coming from the TT on the scope would be pure as the one coming from the CD? God forbbid he coughed or sneazed! Migh be worse than a Keith Jarrett concert! *shok*

Point of order. There is nothing worse than a Keith Jarret concert (except maybe a Glenn Gould concert).

Anyway, I have this whole digital/analogue, distortion thing nailed down tight. It's simple.

Live music generates far higher sound pressure levels than music playback in the home.

Our ears are non-linear, as the spl increases so does the distortion produced by the ear.

A distortion free digital recording replayed in the home does not generate sufficient spl to cause distortion in the ear.

'Pre-distorting' the signal using, say, a record player or an appropriately 'voiced' amplifier replaces the distortion that the ear is not producing due to the lower spl.

The brain interprets the source distortion as being the same as that produced by live music at a higher spl.

It therefore interprets this distortion as making the playback sound more like the live sound, ie more 'real'.

*drinks*

 

[Glacialpath]

Dave that's the most sense I've read from you and is basically what I was getting at in the other thread about Analogue being the better format. I just didn't explain it as well as you did and is why I got so much abuse from others.

It's not that analogue is the better format as we all know Digital is by fat the better one of the 2. But your explination is why a lot of us feel Analogue is better.

Thank you for putting it in such clear terms.

Now can we all leave this subject and put it to bed.

It's all about personal preference.

 

[Glacialpath]

Yeah I was wondering if they were actually going to give visual/audible proof.

I guess they figure that because the have engineering back grounds and have already done the testing that the viewers will be happy with their word. I was as it made sense. I know others would like to see/hear the proof and too right they would be.