|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
118.148.192.127
In Reply to: RE: Yes and no. . . . posted by David Smith on May 25, 2015 at 20:09:01
I agree! However, I think what we are all talking about is with respect to the master tape after the Mastering Engineer has already fiddled with reality... and we should emphasise *MORE* accurate as opposed to more *ACCURATE* relative to the LP.
In practical terms, there is absolutely no way that the waveform on an LP can be deemed to be more accurate (with respect to the mastertape) than a digital copy simply because of the additional EQ, distortions from the power amp and additional mechanical distortions from the cutting head, which is what I was trying to convey in my previous post.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Follow Ups:
Yeah I'm still not in agreement with much of that. I do think an LP can be a good bit more accurate than digital in many areas (ie violin overtones), and while power amps and cutting heads have distortions, so does the translation of an analog signal to a digital one. Both are imperfect, and trying to rate the relative degrees of imperfection becomes an exercise of limited use given that the recording itself is an approximation at best.
Do you maintain that in the case of a digital transfer of an analog master tape that the digital is more accurate (to the master tape) than an lp cut from the analog master tape?
Dave
You'll have to be more specific about what you disagree with. Digital recording has become the norm since the early 80s when LPs bearing the label "Digital Recording" first appeared on record sleeves. You comments about "accuracy" make even less sense especially with LPs today which are cut from a digital source and have gone through an additional D/A stage before the rest of the vinyl mastering process begins.
For now let's ignore digital recordings and the replay side of things and look at only what is stored in the grooves...
You mentioned the example of violin overtones and the OP mentioned brass instruments sounding "bad". I don't understand what is meant by "bad" in this context although I attempted to enquire if "bright" or "harsh" was meant.
Could you clarify in more detail what you mean by analogue having improved accuracy in reproduction of overtones and whether you consider LPs cut in the 50s and 60s to also be superior to digital equivalents? You may be aware that the original recordings were often limited to 15kHz. I have many jazz LPs from that era where the sleeve proudly proclaims a flat response up to "15 000 cps". Would you consider FM Radio to be sufficient in reproducing violin overtones? Again the bandwidth is limited to 15kHz to allow for stereo.
What should be discussed is the relative amplitude of the harmonics relative to the fundamental and whether any additional (undesirable)components are added (and what the relative amplitudes are), NOT the claimed number of harmonics that can be reproduced. The cutting process of an LP is very bandwidth limited on inner grooves (by necessity) and non-linear so the relative amplitudes of the harmonics will be altered during the process and additional IM components are added.
With a digital copy from the analogue master, the relative amplitude is going to remain approximately the same (allowing for the band-limiting analogue filter passband response depending on the bandwidth of the system), therefore the ratios of harmonics with respect to the corresponding fundamentals will be very similar; in which case the encoded signal will be closer to the master.
I am a musician myself so I understand very well what my instruments should sound like so I am just trying to compare the difference in what we are looking for in the reproduced sound. When I master CDs from my high resolution masters, they come out virtually indistinguishable to the source when listened to in isolation. Having said that, I can make them worse by choosing inappropriate levels of dither and noise shaping which will add spurious aharmonic tones to the signal! My LP transcriptions to CD sound indistinguishable from the source both subjectively and objectively via measurement.
My point is that much of the negatives heard in CD compared to a good LP are down to poor choices in the mastering (including dither and noiseshaping). When done properly, the digital version HAS to be more faithful to the master for the reasons given! For starters, once in the digital domain you don't add wow/flutter and you don't add spurious tones due to HF signal modulation from LF signals.
If your LP provides a technically better result than the CD, then the finger would point to your digital replay equipment being inferior to your analogue rig. From a playback perspective there is absolutely no basis to your claim assuming SOTA playback equipment for analogue and digital - cartridges typically give 15 to 20% harmonic distortion at 15k to 20kHz. That level of distortion may "sound" better to you, but don't pretend that it is more "accurate"!
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
"You comments about "accuracy" make even less sense especially with LPs today which are cut from a digital source and have gone through an additional D/A stage before the rest of the vinyl mastering process begins."
That may be true of some, but certainly not all.
"You mentioned the example of violin overtones and the OP mentioned brass instruments sounding "bad". I don't understand what is meant by "bad" in this context although I attempted to enquire if "bright" or "harsh" was meant.
Could you clarify in more detail what you mean by analogue having improved accuracy in reproduction of overtones and whether you consider LPs cut in the 50s and 60s to also be superior to digital equivalents? You may be aware that the original recordings were often limited to 15kHz. I have many jazz LPs from that era where the sleeve proudly proclaims a flat response up to "15 000 cps". Would you consider FM Radio to be sufficient in reproducing violin overtones? Again the bandwidth is limited to 15kHz to allow for stereo."
Sure, I mean that strings, cymbals and brass (in particular, but I would also include bass) are reproduced with more tonal accuracy in analog. I don't think response above 15k is particularly crucial, cetainly not in that regard. With respect to that, I think I'd add that many or most modern recordings, including reissues of older recordings, suffer from exaggerated highs. It sounds impressive and people tend to associate highs with "good sound" for whatever reason. Generally speaking, I consider lps cut in the 50s and 60s to be superior to digital versions of the same recordings.
"With a digital copy from the analogue master, the relative amplitude is going to remain approximately the same (allowing for the band-limiting analogue filter passband response depending on the bandwidth of the system), therefore the ratios of harmonics with respect to the corresponding fundamentals will be very similar; in which case the encoded signal will be closer to the master."
On paper, perhaps.
"My LP transcriptions to CD sound indistinguishable from the source both subjectively and objectively via measurement."
I don't have that same experience, though I do agree that they can sound good and often better than digital reissues.
"My point is that much of the negatives heard in CD compared to a good LP are down to poor choices in the mastering (including dither and noiseshaping). When done properly, the digital version HAS to be more faithful to the master for the reasons given! For starters, once in the digital domain you don't add wow/flutter and you don't add spurious tones due to HF signal modulation from LF signals."
Again, perhaps on paper, looking at the particular variables you are looking at. But the digital version doesn't HAVE to be more faithful, for one it's been converted to digital! (And back of course). That's a pretty big issue you're not including. No doubt digital is superior to analog in terms of wow/flutter type issues, but I don't find that a critical issue in terms of the levels of wow/flutter in high-quality equipment.
"If your LP provides a technically better result than the CD, then the finger would point to your digital replay equipment being inferior to your analogue rig. From a playback perspective there is absolutely no basis to your claim assuming SOTA playback equipment for analogue and digital - cartridges typically give 15 to 20% harmonic distortion at 15k to 20kHz. That level of distortion may "sound" better to you, but don't pretend that it is more "accurate"!"
I'm not speaking exclusively with respect to my rig, it doesn't matter if that's where I make the comparison or at a studio (with SOTA equipment). Your assuming that it's the 20% distortion at 20k that is the basis of the difference I am speaking of, by all means add that to a recording and I'll let you know whether it sounds exactly like vinyl.
I'm not sure exactly where the difference of our experience lies, we probably appreciate different aspects of sound. Different strokes, no biggie. I do disagree with the notion of digital being inherently more accurate in reproducing an analog signal.
Dave
..where you are coming from better.
I have many examples myself of original LPs sounding much better than any digital transfers - a personal example would be Getz Au Go Go. I have the original LP in gatefold sleeve and several CD issues of the same material. The LP does sound better subjectively. However, what I also noticed from the most recent Verve "By Request" version, is that not only has the recording been remastered, but possibly from a different tape master as the intro to Corcovado was different. This highlights that the digital transfer is going to sound different by choice of the Mastering Engineer through EQ adjustments and dynamic range adjustments; therefore not necessarily due to the quantisation process itself.
There are other sources of difference:
i) whether Tube or solid state equipment is used during the transfer from the original master tape. Note that the original master tape is not always available and a subsequent LP reissue may be cut from the digital master (sometimes from CD...)
ii) whether the power amp used for the cutting head is tube or solid state. The distortion characteristics on overload are different and can contribute to the perceived differences.
iii) Whether the tape azimuth is perfectly matched to the original record head
iv) Half -speed mastering can improve the sound by reducing the HF losses from the playback head at the expense of low end. Tonally this will change the subjective tonal balance of a recording.
v) W/F is not zero for the cutting lathe and is independent of the W/F of the tape machine so HAS to be factored in. Given that this alters the waveform in the groove, by definition the waveform is now different to the master tape therefore cannot be deemed to be "as faithful" to the original.
Psychoacoustics plays a big part in our perception. Noise can enable us to hear sounds which would technically be below the noise floor in amplitude. Even though CD has a SNR much greater than LP, if we are talking about a transfer from an analogue master tape with hiss, we should get the same perceived result. From your description of cymbals and brass etc, would you say that transient response is one of the areas that you feel is better handled with analogue? By transient response I mean the attack and decay characteristics?
Also, would it be fair to say that part of your viewpoint is based on a suspicion of the quantisation process itself? We can agree to disagree on that one if you like! However, with respect to the transient response, I am genuinely interested in your views!
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
"There are other sources of difference:
i) whether Tube or solid state equipment is used during the transfer from the original master tape. Note that the original master tape is not always available and a subsequent LP reissue may be cut from the digital master (sometimes from CD...)
ii) whether the power amp used for the cutting head is tube or solid state. The distortion characteristics on overload are different and can contribute to the perceived differences.iii) Whether the tape azimuth is perfectly matched to the original record head
iv) Half -speed mastering can improve the sound by reducing the HF losses from the playback head at the expense of low end. Tonally this will change the subjective tonal balance of a recording."
All true, but my comments are coming both from experience with commercial recordings and also in the studio, where these kinds of variables are constant. Having said that, it's also much more general an issue than one recording or another where the above might apply.
"v) W/F is not zero for the cutting lathe and is independent of the W/F of the tape machine so HAS to be factored in. Given that this alters the waveform in the groove, by definition the waveform is now different to the master tape therefore cannot be deemed to be "as faithful" to the original."
Also true, by definition, but again at the levels in high-quality equipment this simply isn't an issue that shows up in listening (IMO). It certainly isn't part of why I think vinyl sounds more tonally accurate.
"From your description of cymbals and brass etc, would you say that transient response is one of the areas that you feel is better handled with analogue? By transient response I mean the attack and decay characteristics?"
Hmm, kind of a tough one to answer for me. I don't think it has much to do with the tonal accuracy I'm talking about. In different respects I think each has advantages in this regard. The dynamic possibilities of digital cannot be matched by vinyl, however once again I do think for music the dynamic range of vinyl is more than adequate, but the boom of a bass drum can come across more dramatically in digital. I do think, however, that vinyl conveys the physical aspects of the sound better - sticks hitting cymbals, the bow contacting the strings, etc. I do relate that to the physical nature of the playback, and it certainly could be argued that it is a "creation" of the vinyl playback that isn't on the tape, but in the end the result is subjectively closer to the sound of those instruments being played live (again, IMO). That's a separate thing from the tonal thing though, and I'm guessing with respect to the cause, but I do think it's a reasonable guess at that.
"Also, would it be fair to say that part of your viewpoint is based on a suspicion of the quantisation process itself? "
Of course I can't say it couldn't be, but honestly, this is the result of years of experiences and living in a digital world where the vast majority of new recording is digital by default. I really don't think I am forming my opinion based on what I think is going on, I really have no idea what is going on, but I do have a good amount of experience with live instruments, digital recording and more limited experience with analog recording in addition to listening to commercial recordings. I have and have heard some excellent digital recordings, and there is no doubt that there are a number of things that digital does far superior to analog. I'm just not sure it's all related to "accuracy" or that "faithful to the recording" is really relevant, and subjectively by the things I care about in sound analog sounds more "accurate" to my ears, in general.
There are other factors, most notably recording techniques as those of the "analog era" are rather dramatcially different to modern techniques, and so it is rare to get an opportunity to hear a modern digital recording recorded the way analog recordings were done in order to make a proper comparison. The other way around is a little less uncommon, but even then a direct comparison is usually not possible.
Having said all of this, I'm going to get out some recent digital string recordings and give them a listen!
Dave
Edits: 05/26/15
I agree with you about the emotional involvement one gets when listening to analogue. I remember talking to one recording engineer who said that in the early days of digital recording that the replayed master sounded quite removed from the original monitored version.
I believe this and some of your experience may be attributed to jitter on the reference master clock used for the A/D stage. Any jitter on the reference clock will forever "contaminate" the coded signal and can never be removed subsequently. Hence, irrespective of the quality of the playback DAC etc, I can understand some of your perception.
I have used the Apogee Big Ben, which is already very good, but acquired the Grimm CC1 some years ago. The CC1 is claimed to have the lowest levels of jitter possible. However, I have no way of verifying the claim with my test equipment. Suffice it to say, the result of using the CC1 is to give a very "analogue" quality to the sound - totally grain free and preserved timbre of instruments.
I note that you don't permit emails. However, if you are interested, drop me an email and I can share sample files (on Dropbox) of vinyl transcriptions for you to assess.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
I think what you are describing could explain a particular recording, but not the entirety of my experience.
I didn't realize about the email settings, I just changed them, I'd love to hear your examples. If the email isn't working pm me and I'll send you my email.
Dave
FWIW, with my modest equipment, the vinyl LP is still superior to any digital format I've tried, including SACD and DVD Audio, in certain regards. at least with the acoustic classical music I listen to (and play myself).
A friend of mine bought a 1970s professional half-track, 15 ips tape deck from a thrift shop and had it rebuilt, total cost around $600. He then bought a few tapes (for $300 each, but with the tiny quantities they sell, you can't expect mass market prices) from an outfit called The Tape Project, including one that is very familiar to me -- Sonny Rollins' Saxophone Colossus.
Consumer digital audio, at least modestly priced consumer audio, has a long way to go to match that sound.
I believe it.
I'm not trying to knock digital, good digital can be excellent, but it has different advantages than analog/vinyl. Were it simply a matter of one being "accurate" and the other not, it should be easy to recreate the sound of the "inaccurate", but you can add all kinds of distortions to a digital file and it won't sound like a vinyl record, nor will it have any of the positive qualities of a good LP.
Dave
. . . many years ago, I recall tests that were done where LP's were recorded onto CD's and then both were compared. Listeners could not tell the difference between the original vinyl and the CD copy.
Sit a person in front of a system and see if they can listen to a whole recording or concert that way. That is the system one wants.
I'm not a fan of double-blind tests, useful for some things, not so much for others. For a manufacturer making dollars and cents decisions about customer behavior, probably quite useful. For me as a listener, not so much as they are not me and I don't listen to music as in a double-blind test.
Dave
> I don't listen to music as in a double-blind test.
The purpose of a double blind test is to prove that you can hear a difference between two components or two recordings. If you can accurately differentiate one from the other with statistical significance, that is proof that a difference between the two must exist. If your selection between the two comes out in a random fashion, the likelihood is that no difference exists. It doesn't necessarily prove there is no difference, but it certainly suggests that you can't reliably hear any difference.
Anyway, if you can't hear any difference between an LP and its digital copy, then you should be able to derive just as much pleasure from listening to either. On the other hand, if knowing that you're listening to the digital copy gives you less pleasure, then you might find that listening in a double blind situation allows you to derive just as much pleasure either way. Therefore, if you cannot pass a double blind test between analog and digital, you will probably get just as much pleasure out of listening to digital as long as you think you're listening to analog. ;-)
Of course, as you say, you don't listen to music as in a double blind situation.
As for me, I enjoy listening to digital copies of LPs as much as I enjoy listening to the real thing because I can't hear the difference even when I know which is playing.
Best regards,
John Elison
"If your selection between the two comes out in a random fashion, the likelihood is that no difference exists. It doesn't necessarily prove there is no difference, but it certainly suggests that you can't reliably hear any difference."
...in the context of a DBT. Again, not how I listen and so not terribly relevant to me. DBT are useful with respect to short-term listening - buying decisions etc. Not to long-term listening with many repeated listening sessions over a long period of time. Many times I've bought something or began using something because I thought it was either "better" or "good enough" only to find after repeated listening that I didn't continue to have the same experience. Long-term impressions are what I'm concerned with in terms of listening and enjoying music.
Dave
Is double blind testing the basis of all of your system choices? Can you consistently identify each component? Pressings?
Dave
> Is double blind testing the basis of all of your system choices?
No! None of my system choices have been based on double blind testing or any other kind of blind testing. In fact, I've never participated in a double blind test. However, I have made hundreds of digital copies of my vinyl LPs and I listen to digital copies of LPs much more frequently than I listen to LPs played on a turntable. The copies sound the same to me.
Best regards,
John Elison
Just wondering that if I get my favorite LP music and put it on digital media, maybe I would listen to it more often.
One thing that bothers me is that the results come out different with each play, for both digital AND analog!
With LP playback (being mechanical in nature), it is distinctly possible that things may sound different from listening session to listening session given the variations in ambient temperature etc.
This is far less likely with digital playback unless the electronics in your particular are prone to drift even after the internal case temperature has stabilised. Normally electronic drift is irrelevant after a few minutes from switch on.
Any other variations you hear will be undoubtedly related to mood...and the number of drinks consumed since the start of the listening session! (In my case....)
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
> One thing that bothers me is that the results come out different with each play, for both digital AND analog!
I don't think that has any bearing on digital recording of vinyl. Of course, I don't hear this in my system, but if that's what you hear, how would digitizing vinyl have any bearing on it? ;-)
I find this too (in a subtle way to be sure), but I've always chalked it up to factors such as changes in listening position (maybe I'm slouching in my chair!), temperature, humidity, my own psychology. . . ;-)
That's fine, not my experience. Maybe I'll record some more and try it again.
Dave
I've made this comparison many times -- not necessarily with an LP transferred to a CD, but with a CD made and remastered with the same original master tapes used to make the LP. That is a much more fair comparison of the two formats imo. And the CD is sometimes vastly superior, sometimes vastly inferior, to the original LP.
IME a well-made LP will generally have certain advantages over a CD or SACD made from the same original (analog) master tapes, even though it may be worse in other ways.
Of course, all of that is a different kettle of fish from the latest high-rez digital made from today's digital masters, which you seem to listen to much of the time. There, the advantages of digital audio are especially pronounced, again IMO.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: