Why does vinyl sound better?

[SteveFord 1,119]

I've given this matter a lot of thought and my best (?) idea is that the rise and decay of notes is slower in analogue than it is in digital which is the main reason why CDs always sound a little off to me.

Am I more full of it than usual?

[Balok 1,667]

I used the sound of "decaying notes" to select my new CD player. Particularly the sound of piano. THe ARCAM CD17 was simply better than the Cambridge Audio 640. I think LPs frequently sound better because they appear to have better high frequency which is important for detail and imaging. I think newer CD players are quite a bit better than those of even 6 years ago. Better highs, better front to back imaging, better "decay" etc. Then there is jitter, whatever the hell that is.

Bottom line, i don't really know what i am talking about, i just haven't been posting enough lately and I was feeling bad about it.:ddd

[TNRabbit 371]

 

 

I've given this matter a lot of thought and my best (?) idea is that the rise and decay of notes is slower in analogue than it is in digital which is the main reason why CDs always sound a little off to me.

Am I more full of it than usual?

 

Nope, same 'ole amount of "full of it"...

[SteveFord 1,119]

At least I'm consistent!

[stereo_dog 1,220]

 

 

I've given this matter a lot of thought and my best (?) idea is that the rise and decay of notes is slower in analogue than it is in digital which is the main reason why CDs always sound a little off to me.

Am I more full of it than usual?

 

You may be full of it, but you make sense... it MUST be early! :dd

[snarffydoggy 1]

Well I gotta be honest, I don't have a really good vinyl system or a really good cd player to compare them. Maybe some time in the future. What I like most about the CD format is that you can make back ups etc. :dd :dd :dd

[Radioeng2 33]

I think it's an interesting question Steve. Don't know about faster or slower. Digital represents the waveform with it's stairstep pattern, so I don' t know that it could make it faster. And if there was a much difference then you'd think that going back and forth to digital would show it's self up as inaccurate. But digital always seems to be correct....just not quite as real. Whatever real is...

 

I agree that the highs of digital are a flaw and that the newer results are better. The increase smoothness of the different approach to the filtering has helped. And certainly the faster bit rate and wider bandwidth has made it more "analog".

 

But remember that as long as your physically setup well on a TT, that the bandwidth is very good and goes up way beyond the 16 bit bw limitation. Maybe the higher harmonics of naturally recorded music produce beat notes and actually add to the realism of sounds when present. The 16 bit digital, with its cut off, would lack those.

 

And I'm a big believer in the down deep details really adding to the realism of music. We hear this with the subtle changes in a system that shouldn't make much difference, but can make you sit up and notice when you alter some aspect. The many db down sounds get greatly altered with 16 bit as the steps of the digital get way down. The encoding will assign it either one level or the other, and its not entirely accurate. This is audibly better with 24 bit. This is pretty evident in how the end of a note sounds...the decay details.

 

And then the jitter is an influence. Science says that jitter rates shouldn't have to be so incredibly low, but listeners can hear something wrong down to much lower levels than predicted. As I've said many times, the jitter is a big deal for digital data streams. The closest analog analogy is the TT sensativity to the type of platter drive mechanism. The direct drive versus the belt drive. The type of belt even which seems to influence the musical playback. These are kind of cousins to the jitter thing. In analog, it's kind of a solidness to the pitch where in digital its micro tiny timing of each littlest parts of the sound stage and notes.

 

Sometimes I think about how my co-workers still try to describe audio flaws with an analog vocabulary. Then there is the digital terms. But I'm now thinking that there is a third view of the musical event which is the little tiny details that seperate the played back event from the live event. We can tell it in an instant usually, but fail to have the words to be descriptive as to what it is. When we get beyond the more gross distortions, then this is where circuit topology, metal to metal connections, capacitor materials and so forth start to divorce our experience from reality.

 

Mark

[SteveFord 1,119]

Mark,

Do you have some sort of graphic display that will show us exactly what we're hearing on an analog passage and then the same thing on digital?

We're trying to describe something that perhaps we really need to see. Kind of like that old joke with the row of blind men describing an elephant - the one up at the trunk says it's like a huge python, the one holding the leg says it's like a tree, the one in the center says it's like a huge mountain, the one at the tail says that it's like a reed and the one at the very end in the pile of dung says that it's soft and mushy.

[snarffydoggy 1]

8>>> 8>>> 8>>> Steve I Think You Just Described it, All We have to do is determine which end is analog and which end is digital and they both meet in the middle:dd :dd :dd :dd :dd

[elgrau 89]

"Maybe the higher harmonics of naturally recorded music produce beat notes and actually add to the realism of sounds when present. The 16 bit digital, with its cut off, would lack those. "

 

Never really followed or agreed with that statement. The "beat notes" from the "higher harmonics" are in the audible range, correct? (As in the first harmonic of 24,000 Hz would be a 12,000 Hz "beat note"). So that is what is presented to the recording mike in the recording studio: the audible band of frequencies (including any "beat notes" from the higher frequencies of the musical instruments). So these "beat notes" would be present in 16 bit digital recording because (e.g.), 16 bit digital adequately samples a 12,000 hz "beat note" (or 1st harmonic of the 24,000 Hz HF "note")? No? Just because the 12,000 Hz "beat note" came from a harmonic of the 24,000 Hz "note" does not mean that digital media/recording with a c/o frequency of say 20,000 Hz would not pick it up (since the actual "beat note" is at 12,000 Hz). Now if you mean the actual 24,000 Hz "note" as having some subtle effect on what you hear (even if your hearing does not "register" any sounds above say 13,000 Hz), then that's a different (albeit most likely bogus) issue. But that's not what you seem to be saying.

[BillD 239]

Because all we can get on vinyl is what microphones (or other forms of pickups) pick up, and they don't have infinite frequency response (some of the best do get to 20kHz), all we're gonna get on vinyl is that. Now, what we get off the record might be something else altogether. Harmonics created by the needle in groove, etc. could influence what we hear, but it wasn't put on the record. The same is true with digital. Everything needs to be converted to digital from some pickup. The limits of that pickup are what we'll get. Digital playback will not "add" anything (except jitter), like a needle in groove could add useful harmonics (and rumble and other bad stuff).

[Radioeng2 33]

Good question Elg. To back up a little first. When we look for why digital sounds different than analog there has to be reasons that probably are several multiple parts. I'm suggesting that just one of those is related to the issue of why research says people can distinguish some difference, when speakers can produce sound way above the hearing range.

 

Since the wider bandwidth, higher data rates sound closer to analog, by about everybody acknowledgement, then it's not just that digital sounds worse because it is digitally encoded. Some of the differences between them are the larger steps in level between 16 and 20 or 24. Surely the linearity of the lower levels of that 16 bit are one of the biggest culprits. Another being that the filter can be done differently in the wider bandwidth formats along with better filtering by the sampling rate changes in 16 bit. And then the bandwidth of that frequency area above where we supposedly can hear.

 

It seems to be a possibility that while we don't hear the upper frequencies...that maybe it's things up there that wind up back into the audible range. This is really closely related issue to why a single tone or just two tones in an intermod test, don't reveal all that our ears can hear.

 

Just think of all the notes present in one smack of a cymbal. There is the main crash note. Several very strong notes actually that would make that up. The distance from the center to the edge. The overall diameter. The distance from the strike point to the edge which would actually shimmer back and forth. Then the strong notes that resonate against themselves and produce other beat notes. I would expect harmonics of many many multiples of the strongest notes. Wouldn't surprise me if a, let's say main area around 8K, would have harmonics to the tenth multiple or 80K. So if you could freeze on a audio spectrum analyzer a window of maybe 5 seconds, I'd bet on dozens of carriers that make up the main crash, the shimmering, floating beats that follow it along with all the naturally occurring harmonics and beats of them.

 

How could you seperate the naturally occurring, the microphone diaphragm artificial distortions, the recording medium, the playback electronic created products, the relatively high level of distortions created by the driver or even in the ear itself? Pretty tough, but you can change things somewhat in between, like using headphones versus a conventional speaker.

 

Yes, the originally created beats that fall low in frequency would be there to be recorded. As a matter of definition, harmonics as I used them, are multiples up, as distinguished from subharmonics going down. Beat notes refers to the difference between two notes, as in 2000hz and 2500hz having a beat note of 500hz. This would be, as you mention, the audible original event sound that would be recordable even in the 16 bit digital environment. I differ from Bill, a little bit at least, in that I believe mics do pickup in many cases to above the normal hearing range. They graph and spec to the hearing range, but even though it would be drooping, the response continues somewhat. Certainly some mic technologies differ in their ability from each other.

 

As I was trying to say in the earlier post, we tend to ignore details down in level at some point down from main levels. Let's arbitrarily say we think of information down 45-50db to be masked by the higher levels. This is a fundamental part of bit rate reduction decisions for instance. In the case of the mic, what if it is a higher harmonic down 20-30-40db from the hearing range that is the only focus of the graphs. Doesn't mean that this information isn't going onto tape, into the record groove and left off the 16 bit digital.

 

But there has to be some higher frequency stuff that makes it on analog and is not on the digital. The studies as to what actually is on records shows this. So there will be potential for those harmonics to play against other harmonics and against the lower information too.

 

I guess I come into this thinking in part from how you study a transmitter site for intermod products and the multiples and beats that result. It's the reason behind the blanking area around a transmitter site where large signals can overcome the receiver front end and just happen, even though not desirable.

 

In much the same light, I'd wonder if some of the otherwise undesirable distortions in the process don't wind up helping in that they work to replace some of the things that the technology doesn't correctly capture. Like using a tone control to correct for a system that's not flat, it's a distortion to correct it but still possibly desirable absent the original information.

 

So again, this is all a far down phenomena, if it exists at all past my imagination, but there has to be some level of various small things that make up the difference that Steve is asking of. Maybe I'm making it harder than it needs to be.

[Radioeng2 33]

 

 

Mark,

Do you have some sort of graphic display that will show us exactly what we're hearing on an analog passage and then the same thing on digital?

I don't know if we have the resolution to see that deep too well, Steve. There are test procedures in the labs that can show things, but I'm not that capable or knowledgable. But even when you can see some level of difference, it's hard to assign whats more responsible to do altering enough to be audible.

 

For instance, if you run a dual trace scope with one trace showing original audio and the other displaying FM broadcast modulated audio, you see an HUGE difference. Enough so, that it's hard to believe they sound as close to the same as they do. So that doesn't tell you as to what is to blame for the difference you can hear. In that case, bandwidth being limited to 15K is certainly part of the visual difference and noticably audible too. But as far as breaking it down to little fine differences instead of gross differences, it would be hard to say intermod products, pre-emphasis/de-emphasis error, L-R generator error...etc...

 

I have a audio band spectrum analyzer. As to whether you can do a lapse time of display which would tell one form of things, I don't know. Haven't had time to play with it much. It will do some X-Y things, which you might be able to do some difference displays that would tell some things.

 

Gross errors are easier to get a handle on. Little fine differences get so much tougher.

[OBI56 23]

Radioeng, your posts always fascinate me and get me to thinking .... usually a lot! Just thinking in hypothetical terms here, but is it possible that some sorts of gross errors or differences are barely noticeable by the human hearing process but many tiny errors or differences are much more noticeable of bothersome to the human hearing process? We tend to concentrate or assume that gross or large errors MUST be much more noticeable or bothersome than certain tiny ones which would seem to be the wrong assumption to make. I've noticed that the low frequency noise rumble) made by direct drive TTs is far more bothersome to me than that of belt drive units, even when the DD TT measure better than the BD TT; almost like it makes me feel a bit queasy. The old "if the numbers say its better,then it must be better" syndrome. Maybe the whole vinyl VS CD thing could be compared to the tube versus SS debate.

 

Any comments or ideas on this?

[PDR 1,194]

Marc.......kind of like sound masking at grocery stores. Where they play musack to hide the monotone hum of the freezers.

Its not untill they turn the musack off that you notice just how loud the freezers are......you would think no way I should hear that but once the

tunes come on the hum dissapears.......a lot of different low multi tones overiding loud monotones......you focus more on the many... :-k

 

Perry

 

 

[elgrau 89]

No offense, Radioeng, but your last post is (to me) a garbled rambling incomprehensible mass of goobly-gook (and I challenge anyone to make sense of it!). Kind of wish people would stop worshiping (and assigning "magical" properties to) old technology just because it is old. Like BillD said, the analog signal cut into a record groove can only contain what the mic picks up (and few mikes are good to anything close to 20K!). Ergo, vinyl cannot be any better at capturing "higher harmonics" than digital. That being said, there is perhaps aspects of "pure analog" that "sound better" then "digitized analog" (as even digital samples are converted into analog before amplification. And as an aside, the "stairstep" analogy you used to describe digital representation is bogus. The digititized signal is not "flat" over the "stairstep" and then jumps to the next level or "stairstep". Instead, the signal is sampled at discrete intervals and from this sampling, a DAC reconstructs the (very smooth) analog frequency content/signal). But let's stick to reality (as to why CERTAIN aspects of "pure analog" vinyl might sound better (or WORSE in some areas, e.g., dynamic range) than CERTAIN digital methods of recording music).

 

I let a lot of bogus and un-scientific stuff slide on this forum* (as it is just a forum afterall), but at some point the shit just gets too deep! As a recent example, someone recently posted that because their niece or whatever commented how much better a song sounded on the member's system as opposed to their own MP-3 player, that that was "proof" the "tweak" to their system that they had recently added was indeed the cause of this noticed difference in high fidelity(not the much more likely case that the niece was simply noticing that a full blown Carver powered system sounds much better than an MP-3 player!....). Please....simple logic, please.

 

add: went back and re-read your post just to make sure I did not miss anything....I can kind of see where you were coming from here and there, so perhaps my original critique was a bit too "harsh"; but the jist of my criticisms still stand. Too much of people's love of vinyl is more nostalga and worship of the past (as in old cars) than reality. Let's face it, unless you have over $10k to invest in analog systems/media like records, CD's (with the help of a good DAC or SACD, etc.) beat "records" hands down (everyone was gaga over CD's when they first came out - no more pop/crackle/snap and incredible dynamic range, etc.). But it's no longer in fashion to go gaga over CD's anymore. Too passe!

 

*should have said this forum and the "other" Carver forum, as I'm not sure which contained said example!

[RichP714 3,163]

IDK, I think it made sense.

 

another interesting tidbit:

 

http://www.cv.nrao.edu/~rfisher/SamplerErr/harmonics.html

 

 

[elgrau 89]

So where are the conclusions, points, "get off the stage" message? Lot of graphs and arm waving, but what's the point/message?

[snow 240]

Just curious what makes you think Vinyl sounds better?

 

 

 

REGARDS SNOW

[BillD 239]

Well, I'm a neutral party in all this. I've digitized all of my vinyl (24-bit 96kHz WAV) and converted it to Apple Lossless, and it sounds just like the vinyl to me (even with the clicks and pops I didn't process out). This would indicate to me that the real culprit in the CD sound is the engineer behind the console. Surely there were bad LPs made through poor mixdowns, but maybe not as many as there are CDs. Getting the most out of the vinyl medium was a craft, whereas there is little craft in making CDs (just make sure you don't clip, or overflow in digital terms). New vinyl releases are being meticulously made, also. And SACDs (and DVD-As) require a lot of attention because the medium is more expensive, and the artist/manufacturer decided that the work was worthy of the medium - especially if it's 5.1 or better.

 

For me, I'm glad the digital medium came along, especially now that sample rates and word lengths are up. The convenience of having my entire music library on a computer connected to my music system is great.

 

 

 

[RichP714 3,163]

 

 

So where are the conclusions' date=' points, "get off the stage" message? Lot of graphs and arm waving, but what's the point/message?[/quote']

 

What I got out of it was that adding dither can mask quantization noise......

 

Here's a sample (track 93 of Denon Audio Technical CD) of orchestra music recorded at -60dB. The disc has the same passage at -60, -40, -20 and 'normal' levels. You can HEAR the stairstepped quantization noise.

 

http://thecarversite.com/yetanotherforum/userfiles/Track No93.wav

 

 

[snow 240]

 

 

Well, I'm a neutral party in all this. I've digitized all of my vinyl (24-bit 96kHz WAV) and converted it to Apple Lossless, and it sounds just like the vinyl to me (even with the clicks and pops I didn't process out). This would indicate to me that the real culprit in the CD sound is the engineer behind the console. Surely there were bad LPs made through poor mixdowns, but maybe not as many as there are CDs. Getting the most out of the vinyl medium was a craft, whereas there is little craft in making CDs (just make sure you don't clip, or overflow in digital terms). New vinyl releases are being meticulously made, also. And SACDs (and DVD-As) require a lot of attention because the medium is more expensive, and the artist/manufacturer decided that the work was worthy of the medium - especially if it's 5.1 or better.

 

For me, I'm glad the digital medium came along, especially now that sample rates and word lengths are up. The convenience of having my entire music library on a computer connected to my music system is great.

 

 

Well to be honest I dont have a high end Vinyl rig and it would be unfair to compare the old record players I remember as a child to what is available today, I was over at a friends house here several months ago and he put on a record from the 70's and to be honest I was impressed it sounded much better than what I remember records sounding like. But I do think that digital formats will continue to improve and hopefully more studios will be more concerned about quality.

 

 

 

REGARDS SNOW

[SteveFord 1,119]

Crud, I've gotta run to work. I've been thinking along the same lines as Obi and I do believe that BillD is right in that the medium copies things so well that we can't tell the difference BUT something is screwed uip when it just goes to CD.

Our ears and vinyl are imperfect - maybe that's why digital recordings can sound odd? Could things be too precise and that's what the deal really is?

[Radioeng2 33]

 

 

Radioeng' date=' your posts always fascinate me and get me to thinking .... usually a lot![/quote']

Thank you OBI!! I only wish I could have as good information as you pass along!

 

 

 

Just thinking in hypothetical terms here' date=' but is it possible that some sorts of gross errors or differences are barely noticeable by the human hearing process but many tiny errors or differences are much more noticeable of bothersome to the human hearing process? We tend to concentrate or assume that gross or large errors MUST be much more noticeable or bothersome than certain tiny ones which would seem to be the wrong assumption to make.[/quote']

This makes me think of what we've noticed while working on building speaker systems and building crossovers for systems. You can at some point discover that you've left a notch, a hole, in response and have not readily have noticed it. But you can quickly hear some other things like phase errors.

 

 

 

The old "if the numbers say its better' date='then it must be better" syndrome.[/quote']

Oh, say it ain't so!! Surely nobody would think their ears to be wrong and excuse what they hear because the numbers seem to say differently!!

 

 

 

Any comments or ideas on this?

I agree with your comment about tt drive having effect on music in subtle ways. This discussion also makes me think about tape playback, how you can hear tape hiss, but quickly in you mind tune it out and instead focus on the music. Then you may not be even aware of it until the silence between cuts. Or in vinyl how the noise artifacts play on a different plane than the music and then how you can seperate it from the music. Have you noticed these things too??

 

Mark

[Radioeng2 33]

 

 

I've been thinking along the same lines as Obi and I do believe that BillD is right in that the medium copies things so well that we can't tell the difference BUT something is screwed uip when it just goes to CD.

Our ears and vinyl are imperfect - maybe that's why digital recordings can sound odd? Could things be too precise and that's what the deal really is?

Steve, I think that we have to be careful about assigning the masses views onto audio systems. Like that "tubes are about soft clipping and that's all we would listen to them for" kind of thing.

 

Your comment about slow rise and fall times with analog sounds to me like that. Analog audio, when well done is very good and very accurate. While you can certainly say that vinyl is imperfect (all electronic pb could be said to be), it's still closer to real than digital and we hear it that way. Digital is digitally precise, just tough to get as naturally acceptable to the ear.

 

When you take some very well recorded music and listen to it at a higher level and just let yourself be immersed in the sound. Then switch from CD to the same thing on vinyl. You can very quickly feel the ability to relax into the analog playback and tell how more naturally "right" it is.

 

Wynton Marsalis, The Death of Jazz cut from his Majesty of the Blues release is a really good recording that we've used many a time and illustrates this, oh so well! (trivia note...the next cut is a sermon by none other than Obama's shame on America preacher!!) The Death of Jazz is one of the best recorded cuts and extremely musical at the same time I've ever heard. And you can find it on both formats and you don't need an extreme analog rig to hear how the quality thing plays out!