Crosstalk: digital more like Vinyl?

[Phelonious Ponk]

Can someone link me to J J's paper? It's getting referenced a lot here these days and I've lost track of where, in what thread, it was linked.

Tim

 

[jkeny]

There is less HF energy, less LF energy in vinyl.

That isn't the case theoretically, is it? How does this occur in practise then assuming that the RIAA playback curve is accurately adhered to?

There certainly is the things mentioned above about crosstalk and intensity differences that are frequency dependent, and as such, less lows and highs is a sort of compression WHEN compared to what digital is capable of. I hear distortions in the low bass in vinyl, but the interactions of crosstalk and intensity and L/R issues make a more pleasant sounding mid to HF (what limited amount compared to digital there is) sound field. And I am talking playback here, not concerned about what is theoretically possible at the cutter but what is the reality at the consumer.

I'm not so sure about the compression that is mentioned but yes, the mid & highs is where vinyl sings.

 

[jkeny]

BTW, at higher frequencies ITD is still a factor although the information is encoded in the envelope rather than in the carrier. So don't make the mistake of discounting ITD.

 

[f1eng]

In principle, if one had the transfer function of a given turntable/arm/cartridge/phono stage in its environment (i.e. including all mechanical and acoustic effects) it would be possible to completely replicate its sound using DSP. The problem would be actually getting the transfer function and whether there is a DSP engine powerful enough to run the resulting algorithm in real time.

But what is the point? Analogistas either wouldn't believe it was happening or that the digital must be screwing it up in some as-yet-undiscovered-by-man way, and you can get the effect you seek by running an LP rig tuned to your taste anyway and it would be reassuringly more expensive

My experiments show a lot of the nice sound of LPs is generated in the replay system rather than revealed by it, but that doesn't stop it sounding really nice and enjoyable to me...

 

[esldude]

Are you saying that a small amount of compression that slightly squashes dynamics has the effect of perceptually increasing the dynamics? That's an interesting idea - is there any background support for it?

Yes, that is what I am saying, and have read research on the matter. I don't have a link to provide at the moment. I didn't read it online anyway, but in texts on hearing.

 

[f1eng]

Yes, that is what I am saying, and have read research on the matter. I don't have a link to provide at the moment. I didn't read it online anyway, but in texts on hearing.

It is a long time since I was in the business (left in 1976) but low level detail was always amplified when cutting the master for LP to keep it away from the surface noise. I don't know about dynamics but I would expect that is why the low level detail is more obvious on LPs, it really is louder relative too the average sound of the rest of the music.
CDs tend to be limited by soft (or often not so soft) limiting the high levels which does not have this benefit.

 

[RogerD]

It is a long time since I was in the business (left in 1976) but low level detail was always amplified when cutting the master for LP to keep it away from the surface noise. I don't know about dynamics but I would expect that is why the low level detail is more obvious on LPs, it really is louder relative too the average sound of the rest of the music.
CDs tend to be limited by soft (or often not so soft) limiting the high levels which does not have this benefit.

Ok, I can agree with that. In my system at least,the digital version always has a smaller perimeter or a different perspective,but mostly it seems 'letterboxed" to me versus the analog copy. It seems like the analog version is not limited in this way and whether the original was recorded tube or solid state that doesn't matter. This perimeter is especially noticeable between AAD and DDD copies. It is not psychoacoustic related,it is fundamental betwen the two formats,I think. What would cause this? I don't think it is distortion related at all.

 

[jkeny]

Yes, that is what I am saying, and have read research on the matter. I don't have a link to provide at the moment. I didn't read it online anyway, but in texts on hearing.

OK, thanks, if you happen to remember the texts, I would appreciate the titles - I'm interested in reading them.

 

[jkeny]

It is a long time since I was in the business (left in 1976) but low level detail was always amplified when cutting the master for LP to keep it away from the surface noise. I don't know about dynamics but I would expect that is why the low level detail is more obvious on LPs, it really is louder relative too the average sound of the rest of the music.
CDs tend to be limited by soft (or often not so soft) limiting the high levels which does not have this benefit.

Right, I didn't know this. So we are getting a signal which has been conditioned in more ways than one (RIAA curves) to compensate for the nature of the playback system & delivery substrate. Very interesting. I've never seen this info before (but that's no surprise - I haven't been in the mastering business).

 

[FrantzM]

It is a long time since I was in the business (left in 1976) but low level detail was always amplified when cutting the master for LP to keep it away from the surface noise. I don't know about dynamics but I would expect that is why the low level detail is more obvious on LPs, it really is louder relative too the average sound of the rest of the music.
CDs tend to be limited by soft (or often not so soft) limiting the high levels which does not have this benefit.

Ok, I can agree with that. In my system at least,the digital version always has a smaller perimeter or a different perspective,but mostly it seems 'letterboxed" to me versus the analog copy. It seems like the analog version is not limited in this way and whether the original was recorded tube or solid state that doesn't matter. This perimeter is especially noticeable between AAD and DDD copies. It is not psychoacoustic related,it is fundamental betwen the two formats,I think. What would cause this? I don't think it is distortion related at all.

f1eng

I agree with you .

@roger

How can you be absolutely be certain that it is not a matter of mastering? AFAIK there are few if any digital and analog with exact same mastering . I actually would like to be pointed toward such. Additionally and there I know I am threading a sensitive issue.. Under what conditions? I can safely bet knowledge wasn't removed.

 

[jkeny]

Ok, I can agree with that. In my system at least,the digital version always has a smaller perimeter or a different perspective,but mostly it seems 'letterboxed" to me versus the analog copy. It seems like the analog version is not limited in this way and whether the original was recorded tube or solid state that doesn't matter. This perimeter is especially noticeable between AAD and DDD copies. It is not psychoacoustic related,it is fundamental betwen the two formats,I think. What would cause this? I don't think it is distortion related at all.

I just wanted to pick up on this highlighted phrase as I see it used often as meaning that "it's not a delusion". I find this a confusion of what the term "psychoacoustics" means & leads possibly to some confusion. I believe that psychoacoustics is the study of the perception of hearing more focussed on the signal processing from the auditory nerve onwards. It doesn't mean "delusional". Sorry, I just wante dto clear that u if that's what you meant?

 

[RogerD]

f1eng

I agree with you .

@roger

How can you be absolutely be certain that it is not a matter of mastering? AFAIK there are few if any digital and analog with exact same mastering . I actually would like to be pointed toward such. Additionally and there I know I am threading a sensitive issue.. Under what conditions? I can safely bet knowledge wasn't removed.

Hi Frantz,

It is consistent or far too noticable im my experience. I am not saying digital is inferior,just that this difference effects the sonic presentation of the material. I'm just looking for a explanation. I don't think I am alone in this. I think it has to do with bandwith or linearity of the process. Maybe it's a classic approach of not measuring the right thing. Anyway I didn't mean to hijack this thread,but it is about digital sounding like analog.

 

[RogerD]

I just wanted to pick up on this highlighted phrase as I see it used often as meaning that "it's not a delusion". I find this a confusion of what the term "psychoacoustics" means & leads possibly to some confusion. I believe that psychoacoustics is the study of the perception of hearing more focussed on the signal processing from the auditory nerve onwards. It doesn't mean "delusional". Sorry, I just wante dto clear that u if that's what you meant?

I meant it is(I think)a attribute or inherent of the format. Being delusional is not a description I would use. Now if I'm the only one that hears this,well I won't go there.

 

[jkeny]

The needle can not follow a high energy groove, its mechanics, digital is not restrained by mechanics,,,,and I talk of the playback consumer side. You can cut a disc with some pretty high energy at say 15Khz but no needle will track it.

I know that - that is the RIAA curve I mentioned but it only deals with attenuation & boosting signals based on frequency, not signal level i.e there is no boost for low signal levels contained in the RIAA spec, is there? I know there is a calculation done about how tightly the grooves can approach one another & this can dynamically alter this depending on the signal level & frequency but AFAIK it was not changing the signal itself just how it was cut?

The system in that way is compressed so the consumer can play it back, when cutting you have POWER AMPS pushing the cut blade sideways but you got to get some recording time on the record to hear more than one song. Speed of the record and width of grooves (energy) is the mechanical limiting factors, fixed by the rpm and play length of the record. Long play records, developed at a slower speed (33.3) result in less energy as well so you can put more songs on it.

I think I'm saying the same thing above as you are here?

And do some research in the difference between a mono cut record and how we use one needle now to convey two channels and you will begin to see why analog via vinyl is absolutely different than via digital. Yes, mids and highs "sound" good, and sometimes better, and it is due to the record, while lows and low mids are not up to digital and it goes back to the cross effects with frequency etc on a record due to the cartridge and the "stereo" cut.

 

[jkeny]

This would seem to be an interesting Ph. D thesis on this subject that I just came across from a guy in Trinity College Dublin (my home town) "The Composition and Performance of Spatial Music"

He has this to say

You are also basically correct in that a phantom source
> will be a little broader than a real source, largely due to the fact that
> the localization of phantom sources is somewhat frequency dependent.
> Straightforward amplitude panning will produce a slight discrepancy between
> the localization at low and high frequency ranges, which leads to a slight
> change in timbre (manifests as a slight spectral dip at approx. 2kHz) and
> an increase in the Apparent Source Width (ASW) of the source when it is
> reproduced as a phantom image.
>
> This is a big problem when dynamically panning a sound as this change in
> timbre tends to highlight the loudspeakers in a negative way and it
> significantly reduces the smoothness of the spatial trajectory
> (incidentally, one the main goals of the Ambisonics spatialization system
> was to eliminate this issue, which it largely does but not without some
> costs).
>
> Ville Pullki has done a lot of research in this area, particularly in
> terms of his amplitude panning system VBAP. He has also conducted research
> on panning using three loudspeakers (again with VBAP). My recollection is
> that this will increase the ASW even further, and similarly reduce
> localization accuracy. Of course there is a strong relationship between the
> two factors, and the concept of locatedness is useful in this regard.
>
> Anyway, chapter 6 my PhD thesis has lots of specific references in terms
> of listening tests with phantom sources and the relationship between
> localization accuracy and apparent source width, both for stereophony and
> ambisonics. You can find it here -> http://endabates.net/academic.html

 

[RogerD]

This would seem to be an interesting Ph. D thesis on this subject that I just came across from a guy in Trinity College Dublin (my home town) "The Composition and Performance of Spatial Music"

He has this to say

Meat and potatoes......thank you.

 

[slowGEEZR]

So, does vinyl sound better using cartridges with more crosstalk?

 

[Atmasphere]

Can't speak for j_j, but it would be my opinion that low level details are more apparent due to some slight compression needed for making a pressed record playable. Small amounts of compression elevate moderately low level details and increase apparent dynamics without being noticed as such. Not the whole answer mind you just a part of it.

Most of the time you don't have to compress the signal. So far in all the mastering we have done, we have found that what you really need to do is look at the recording and see if there are any trouble spots. A lot of the time you can increase groove depth slightly or decrease the overall signal by a db and avoid and limiting or compression altogether. Compression is/was often used though because record labels didn't want to pay the mastering engineer for the time to do this. These days recordings are often compressed because digital sounds best when you can use as many bits as are available- the normalizing process is used for the same reason. If you make an LP from such files, they have the same compression as the CD.

There is less HF energy, less LF energy in vinyl. There certainly is the things mentioned above about crosstalk and intensity differences that are frequency dependent, and as such, less lows and highs is a sort of compression WHEN compared to what digital is capable of. I hear distortions in the low bass in vinyl, but the interactions of crosstalk and intensity and L/R issues make a more pleasant sounding mid to HF (what limited amount compared to digital there is) sound field. And I am talking playback here, not concerned about what is theoretically possible at the cutter but what is the reality at the consumer.

I have to take issue with this- we can record easily enough at 30KHz and play it back on some pretty pedestrian equipment (Technics Sl1200/ Grado Gold cart/ HK 430 receiver). The reason the LP seems to have less energy is that there is less phase shift due to filter effects (digital being bandwidth limited by comparison...); the human ear often converts phase shift problems into toanlity- I suspect that is what you are really hearing.

Right, I didn't know this. So we are getting a signal which has been conditioned in more ways than one (RIAA curves) to compensate for the nature of the playback system & delivery substrate. Very interesting. I've never seen this info before (but that's no surprise - I haven't been in the mastering business).

The only EQ involved is the RIAA. There is no process that equalizes low level signals differently from high level signals and this has nothing to do with the substrate. Compression of course will make quiet passages louder...

The needle can not follow a high energy groove, its mechanics, digital is not restrained by mechanics,,,,and I talk of the playback consumer side. You can cut a disc with some pretty high energy at say 15Khz but no needle will track it. The system in that way is compressed so the consumer can play it back, when cutting you have POWER AMPS pushing the cut blade sideways but you got to get some recording time on the record to hear more than one song. Speed of the record and width of grooves (energy) is the mechanical limiting factors, fixed by the rpm and play length of the record. Long play records, developed at a slower speed (33.3) result in less energy as well so you can put more songs on it. And do some research in the difference between a mono cut record and how we use one needle now to convey two channels and you will begin to see why analog via vinyl is absolutely different than via digital. Yes, mids and highs "sound" good, and sometimes better, and it is due to the record, while lows and low mids are not up to digital and it goes back to the cross effects with freqeucny etc on a record due to the cartridge and the "stereo" cut.

Again, I think you are working with some mythology here. Think back to the 1970s when CD-4 LPs were cut- they used a 50KHz carrier that was modulated in FM stereo for the additional 2 channels. After that, RCA developed a color video disk media that used a needle to track the signal- to do color you need 2MHz response. Almost any LP system has 30KHz bandwidth between record and playback and some, between record and playback exceed 40KHz. I don't know of any CD format that has got that sort of routine bandwidth yet. On top of that, LF bandwidth goes well below 20Hz. Today we recorded a direct-to-disk session (will be on Nero's Neptune, a Twin Cities local label); we were seeing information recorded in the grooves caused by a train passing the studio that was only about 8 Hz. I had no idea the mics could go that low... we don' t get this sort of bandwidth on the analog tape!

At any rate I find that while digital is supposed to be better in the bass, in practice I get better bass with the LP of the same recording. A lot has to do with how well your tonearm can track the cartridge, how dead the platter pad is relative to the LP, and how dead the plinth and platter are. If there are problems in any of these areas, bass playback will suffer.

 

[jkeny]

Thanks for clear up the mythologies, Atmasphere - it's great to have real experts on the forum - helps to avoid going down blind alleys

So compression is/was used due to laziness, economics &/or bad technique - I can believe that. But I never heard that small amount of compression would perceptually increase the dynamic range. Have you experience of this?

 

[f1eng]

Ok, I can agree with that. In my system at least,the digital version always has a smaller perimeter or a different perspective,but mostly it seems 'letterboxed" to me versus the analog copy. It seems like the analog version is not limited in this way and whether the original was recorded tube or solid state that doesn't matter. This perimeter is especially noticeable between AAD and DDD copies. It is not psychoacoustic related,it is fundamental betwen the two formats,I think. What would cause this? I don't think it is distortion related at all.

OK Roger, there is no doubt that digital is "letterboxed" in as much as the frequency response is defined by the sampling frequency and, in principle the signal-to-noise ratio is defined by the bits, although with noise shaping and high sampling rates the noise can be moved away from the audible range increasing the signal to noise ratio.
CD is "letterboxed" such that the minimum frequency is DC and the maximum 22.05kHz. The signal to noise ratio is 96dB.
Analogue systems have fuzzy boundaries, but boundaries none the less. Since few people use reel-to-reel tape these days we are pretty well stuck with LP and FM radio for high quality analogue listening. Here in my home the highest quality analogue sound I ever hear is, and always has been, live BBC concert broadcasts. There used to be one at 7:30 every evening back in the day, and I used to tape them all for my education, keeping those I liked and recording over those I did not. FM stereo carries the difference information on a subcarrier and is limited by the signal to noise ratio of the broadcast medium such that its "letterbox" is slightly fuzzy, but is smaller that the CD letterbox. A parcel which would just fit through the FM stereo letterbox would easily fit through the CD letterbox with plenty of room all round.
With LPs it is less well defined and more fuzzy.
Firstly there is the limitations of the manufacturing process. The best engineers always pushed the limits here but plenty took it easy and did not! Firstly, in order to have a continuous groove the bass must be monoed. Low levels of bass can just about be cut on a record but with the risk of wavering stereo imagery (at what point does bass really become non directional?) it may as well all be monoed. Secondly no cutters can record high frequencies at maximum level. If the master is analogue that is OK since neither can tape recorders, but otherwise some sort of limiting is necessary.
Some otherwise excellent cartridges don't track high levels too well, so the maximum level one should cut on a disc is limited (not that everybody does so and there are several discs which only a very few pickups can track, remember the famous Telarc 1812?). Pickup cartridges are seismic vibration transducers and their lowest frequency of accurate transduction is when the cartridge body is first stationary , which will be at around 2x the natural frequeny of the arm/cartridge mass on the cartridge suspension depending on damping. This is very varied from one installation to another. Cartridges were made for quadraphonics which could trace a high frequency carrier, but producing them with consistent performance was one of the reasons the system never worked properly.
When I used Bruel and Kjaer 20Hz to 20kHz frequency sweep records for measurements B&K only advised them to be used 5 times for precise measurement because of wear at high frequency. One component which has massive variation from one design to another is the cartridge. Some of the finest current cartridges roll off below 20kHz, some well below. Some others have resonant peaks in the last audio octave or at above audible frequencies. Most have very high distortion in the top octave. This means that some LP decks won't have much output above 15kHz but it will be accurate and some will have lots and it won't. Most of that highest frequency stuff will be resonant peakiness and/or distortion products.
When Meridian measured the dynamic range of LPs many years ago (somebody from Meridian please correct me if I have remembered this wrong) the highest they found was the equivalent of 11-bit. Most were much less. I have measured most of the LPs I enjoy listening to still and 50dB is the most I have on any of them.
So the LP "letterbox" is much fuzzier than the FM one but yet again, with the exception of a few cartridges and LPs with a fuzzy spike above 22.05kHz, will fit through a CD letterbox without touching the sides.

Now I am also somebody who enjoys the sound of LPs at home, and believes that they sound great. I do not believe that they are superior in any as-yet-to-be-discovered-by-man technical way. Nor can I swallow any of the frequently re-stated erroneous stories about shortcomings of digital recording. Repeating the same mistake thousands/millions of times will never make it correct. No, digital recording isn't perfect, but it is very much nearer to being perfect than any analogue method I have used.
With an expert engineer, frequent re-calibration and carefully set levels it is possible to produce a really nice sounding recording on tape, but IME one can -always- tell the off-tape sound from the microphone feed. With a modestly priced digital recording system such as a Metric Halo device it is close to impossible, on music, to tell the microphone feed from the output of the recorder. In fact, with my old ears I never have heard a difference on any of the music I have recorded on it. My Stelladat recorder was pretty well transparent almost 20 years ago, input to output.

The difference is that the analogue alterations, particularly modest overload, are euphonic. I know people who record a CD onto tape and prefer the sound from the tape to the original. This is obviously not a digital shortoming, rather a pleasant analogue addition.
Record players have the benefit that pretty well all the additions they create in their sound output, mistracking and speed fluctuation being exceptions, are euphonic.
Non suspended decks pick up a bit of environmental vibration via structure and air which is just like a bit of extra reverb. Even the finest pickup cartridges have midband distortion in the 1% to 5% range, mainly second harmonic, and only the most deranged would claim that this level of distortion is inaudible

So I would say that the sound quality of record players can be explained by existing techical knowledge, most of which was alredy well known and understood 50 years ago.

My belief is that the reason many modern recordings sound bright, harsh, unpleasant, what-have-you is down to microphone choice and positioning, not the method of recording/distribution.
Back in the day we had 2 large sensitive microphones positioned almost as far from the performers as the audience would be at a concert. The sound balance was largely achieved by moving microphones and performers. The sensitive microphones had large capsules and rolled off pretty early, so didn't sound harsh. They weren't shoved right against the instrument either.
Nowadays multiple super high bandwidth microphones are shoved right against the performers or instruments to be mixed up later. What have we lost? A natural treble roll-off. Phase coherence. A natural acoustic. What have we we gained? A totally un-natural perspective. Un-naturally high levels at high frequencies. A mish-mash of mixed together manipulated phase inconsistent mess. Harshness.

Just IMHO.