"Digital ... don't got no balls!!!."

[ben.]

Well, the guy who really pushed for an answer to the question of what vinyl does objectively better than CDs hasn't noticed. Maybe later...

As I've always said since joining this forum, these measurements only take you so far, and sometimes in the wrong direction. As an engineer it will be very important to understnad what they mean, as well as their limitations as to the corresponding perception.

I'm not looking for a mediium that's "correct". I'd rather it be "right".

Who, a properly calibrated pro deck at 15ips is pretty flat down to 20 Hz (distortion is another story, but it's pretty benign distortion) and below if the heads are in good condition. Folks that liked to run at 30ips would run into uneven response due to "headbumps", which could be overcome by changing to a head with gaps optimized for 30ips. Of course, much depends on the tape itself. There used to be options...

[sfogg]

" other phase relationships are created."

Exactly... it is *created*, in other words added by the medium. Phase distortion by vinyl in other words. That that also happens to be a pleasing sounding artifact is one of the reasons some prefer the 'sound' of vinyl.

By any objective test of phase coherence to the source digital wins.

"That is where the advantage of analog lies and why analog produces a better soundstage image, phasing."

Specifically a *roughly* 45 degree phase shift (though it varies and isn't constant during playback) between L and R on playback. That phase shift acts as a short of low level crude 'cross talk cancelation' circuit which acts to slightly cancel some of the sound from the left speaker at your right ear and vice versa with your other ear. It in effect increases the channel seperation at your ears. That makes the soundstaging and depth larger and why a CD can sound flat in comparison. For a similiar effect hold a cushion off your couch vertically up to your face so it acts as a barrier to your ears. The soundstage will grow/deepen fairly dramatically if you try this.

Add the same phase change to digital (and some have done this... specifically Carver with his 'Digital Time Lens') and you get the same sort of effect to the soundstage. And once one understands this they can take it a lot further by actively trying cross talk cancelation circuitry.

"Instead "new" digital formats are constantly being developed in an attempt to equal vinyl reproduction."

The new digital formats were developed because the patents on CD were expiring. Sony/Philips made money on every CD (and every device that could play them) ever created as everyone paid them liscensing fees. Once the patents expired those revenue streams expired too. New formats have new patents and new liscensing....

Shawn

[ben.]

Good points & information as always, Shawn.

I think that this added phase incoherence may contribute to the perception that well done vinyl playback is more lifelike, as there's nothing in our daily lives that is really phase coherent...

[JJkizak]

Without good oscilloscopes there would be no vinyl or digital anything.

JJK

[garymd]

And for what it's worth, I'm dying to hear something that sounds better than the best of digital. Greece is a bit far of a trip, but if I ever make it over to the east coast I know there are a bunch of guys dying to prove me wrong []

(hey, maybe I can catch a ride after the pilgrimage and then just fly back to Indy...with student discounts a single one-way flight should be rather cheap).

Hey Doc,

I might be driving to Indy this year. You can keep me company on the ride home and fly back from Baltimore. It would be a gas.[]

[sfogg]

Ben,

"I think that this added phase incoherence may contribute to the perception that well done vinyl playback is more lifelike"

It can as it imparts that sense of depth to the playback in two channel. The wandering phase also likely imparts a little 'air' to the high end since it makes everything more diffuse up there to give it that shimmer.

Of course if phasing between L/R gets too bad things will sound lousy. For example wiring one speaker out of phase with the other (180 degrees out of phase) sounds very bad though strongly demonstrates the diffusion effect phase can have.

" as there's nothing in our daily lives that is really phase coherent... "

Any actual instrument is the reference point... it is creating the sound/phase so it is of course 'phase coherent' to itself. Alterations from that later on are alterations of the source.

A huge difference been our daily lives and reproduced music though is simply in the number of sources of sound we hear. In our daily lives we are hearing thousands of individual discrete sound sources. Each with a phasing true to itself.

When trying for reproduced music (in two channel) we now have two sound sources that by phasing/amplitude relationships between them are trying to phantom image the sound of the original multiple sources of sound that were originally each radiating from a single point in space. Our brain is trying to reproduce that based on correlations in amplitude/phasing betwen the two sources of sound (L/R speakers). Two sources of sound can never exactly reproduce the timing/amplitude/phase of the original single source of sound. For example just think of the reflections in your room if you had a singer dead center in front of you. Then think about how those reflections are different if you are listening to that being reproduced by L/R. Ditto with arrival times to each ear for the two different situations.

Shawn

[maxg]

"The new digital formats were developed because the patents on CD were expiring. Sony/Philips made money on every CD (and every device that could play them) ever created as everyone paid them liscensing fees. Once the patents expired those revenue streams expired too. New formats have new patents and new liscensing.... "

Whatever the reasoning behind the introduction of the new digital media the main selling point (certainly for 2 channel SACD - which is probably the most successful implementation of a largely unsuccessful bunch) is how much closer it is to analogue.

Has to be a reason for that marketing approach IMHO.

It is also interesting that we are now moving the goalposts largely, from trying to say CD sounds better than vinyl, to dissecting the "faults" of vinyl that produce that better sound.

It seems that this 45 degree phase shift between left and right is exactly what you need to create a convincing soundstage from 2 speakers. One wonders why they did not introduce it into CD's in the first place.

[Audio Flynn]

"What does the vinyl "do right" that analog tape doesn't do right? The reason I ask is because my experience with high end reel to reel still doesn't always better mediocre digital."

Home analog 1/2" tape players are in no way the equivelant of studio 1" mastering recorders. The magnetic gap might be different, the tracks narrower, the saturation levels lower, the speed slower. The more tracks that were put on a home recorder the lower the ultimate fidelity. A unidirectional 1/2" machine has two tracks nearly 1/4" wide ( broadcast standard), 4 track bi-directional stereo 1/8", a quad track stereo bi-directional has 1/16" tracks about the same as an audio cassette. It's only advantage is tape speed.

Vinyl and remixed CDs are taken from the wide fast master tape. Most likely on a 30"/sec machine with 8 1/4" tracks and narrow highly saturated magnetic gaps.

Why can vinyl sound better than CDs? CDs are two dimentional. Their strenght is the large dynamic range up and down, 1s and 0s. LPs are, at the least 4 dimentional. Up, down, left, right and also to a degree as the cutter head swings in its arc, other phase relationships are created. That is where the advantage of analog lies and why analog produces a better soundstage image, phasing. Things like decay, ambiance and tonal interactions between metal and wood are better captured on vinyl by their various phase relationships.

I have demo'ed my system for a number of people by sync'ing an LP and the SACD version of an album like Santana's "Abraxas" or Brubeck's "Take Five". Without a doubt, the vinyl version, even with a higher noise floor, is superior with the sole exception of a click here and there. The noise floor is not evident as it is below the background level of both LP and SACD.

If digital recording was in fact superior to vinyl, all the money spent and effort by audio companies to equal the sound produced by analog would cease immediately. Instead "new" digital formats are constantly being developed in an attempt to equal vinyl reproduction.

Rick

Craig (NOSVALVES) and I did a demo a the 2004 AudioKarma meeting with Weather Report "Heavy Weather" SACD to vinyl.

SACD clearly had more bass weight and defineiton under 70 or 80 hz. Realism and soundstage for other frequncies appeard to be superior on the vinyl.

Setup may have been:

LaScalas

Blueberry

VRDs

MMF7

Sony ES 333

zip cord

[3dzapper]

"It seems that this 45 degree phase shift between left and right is exactly what you need to create a convincing soundstage from 2 speakers. One wonders why they did not introduce it into CD's in the first place."

Max, It could be done just not with a single 16 or 24 bit "word". The ambiance/phase information would need to be encoded in other words then the phase relationships encoded in those words would be decoded into an AC voltage varing in phase to the music word. These could be added in the analog section reproducing the true ambiance of the performance.

The downside of this is reduced recording time per disc due to the increased data and, more importantly to the industry, an increase in costs for technology.

The new "blue disks" have the storage capacity to accomplish the task. Especially if video is not involved. We'll see. Another step "closer" to analog may be on the far horizon.

Rick

[sfogg]

"Has to be a reason for that marketing approach IMHO."

Yeah... to sell the format.

If they explained that the higher sampling frequency makes *zero* difference below 22kHz people wouldn't exactly be running right out to buy it would they? Or that the DSD process has large amounts of suprasonic noise in it that isn't exactly a selling point.

Remember... these are the same marketers which gave you 'Perfect Sound Forver.' If you don't trust them WRT CDs why trust them for DVD-A/SACD?

" from trying to say CD sounds better than vinyl,"

CD sounds better than vinyl.... to me.

To you the opposite is the case.

Obviously what 'sounds better' is subjective to the listener.

The wandering phase response of vinyl doesn't work well with my prefered method of listening to music.... in surround sound. The difference in depth and soundstaging in my system between CD and vinyl is profound... with CD easily trumping vinyl. The ambiance extraction requires good phase relationships between channels to be able to determine what is ambiance in the recording and what isn't. It simply works far better with CDs then with vinyl. One of the reasons Quad (matrix encoding) failed (there were many) was the typical delivery format (vinyl) wasn't up to the task because of its phase response.

" to dissecting the "faults" of vinyl that produce that better sound."

Isn't it important to understand why something sounds the way it does? If you recognize what causes you to prefer one medium over the other it can introduce one to ways of maybe achieving that sound they like on the other format(s) while perhaps also gaining the other benefits of that new medium. (Dynamic range, SNR, convience...etc...etc..)

" One wonders why they did not introduce it into CD's in the first place. "

Because it is a delivery format. A delivery format is supposed to deliver what was put into it as accurately as possible.

If a mixer/engineer wanted that sound they could simply mix the CD to have that phase shift between channels. If they added that phase shift the CD would simply 'deliver' it as they created it without further altering the phase response between channels.

Shawn

[sfogg]

Rick,

"Max, It could be done just not with a single 16 or 24 bit "word". "

A single word is only one channel.... but other then that of course it could be done today on CD. Go get a Stereophile test CD and listen to the 'phase test' track.

It goes from in phase to 180 degrees out of phase. It would be trivial to do the exact same thing but with the channels only 45 degrees out of phase with each other.

"The ambiance/phase information would need to be encoded in other words then the phase relationships encoded in those words would be decoded"

Not true at all. The ambiance/phase information is already in the recording. One just needs a better process then two channel to retrieve it. Also if someone wanted to specifically 'encode' ambiance into a recording they can do that too with two channel CD. It is called matrix encoding and has been around for a long time. And it works far better with digital then it does with vinyl because of the phase accuracy of the system.

" Another step "closer" to analog may be on the far horizon."

Nope. Another step further away.

Shawn

[ben.]

Ben, "I think that this added phase incoherence may contribute to the perception that well done vinyl playback is more lifelike" It can as it imparts that sense of depth to the playback in two channel. The wandering phase also likely imparts a little 'air' to the high end since it makes everything more diffuse up there to give it that shimmer. Of course if phasing between L/R gets too bad things will sound lousy. For example wiring one speaker out of phase with the other (180 degrees out of phase) sounds very bad though strongly demonstrates the diffusion effect phase can have. " as there's nothing in our daily lives that is really phase coherent... " Any actual instrument is the reference point... it is creating the sound/phase so it is of course 'phase coherent' to itself. Alterations from that later on are alterations of the source. A huge difference been our daily lives and reproduced music though is simply in the number of sources of sound we hear. In our daily lives we are hearing thousands of individual discrete sound sources. Each with a phasing true to itself. When trying for reproduced music (in two channel) we now have two sound sources that by phasing/amplitude relationships between them are trying to phantom image the sound of the original multiple sources of sound that were originally each radiating from a single point in space. Our brain is trying to reproduce that based on correlations in amplitude/phasing betwen the two sources of sound (L/R speakers). Two sources of sound can never exactly reproduce the timing/amplitude/phase of the original single source of sound. For example just think of the reflections in your room if you had a singer dead center in front of you. Then think about how those reflections are different if you are listening to that being reproduced by L/R. Ditto with arrival times to each ear for the two different situations. Shawn

Sure a live sound source is coherent in a sense, but there's all kinds of interactions in the environment going on... Besides, every step along the way in electronic recording is some type of distortion. I just don't let it bother me. I'm into making things sound good and listening to things that sound good. I'm not hung up on the fact that my brain is being fooled. I rather appreciate the fact that it can be.

[Rivendell61]

Rick,

A couple of examples of CDs using phase information for spatial/ambient cues would be the Roger Waters 'Amused to Death' CD (Q sound) and the Orb's live album ('Orb 93', using Roland RSS10s).

Mark.

[DrWho]

Hey Doc,

I might be driving to Indy this

year. You can keep me company on the ride home and fly back from

Baltimore. It would be a gas.[]

That would be way cool...If you wanted to save on shipping I could

deliver your amp switcher at that time too...(Now that I'm back on

campus with all my tools I will be finishing it as the parts show come

in).

If I can arrange it, I will try to bring some AD/DA converters with me

as well so that we can AB the difference that dropping these devices

into the middle of the signal path can make. If there is no audible

difference dropping the devices into the middle of the line, then there

would be no audible difference with a CD either (the digital data

feeding the DAC would be identical in both cases).

[3dzapper]

Mark, Don't make me name the tens of thousands of analog recordings that are "encoded" with phase information.[]

"It goes from in phase to 180 degrees out of phase. It would be trivial to do the exact same thing but with the channels only 45 degrees out of phase with each other. "

The high tech feat of adding a capacitor to one channel?

"Not true at all. The ambiance/phase information is already in the recording. One just needs a better process then two channel to retrieve it. Also if someone wanted to specifically 'encode' ambiance into a recording they can do that too with two channel CD. It is called matrix encoding and has been around for a long time. And it works far better with digital then it does with vinyl because of the phase accuracy of the system."

Huh? Most of the spatial information is contained in the HF area. Let's say 15KHz to pick an arbitary number. Sampling 15K at 44.1KHz would result in a net of three sample points per wave. The instrument itself plus the room in which it is recorded will provide both higher and lower harmonics in variable phase relationships to the original. 3 sampling points along a 360 degree wave gives odds that less than 1 in one hundred phase additions or cancelations will be sampled. DSD has a greater potential to capture this information but as yet fails to do so.

I have no desire to spend ten grand for a device that can "find" the hidden ambiance that is not really there in a CD. I am old enough to remember the Quadraphonic rage and the LPs encoding that went up into 45KHz range. (Needless to say, the Quad info didn't last many playings.)

With the recording industry losing marketshare to online MP3 sales, I would not expect them to embrace high end audio's miniscule market. Therefore, accept it, analog is the future of high quality playback systems.[]

Rick

I will grant that CDs are convient and superior to cassettes for portable play and some don't sound "bad". A great CD is able to near the listenabity of a good LP.

[DrWho]

Oh...I forgot to mention that my experience with digital versus analog

comparisons were done in the studio...somebody mentioned home gear

which I didn't even know existed...I have messed around with 2", 1" and

I think 1/2" - it's been so long and I was so young at the time (I was

pretty much just the ears for the old guys running the comparisons -

they wanted to have the unbiased opinion of a young kid with good ears).

And btw, you only need TWO samples to perfectly reconstruct a

freqeuency that is within the bandwidth of the medium. I would show you

the mathematical proof for this, but I don't think you would understand

the math (I don't think I understand it completely either as it

requires a very very high understanding of differential equations). But

to make a simple analogy it is the same reason that a system of two

equations with two unkowns has only one solution. (ie, for 2x+3y=4 and

x+y=2 there exists one solution for both X and Y where both equations

are true). It's the exact same concept when you're talking about

sampling frequencies. This is the most fundamental concept behind

discrete signal processing (which is the basis for all digital mediums).

Or to put it another way....let's say you were sampling a 11kHz tone at

44.1kHz and at 100000000000000000000000kHz. Because 11kHz falls within

the bandwidth limitations of both systems, the output of both will be

100% identical. In other words, 20kHz does not have less resolution

than 20Hz when sampled at 44.1kHz.

[Rivendell61]

Rick,

I was responding to your (or Max's) question about the ability/ease of encoding phase information on a CD--yes, no problem. But it generally is not thought desirable to add what is not heard in the 'hall'--so it is not often done. [The CDs I mentioned (Waters, etc) are still interesting for the amazingly wide sound stage--sound coming from 90 degrees to left and right of the listening position.]

Regarding capturing the 'ambience' information that is naturally present.....a CD will capture everything that was audible in the recording venue--and picked up by the mics. No lost harmonics, etc. And DSD will not catch it better. It is already caught.

Mark

[sfogg]

"The high tech feat of adding a capacitor to one channel?"

No, the phase change is in the CD itself. Surely you don't think the CD tells the player to put a capacitor in the signal path?

Any phase difference you want can be put onto a CD between the two

channels. There is absolutely no need for other samples or channels as

you earlier claimed to need.

"Let's say 15KHz to pick an arbitary number. Sampling 15K at 44.1KHz would result in a net of three sample points per wave. "

And that is more then is needed to properly capture/reproduce that

wave. You only need more then two points per period to capture

amplitude and wavelength. Phase is timing.... you automatically capture

that as you capture the other two over a period of time.

"Most of the spatial information is contained in the HF area. "

Spatial information is full bandwidth. Your ears can tell information

about a room even deep into the bass range... down to around 40hz.

"The instrument itself plus the room in which it is recorded will

provide both higher and lower harmonics in variable phase relationships

to the original."

That is captured too as long as there are more then two samples per

period.... in other words on CD it will be captured as long as the

frequency is below 22kHz.

Remember *any* complex signal is nothing more then a bunch of sine waves.

"3 sampling points along a 360 degree wave gives odds that less than 1

in one hundred phase additions or cancelations will be sampled."

3 samples per period give 100% odds that the signal will be accurately

captured. The samples aren't trying to draw the wave(s)... they are in

effect a mathematical derivative of the wave(s).

"I have no desire to spend ten grand for a device that can "find" the hidden ambiance that is not really there in a CD. "

So don't spend ten grand on it. The ambiance *is* really there on most

CDs BTW. It isn't creating ambiance... it really is using what is

already in the recording.

The device that is artificially imparting phase changes to the recording is the one that is creating artifacts.

"I am old enough to remember the Quadraphonic rage and the LPs encoding

that went up into 45KHz range. (Needless to say, the Quad info didn't

last many playings.)"

That was one type of Quad that used the FM carrier for the signal. That

version can't be used on CDs as they don't have the bandwidth. All the

other versions of quad used matrix encoding to encode four channels of

information down to two channel delivery. That matrix encoding is done

with amplitude/phase relationships between L/R. The carrier of the time

couldn't preserve phase very well so the matrix encoding/decoding

didn't work very well. Quad had plenty of other problems too like a

lousy speaker layout, bad psychoacoustics behind it and just gimmicky

mixing too.

Shawn

[sfogg]

"

And btw, you only need TWO samples to perfectly reconstruct a

freqeuency that is within the bandwidth of the medium. I would show you

the mathematical proof for this,"

Actually better re-read Nyquist... that is wrong.

You need more then two (even 2.00001 is fine) samples per wave to

reconstrunct the wave. You no longer capture properly (alias) when you

have 2 or less samples per period.

Shawn

[Erik Mandaville]

Shawn:

"Obviously what 'sounds better' is subjective to the listener."

No sir! MY system is the very best! I listen to it all the time, so I know this to be the case. I've compared it to others, and mine was still always the best. For some reason, other people didn't agree with me, but...I COMPARED THEM WITH MY OWN EARS, AND MINE WAS STILL THE BEST EVERY TIME. Believe me, I know what I like!

Erik

edit: You know I couldn't agree more.