DVD Audio & SACD

[Rob Babcock]

C'mon, grow up. Don't throw that straw man crap at me- I never said upsampling added info, and I'm pretty sure you know that. Disagree with me if you want, but don't flat out lie. I stand by my previous advice; take time out from your wise *** comments and read the article if you honestly don't understand what upsampling does. I'm not going to bother trying to explain it to you as you obviously just want to fight.

As far your previous posts, I have a lot of hi rez discs. Some are made from the same master and some aren't. A goodly amount are DSD transfers of previous analog recordings. Comparing the hybrid layer of these with the DSD layer gives you a taste of the difference. But a better way is to listen to discs made from Hi rez masters. If that doesn't satisfy you, so be it. It's simply not possible to compare every recording ever made with a hi rez version. Maybe in a decade, but of course you can't dig up Miles Davis and make him record new material. My best SACD & DVD-A discs sound much better than any of my CDs, which I think is telling since I have twenty times more CDs. The ones that don't really sound that hot are made from lower res masters (some of them regular 16 & 18 bit recordings).

In my opinion, the diff is black & white. Don't look at it as apples & oranges- that's a metaphorical statement of the stark differences I feel exist. This is a case where I just don't feel the need to try to test them. For example, if someone handed me two buckets and asked me which was heavier, I certainly wouldn't have to weigh them if one was twice as heavy. If they were close I'd weigh them. That's how I view Hi Rez vs 16 bit: it's just not close enough to have to try to measure them. Anyone with average powers of perception can usually hear the diff in my experience. Again, logistics dictate that I can't come to your house to perform a DBT that will persuade you. I've performed my own DBTs, and I've found even decidely non-audiophile friends and family can hear the diff. YMMV.

The one thing you either never answered (or I missed, my apologies of the latter is the case) is whether or not you've ever listened to SACD or DVD-A.

Those are probably fascinating papers you list- since you didn't link them I'll have to Google for them if I have a chance.

I may as well throw some gasoline on the fire and state that my tentative early opinion is that DVD-A sounds better than SACD. I can at this point only base my opinion on the discs I have, but I don't think DSD offers as good of treble response as hi rate PCM.

Enjoy your Redbook CDs. Despite their less-than-optimal resolution I still find them enjoyable.

 

[WmAx]

C'mon, grow up. Don't throw that straw man crap at me- I never said upsampling added info, and I'm pretty sure you know that

.

I do apologaize for this one thing, I did misread your statement of this. However, increasing the sample frequency can not retrieve any low level detail audibly. SOunds like you are talking about something else, such as lowring the noise floor through using a larger bit space, allowing for some tolerance of processing errors in the noisefloor. This has nothing to do with sample frequency.

As far your previous posts, I have a lot of hi rez discs. Some are made from the same master and some aren't.

I presume you have actually analysed the waveforms side by side from each, in order to verify?

Comparing the hybrid layer of these with the DSD layer gives you a taste of the difference. But a better way is to listen to discs made from Hi rez masters. If that doesn't satisfy you, so be it.

What controls did you establish in these listening tests(after verifying indeed they wre the same master with no modifications by wave analysis)? DBT? Level matched? Verification of hardware for non linearities? Number of trails? The scores? These are all required to insure that the test is reliable.

In my opinion, the diff is black & white. Don't look at it as apples & oranges- that's a metaphorical statement of the stark differences I feel exist. This is a case where I just don't feel the need to try to test them.

Well, you are entitled to your opinion. But you really should consider the variables as related to the actual format limitations. There is no substantiated reason that 44.1/16 is inherantly inferior, audibly for playback, than hi-rez.

Anyone with average powers of perception can usually hear the diff in my experience.

Well, considering these are probably non-level matched, uncontrolled conditions without even ensuring the are equal masters -- not a suprise that its easy to hear/percieve teh difference.

The one thing you either never answered (or I missed, my apologies of the latter is the case) is whether or not you've ever listened to SACD or DVD-A.

Those are probably fascinating papers you list- since you didn't link them I'll have to Google for them if I have a chance.

Sorry, those are available from aes.org and are not online. But, I think one of them is published online(the 2nd one). Just search for the paper title on google.

-Chris

 

[Rob Babcock]

Wow! You're on a roll- first you ask me under what conditions I performed my DBT, then without waiting for any reply, you simply made answers to attack! Bravo! How kind of you to presume I analyzed the waveforms, even though you imply I'm not capable of level matching! You give me too much credit. I suppose I could outline my test conditions for you, but then you seem to be doing just fine making them up.

At any rate, the source of the various masters can often be ascertained by simply reading & doing some research online. Certainly not in all cases, but in many. Many of the rags & online publications occasionally discuss this.

I really am curious whether or not you've ever listened to SACD or DVD-A. Of course, you may choose not to ever answer despite my asking repeatedly, as is your perogative. If you ever bother to actually set the agenda aside and listen, let me know what you hear. I actually do fall closer to the objectivist side than the subjectivist camp, but ultimately it all comes down to the sound. I'll give you points for being the only person I've conversed with that claims there's no substantial difference. The fact that someone would dispute this really seems to me like asserting the world is flat- I'd really not be able to link you any study on that one either.

At least we agree on one point; the right to ones own opinion. It's really nothing to me what you believe. I'll tell you though that my (totally unsubstantiated, non-empiracle, unscientific) hunch is that in a few years there'll be a lot better understanding of PCM & DSD, and a lot more evidence of the superiority of the higher bitrate. Assuming we're both around (and this site is, too) in five years, we can get a good yuck by checking back and seeing how good our guesses were.

To anybody who actually has bothered to read this, I say simply this: listen for yourself. And as I stated previously, the extended resolution is only part of the reason to dive into the new formats. I personally feel the MC presentation is a larger factor in the improved sound quality than the greater resolution. In fact, a well recorded DTS mulitchannel disc can often rival the perceivec quality of a CD even though it's vastly lower res.

WmAx, you do seem to have researched this to a great degree. What I haven't really gathered is whether or not you think the new formats are worthwhile. What do you say on the matter. I apologize if this has gotten overly contentious- as a site moderator I really should dial it down a notch. I just get peeved when someone takes me out of context or misrepresents my statements. I'll grant that you do seem to be sincere, and it's been an interesting discussion. I'll give you the last word; I've had my say!

 

[Rob Babcock]

Btw, and I don't say this to pick on you, WmAx, but why is it that if we disagree with someone our first instinct is to attack the methodology or competence of the person? The assumption seems to be that their percieved shortcomings had to have influenced them to a wrong conclusion. But why would an incompetent person necessarily always do the same thing incorrectly in every circumstance? If I couldn't level match, for example, wouldn't it be just as likely that I'd have the CD in the test set louder than SACD, at least some of the time?

On the aforementioned subject, I've found that the high rez has been chosen as the better even when it's not played back as loudly as the PCM. I've sometimes tried this simply to see if volume was the bias.

For the record, I'm agnostic on some of the Great Debates of audio. Cables, for one. To my understanding, the only way a cable should make much diff would be a negative difference, ie if it really screwed something up. Therefore, I'd expect a very crappy cable to sound differnent, but I'd expect a very few to sound bad and equally few to sound noticeably better. Same with amps- I think most of the differnces would be audible only at the extreme edge of the power envelope, under conditions where an amp is pushed to its limits.

Yet, I also think that we're in our infancy with regards to our understanding of the science of sound. In many cases, I believe that there are audible differences in things that we just haven't learned to measure yet. To have the attitude of "if I can't measure it, it doesn't exist" seems to me arrogant and shortsighted. There's a lot of measurements that don't correlate to sound, and probably some things that likely will be quantifiable one day that we don't even consider now. (I'm not aiming this at anyone, so don't be offended). I'll allow that some of the things I can't verify in DBT may still be real, even if my science doesn't yet understand it.

I will confess to one touch of religious fervor: I truly believe that we one day will figure it all out. I think that anything that doesn't flatly defy the laws of physics & quantum mechanics will one day be done. Everything that's not totally impossible is inevitable.

Rant over! I now return you to our regularly scheduled dust-up!

 

[WmAx]

Wow! You're on a roll- first you ask me under what conditions I performed my DBT, then without waiting for any reply, you simply made answers to attack! Bravo!

Lol. I'm sorry, really. I'm used to debating way toooo much. I dont mean to come as a total jerk. It's just me. lol

How kind of you to presume I analyzed the waveforms, even though you imply I'm not capable of level matching! You give me too much credit. I suppose I could outline my test conditions for you, but then you seem to be doing just fine making them up.

t any rate, the source of the various masters can often be ascertained by simply reading & doing some research online. Certainly not in all cases, but in many. Many of the rags & online publications occasionally discuss this.

Unfortunately, I belive this can be misleading.Did these sources specifically state that it was the same master, directly converted to 44.1/16 with NO other alterations? Let's review the Telarc example I was discussing earlier. Tierney Sutton - Dancing In The Dark. Let's look inside and see what it says on the ablum liner:

This CD was recorded using the Direct Stream Digital recording system DSD is a new and improved method of coverting music into the digital domain, sampling at 2.8224 mHz. This results in a frequency response from 0 Hz to beyone 100 kHz, and a dynamic range greater than 120dB. Much of the added resolution afforded by the DSD process is retained in standard CD production by using a dedicated DSD conversion processor.

If I was to make assumption, i would have perhpas though they just converted to RBCD format as simply as possible. However, I communiated directly with Michal Bishop(one of the head recording engineers at Telarc) and he told me that the RBCD and SACD wre not the same. He said the RBCD had been comprimised via compression so that it would remain competitive with mainstream RBCD releases from other artists. So, the RBCD music was compressed without any notice of this and not only that, it was pushed so far in absolute level that it was clipping waveforms at 0 magnitude when analysed in a waveform editor. INteresting?

I really am curious whether or not you've ever listened to SACD or DVD-A. Of course, you may choose not to ever answer despite my asking repeatedly, as is your perogative.

I have not made any claims of personal audbility. Therefor, it is irrelevant.

I'll give you points for being the only person I've conversed with that claims there's no substantial difference. The fact that someone would dispute this really seems to me like asserting the world is flat- I'd really not be able to link you any study on that one either.

Well, I don't buy the speculation and testimonials, that's all. No matter who it's from. I looked into the matter and found ZERO studies that confirmed the audiblity difference with proper scientific methdology. Only speculations, loose test scenarios, and testimonials. I then backtracked, researching the information in reverse, and found that perceptual studies have established the requirements for optimal transmission already. Did you know that massive levels of perceptual research went into picking 44.1/16bit? While I could find no important research in perceptual studies to base hi-rez playback on. With lack of substantiation taht is really is better/superior, when the prexisting format was already optimal in study, is something to be curious about.

At least we agree on one point; the right to ones own opinion. It's really nothing to me what you believe. I'll tell you though that my (totally unsubstantiated, non-empiracle, unscientific) hunch is that in a few years there'll be a lot better understanding of PCM & DSD, and a lot more evidence of the superiority of the higher bitrate.

Well, no evidence that is reliavle is around yet....

A higher signal/noise ratio is definately a value in recording, though. It allows much looser tolerances for recording levels, allowing for greater margin of error. In software editing, the levels can be adjusted and optimally placed to the 16bit space along with a high quality dithering process to optimze signal/noise ratio. A higher recording sample rate can be beneficial, too, though audibly is a question. With higher bandwidths, theoretically teh sampling anti-alias filter does not have to be as proportinaly aggressive and close to the nyquist limit as 44.1 kHz A-D would require. So, with the higher sample rate data, it is technically possible to apply a customized FIR filter via sotware donwsampling to lower rates like 44.1. Though, the filter rate/toplogy is almost always preset and unadjustable in software. But that is not required -- it could be variable. Their is a very expensive software application taht does allow custom filters when downsampling, but the name escapes me at the moment. The reason for cusomization? PUrely theoretical, allowing variation/experimentation.

I personally feel the MC presentation is a larger factor in the improved sound quality than the greater resolution.

I agree.

WmAx, you do seem to have researched this to a great degree. What I haven't really gathered is whether or not you think the new formats are worthwhile.

Multichannel is definatley worthwhile, and a verifiable audible benefit. DVD-A is at least practical from a recording/production standpoint, it's using standarized PCM process. SACD, I feel is compltely counterproductive. DSD system is not compatible with PCM, requiring studios to retrofit with this new equipment and expense that has not been substantiated with controlled listening tests.

I just get peeved when someone takes me out of context or misrepresents my statements.

That was my fault. I did misread that statement. I apologize.

If you want to recieve any of the papers discussed, please email me(wmax@linaeum.com). I'm not sure what the copyright matter is on these papers, though. I had to pay $10.00 each at the aes.org site for each one.

-Chris

 

[WmAx]

Btw, and I don't say this to pick on you, WmAx, but why is it that if we disagree with someone our first instinct is to attack the methodology or competence of the person? The assumption seems to be that their percieved shortcomings had to have influenced them to a wrong conclusion. But why would an incompetent person necessarily always do the same thing incorrectly in every circumstance? If I couldn't level match, for example, wouldn't it be just as likely that I'd have the CD in the test set louder than SACD, at least some of the time?

Well, it's not nescarrily about competance, but an assumed ignorance. That is not meant as a negative use - just that most people do not realize the multiple controls that must be established in such testing or the various distortions and/or non linearities that may be an issue with hardware, contaminiating the tests.

On the aforementioned subject, I've found that the high rez has been chosen as the better even when it's not played back as loudly as the PCM. I've sometimes tried this simply to see if volume was the bias.

Without extensive analysis of the actual audio content and the hardware, this can not be known as to 'why'. However, if the CD is more compressed(which seems to be the case on every analysis I've seen to date), then it would not be an equal comparison. The short term preferred sound quality difference to to SPL difference of higher amplitude is only valid if the source material is an all other ways equal.

Same with amps- I think most of the differnces would be audible only at the extreme edge of the power envelope, under conditions where an amp is pushed to its limits.

Or a really expensive SET amp that produces 4% THD at 10 watts and has an output impedance of 5 ohms.

Yet, I also think that we're in our infancy with regards to our understanding of the science of sound. In many cases, I believe that there are audible differences in things that we just haven't learned to measure yet.

Ah, here is the thing..... their is no substantial evidence of such thing(s). Rememver, we only hear sound pressure variations. Easily measurable. Easily measurable beyond our limits. The only complex of not so well understtod is tha actual psychoacoustic components. That is, how the brain specifically interprets the various sounds themselves.

To have the attitude of "if I can't measure it, it doesn't exist" seems to me arrogant and shortsighted. There's a lot of measurements that don't correlate to sound, and probably some things that likely will be quantifiable

Arrogant? Well, that is certainly counter to scientific protocol. What happened to requirement for substantiation? 1st, prove something exists. 2nd, investigate/quantify it. It does not work the other way around --- that is inefficient and counterproductive. How do you pick random things not proven to exist and investiage them? This is a statistical nightmare and would certainly retard progress if embraced in place of scientific methodology.

-Chris

 

[Rob Babcock]

Just don't fall into the trap of thinking that what you are measuring now is all their is. As the old saying goes, "It's not what we don't know that hurts us. It's all the things we do know that ain't so." We hear amplitude, yes. We hear frequency too. Do we hear phase? That's a touchy one. There's a generally accepted level at which we do, but that's not perfectly understood, either. As a fellow audiophool told me, "If something measure badly but sounds good, you're measuring the wrong thing." I'm only saying it's niave to assume a great understanding of the physics and the psychoacoustics than we currently weild.

I can't recall the name right now, but one prominent engineer brought in to consult urged Philips & Sony right from the beginning not to adopt any standard that held the sampling rate under 100k. I'm not in the mood to Google for dueling links, but I know that it was never universally agreed that 44.1/16 bit was optimal. "Nominal" is more like it, and is marginally transparent. The thing about studies, tests & stats is that I've seen a study or test confirming just about any point you'd care to make. I

As I said, I put a fair amount of weight in DBT, but I think it has failings. A DBT can be "rigged" (not the best word, but it's getting late) to produce any result you need. I think my comments about MP3 vs CD are dead on relevant here as it's the same phenomenon. Back in the early days of the wax cylinder, Edison conducted tests where many subjects couldn't reliably tell a live voice from the recording, as incredible as this seems today. The same fallibility in perceptions trip up ABX style tests as subjective listening. Not to mention the extremely wide range of human perceptions. Frequency response is one of these areas: at my age, in all likelihood 16-17k is the best I could hope to hear (I definately can't hear as high a range as I could at age 20). 20-20khz is the range we consider audible, but it's well documented that women and young girls can often hear far higher. That's one area CD has long been knocked on- the brickwall filters can not only trim the top end but cause nonlinearities within the audible range. And the ability to percieve harmonics well above this (not hear, per se, but perceive) has not been exhaustively studied.

In the end, I guess we're just chasing our tails. I can spend the next couple days digging up studies, you'll counter with another, etc etc. But as fallible as it can sometimes be, in the end have have to trust your ears. I've heard direct 24 bit recordings and I hear a whole 'nother layer of reality, if you will. If you don't, that's cool. I used to be a slave to your "Stereo Review" style objectivism, but experience has shown me there's more there than we're measuring & quantifying with the level of science we have now.

Sorry if I came across as a little harsh. I guess I'm used to debates and ad hominum attacks, too. Years of flogging the forums makes me rise to the bait a little faster than I should. I get a little fiesty from time to time, I'll admit! You've shown class and this has been pretty interesting.

Maybe we'll have to open thi

s up a little for the rest of the class!

 

[WmAx]

Just don't fall into the trap of thinking that what you are measuring now is all their is. As the old saying goes, "It's not what we don't know that hurts us. It's all the things we do know that ain't so."

First, I'm not fond of 'sayings'. They are typially misapplied to the specific case they are applied.

Their are zero reaasons to believe their 'something else' besides sound pressure level and it's associated variables is audible. (one might also consider a structure/physical vibratinon that is conducted through your body in sufficienty amplitude to reach your ears as 'audible')In fact, to suggest their is somehting else besides sound pressure difference taht is aubible is suggesting something that qualifies for the 'unknown', similar to psychic claims. If someone can prove a currently 'misundertood' thing that is audible, they are qualified to win 1,000,000.00 dollars. The JREF prize a real one, that is one offered to anyone that can prove the existance of the 'unknown', whether it is pyschic phenomena or audio things(the JREF was recently updated to include audiophile type things). The catch? It must be demonstrated in strict controlled, scientifically valid conditions. If you are you know of anyone else that can reliabley detect an unmeasurable thing that you can hear, contact JREF immediately and start the process.

You can find the nescarry links to the entry application here:

www.randi.org

"If something measure badly but sounds good, you're measuring the wrong thing." I'm only saying it's niave to assume a great understanding of the physics and the psychoacoustics than we currently weild.

Today, their are no such mysteries. Thes statements are not logical. Ths is not the 60s or 70's where it was difficult to completely measure a device for all relevant parameters. With modern signal analysis equipment, a complete set of data can be produced in minutes, what would have taken days to do back then.

I can't recall the name right now, but one prominent engineer brought in to consult urged Philips & Sony right from the beginning not to adopt any standard that held the sampling rate under 100k.

Here's the thing: someone's title is irrelevant. If they can't demonstrate substantiaon for their claim -- it is not taken seroiusly.

I'm not in the mood to Google for dueling links, but I know that it was never universally agreed that 44.1/16 bit was optimal

I am fully aware of the 'debate'. However, one side of this debate is not founded in an u nderstanding of the perceptual issues that have been established via peer reviwed, scrutinized studies.

"Nominal" is more like it, and is marginally transparent. The thing about studies, tests & stats is that I've seen a study or test confirming just about any point you'd care to make. I

But what is the value of each study? I only respect studies that have taken geat care to achiev a valid result. Not all studies are created equal!

As I said, I put a fair amount of weight in DBT, but I think it has failings. A DBT can be "rigged" (not the best word, but it's getting late) to produce any result you need.

Yes, any test can be rigged. That's why the material is publshed in JAES for peer review, scrutiny and replication for confirmation.

I think my comments about MP3 vs CD are dead on relevant here as it's the same phenomenon.

This is not about MP3, and I'm not prepared to reference specific accredited studies on this issue.

Back in the early days of the wax cylinder, Edison conducted tests where many subjects couldn't reliably tell a live voice from the recording, as incredible as this seems today.

As I said, do not mistake any test as credible. What were the specific conditions of this test? The data? The statisticl value?

The same fallibility in perceptions trip up ABX style tests as subjective listening.

DBT tests(ABX is a variant of DBT) do not screw up perceptions. In fact ABX is one of the most sensitive tests that can allow one to pick out the most subtle differences not possible via casual/sighted listening. BTW, ABX is not in itself time limited, either. Tom Nousaine has applied long-term ABX tests in people home's(hooked variable items to he ABX equipment - person can switch to A B or X any tie they want for as long as they want). ABX is a highly versatile testing format.

20-20khz is the range we consider audible, but it's well documented that women and young girls can often hear far higher.

Reer to the 2nd reference 'very high frequency sounds'. In all of these, highly experience audio experts were used as test subjects. In the 2nd paper, their were some young people, including a young girl. Like in all previous auditory studies of credibility, the introduion of ultrasnoic information did not create any discernable differnce. Have you analysed the spectrum of music? In most music, anyways, when ultrasonic material is recorded along with the below 20khz material, the relative amplitude is typically 20dB and falling or more under the average levels.

That's one area CD has long been knocked on- the brickwall filters can not only trim the top end but cause nonlinearities within the audible range. And the ability to percieve harmonics well above this (not hear, per se, but perceive) has not been exhaustively studied.

The studies references used filter slopes that equal RBCD. The first one used specific test signals that are easier to discern than music signals, and still not positive results. The 2nd study used various music recorded with special lab grade equiment to preserve the harmonics as well as possible. No positive results. So far, the only study that has achieved andy 'positive' results was a highly questionable one, that used brain scans. But this in itself was highly qeustionable due to some odd results. Still no credible results were achieved for audibility.

The credibile studies used various audio professionals and musicians with hearing that is trained to detect small differences. It can not be assumed their attenton or ability to discern is typioca of the average person. Combine this with the failure to achieve any positive results, the probability of some random being able to actually hear a difference is extremely small. Let's say that 1 in 20,000 people could hear the difference(and if they could, it would still be very subtle and only discernagle in quick comparisons between the two), just for argument's sake: this would not justify a special format with a higher bandwidth. In this 'purely theoretical circumstance', the 0.005% rate of occurence is not significant.

In the end, I guess we're just chasing our tails. I can spend the next couple days digging up studies, you'll counter with another, etc etc.

What is the value of these studies? WHat controls?

But as fallible as it can sometimes be, in the end have have to trust your ears.

While this is true to an extent, it's not applicable to the circumstances and scope of accurately determing whether something is transparent. Keyword: accurately

I've heard direct 24 bit recordings and I hear a whole 'nother layer of reality, if you will. If you don't, that's cool. I used to be a slave to your "Stereo Review" style objectivism, but experience has shown me there's more there than we're measuring & quantifying with the level of science we have now

Inadequately controlled listening leaves open many variables. Not reliable method for determination.

Sorry if I came across as a little harsh. I guess I'm used to debates and ad hominum attacks, too. Years of flogging the forums makes me rise to the bait a little faster than I should. I get a little fiesty from time to time, I'll admit!

Well, I'm also a little upset, personally. So it's my fault. My favorite forum(audioreview.com) has a new administrator and he is heavily implying that he will removing the requirment for people to back up claims on the forum(as is currently imposed by the majority of members). Basicly, he want to enforce a similar rule that you might find at the Cable Asylum on AUdio Asylum. A safe haven for people with little to no concern for objectivity.

You've shown class and this has been pretty interesting.

I've shown class? I did not realize I had ever done this.

-Chris

 

[Unregistered]

"But as fallible as it can sometimes be, in the end have have to trust your ears. "

The simple fact is that we CANNOT trust our ears. The ear and brain are easily fooled and there have been tons of studies that draw that conclusion. The Double Blind and/or A/B/X tests are meant to eliminate biases, preconceived notions and other issues such as group psychology (ie if everyone else claims to hear a difference, you will feel pressured to agree with the group).

Google for links on mastering and you will find a wealth of studies, facts, and even opinions of well known mastering engineers that are not fully convinced that anything beyond 44.1/16 is necessary at all. Higher sampling rates and bit depths have some advantages but they are not the end-all be-all that many would have you believe.

 

[dsa220]

If I was to make assumption, i would have perhpas though they just converted to RBCD format as simply as possible. However, I communiated directly with Michal Bishop(one of the head recording engineers at Telarc) and he told me that the RBCD and SACD wre not the same. He said the RBCD had been comprimised via compression so that it would remain competitive with mainstream RBCD releases from other artists. So, the RBCD music was compressed without any notice of this and not only that, it was pushed so far in absolute level that it was clipping waveforms at 0 magnitude when analysed in a waveform editor. INteresting?

Well, would you not want a format that does not require that you have to compress and therefore compromise the original recording? It would appear that (according to what was written) that the RBCD 44.1/16 bit is a compromised format for todays higher resolution reording equipment. Would it therefore not make sense that a format that can hold a wider bandwidth would [hopefully] sound different?

The simple fact is that we CANNOT trust our ears. The ear and brain are easily fooled and there have been tons of studies that draw that conclusion. The Double Blind and/or A/B/X tests are meant to eliminate biases, preconceived notions and other issues such as group psychology (ie if everyone else claims to hear a difference, you will feel pressured to agree with the group).

Then why do we even bother? If I can not trust my ears, then maybe my old 8-track player wasn't as crappy as it sounded! I have assembled my audio system not to please others, but to please me. To me, it's not a matter of whether or not I trust my ears, but a matter of whether or not I like what I hear, which is why I find this 'discussion' so interesting yet ultimately futile.

I wrote the editor of 'The Tube Cad Journal' and asked his permission to copy portions of one of his editorials that may apply here, at least with respect to DBT's:

The whole of high-end audio lives in the realm of marginality: the $30,000 vacuum-tube amplifier's output is only marginally different from the average $200 Chinese receiver's output; as seen on the scope, no difference can be discerned. But, like the extra letter revealed by Vanna, small audible differences can have large effects.

But are those differences audible? No scientific proof has been offered that shows that they are. However, proving a negative is much more difficult than proving a positive. If someone shows us a pink crow, then we have proof of its existence; but how do you prove that no pink crows exist, without collecting every single crow in the world and checking for pink? Second, when we test for audibility we are not testing the sound, but the listener. If we were testing the sound, we would need only a microphone and no listener. Now what happens to people under test? Sometimes the act of testing influences the results. Imagine that you are being tested to see if you can tell the difference between your wife’s kissing and another women’s kissing. No problem. You intimately know your wife’s lips, their texture and temperature, their pliancy, her smell and feel against you, and her technique from years of avid kissing. The differences are not subtle.

Yet, I promise you, if you were blindfolded, placed on stage with a huge, loud audience, and Regis making jokes at your expense, you could not tell whether your wife or Regis was kissing you. This example is extreme, but the principle is the same. I know from experience how the act of being tested can alter the results.

Try this test: take two diodes (1N4001) and solder each of them in parallel, but opposite in orientation, with a 10-ohm resistor. Now place a soldered pair in series with the leads to each loudspeaker. Now listen. If the no difference can be heard, or if it can be heard too readily, increase or decrease the number of diodes or the value of resistors. Once a very small difference is achieved that you can easily hear, have someone blind-test your ability to discern the difference.

So is scientifically rigorous testing worthless in evaluating the audiophiles claims? No, of course, not. But such testing must be careful to test apples and only apples. But then even if the tests were perfect and showed no positive results, why shouldn’t someone be allowed to remain adrift in the sea of subjectivity, unlike the Arts, their stereo systems are not government funded. And surely, the whole point to high-end audio is the sound of music, which, like drinking wine, is a truly subjective enjoyment.

John E. R. Broskie - Editor

I would like to thank Mr. Broskie for his permission to copy portions of his article. If you would like to read the full article, here is a link: http://www.tubecad.com/index_files/page0030.htm

 

[WmAx]

Well, would you not want a format that does not require that you have to compress and therefore compromise the original recording? It would appear that (according to what was written) that the RBCD 44.1/16 bit is a compromised format for todays higher resolution reording equipment. Would it therefore not make sense that a format that can hold a wider bandwidth would [hopefully] sound different?

No, this is not the issue. The issue is that the music was compressed not to faciliate the limitations of the format itself(wich, btw, the recording was only a small fraction of the range availabe on RBCD), but a compromise that was and is artifically introduced to the RBCD version.

Then why do we even bother? If I can not trust my ears, then maybe my old 8-track player wasn't as crappy as it sounded!

Well, since your 8 track had several audible problems(well known by way of JNDs established in research) such as insufficient signal/noise ratio and audible levels of THD, the medium degrades with each play/rewind, etc.. It is hardly a question that 8-track was audibly inferior.

I have assembled my audio system not to please others, but to please me. To me, it's not a matter of whether or not I trust my ears, but a matter of whether or not I like what I hear, which is why I find this 'discussion' so interesting yet ultimately futile.

Futile? No. If you appreciate the percieved(<--KEYWORD) sound that hear, then I'm glad for you. Just do not assume that all things that are responseible for what you percieve are due to actual isolated audible differences.

I wrote the editor of 'The Tube Cad Journal' and asked his permission to copy portions of one of his editorials that may apply here, at least with respect to DBT's:

Ok, it's a bit odd to respond to the text of someone that is not here to reply. But since you posted it....

The whole of high-end audio lives in the realm of marginality: the $30,000 vacuum-tube amplifier's output is only marginally different from the average $200 Chinese receiver's output; as seen on the scope, no difference can be discerned. But, like the extra letter revealed by Vanna, small audible differences can have large effects.

First, a simple oscilliscope analysis is difficult to thoroughly analyse a piece of audio equipment for all relevant characteristics. It is a tool that is better served used to analyse specific charateristics/problems, not a full analysis of all relevant tings to sound. In order to manually plot a complete broadband THD plot, it would take a great deal of time for this one thing. Then, you need to do the same thing over an over at different loads and amplitudes. A modern hi precision network analyser such as an Audio Precision analyser makes such measurements quickly, automaticly and efficieintly. 2nd, you must correlate the differences you do find to perceptual research -- you can't just look at a diffference and assume it's audible or to what degree it may be audible with no reference data to coorelate it to.

However, proving a negative is much more difficult than proving a positive. If someone shows us a pink crow, then we have proof of its existence; but how do you prove that no pink crows exist, without collecting every single crow in the world and checking for pink?

Refer to posts to Rob above. I have addressed this already. Short version: Possibility does not nescarrily equate probability. You can not assume something exists until such is demonstrated. To do otherwise is inefficient and illogical.

Second, when we test for audibility we are not testing the sound, but the listener.

If this is an uncontrolled listening test, then it is not just sound you are evaluation. This is true. It's also the psychological components that are due to extraneious stimuli. In that case, you have moved off topic of the disccusion, as the subject in this thread is discussion of verifiable real differences, not those that are due to psychological reasons.

Yet, I promise you, if you were blindfolded, placed on stage with a huge, loud audience, and Regis making jokes at your expense, you could not tell whether your wife or Regis was kissing you. This example is extreme, but the principle is the same.

This example is bogus. No credible listening test allows such distraction from the test, nor does any such test put the subject under such audience, nor does it allow social interaction(unless social interaction is the intended variable). Rediculous example. Not relevant to any proper DBT.

Try this test: take two diodes (1N4001) and solder each of them in parallel, but opposite in orientation, with a 10-ohm resistor. Now place

I don't know what this is supposed to do or prove -- but I saw no data to back up an impliciation that it causes an audible but unmeasurable difference in a proper controlled test.

But then even if the tests were perfect and showed no positive results, why shouldn’t someone be allowed to remain adrift in the sea of subjectivity,

If they wish to do this, they every right. But how many realize this is the case? It's a different matter to do so ignorantly vs. knowingly.

-Chris

 

[dsa220]

Quote:
Then why do we even bother? If I can not trust my ears, then maybe my old 8-track player wasn't as crappy as it sounded!


Well, since your 8 track had several audible problems(well known by way of JNDs established in research) such as insufficient signal/noise ratio and audible levels of THD, the medium degrades with each play/rewind, etc.. It is hardly a question that 8-track was audibly inferior.

WmAx, maybe the word sarcasm should be entered into your general vocabulary.

 

[Polkfan]

I say enjoy the music. If it sounds good, then it sounds good.

 

[Khorn]

I say enjoy the music. If it sounds good, then it sounds good.

Exactly but, some people can't accept that. They tell you that in their opinion you are wrong and you should have to prove to them why you like things the way you do. I think it's a given that some kinds of distortion actually add to the musical experience eg: tube 'ringing'. To most of us, this is about music and not scientific experimentation. Are all systems flawed?.....on an absolute basis...sure. But some flaws can be preceived as more musically pleasing than others. An absolutely scientifically accurate and "clean" system woud most likely sound "dead" and "listener fatigue" you to death.

Measurement may provide an insite into accuracy and faults but again, those very faults might be exactly what many perceive as musically pleasing.

We may be able to accurately measure many parameters related to sound reproduction but, there are still many things that we do not know or understand yet. Every component that an audio signal passes through most likely alters it (the signal) in some way, even some we might perceive but not show up in a way that we are able to measure at present. This doesn't mean we won't be able to measure these things in the future.

Point is after stating the above IMHO comes it down to the fact that most people don't really care about WHY something sounds the way it does, they just care if they like it or not and if you stop to think about it that's the way it is with most things.

 

[WmAx]

Exactly but, some people can't accept that. They tell you that in their opinion you are wrong and you should have to prove to them why you like things the way you do.

No one has to justify why they like or do not like something. I'm certaily not 'for' such a thing. I, for example, would only demand substantiation if such person claimed that a specific 'mysterious' differences was fact, as a fact, or any other statement or circumstance that they stated it as an absolute certainty of physical cause relationship. I make various correction so some statements, too, but this is just to reduce errors. Otherwise, I'm happy go lucky.

Every component that an audio signal passes through most likely alters it (the signal) in some way, even some we might perceive but not show up in a way that we are able to measure at present.

Try reformatting this to: Every component that an audio signal passes through may alter it(the signal) in some way, even some we might percieve but not show up in a way that we are able to to meaure at present.

"Most likely alters it" is the same as claiming high probability. No credible research that I am aware of would qualify this is probable. However, just chaging this to 'may' removes this burden that is required to support 'most likely'.

-Chris

 

[av_phile]

Are all systems flawed?.....on an absolute basis...sure. But some flaws can be preceived as more musically pleasing than others. An absolutely scientifically accurate and "clean" system woud most likely sound "dead" and "listener fatigue" you to death.

Measurement may provide an insite into accuracy and faults but again, those very faults might be exactly what many perceive as musically pleasing.

Sorry, but I find this a contradiction in terms. How can an system that by all measurements is "absoultely accurate and clean: would sound "dead" and "listener fatigue" you to death. It probabaly will if the RECORDING is dead and full of disstrotions to begin with, as that is what an accurate and clean system will reveal. Experiments have already indicated that listening fatigue is caused by high order distortions introduced by a gear into the signal. Conductors and musicians do not experience listening fatigue. As there are no THDs with real instruments. So a system that is absolutely accurate and clean will reveal accurately what an excellent recording sound like. Nothing more nor less. OTH those tubed gears that have abundant even-ordered hamronic distortions may sound euphonic but by adding such euphonic THDs into the signal they are anathema to audiophile purists. So you may like what you hear, mistakingly conclude it sounds full-bodied and realistic, but these are THD that bloat the signals and are not supposed to be there with the recording. What is "musically pleasing" has nothing to do with hi-fidelity. Even a discman or a walkman can be musically pleasing to some, but they are nowhere near Hi-Fi.

 

[Khorn]

Sorry, but I find this a contradiction in terms. How can an system that by all measurements is "absoultely accurate and clean: would sound "dead" and "listener fatigue" you to death.[quote It probabaly will if the RECORDING is dead and full of disstrotions to begin with, as that is what an accurate and clean system will reveal. Experiments have already indicated that listening fatigue is caused by high order distortions introduced by a gear into the signal.

I guess I'm basing this on my experience with early eighties digital gear and recordings. They were touted as "clean and accurate" but in many cases sounded "dead and fatiguing".

Conductors and musicians do not experience listening fatigue. As there are no THDs with real instruments. So a system that is absolutely accurate and clean will reveal accurately what an excellent recording sound like. Nothing more nor less. OTH those tubed gears that have abundant even-ordered hamronic distortions may sound euphonic but by adding such euphonic THDs into the signal they are anathema to audiophile purists. So you may like what you hear, mistakingly conclude it sounds full-bodied and realistic, but these are THD that bloat the signals and are not supposed to be there with the recording.

I fully agree on this point. I have used both Tube and SS gear many times over the years and yes the tube sound can be very "pleasing", specially a combination of a tubed pre-amp and SS amp. I am now back to an all SS system due to practical considerations (mainly S/N among other things) in my particular situation.

What is "musically pleasing" has nothing to do with hi-fidelity. Even a discman or a walkman can be musically pleasing to some, but they are nowhere near Hi-Fi.

Yes and no. Sure you can really enjoy music listening to a portable radio and I had to do so for the past half year or so while my system was down due to moving and renovations. However, when I got the system up and running it took the "musically pleasing listening" to a whole other level due to the systems ability to present the music with far more audible information which in turn makes it more "musically pleasing." If you are able to hear more of something you are also able to like or dislike it to a greater degree.

 

[WmAx]

Even a discman or a walkman can be musically pleasing to some, but they are nowhere near Hi-Fi.

What exactly is your defnition of 'hi-fi'?

Many DIscmans measure very well. I can not point to a specific resource at the moment, but I have seen quite a few involved measurements of fequenc response, THD vs. frequency, IMD, S/N, and varying distortion into different loads of various current and past model Sony Discman units. Performance was typically very good, with no reason for any significant audible differecne from a well made component cd player.

-Chris

 

[Rob Babcock]

Interesting discussion, guys, but this is getting a little off topic. The subject is your experiences with SACD & DVD-A. The dissertations on digital theory & walkmans is interesting but not the point of the original question.

Does anyone have any experiences of their own to discuss? I stated earlier my opinion that the sonic improvement of SACD & DVD-A had more to do with the MC presentation that the increased resolution. Does anyone here have any opinion on that?

 

[WmAx]

I stated earlier my opinion that the sonic improvement of SACD & DVD-A had more to do with the MC presentation that the increased resolution. Does anyone here have any opinion on that?

I should be whipped with chains.

As for multi-channel, hmm, opinions? It is fact that the greater number of channels provide for a more stablie center image(center channel) and provides for a consistant rear ambience fill, to assist in the auditory illusion. As for opinion .... well in 'some' limited examples of specific recordings with specific 'rare' 2 channel systems, some people have stated an opinion that it has sounded extraordinarly realistic. I would be one of those; but I realize the physical limitations and inconsistant behaviour of 2 channel recordings/systems. Multichannel has the advantage in reliability. A system such as Tom Hollmans's 10.2 playback/recording standard(still a prototype/experimental) has elicited consistant comments of 'extraordinary realism' from people forutnate enough to have demoed the system. One can only hope that such a standarized recording/playback system is eventually adpopted into consumer audio.

-Chris