Let's Talk Computer Audio

[RBFC]

Perhaps those who don't feel that headphones can illuminate the small differences between sources feel this way just because the headphone experience is so different from that of room/speakers. For instance, is a small volume difference of one instrument in the recording due to its physical placement in the soundstage, or is it simply because that instrument is being played a bit softer? Those of us (me included) who don't have the listening experience with headphones may find this distinction impossible to sort out. While I'm sure that one may hear "more" of the recording on headphones, it does not necessarily follow that headphones provide a more believable recreation of the musical event. So, only those folks who can translate the small "cues" heard on headphones into what sonic effect would be produced via speakers are qualified to comment. That leaves very few of us, from what I'm able to discern.

Lee

 

[JackD201]

Tim,

You would be right if the comparably poorer channel separation of vinyl was full spectrum. It isn't. Positioning cues are in the upper ranges, only venue size cues (long reverberations) are down below. The best digital is very spacious, poor digital is flat as a pancake. Both have the same channel separation. Reproduction and subsequent processing of recorded ambient information, regardless of medium, is not as simple as you make it out to be. The medical and independent academic studies aren't on your side on this one.

 

[Julf]

Well said!!!! The subjective nature of our hobby is the very thing that keeps things interesting. We can all like something different!!! And actually hear - something different!!! Does it make one person better than another - no way!!!!

As long as we don't start to assume that our subjective impression is true for everybody else - and as long as we acknowledge that our subjective impression doesn't always correspond to physical reality.

Computer audio is no different - change a DAC, soundcard, with prepro, no prepro etc...you'll more than likely change the sound.

Computer audio is different - up to the DAC. From the point where the signal becomes analog, it is of course analog, and subject to the vagaries of an analog signal. But until then, bits really are bits.

I'm wondering can changing the way a program handles the 0's and 1's change the sound in the time domain (not the data). Would we tell the difference some say yes others say no, and you won't be able to tell either differently because its etched in their minds. Put black and white lines closer and closer together on a red backround you will swear that where the black and white lines almost touch - is green. Its a mind trick - even though we know its black and white.

And that is a valid analogy as long as the recipient is the human sensory system. But if the recipient is another digital system (as in a computer passing data over an asynch USB connection) the recipient can reconstruct the accurate timing (in your analogy, the spacing of the lines).

 

[Phelonious Ponk]

Tim,

You would be right if the comparably poorer channel separation of vinyl was full spectrum. It isn't. Positioning cues are in the upper ranges, only venue size cues (long reverberations) are down below. The best digital is very spacious, poor digital is flat as a pancake. Both have the same channel separation. Reproduction and subsequent processing of recorded ambient information, regardless of medium, is not as simple as you make it out to be. The medical and independent academic studies aren't on your side on this one.

Are you saying that vinyl has better channel separation in the upper ranges than it does in the lower? Or that digital has poorer channel separation in the lower ranges? I'm not sure I'm following you, and I'm not trying to make it simple. The elements that contribute to it are, I think, a few, and pretty easily understood. The cumulative affect is incredibly subtle and varied and not easy to get a handle on.

And what you're hearing is not always clear. Are you sure the spaciousness you hear in vinyl always comes from the recording at all?

Tim

 

[Bruce B]

Are you saying that vinyl has better channel separation in the upper ranges than it does in the lower?
Tim

yes it does...

 

[JackD201]

Look up how the grooves are cut in a V and how a stylus interacts mechanically with it and you'll understand what I mean. LP has no problem with image localization from around 60 to80Hz on up. Very stable and dense images are common in properly set up tables. Poor channel separation in these upper ranges would result in soft delineations and a collapse towards the center. Since in practical use most bass information is panned to the center anyway, practically mono in nature, it's not as big a deal as one would think just looking at the numbers. Especially when considering that below these frequencies sound waves become omnidirectional in nature. The question is why has this been the habit for decades of stereo? Analog or digital, panning center just gives a more stable foundation when you stack a bass on top of a kick. It's also what listeners have become accustomed to. Even in classical where tympanies aren't panned center, the strike has enough high frequency content to place them rear right but try following the decay and at some point it just spreads out, again, analog or digital.

I would put forward that channel symmetry in terms of amplitude and FR, electrical and acoustical, is more important than channel separation per se when it comes to the creation of solid phantom images and their resulting localization. You can get away with a lot of bleeding and crosstalk before an image collapses. One channel being a few dB down skews images pretty quickly.

 

[mep]

Are you saying that vinyl has better channel separation in the upper ranges than it does in the lower? Or that digital has poorer channel separation in the lower ranges? I'm not sure I'm following you, and I'm not trying to make it simple. The elements that contribute to it are, I think, a few, and pretty easily understood. The cumulative affect is incredibly subtle and varied and not easy to get a handle on.

And what you're hearing is not always clear. Are you sure the spaciousness you hear in vinyl always comes from the recording at all?

Tim

No Tim, I'm sure it doesn't come from the recording. By now we all know that according to the naysayers and analog haters, anything that sounds good coming from vinyl is a result of distortions. LPs can't possibly sound good as we have discussed before. Vinyl has countless distortions and somehow they all add up together and form into a ball of *goodness* and great sound comes from them. And DSD apparently has tons of different distortions as well because it sounds very close to analog which is why I like it the best out of all the digital formats I have heard and own.

 

[microstrip]

I readily test and hear "feeling of the room" with headphone testing. Indeed when testing for high resolution audio, that is precisely what I focus on. How well notes decay into silence -- something that is actually harder to do with speakers since you can't isolate noise unless you listen at reference levels. It is those smaller "bits" that provide a sense of being there. Digital is superbly good at loud signals so testing for low level signal is where it is at, and where headphones shine.

Presence in the room is not "feeling of the room". It is related to the proper size, body and interplay between players and groups of players. I was never able to feel it using headphones - congratulation if you manage to do it. And the decays perceived through headphones, without the help of the room and reflection on the walls most of the time sound very non -natural. IMHO, only in the presence of an exceptional system and conditions you can fully test anything properly. Otherwise you are too far from reality to use the support of the feeling correct to help you in your comparison. And I have the feeling that it is not only the small bits (if we know exactly what this means ) that provide a sense of being there - I have felt it using NOS DACs that no one thinks that have good small bits.

 

[jkeny]

Just taking a break from milking those snakes - damn that snakeoil is difficult to extract

First apologies to Tim for not reading his post carefully enough - when he mentioned reclocker, I assumed he meant the "usual" use of that term, i.e resampling of SPDIF signal. Which, in my experience always results in masking of the qualities of low jitter sources while, on the other hand, being beneficial to the sound of higher jitter sources.

My comment about DBT BS is exactly as picked up by Tim - the sort of challenge that is often put forth on audio forums to "prove" that subtle differences exist between the sound of playback devices/software/etc. Yes we all use "personal" blind tests as sanity checks which is usually enough for our personal needs in the context of this hobby. Getting into DBTs as a serious test of anything in audio requires far more scientific rigour & expense.

Flicking through the various posts, I agree with the notion that somehow headphone listening is not as revealing of these subtle differences as speaker listening. The most noticeable area where differences are picked up is in depth & solidity of sound stage. This effect I believe is to do with more stable inter-channel timing in the audio reproduction right down to a low signal level - the lower the signal level reproduction & timing, the more "realistic" the sound. This is often referred to as more intelligibility, better emotional connection with the music. This sound stage illusion, produced from 2 speakers, is a tenuous thing which is easily upset by a number of factors in the music & in the room itself. Remember this illusion is a perception that requires no small amount of brain processing - any shifts in the cues that create this perception can cause fatigue & disinterest rather than engagement & relaxation. I believe that this is at the heart of what is meant by intelligibilty & emotional connection with the music - our brain has less processing to do & we can focus more on the actual performance & interactions within the music - it's more of a whole!

So what, at first, seems to be subtle small improvements actually translate into a much more interesting enjoyent of the music - something not subtle & insignificant at all, if one's goal in listening to music is relaxation & the illusion of "being there"

Edit: Something related to the headphone Vs speaker experience may be the factors that the brain uses to judge distance of a sound:
Three physical mechanisms, and a fourth cognitive one.
1) Spectral balance: treble attenuates quicker in air.
2) Absolute level.
3) Distribution of direct and reflected sound in time and spectrum.
4) Knowledge of the room one is in, and prior experience, mapping 1-2-3 quite accurately on distance.

Now back to those snakes!

 

[Phelonious Ponk]

Look up how the grooves are cut in a V and how a stylus interacts mechanically with it and you'll understand what I mean. LP has no problem with image localization from around 60 to80Hz on up. Very stable and dense images are common in properly set up tables. Poor channel separation in these upper ranges would result in soft delineations and a collapse towards the center. Since in practical use most bass information is panned to the center anyway, practically mono in nature, it's not as big a deal as one would think just looking at the numbers. Especially when considering that below these frequencies sound waves become omnidirectional in nature. The question is why has this been the habit for decades of stereo? Analog or digital, panning center just gives a more stable foundation when you stack a bass on top of a kick. It's also what listeners have become accustomed to. Even in classical where tympanies aren't panned center, the strike has enough high frequency content to place them rear right but try following the decay and at some point it just spreads out, again, analog or digital.

I would put forward that channel symmetry in terms of amplitude and FR, electrical and acoustical, is more important than channel separation per se when it comes to the creation of solid phantom images and their resulting localization. You can get away with a lot of bleeding and crosstalk before an image collapses. One channel being a few dB down skews images pretty quickly.

That sounds right to me.

Tim

 

[Phelonious Ponk]

Just taking a break from milking those snakes - damn that snakeoil is difficult to extract

First apologies to Tim for not reading his post carefully enough - when he mentioned reclocker, I assumed he meant the "usual" use of that term, i.e resampling of SPDIF signal. Which, in my experience always results in masking of the qualities of low jitter sources while, on the other hand, being beneficial to the sound of higher jitter sources.

My comment about DBT BS is exactly as picked up by Tim - the sort of challenge that is often put forth on audio forums to "prove" that subtle differences exist between the sound of playback devices/software/etc. Yes we all use "personal" blind tests as sanity checks which is usually enough for our personal needs in the context of this hobby. Getting into DBTs as a serious test of anything in audio requires far more scientific rigour & expense.

Flicking through the various posts, I agree with the notion that somehow headphone listening is not as revealing of these subtle differences as speaker listening. The most noticeable area where differences are picked up is in depth & solidity of sound stage. This effect I believe is to do with more stable inter-channel timing in the audio reproduction right down to a low signal level - the lower the signal level reproduction & timing, the more "realistic" the sound. This is often referred to as more intelligibility, better emotional connection with the music. This sound stage illusion, produced from 2 speakers, is a tenuous thing which is easily upset by a number of factors in the music & in the room itself. Remember this illusion is a perception that requires no small amount of brain processing - any shifts in the cues that create this perception can cause fatigue & disinterest rather than engagement & relaxation. I believe that this is at the heart of what is meant by intelligibilty & emotional connection with the music - our brain has less processing to do & we can focus more on the actual performance & interactions within the music - it's more of a whole!

So what, at first, seems to be subtle small improvements actually translate into a much more interesting enjoyent of the music - something not subtle & insignificant at all, if one's goal in listening to music is relaxation & the illusion of "being there"

Edit: Something related to the headphone Vs speaker experience may be the factors that the brain uses to judge distance of a sound:
Three physical mechanisms, and a fourth cognitive one.
1) Spectral balance: treble attenuates quicker in air.
2) Absolute level.
3) Distribution of direct and reflected sound in time and spectrum.
4) Knowledge of the room one is in, and prior experience, mapping 1-2-3 quite accurately on distance.

Now back to those snakes!

I don't doubt any of what you've said, John, I'd just add that hearing and judging cues through headphones is just context. It's not that much different than listening to your speakers in a differernt room, or changing from conventional front-firing speakers to bi-poles or di-poles. You get used to the "headstage" as opposed to the sound stage. The cues manifest themselves in different ways, but they're all there, and arguably more audible in the isolated space. Is it the same as a speaker system's sound stage? Of course not. Neither is a panel's sound stage the same as box. That doesn't mean the information isn't all there and can't be assessed in those varying presentations. No oil. Just logic.

Tim

 

[microstrip]

(...) I'd just add that hearing and judging cues through headphones is just context. It's not that much different than listening to your speakers in a differernt room, or changing from conventional front-firing speakers to bi-poles or di-poles. You get used to the "headstage" as opposed to the sound stage. The cues manifest themselves in different ways, but they're all there, and arguably more audible in the isolated space. Is it the same as a speaker system's sound stage? Of course not. Neither is a panel's sound stage the same as box. That doesn't mean the information isn't all there and can't be assessed in those varying presentations. No oil. Just logic.

Tim

Tim,

Sorry but hearing and judging cues through headphones is much more than context. And your analogy trying to compare different types of speakers is also misleading. If you listen through a system that suppresses perceived information you loose something. IMHO you can not reduce listening to a mechanical and instrumental action, rejecting psychoacoustics and perception. In my view your logic is completely flawed.

Anyway if you carry listening tests using different speakers and rooms you can reach different conclusions about audibility of differences. This just teaches you you must carry the experiment using different types of systems, nothing else. I am sure that as Amir outlined headphone listening is excellent and adequate for some type of differences - but not enough to be the definitive test in audiophile matters.

 

[Phelonious Ponk]

Tim,

Sorry but hearing and judging cues through headphones is much more than context. And your analogy trying to compare different types of speakers is also misleading. If you listen through a system that suppresses perceived information you loose something. IMHO you can not reduce listening to a mechanical and instrumental action, rejecting psychoacoustics and perception. In my view your logic is completely flawed.

Anyway if you carry listening tests using different speakers and rooms you can reach different conclusions about audibility of differences. This just teaches you you must carry the experiment using different types of systems, nothing else. I am sure that as Amir outlined headphone listening is excellent and adequate for some type of differences - but not enough to be the definitive test in audiophile matters.

Micro, I haven't a clue what you're talking about, but as is often the case, it doesn't appear to be related to anything I actually said.

What I'm talking about is different presentations of sound space -- bipole, dipole, conventional front-firing box speakers, true omni-directionals...headphones. If we want to get even more granular there are open phones, closed phones, IEMs. All are different. All present the same sonic information. Put your ears on, get your set-up right (because if you don't, your room will profoundly screw things up), pay attention, and you can ascertain as much about the quality of the information and its reproduction through a good headphone system as you can through a speaker system. More in many cases.

There was no analogy. I'm not reducing anything or rejecting the role of perceptions. I am simply allowing my perceptions to fully function in another space. One you can't seem to get your head into, if you'll excuse the pun.

Tim

 

[RBFC]

Please refrain from personal attacks, guys.

Lee

 

[amirm]

Presence in the room is not "feeling of the room". It is related to the proper size, body and interplay between players and groups of players. I was never able to feel it using headphones - congratulation if you manage to do it.

??? Seems like we are discussing two different topics. I am talking about testing to hear differences. In that regard, headphone is a great *instrument*. I am not talking about enjoying music that way. So congratulation is not the right answer to what I was describing. Here is the simple math:

What you are hearing is Channel A + Channel B = your experience with speakers. It is not possible for that sum to change without the other two factors that make it to stay the same. To the extent we are trying to detect if there is a difference in high res vs not, then listening to even one channel, using headphones will have to work. It is not possible for it to not do that.

And the decays perceived through headphones, without the help of the room and reflection on the walls most of the time sound very non -natural. IMHO, only in the presence of an exceptional system and conditions you can fully test anything properly.

We really are talking about two things then per above. Non-natural? What does that have to do with detecting if there is a difference?

Otherwise you are too far from reality to use the support of the feeling correct to help you in your comparison. And I have the feeling that it is not only the small bits (if we know exactly what this means ) that provide a sense of being there - I have felt it using NOS DACs that no one thinks that have good small bits.

I am not after detecting feelings. I am after detecting if additional bit depth or sampling rate make a difference. If changing these makes no difference in headphones then they absolutely have no audible effect of importance. It would prove that the analog output in each channel did not change. How the sum total of two zero difference signals will amount to something new is what I like to see explained.

And let me say again: none of this is theoretical for me. I have done thousands of hours of testing using headphones and considerable amount using speakers. Speaker testing is hugely limited compared to headphones.

And so that we don't keep talking in theory, here is a formal double blind tests of SACD vs DVD-A published at AES conference: "DVD-Audio versus SACD Perceptual Discrimination of Digital Audio Coding Formats Listening Comparison Test between DSD and High Resolution PCM (24-bit / 176.4 kHz)"

"The four highest scores fell into the region of “critical probability.” This amounted to only 2.76% of all the tests. These four tests were carried out by four separate listeners, all of whom chose stereo music examples, and in all four cases headphones were used—thus excluding the influence of the listening environment to the greatest possible extent."


No other tester managed to do better than chance. They headphones they used were Stax which is the high-end standard for headphone tesing (same as what I have).

So let's stay focused on headphone use for detecting differences in audio, not what is more enjoyable. The latter is a different topic and unrelated.

 

[microstrip]

??? Seems like we are discussing two different topics. I am talking about testing to hear differences. In that regard, headphone is a great *instrument*. I am not talking about enjoying music that way. So congratulation is not the right answer to what I was describing. Here is the simple math:

What you are hearing is Channel A + Channel B = your experience with speakers. It is not possible for that sum to change without the other two factors that make it to stay the same. To the extent we are trying to detect if there is a difference in high res vs not, then listening to even one channel, using headphones will have to work. It is not possible for it to not do that.


We really are talking about two things then per above. Non-natural? What does that have to do with detecting if there is a difference?


I am not after detecting feelings. I am after detecting if additional bit depth or sampling rate make a difference. If changing these makes no difference in headphones then they absolutely have no audible effect of importance. It would prove that the analog output in each channel did not change. How the sum total of two zero difference signals will amount to something new is what I like to see explained.

And let me say again: none of this is theoretical for me. I have done thousands of hours of testing using headphones and considerable amount using speakers. Speaker testing is hugely limited compared to headphones.

And so that we don't keep talking in theory, here is a formal double blind tests of SACD vs DVD-A published at AES conference: "DVD-Audio versus SACD Perceptual Discrimination of Digital Audio Coding Formats Listening Comparison Test between DSD and High Resolution PCM (24-bit / 176.4 kHz)"

"The four highest scores fell into the region of “critical probability.” This amounted to only 2.76% of all the tests. These four tests were carried out by four separate listeners, all of whom chose stereo music examples, and in all four cases headphones were used—thus excluding the influence of the listening environment to the greatest possible extent."


No other tester managed to do better than chance. They headphones they used were Stax which is the high-end standard for headphone tesing (same as what I have).

So let's stay focused on headphone use for detecting differences in audio, not what is more enjoyable. The latter is a different topic and unrelated.

Amir,

I have no doubt that you have carried thousands of tests using headphones. My point is that listening through speakers will expose information it is not shown using through headphones. This information is important for sound reproduction, but it is not easy to valuate in a consistent and reliable way. And many effects that are not easily heard in single channel isolated are easily noticed when listened in stereo.

BTW, the outcome summary of the study you refer was "These listening tests indicate that as a rule, no significant differences could be heard between DSD and high-resolution PCM (24-bit / 176.4 kHz)", something that many people on this forum do not accept. So IMHO this study does not validate headphone listening as a the ultimate tool for checking differences.

 

[Julf]

BTW, the outcome summary of the study you refer was "These listening tests indicate that as a rule, no significant differences could be heard between DSD and high-resolution PCM (24-bit / 176.4 kHz)", something that many people on this forum do not accept. So IMHO this study does not validate headphone listening as a the ultimate tool for checking differences.

So any study that arrives at results you don't agree with is invalid?

 

[Phelonious Ponk]

Pretty much.

 

[FrantzM]

So any study that arrives at results you don't agree with is invalid?

Pretty much.

 

[jkeny]

When using headphones could it be that taking away the room distribution of direct Vs reflected sound puts us in a different acoustic space - one that is less natural & more difficult for our perceptions to interpret? Our brain is always interpreting the acoustic space we are in - with speakers the interaction with the room is the acoustic space through which we percieve the reproduced musical event. With headphones, there is a very different & unnatural acoustic space which may be a factor?