Some of my evolving thoughts are here: http://www.diyaudio.com/forums/blogs/pano/620-more-distortion-yes-please.html
Don't get me wrong, I LOVE to measure - shootz, I even flew 1000s of miles last month just to do some measurements. But getting what is measured to correspond to what I hear is sometimes difficult. There is always a subjective note of "do I like it?" And that's an important point.
The subjective side comes when I sit down, put on a piece of music and hear "The Real Thing" or just some speakers and other gear. I.E. - does it sound like what I might hear live? Every recording is different and so is every live event, but there is certainly an overall feel and sound that one can quickly identify as "real." Not many systems I've heard can do it, but it is possible. That's my goal - it has to sound like real music played by real people on real instruments in a real space. That's what I really enjoy, so that's my goal.
Objective design choices, measurements and techniques are simply tools to get me to my ultimate subjective goal.
Hard to come up with a measurement with zero correlation to how something actually sounds. I thought a bit and wondered about weight of amplifier. Nope, almost certainly on average a heavier amp will sound better. How about a measurement of width of amplifier front panel? But even that probably has a non-zero correlation with sound (probably slightly positive). How about 'reflection coefficient of amplifier front panel' ? Do amps with matt black front panels sound better than those with high gloss ones?
I'm forced to the conclusion that the whole notion is a straw man.
Yes, Pano, I think that in fact digital audio has been the genesis of the chasm that seems to be growing between the subjectivist & objectivist viewpoint. Why? Because I think it introduced a mathematical basis for the recording of audio that most people can't understand & at the same time it was advertised as perfect & the public bought into it, by & large.
This has led to the situation where a lot of people believe that what they hear from their digital audio systems is perfect because it's not obviously flawed i.e no hiss, no scratches, no clips, no wow & flutter, etc. Maybe it doesn't sound like the original event - it sounds even "better"
. This commonly encountered perception of "better" has to be factored into subjective reports of the sound of a device. I'm not sure a lot of people use live performance as the criteria for evaluating an audio system.
What I'm trying to say is that a lot of people are listening to the distortion of digital audio in the form of jitter, CM noise, etc. & thinking/assuming that it is the way it should sound - ultra-realism with lots of attack, etc..
The argument against all that I've said is that it measures better (according to the measurements that are currently used) & this is why it is more insidious & more detrimental - it is more subtle, less obvious & therefore easier to defend as correct.
In the end, this conflict will be to the betterment of audio IFF we actually do strive towards a set of existing or new measurements that more closely match what we perceive as "better" sounding.
This has led to the situation where a lot of people believe that what they hear from their digital audio systems is perfect because it's not obviously flawed i.e no hiss, no scratches, no clips, no wow & flutter, etc. Maybe it doesn't sound like the original event - it sounds even "better"
. This commonly encountered perception of "better" has to be factored into subjective reports of the sound of a device. I'm not sure a lot of people use live performance as the criteria for evaluating an audio system.What I'm trying to say is that a lot of people are listening to the distortion of digital audio in the form of jitter, CM noise, etc. & thinking/assuming that it is the way it should sound - ultra-realism with lots of attack, etc..
The argument against all that I've said is that it measures better (according to the measurements that are currently used) & this is why it is more insidious & more detrimental - it is more subtle, less obvious & therefore easier to defend as correct.
In the end, this conflict will be to the betterment of audio IFF we actually do strive towards a set of existing or new measurements that more closely match what we perceive as "better" sounding.
True up to a point
I have no problem with there being confrontation (as long as it remains civil) . Sometimes it is pointless, but sometimes it results in a different perspective, or perhaps even an aha moment. If nothing else it allows one to get a better understanding of the views of others (provided we read what is written rationally and without presupposition). The reason I started this thread was to try and steer the conversation to an area where it *was* the topic rather than degrading the S/R in other threads where it might rightly be brought up, but tends to drag on and detract from the original threads intent.
Tony.
OK, I'll bite...(putting on flame proof suit)
Anything that can be heard, can be measured. After all , your ears are measuring devices, just poorly calibrated ones. There's also the fact that many different things, internal or external to the listener can affect what is heard (or what one believes they hear). I can listen to my system one day, and it sounds relatively crappy, and the following day, without changing anything, will sound good. Same exact program source, no change in volume setting or anything else. So why the difference? Could be anything from a change in mood, a chunk of ear wax that fell out, or maybe an increase in noise on the power grid. My point is this; the human hearing system is falible and easily fooled, especially when one is "splitting hairs" trying to hear the difference, say, between this resistor and that one. And I'm also not saying no differeces are heard, just that we need to realize that there's no magic involved, if a real difference is heard, then something must be different, and therefore must be measurable in some way, like noise, frequency response, phase change, whatever.
I was somewhat amused following a recent thread here where it was claimed that two separate digital audio files with identical checksums could sound different, not possible. I'd say the percieved differece most likely occured after the sound left the speaker.
The most important thing for me is to just relax and enjoy, and if it isn't sounding good today, it will probably sound better tomorrow. I guess that puts me more on the objective side of this discussion, but it really comes down to every individual's biases and motivations.
Mike
Anything that can be heard, can be measured. After all , your ears are measuring devices, just poorly calibrated ones. There's also the fact that many different things, internal or external to the listener can affect what is heard (or what one believes they hear). I can listen to my system one day, and it sounds relatively crappy, and the following day, without changing anything, will sound good. Same exact program source, no change in volume setting or anything else. So why the difference? Could be anything from a change in mood, a chunk of ear wax that fell out, or maybe an increase in noise on the power grid. My point is this; the human hearing system is falible and easily fooled, especially when one is "splitting hairs" trying to hear the difference, say, between this resistor and that one. And I'm also not saying no differeces are heard, just that we need to realize that there's no magic involved, if a real difference is heard, then something must be different, and therefore must be measurable in some way, like noise, frequency response, phase change, whatever.
I was somewhat amused following a recent thread here where it was claimed that two separate digital audio files with identical checksums could sound different, not possible. I'd say the percieved differece most likely occured after the sound left the speaker.
The most important thing for me is to just relax and enjoy, and if it isn't sounding good today, it will probably sound better tomorrow. I guess that puts me more on the objective side of this discussion, but it really comes down to every individual's biases and motivations.
Mike
Just a quick question, based on these two statements. Most if not all of us are in audio for the enjoyment. So why do we focus so much on the sound, and not on the emotional response to said sound? Emotions are objective, they are chemicals whose concentration can be measured. MRIs can show up whether someone's enjoying themselves or not. So can GSR to some extent and facial expression metrics (FACS by Ekman). There are plenty of methods available so has anyone tried using these?
Good question. For me, I love the sound just as much as I love the music. Having grown up playing and hearing live music, then going on to work in the biz, the sound itself has real emotional weight for me.
That's why I chase it.
To expand a little on my previous post, I don't get myself caught up in an endless persuit of perfection. I design and build my gear to satisfy my desire to enjoy the music, and hopfully have a little fun while doing it. When I've reached a point where I feel it's good and I've done the best I can with what I have to work with, it's time to relax and enjoy. I can't afford $30.00 resistors, and even if I could, probably wouldn't use them anyway. We can drive ourselves looney striving for the last little bit of perfection, or realize that at some point you reach a point of dimishing returns, and let it be. Perfection is made of unobtainium.
Just relax and enjoy the music.
Mike
Just relax and enjoy the music.
Mike
Hi,
We use the empirical methode, in some cases we simply fully rebuild and scrap teh boards we cannot find the cause.
Ciao T
We use the empirical methode, in some cases we simply fully rebuild and scrap teh boards we cannot find the cause.
Ciao T
Hi,
I likes good SNR. Lets keep the SNR in most threads high.
I would BTW give a salutary mention to Bob Cordell who in his Power Amplifier Book, instead of completely rubbishing anything from the subjective corner not only examines seriously much of it, find's some interesting things and actually even spends a fair few pages on how to design zero feedback amplifiers even though he clearly does not "believe" in them. I think his approach would behove some of the other supposed objectivists well.
Samuel Groners work (including his commentary on D. Selfs Book) also illustrates an approach that is by far more reasonable and approaches synthesis than the kind of debate more commonly observed.
Ciao T
True up to a point
The reason I started this thread was to try and steer the conversation to an area where it *was* the topic rather than degrading the S/R in other threads
I likes good SNR. Lets keep the SNR in most threads high.
I would BTW give a salutary mention to Bob Cordell who in his Power Amplifier Book, instead of completely rubbishing anything from the subjective corner not only examines seriously much of it, find's some interesting things and actually even spends a fair few pages on how to design zero feedback amplifiers even though he clearly does not "believe" in them. I think his approach would behove some of the other supposed objectivists well.
Samuel Groners work (including his commentary on D. Selfs Book) also illustrates an approach that is by far more reasonable and approaches synthesis than the kind of debate more commonly observed.
Ciao T
One of the suggested explanations evokes the Schumann electromagnetic resonance and it's daily variations (see Wikipedia). They are supposed to create a brainwave entrainment and then facilitating our access to Audio Nirvana. But this depends of the resonance peak. Ouch.
Of course, there is the usual commercial exploitation and all the hoax. Schumann resonance exists. Brainwaves can be entrained by external factors. But the relation between both is far from being established.
Hi,
To be clear, are you saying that anything that can be heard is covered by measurements that are extant, fully developed and in wide or use?
Or are you merely stating that if that something can be heard it must have a origin that is within the laws of nature (or in other words is not supra-natural in nature) and hence it is possible to devise methods that allow these effects to be quantified (measured)?
Why would you say "not possible"?
For arguments sake, lets have in example the file widely distributed across the HDD surface (read the file is heavily fragmented) and a second that is a single contiguous file. The fragmented file will cause more disk activity, you can easily measure the effect from that with a scope on the PSU lines in the PC and with considerably more difficulty on the output from the PC.
The problem is that we have a "hidden variable" not accounted for, which can cause precisely the same data, on the same system to cause vaied results.
The most important thing for me is to just relax and enjoy, and if it isn't sounding good today and no better tomorrow and the day after, it sounds bad and I will not attempt to talk it to good (which really would be magic if it could be done). I guess that puts me more on the subjective side of this discussion, but it really comes down to every individual's biases and motivations.
The funny part, the goals are the same, however I am more of a realist and gnostic, who accepts the reality, the differences and deals with them, instead of an optimist and believer, who wishes that there are no differences and wishes reality would reliably conform to an idealised and simplified theory and then talks himself into believing it is really so...
Ciao T
To be clear, are you saying that anything that can be heard is covered by measurements that are extant, fully developed and in wide or use?
Or are you merely stating that if that something can be heard it must have a origin that is within the laws of nature (or in other words is not supra-natural in nature) and hence it is possible to devise methods that allow these effects to be quantified (measured)?
Why would you say "not possible"?
For arguments sake, lets have in example the file widely distributed across the HDD surface (read the file is heavily fragmented) and a second that is a single contiguous file. The fragmented file will cause more disk activity, you can easily measure the effect from that with a scope on the PSU lines in the PC and with considerably more difficulty on the output from the PC.
The problem is that we have a "hidden variable" not accounted for, which can cause precisely the same data, on the same system to cause vaied results.
The most important thing for me is to just relax and enjoy, and if it isn't sounding good today and no better tomorrow and the day after, it sounds bad and I will not attempt to talk it to good (which really would be magic if it could be done). I guess that puts me more on the subjective side of this discussion, but it really comes down to every individual's biases and motivations.
The funny part, the goals are the same, however I am more of a realist and gnostic, who accepts the reality, the differences and deals with them, instead of an optimist and believer, who wishes that there are no differences and wishes reality would reliably conform to an idealised and simplified theory and then talks himself into believing it is really so...
Ciao T
Yes, you can generate an artificial and hypothetical situation that would void the example, but you have to ADD a variable. The quoted statement said nothing about the playback or storage mediums and so is 100% correct - identical files will play identically ALL ELSE BEING EQUAL
The capitalised rider could be pedantically added to every statement, but why do that just to tame the pedants in our midst?
It might be possible to measure anything that can be heard - after all as ThorstenL hints above, if it originates in nature it must be measureable. However, what I'm not sure of is whether the ability to measure everything currently exists.
An example (and someone please jump in if I've missed how to do this): what is the parameter/set of parameters to measure soundstage?
This is one of the reasons why (again as ThorstenL has pointed out) a designer will use a combination of measurements and listening.
The flame wars in other threads seem to occur when someone says "show me the measurements to prove that it has a deeper soundstage" or some such......
Fran
Very good question fran and something I'd be very interested to see the answer to!
As soundstage is basically an auditory illusion (ie our brain is part of the equation) I would think that devising measurements that could predict soundstage would be very difficult!
My current room and setup unfortunately does not provide me with much in the way of soundstage. I can hear stereo effects but not the 3D soundstage which people strive. In fact I don't think I've ever had a system where this was present on a consistent repeatable basis.
I have on three occasions been completely fooled by my speakers.
The first time was many years ago, I was watching a movie and thought that's odd there seems to be a helicopter outside up quite high, turned out it was in the movie, as as the position of the helicopter changed it came into view on the TV screen.
Second time was listening to Radio Kaos (Roger waters) there is a dog bark on the track. I could have sworn the dog was outside at least 20M to the right of the right speaker (litterally through the wall). Never managed to reproduce that though.
Third time I was watching a TV commercial and a child said something behind me, the camera then panned around to reveal the child behind the camera position . That really freaked me out! The listening position was in the middle of the room with a good 4 to 5 meters between me and the back wall.
This was all with stereo source material and two (not particularly good) speakers. The two TV instances were standard analog stereo TV broadcast!. No mind altering substances were involved
These are all cases where I would swear black and blue that I heard it, and there were no visual clues to help with the illusion. I have of course no way of proving it. The part that I've always been baffled by is that I didn't think that effects outside the width of the speakers, or even more bizarrely above or behind me, were actually possible.
Tony.
Something Pano introduced in his blog linked to above which is worth reading, btw. Distortion is often introduced into a system to improve the final perception - dither in digital audio, the Globe effect in perception. These are introduced to compensate for a less than ideal characteristic in the system
Could it be the case that we have some form of non-ideal behaviour in the audio reproduction process, from microphones through to speakers & ending finally in the non-ideal behaviour of the ear which is the target instrument for our audio reproduction system. The ideal goal would be to deliver to the ear exactly the same signals that it encounters at the live event but this is impossible so we are left with trying to create this illusion.
Does anybody have a handle on how this illusion is created. Is it even possible to do so?
Continuing the analogy with visual perception, it strikes me that we are at the stage in audio where art was before perspective was discovered. At the moment we seem to be trying to get each & every colour & tone as accurately rendered as possible without the more holistic view of what we are trying to portray.
Is digital audio the equivalent of the Pointilism movement in art? Whereas Pointilism was trying to recreate an overall effect such as shimmering sunlight, digital audio pointilism is too focused on the individual colours, size, density & exactness of the points & losing the overall picture? Anyway, that's not to say that reducing or eliminating distortion at each stage is not a worthwhile & necessary goal but not at the expense of the final goal of a natural portrayal of the original event.
Who knows maybe we are being completely unrealistic in our expectations that this portrayal can be successfully achieved with 2 speakers? Maybe we need audio cubism
Could it be the case that we have some form of non-ideal behaviour in the audio reproduction process, from microphones through to speakers & ending finally in the non-ideal behaviour of the ear which is the target instrument for our audio reproduction system. The ideal goal would be to deliver to the ear exactly the same signals that it encounters at the live event but this is impossible so we are left with trying to create this illusion.
Does anybody have a handle on how this illusion is created. Is it even possible to do so?
Continuing the analogy with visual perception, it strikes me that we are at the stage in audio where art was before perspective was discovered. At the moment we seem to be trying to get each & every colour & tone as accurately rendered as possible without the more holistic view of what we are trying to portray.
Is digital audio the equivalent of the Pointilism movement in art? Whereas Pointilism was trying to recreate an overall effect such as shimmering sunlight, digital audio pointilism is too focused on the individual colours, size, density & exactness of the points & losing the overall picture? Anyway, that's not to say that reducing or eliminating distortion at each stage is not a worthwhile & necessary goal but not at the expense of the final goal of a natural portrayal of the original event.
Who knows maybe we are being completely unrealistic in our expectations that this portrayal can be successfully achieved with 2 speakers? Maybe we need audio cubism
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.