THD is a key measurement when designing an electronic HiFi device, but I guess it is less important as soon as you connect a loudspeaker which has a distortion of about, hopefully, 1%... and then you listen to music in a room without "room correction" or another treatment to correct room modes and RT.
So... it doesn't matter if it is a good metric or not, at least in the human hearing context,, because it isn't the only "distortion" in the game, and it isn't the most significative one 😱
It's far more complex to analize the full context, but - I guess - it is the only way to do that...
Distortion in the electrical domain is the most important element, IME. In the earlier post, of a grab bag of audio jargon, which I think most people can relate to, I can get a system which is suffering from the "bad" stuff to head over to the good side by doing absolutely zero about the room and the physical aspects of the speakers - all the solutions lie with parts involved in electrical behaviour.
Yes, the speaker may have high levels of distortion at certain, typically low, frequencies, but subjectively this doesn't stop the sound from "working" - it has been many years since I last tried a driver, relatively cheap, which had a definite, disturbing, distortion signature which wouldn't go away when fed with clean sound ...
You must see all the problem at once 😉
Electrical distortions IS important as an electrical parameter, but once it has been taken under 0.01% or even lower (just see D. Self designs), it becomes completely buried into the "distortion floor 😱" of loudspeakers and you should forget it for all uses and purposes.
Subjectively ... it stops nothing at all 😀
Subjectively ... is just inside your brain, so nobody knows what you hear neither why you hear that way. The driver can be cheap or expensive, but what really matters is its distortion measurements and the effect of the environment surrounding it. After all, this is what you share (other than the music) if you are listening with a partner, because he/she isn't inside your head or inside your ears.
Perhaps we could talk about a couple of psycho effects which can take place in the group, but I guess it is a different story...
That thinking is why the audio industry is still a mess: 20 years ago I made a habit of visiting hifi dealers to see if anyone "got" how to achieve decent sound without going the extra yards required that I used. I got tired of rolling my eyes and listening to the dribble spouted by the sales chappie about how great the sound was, and gave up the exercise.
The most recent hifi show demonstrated that significant movement forward has occurred: there were at least a few systems that weren't totally dreary and instantly forgettable. In each and every case the non acceptable sound was obviously due to poorly organised and tweaked electronics - the poor speakers just had to deal with the crud fed to them as best they could ...
Yes, there is always the niggling thought that what you hear is not what others hear, and I'm certainly aware that people listen for certain types of sound elements in a track, so that they can rank the performance of a system. However, I now believe that many people really think their own system is "the one" - that's actually got the sound right - and what they're really listening for is a "better" version of exactly how the track sounds on their own gear; if it sounds significantly, or radically different then the unknown system is getting it "wrong", and will be marked down!
That said, I've gone through the experience too many times, with too many people, of having a system doing what I call "good" sound vs. "bad", and them being in agreement with my perceptions, to believe that I've got it totally wrong ...
Last edited:
An interesting, other experience of what can happen when it comes together: My Experiences Building a Total Shunyata Zitron System--A Paradigm Shift. Now, this chap's done it by the throwing of expensive, pixie dust add-ons into the mix, so one has to just bypass that rhetoric; what counts is the outcome, which can come about, always, if one happens to get the right mix: key to the whole she-bang is the using of the Bryston components, these are technically good enough for the sound to fall into place once the disturbing distortions are quelled through a happy marriage of the other bits and pieces.
Essential to note that the big money is not necessary to get it to gell; knowledge and understanding are the most important ingredients, and successful sound can occur with cheap-as-chips setups if the right approach is taken ...
Essential to note that the big money is not necessary to get it to gell; knowledge and understanding are the most important ingredients, and successful sound can occur with cheap-as-chips setups if the right approach is taken ...
Geddes doesn't seem to think "transparent" audio electronics is hard to do, and good enough electronics can be had very cheaply
his Summa speakers were last sold for $6k each, he spent < 2% of a pair's price on his demo amp
it is hard for me to parse what fas42 is really saying - but to elevate elctronics over loudspeakers, directivity, diffraction, early reflection control, room influences is directly opposed to Geddes work - and many other with published peer reviewed studies - hard to understand how fas42 isn't peddling snake oil himself
his Summa speakers were last sold for $6k each, he spent < 2% of a pair's price on his demo amp
it is hard for me to parse what fas42 is really saying - but to elevate elctronics over loudspeakers, directivity, diffraction, early reflection control, room influences is directly opposed to Geddes work - and many other with published peer reviewed studies - hard to understand how fas42 isn't peddling snake oil himself
I can appreciate that. It's an epiphany moment when one gets it, and until you experience and appreciate the quantum jump in subjective quality that occurs when the electronics function sufficiently well, then I would suppose it comes across as extremely fanciful. The simple answer why Geddes and others don't use this "measure" is because they've not experienced it themselves, the electronics they were using were not sufficiently competent -- in some ways it's like breaking the sound barrier, how many "experts" considered it was not appropriate to try to do this, because they had "calculated" that too many issues would occur ...?
Once this level of quality is reached, then one automatically compares everything you listen to against that, to a large degree you can assess how sufficient the quality is in key areas, what extra is needed to able to go through the "barrier" -- how close to the speed of sound the system is currently running at ...
Edit: Didn't appreciate the quoted remarks of Geddes on first read -- that's fine, he's found a low cost amplifier that performs very well by a certain metric he's devised. Now, that doesn't guarantee that that it will easily perform to a sufficient standard - it does probably mean that it's running fairly close to what's required in raw form ...
Last edited:
There's a simple "big picture" way of looking at what's going on:
1. The aim of enthusiast's audio playback is create a convincing illusion of a musical event
2. This has been done many times over the years so people know that it is possible
3. IME, key to this happening is that the subjective audibility of certain types of distortion generated by the playback mechanism is sufficiently low
4. The conventional, highly researched method of minimising this subjective audibility is to work with the speakers and their interaction with the listening space. A whole suite of finely crafted techniques have evolved, which if correctly implemented are very effective at doing this.
5. However, there is another technique: reduce the key distortion prior to it being fed to the speaker driver. If this is done adequately then the application of the conventional methods is not necessary; yes, they will enhance the illusion to some degree if also applied -- but, they are no longer a necessary condition ...
1. The aim of enthusiast's audio playback is create a convincing illusion of a musical event
2. This has been done many times over the years so people know that it is possible
3. IME, key to this happening is that the subjective audibility of certain types of distortion generated by the playback mechanism is sufficiently low
4. The conventional, highly researched method of minimising this subjective audibility is to work with the speakers and their interaction with the listening space. A whole suite of finely crafted techniques have evolved, which if correctly implemented are very effective at doing this.
5. However, there is another technique: reduce the key distortion prior to it being fed to the speaker driver. If this is done adequately then the application of the conventional methods is not necessary; yes, they will enhance the illusion to some degree if also applied -- but, they are no longer a necessary condition ...
Last edited:
Yes, but this condition can be met by creating the illusion of a musical event, but not necessarily the one that was recorded.
Something like this, that fed with a suitable recording (of a violin probably!) can possibly create a plausible holographic 'sound field' in your living room by resonating just like a violin in your living room:
Violinspeaker - Product
Do we know that many audiophiles are not really attempting something half way similar with their open baffles and on? Trying to take a dry recording and boost their living room's ambience to the extent that it sounds like a performance in front of them. A neat trick, but really only good for girl-and-guitar type music, perhaps.
A player piano or fairground organ might also be set up to play back MIDI files for a similar lifelike illusion of a musical event!
Attachments
...and back to snake-oil! Do I also need a cello speaker, a flute-speaker, and a kettle-drum speaker? Do I need to tune my speakers before every "performance"?
(Me) "It's the 'half' where distortion would be at about 100% on a vinyl LP. "
I suppose I should have said THD+N: the typical SNR of vinyl is supposedly 50dB.
So a signal at -48dB from peak ('half' of 16 bit dynamic range) on an LP is reputedly at a distortion level of hundredths of a percent... I dare say the same could be claimed for a signal broadcast over AM radio. But does that actually mean anything in amongst the noise, pops and clicks, and the inherent 'grain' of the analogue medium?
I suppose that it is possible to take any signal buried in any level of un-correlated interference and, if you integrate for long enough, find its average level of distortion. But you'll never actually hear the purity of the sine wave yourself, because at virtually no point will you ever hear the 'correct' level: it will always be subject to a massive random error.
The digitally sampled waveform cannot claim its 'noise' is un-correlated, but it can claim that the amplitude is low. (As you say, 1% on a signal already at -48dB). But then there's dither...
Well, should be no need for anyone to go into silliness to make sound, well, sound good ...
The truth is, that there is no such such thing as a completely dry recording -- in spite of the best efforts of the sound engineers, the hallmark cues from subtle echos off the surroundings when recording are still picked up by the mic's. Unless they record in a 100% perfect anechoic chamber, with every surface of themselves and their instruments loaded up with sound absorbing material there will still be some reverberation occurring. Yes, very low level, subtle, but it is there. Which is why one should work hard, very hard to improve the "transparency" of the playback, because then those telltale clues will be clearly audible, and acoustically the sound as it emerges from the speaker drivers makes sense to your hearing mechanism - it sounds 'real'.
When a system works well every completely unknown recording is an adventure, you're drawn in to explore the acoustic space revealed: it may be huge, or have very small spaces around each sound element in the mix; but they're all interesting, and it's satisfying visiting the spaces in each recording ...
A good example of a fascinating soundscape is Michael Jackson's "Bad": the tracks on this demonstrate a vast labyrinth of sound that makes one feel you can crawl around for hours, exploring it all ... 🙂
Last edited:
hmmm..... either your tastes are very different from mine, or you haven't heard any well implemented Open Baffles. Or maybe you're just guessing.
Not all Open Baffle speakers are wimpy and defused.
Girl and guitar music is about the last thing I would feed my OB's. They prefer Pink Floyd, Depeche Mode and Metallica! Last night my neighbors got to listen to some dubstep.....
When the girl wants to play her guitar and sing, she prefers some full range horns and an SE tube amp.
I am one of those EE types. I build my own equipment, I measure it with the finest HP or Agilent equipment, then ignore the numbers. My cables....
Walmart! Due to the large population of thumpa-thumpa cars in South Florida you can get good 12 gauge speaker wire for $12 a roll. They have decent interconnects and even 1 Farad caps.
I work in a large electronics plant that had about 1000 engineers before the layoffs. There were about 5 of us who built audio stuff (a lot more back in the late 70's), and 3 of them were into the numbers. You know the ones on the front of that fancy Agilent stuff. The two of us who knew better both play guitar through tube amps.....coincidence????
The thing is, to get your sound right, and then add the ability to go as loud as you'll ever likely want to go, for a peak transient, just as cleanly. At one stage I was thinking 120dB, but this has evolved: I would be aiming now for the ability to hit 132dB as a transient SPL. Your ears can handle this sort of thing, because they're doing it all the time in everday living, when there's a sudden impact sound ...
I wasn't suggesting specifically that open baffles are wimpy, more that they (and others I have seen with rear- and side-firing drivers..?) are harnessing the ambience of the room more than a conventional box design. I was thinking along the lines that if the room's ambience dominates over the recording's, it might help in creating the illusion of a musical performance in your living room, but that the choice of material would be important in how strong the illusion is.
My only experience of these unconventional speakers has been at a show where I wasn't impressed: the 'phasiness' made me feel a little queasy. As always, they were playing girl-and-guitar, jazz, or Dave Brubeck's Take Five. The one exception was a room where they were playing some orchestral waltz music, and this didn't work for me at all.
Reading about these unconventional speakers in general, a common sentiment seems to be that placement is critical - which I automatically interpret as meaning there is no "critical" best position, merely several positions that sound better or worse with this recording or that recording (I think there's a common audiophile belief that there is always a perfect setup that will be best with all music if only it can be found). I see the speaker-and-room interaction as being one with many dimensions, and for which it is highly likely that there is no perfect solution; merely several not-too-bad compromises. Compared to box speakers, the open baffle/rear-firing drivers type of thing increases the number of dimensions in the problem does it not?
The levels I use when listening would never expose my ears to 132dB, and nor would I want it to. Can't think of anything in my everyday living that would be that loud either. 😕
Obviously it can't be done with normal equipment, the maximum I can do currently is about 106dB. But it is a worthy goal, because then compression at any stage is never an issue. Remember, the intent is never to overload the hearing mechanism on a sustained or ongoing basis, merely to have the capacity to reach that sound level for a split second, so that completely realistic sound can be reproduced.
In terms of everyday living, if one solid object hits another at reasonably high speed, at the precise moment of the initial impact the extremely short pulse of impact sound can reach those sorts of levels - this is why serious microphones are typically spec'd to handle volumes with these sorts of numbers.
Normal music recordings are somewhat compressed, so it would be silly to run them such that the peak of 132dB is reached several times. However, if one were to have a uncompressed recording of a natural event where true peaks at this level occurred, then one could faithfully reproduce the experience.
It's highly unlikely at the moment that I would actually assemble such a system, but it would be an interesting technical challenge to do this at a moderate cost, while retaining the quality of reproduction. Large studio monitors and more substantial PA systems can do this without much difficulty, but they cost a bit or the quality's not too brilliant ...
In terms of everyday living, if one solid object hits another at reasonably high speed, at the precise moment of the initial impact the extremely short pulse of impact sound can reach those sorts of levels - this is why serious microphones are typically spec'd to handle volumes with these sorts of numbers.
Normal music recordings are somewhat compressed, so it would be silly to run them such that the peak of 132dB is reached several times. However, if one were to have a uncompressed recording of a natural event where true peaks at this level occurred, then one could faithfully reproduce the experience.
It's highly unlikely at the moment that I would actually assemble such a system, but it would be an interesting technical challenge to do this at a moderate cost, while retaining the quality of reproduction. Large studio monitors and more substantial PA systems can do this without much difficulty, but they cost a bit or the quality's not too brilliant ...
- Status
- Not open for further replies.