|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
160.62.7.250
In Reply to: RE: Psychoacoustics? posted by Frihed89 on November 25, 2016 at 16:05:56
And then using that relationship to design products that fit better with the way we hear in a more scientific, engineering based approach.
starting back in the 1940s (at least if not before) some engineers were looking into the relationship between what most people find the best sounding (whatever that means but hopefully means more like a real instrument in real space) and the measurements of the gear that is generating that more pleasing sound.
Starting with Shorter from the BBC and later Hiraga people have tried to generate corrlations between distortion and listener impact. More sophisticated recent efforts to come up with a quantitative assessment of sound quality can be found in the Thesis of Cheever and the papers from Geddes. Their equations can (in theory at least) be used to evaluate if a design will likely be favorable to listeners or unfavorable sonically.
Keith Howard also looked empirically at what adding different distortion patterns to a digital file would have on the sound quality. Unsurprisingly, he found that no added distortion sounded the best but that a monotonic pattern (exponentially decreasing amplitude of harmonics as the harmonic order increases) was the least offensive and that all odd harmonic (typcial of most push/pull circuits...by design) patterns were the most offensive.
For speakers, work has been done by O'Toole and Geddes and many others where things like driver break up, cabinet resonance, dispersion directivity, dynamic compression etc. are probably more important than harmonic distortion, which is mostly low order anyway in a speaker.
When you take these things into account, and yes it is statistical rather than absolute, you can begin to design products that might in an absolute objectivist sense fall short but will satisfy more people more of the time.
One must never forget that with humans there will always be exceptions that will leave room for dissent and alternative preference for something other than the most psychoacoustically correct designs.
However, for the most part, psychoacoustically correct designs largely do not exist except for trial and error where the designer hit upon some principles leading to a good sound. Only Vladimir Lamm, to the best of my knowledge, claims to use a hearing model to optimize his product designs. Apparently, adherence to the model, rather than listening tests are the key to his products sounds. Based on what I have heard from Lamm products his model must be pretty good.
I don't see it at all as someone trying to "Jam" linearity down anyone's throat. In fact, it is often the case that the less linear "objective" measuring piece of gear is in fact the one that does best psychoacoustically. Read Cheever, he does a good job exposing this.
Follow Ups:
That was my original understanding on the basis of limited reading.
I would love to see an argument that poorer linearity sounds more 'real'.
Easy, when listening to a subwoofer, people nearly always pick the one that "sound most like" a kick drum.
The problem is, a subwoofer is ONLY supposed to produce what comes out of the low pass crossover which is just the lowest parts of the drum.
The subwoofer with greater distortion and / or least low extension will sound most like the drum because it has a wider spectrum, even though it's not part of the input signal.
Another example, in recording tweaked electronics were used to add warmth (Beatles white album was the first one i knew of). Even harmonics are musically related and can add that warmth but is another non-linearity.
Now, what is the faithful reproduce is the issue and ears are not that reliable while a generation loss recording can make warts standout like soot on a white carpet.
I don't think the poster you replied to was talking about what sounded real, but what they liked.
SET amps have higher THD than most well-engineered SS amps, but they still have a following.
They have high second harmonic primarily because they are non-linear. They amplify the positive side of a signal differently than the negative side which is compressed relative to the positive side. Second harmonic is very pretty and a very musical distortion.
It is also inaudible up to about 2%...with a pure sine wave. This means that with real music, even higher levels are likely to go unnoticed.
Keith Howard reported that even predominantly 2nd harmonic was not better sounding than an undistorted file when he made digital files with added distortion of different types. He did find; however, that a monotonic pattern (i.e. with even and odd harmonics falling away exponentially) was the least damaging of the file. This is totally consistent with what Hiraga was claiming and also what Cheever found.
All odd harmonic patterns, even at low levels, were the least pleasant and all even patterns were in the middle somewhere. Most well designed push/pull amps will have predominantly odd harmonics.
The human ear/brain is accustomed to a monotonic pattern, which is what the ear mechanism generates on it's own. Therefore, masking is pretty good if that pattern, at a given SPL, is followed closely.
Since there is no electronic equivalent to the undistorted file, the best we can do at this time is to have the least damaging pattern.
Note, Mr. Howard did not find ANY of the patterns Euphonic and preferrable to the undistorted original.
Yes, but still nonlinear based on THD. To me they present the "illusion" of a live performance, which is fine with me.
I think a lot of that has to do with the relative lack of distortion at high frequencies, which is superior with SETs, at least at modest powers. This definitely helps with the 3d effect of images and soundstage. HF distortion flattens both due to how these high frequency distortions are perceived.
Hi... new to this.
Interesting discussion about something I have no idea about.
Does that mean that a hybrid bi-amp situation where a tube amp powers the tweeter and a solid state amp the woofer could be worth the effort?
It could be worth the effort but this does not mean that.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: