Surprisingly, John appeared at my booth at RMAF today. Apparently Wilson Audio flew him out. He looked good (John did).
Jan
The web link says the "dynamic range of the AHB2 approaches 130 dB".
I like that word "approaches". Use it myself sometimes...
I have seen dynamic range described as peak output divided by noise floor
(in square root Htz). In that case, an amplifier with a 100V peak need only
have a 30 uV floor, which is not difficult to achieve.
Probably this is not what they mean.
I do have a couple of examples that can do 100V peaks with 30 uV noise
(20 - 20 KHz unweighted), and a couple of listeners who appreciate that, but
they do have very sensitive loudspeakers....
Perhaps it's time for you to let it be known what kind of wine you like.
Works for me...
The web link says the "dynamic range of the AHB2 approaches 130 dB".
I like that word "approaches". Use it myself sometimes...
I have seen dynamic range described as peak output divided by noise floor
(in square root Htz). In that case, an amplifier with a 100V peak need only
have a 30 uV floor, which is not difficult to achieve.
Probably this is not what they mean.
I do have a couple of examples that can do 100V peaks with 30 uV noise
(20 - 20 KHz unweighted), and a couple of listeners who appreciate that, but
they do have very sensitive loudspeakers....
The back of the envelope calculation I did was based on the JAES Dynamic range spec (used only for DACs as far as I know) THD+N dB below a -60 dB reference, A weighted, to 100 % or digital full scale sine wave.
I don't think its that hard to meet if you are serious or have ever sold to Japanese distributors who will tell you immediately you have a problem.
I think the crossover distortion issue they addressed may be a bit of a red herring. I suspect THX won't pay for the patent from their royalties, or any others. There is still no opportunity any more for linear amplifiers in high volume products.
It would be very excellent if someone could give an intuitive explanation of how this circuit accomplishes what it does.
In particular, how does the gain set resistor work ?
I'm in there as the last reference.
Scott
Your patent 2002/0153940 applies to analog amplifiers.
I think your invention scheme can be in good use for varying the psu voltage on the output stage power switchers of class D amplifiers as well, no?
The command signal may track the signal amplitude profile or the position of the volume control
George
Last edited:
I would not argue about his honesty nor that his staff wouldn´t have been "by the book".
But obviously they didn´t follow a book about serious psychoacoustic testing nor analytical sensory testing at all.
Otherwise this test approach would not have included a price to win, and because of this price to win, raise the bar to such an absurd level.
Choosing 5.96E-8 as accepted alpha error leads to an astronimical high risk of getting beta errors.
The reasoning behind this? Well, the money prize was high and people told that the audible differences were like "night and day".
Second oberservation was the log of results; according to the linked website Clark said no one ever got more than 65% correct response.
Quick calculation of the probability for no one getting 16 correct responses out of 24 trials and 900 tests overall gives:
p(X=0) = 3.003E-18
Sounds quite unlikely; so maybe, most people didn´t master the pretest, but we don´t have the information.
Serious lack of information and quite questionable methodology.
Am i surprised that the nil hypothesis couldn´t have been rejected under these test conditions? No.
Well, me and my colleagues could.
As said several times before, i started with controlled listening tests back in the beginning of the 1980s after having read about Shanefield´s arguments.
And i have noticed quickly that rejecting the nil hypothesis in such a test is quite difficult even for big differences if a listener is not used to participate in such tests.
Of course science knows about these difficulties, and therefore tests have to be objective, valid and reliable.
Or to cite JJ on this:
"Do you need controls? Of coures you do"
The alternative is well described in here:
Source:
Søren Bech, Nick Zacharov.: Perceptual audio evaluation : theory, method and application. JohnWiley & Sons Ltd,2006.
Once you step into perceptual/subjective audio evaluation how do you separate more accurate from more pleasing? If you start with comparing two sources you are left with a preference. If you add a third its then which two are closer to each other. Perhaps that is a start but as the references show this is a difficult task.
The lossy audio codecs have been tweaked to a degree where they are really quite good. I'm sure there is a lot knowledge on what we are sensitive to and what we don't hear in that work. I don't see efforts toward applying that knowledge to speakers or electronics.
What I find interesting is that the amplifiers that are least likely to pass such requirements would be amplifiers from the 1930's through the 1960's. However those designs have resurfaced with a strong following and are capable of commanding extravagant prices, possibly because of their flaws. They exhibit clearly audible departures from the parameters known to represent audible differences, frequency response, noise, distortion, etc. however are revered for have exceptional lifelike qualities. Meanwhile amplifiers that are free of known errors often are regarded as lacking in audio quality. Is accuracy a false goal?
The lossy audio codecs have been tweaked to a degree where they are really quite good. I'm sure there is a lot knowledge on what we are sensitive to and what we don't hear in that work. I don't see efforts toward applying that knowledge to speakers or electronics.
effects amps should be chosen by the artists for the genre, their goals in the studio, mastering process
most of us want to play back recordings from many styles, artists, studios in the home with good results, hearing the artists intent - as represented on the recording - not make a Segovia recording sound like Hendrix
psychoacoustic codec development has given us man-decades of controlled blind listening tests, tuning against hearing models - and required many to relearn/reverify the caveats for valid listening tests
the results however haven't required anything new from amp designers - at best gives limits from the masking model that provides cover for some types of distortion relative inaudibility
most of us want to play back recordings from many styles, artists, studios in the home with good results, hearing the artists intent - as represented on the recording - not make a Segovia recording sound like Hendrix
psychoacoustic codec development has given us man-decades of controlled blind listening tests, tuning against hearing models - and required many to relearn/reverify the caveats for valid listening tests
the results however haven't required anything new from amp designers - at best gives limits from the masking model that provides cover for some types of distortion relative inaudibility
Last edited:
pointless objection Steve - by that reasoning then commercial music recording, distribution, home listeing has been a fraud, failure from the start?
speakers and rooms, their interaction are certainly the large, "clearly audible" variables in home listening
heck, in the past decade a few hundred million have been listening with earbuds
but I think the recent topic here is amplifiers/preamps electronics's "audibility" - do John, Charles, Nelson's products "sound better"
even could be refined to - do their electronics sound better because they avoid some flaw that we can't seem to measure - or hear in blind listening tests
that some prefer euphonic distortions isn't to many objectivist's very interesting
since we are pretty sure there are some measurable deviations from "accuracy" that can be heard
but at least John here is adamant that he never increases distortion with his designs
speakers and rooms, their interaction are certainly the large, "clearly audible" variables in home listening
heck, in the past decade a few hundred million have been listening with earbuds
but I think the recent topic here is amplifiers/preamps electronics's "audibility" - do John, Charles, Nelson's products "sound better"
even could be refined to - do their electronics sound better because they avoid some flaw that we can't seem to measure - or hear in blind listening tests
that some prefer euphonic distortions isn't to many objectivist's very interesting
since we are pretty sure there are some measurable deviations from "accuracy" that can be heard
but at least John here is adamant that he never increases distortion with his designs
Last edited:
There is a method which is very effective, which relies on using musical tracks as the key determinant. The usual technique of listening to 'high quality' tracks is about the worst approach one can use, IME, because it's equivalent to comparing cars by driving them at cruising speeds on a new highway - you need a few nasty bumps, to separate the men from the boys.
The iterative process I use is that if I think or am cajoled into perceiving that a system is up to scratch, is to put on a more testing track. If this shows up to be unpleasant, or bland, then I'm back at the coalface again, this is showing me where the problems are.
How do I know that I'm going in the right direction? Because these are tracks that I know can present at a certain standard, because I've heard them perform at that higher level on a number of occasions. And why do I think what I heard then was in fact a "higher" quality, not just made more "pleasant"? For the same reasons one assesses wines tasted as being superior or not - a higher class wine sends signals in a myriad of ways that one is dealing with a different animal, a "roughie" can't be magically transformed into such by adding sugar, or watering it down; when something is genuinely 'better' a greater complexity within is clearly revealed, without bitterness, off flavours, or unpleasant aftertastes ...
Last edited:
- Status
- Not open for further replies.