Mike Reed
pfm Member
That about encapsulates how I feel about this 'fidelity' issue. Monkeys over measurements any day!
Naim = bright?
- Thread starter zarniwoop
- Start date
- Status
That about encapsulates how I feel about this 'fidelity' issue. Monkeys over measurements any day!
sergeauckland
pfm Member
I'm flattered to be compared to such a luminary, but sadly, I'm not the reincarnation of JH.
If only there were more reviewers like him around, the HiFi industry might not be the foo-ridden swamp it now is.
S.
sergeauckland
pfm Member
I'm not at all sure if the foo-ridden swamp is actually better (that word again) than no swamp at all.
However, there would still be player and amplifier and loudspeaker manufacturers, but perhaps fewer of them, and selling more on design and style than foo.
S.
Steven Toy
Accuphase newbie
I said more not all.
The room at Scalford was even smaller, especially for being packed for most of the day
You gave away free chocolate? ;0)
Only kidding dude, I'm sure your stereo sounds very nice.
russel
./_dazed_and_confused
Yes, you drive the output pair with a current source ISTR there was non switching design in Wireless World that did that. All I can say is I saw both switch off in an amp I measured, I know the theory says they both should not but both did. It was however a long time ago, tens of years, I cannot for the life of me remember the amplifier it would be interesting to see the circuit.
Back to the emitter resistors and as far as I am aware the main reason is for thermal stability, I am not sure it would be possible to have thermal stability in an standard amplifier without them. The fact it can be used to monitor the output current for protection is a nice bonus.
"The result, as so often is the case with blind testing, was inconclusive, the panel, overall, being unable to distinguish between one amplifier and another. (Though examining the results on an individual basis indicated that perhaps one or two listeners had, in fact, done soat least on those tests. Whether they would do so again is open to question.)"
Perhaps not the most convincing citation of ABX testing suggesting differences can't be heard - it qualifies the conclusion somewhat.
Colour me ill-prepared if you like - I haven't done extensive research on ABX blind testing in general but after some reflection, doing such testing with amplifiers would take a lot of effort in planning. One problem to overcome is anyone with the preconceived notion that amplifiers of very similar measured results will sound the same can skew the results in their favour even if done subconsciously. It is statistically more likely to prove a negative. To get around this skew, I'd propose including some fairly obvious differences such as dropping the level of one channel, introducing distortion, limiting bandwidth, placing a notch filter in the test etc, as part of the study. Doing so would expose any participant saying no to each test sequence as being either cloth-eared or having an agenda. Such testing would probably be rather expensive & I would have more faith if carried by a university rather than a HiFi magazine for example. A "statistically significant" (how that is derived, I'm not qualified to comment on) number of tests and people would be needed. One would also need a mix of enthusiasts (horrid term!) & normal people!
One aspect of this thread that I've found interesting (when I haven't fallen asleep) is the discussion of what HiFi means. I think it was Quad's mantra that stated "The closest approach to the original sound". That has stuck in my mind (as being a goal I can live with) as opposed to a sound people actually like - implying that auditioning is actually pointless. I find that idea almost wilfully bizarre if consistent! I certainly don't think listening is invalid merely because doing so is "subjective". I tend to be rather sceptical so need to repeat to draw conclusions.
Although blind ABX testing is the best of a bad bunch (formal sighted testing are laughably biased), it is far from perfect as it relies on memory you cant pump A into the left speaker & B or X into the right to draw any conclusions!
As for Serge, I think he masquerades his opinions as facts at times but I also have considerable empathy for his PoV. The HiFi world is littered with charlatans offering snake oil, whose produces are at best based on pseudo-science. These people are Alchemists-on-the-Make. Where we differ is that Im willing to trust my ears whilst acknowledging that doing so can be deceptive.
If I'm proved wrong then it's a win win - I can choose products on looks, specs build & MTBF figures!
ultrawomble
Sith Lord
Not convincing? It would seem the panel went into the test believing that they could hear differences....... except it appears they were mistaken.
Robert
Tapehead
However he can certainly prove that certain elements within the system are transparent, at least where the electronics are concerned.
If we take say 20 competently designed amplifiers or dacs and determine via a blind test that they all sound the same (not difficult), we have a pretty good reference point for comparison. So if Serge's amplifier were to also sound indistinguishable from the other 20, we've demonstrated transparency beyond any reasonable doubt.
No loudspeaker is transparent, but i recall back in the day plenty of attempts to compare loudspeakers with live sound. The very earliest Choice reviews used this method and while it is far from perfect it did highlight a number of performance issues, mostly around dynamic ability and colouration of the models under test.
Nothing like this exists today and instead we have reviews, in both the formal press and on forums such as this, where individuals will be 'reviewing' with often no direct comparison available, not even another loudspeaker let alone any chance of comparing with live sound.
Darth Vader
From the Dark Side
That would be Class B operation. The difference between the Class B amplifier and the Class AB amplifier is the biassing level. The Class B amplifier is biased into cutoff and this introduces crossover distortion. This distortion occurs during the time that neither transistor is conducting and is caused by the transition time required for the transistor to come out of cut off into the active region of operation. The Class AB amplifier is biased at soft cut off thus eliminating cross over distortion in good designs.
I remember the early trannies that operated in class B with outputs aound 125/250mW. With such tny speakersit really didn't matter.......
Cheers,
DV
Brian
Eating fat, staying slim
All well and good, except the majority of people who listen to music in the home don't give a damn whether something is transparent or not, so that's all rather pointless really.
The repeated use of the word "transparent" is sadly now making me cringe almost as much as the "tory" word.
Mike Reed
pfm Member
The repeated use of the word "transparent" is sadly now making me cringe almost as much as the "tory" word.
Would that be 'conserve' or 'lava'?
Barrymagrec
pfm Member
Would that be 'conserve' or 'lava'?
It means s**t either way...
sergeauckland
pfm Member
Not necessarily. The Peter Blomley Class B amplifier had no crossover distortion even though each transistor only carried one half-cycle. The difference was that each output transistor was biased above cut-off, but the base drive was a uni-directional current drive which allowed damn-near perfect linearity (THD 0.01% at all power levels and IMD <0.003%).
However, I agree that conventional Class B amplifiers must distort at crossover, which can't be fully corrected by any amount of feedback.
S.
sergeauckland
pfm Member
I remember loudspeaker reviews in the '60s and early 70s before all the subjectivist nonsense came about, which, after the technical evaluation, compared the loudspeaker under review with a transfer standard, usually the Quad ESL. This was on the basis that the ESL at the time was about the least coloured loudspeaker, so making side-by-side comparisons was of value. Nothing like this happens in modern reviews, it much more about night and day, trouser-flapping bass, jaw-dropping whatever, inky-black silences, removal of veils and so on.
S.
Yank
Bulbous Also Tapered
No doubt there are charlatans, but it's a mistake to view the entire industry that way.
Paul R
pfm Member
I don't know what 'biased into cutoff' means.
A Class B amp is biased to minimise crossover distortion. A Class AB runs in Class A up to some power. Not the same.
Most popular Class A amps are actually AB because it will come out of A under some load or signal conditions rather than current limit.
Paul
sergeauckland
pfm Member
The usual Class B circuit has no bias, so can't minimise crossover distortion. Peter Blomley's circuit, which I've not heard of any commercial application, has a bias current, like a class AB amplfier, but drives the output transistors with unidirectional currents, hence keeping to the class B definition, but with no crossover distortion.
Some (most?) class A amplifiers do run in Class AB into more severe loads when the output current exceeds the Class A standing current, but some early class A SS amplifiers didn't, so if given more severe load, would just clip.
The difference between a Class A amplifier and a Class AB is the amount of standing current compared with the load current. Some amplifiers, as I understand Krell do, have dynamic control of the rail voltages and standing current so that class A operation is maintained under all conditions up to a maximum, and by adjusting the rail voltage avoid excessive dissipation into severe loads when the standing current would have to be very high even if the output voltage was lower.
S.
- Status