John Curl's Blowtorch preamplifier part II

[Jakob2]

Originally Posted by jcx

Kunchur’s 2007 paper has a number of problems with the signal theory which I can pick out quite easily

there is no problem getting a periodic signal from a DAC
the assertion to the contrary is just bizarre to anyone familiar with analog/digital Signal Theory – “looking at the points” in a sound editor is a logical fail – the reconstruction/anti-imaging filter output is the waveform to look at

I think you were referring to his Proceeding Paper; afair he didn´t say, that it is impossible in general to get a periodic signal from a DAC, but was pointing to a well known property of dicrete time signals, that has to be taken into account when synthesize waveforms.
More a reminder why it is sometimes not sufficient to use a soundcard for the generation to test waveforms, because without `special´ algorithms the waveform may contain additional unwanted spectral content due to the periodicity problem.

The later slide presentation’s psychoacoustic model proposal goes beyond my meager psychoacoustics knowledge ability to judge credibility
but I can ceratainly spot the correlation argument he uses

What do you mean by that?

[PMA]

Originally Posted by Bonsai

I know, but I wondered why PMA raised this point on a recording that has little content above a few KHz.

Try audibility of noise corner (BW) modulation.

[john curl]

In any case, the recording sounded very good for being so old and recorded on such 'primitive equipment'. As far as analog audio is concerned, there are lots of areas capable of discussion, such as: passive vs active preamps, buffers, complexity vs simplicity, etc., etc. . Digital sound quality problems and solutions are also quite interesting in this context, since digital is the primary medium used today.

[Capt Grogg]

Scott Wurcer
OT email and PM sent

[wayne]

Originally Posted by MikeBettinger

Where these two amplifiers different designs toally or were they built similarly to compare circuit performance? An apples to apples experiment or more of an informal comparison of two different amps?

Mike

Two different designs. One SS the other glow Fet. The response similar and low noise floors on both. The feedback was about about 20 dB in both cases. The point being one had distortion an order of magnitude higher. I would say nothing new here but both were very good. Speakers 100 dB also.

So what can we measure? John is quite correct to use known good components and they are audible but what can we measure to move forward?

[ThorstenL]

John,

Originally Posted by john curl

Well, we appear to have wondered off track again. I hope that we can continue to talk about how to make better audio electronics.

This is bound to happen, this wandering.

If I may take a step back on all of this and give my own view on what we need to do to build better audio electronics? Thank you for the indulgence, Gentlemen.

If we wish to build "better" audio electronics we first must define clearly what is "better".

Some would argue that "better" means better in the traditional measurements hat are being promoted. The fallacy of this should be patently obvious and should need no explanation. As one example, Speakers are almost never characterised for distortion, because the levels of distortion in a speaker at rated power are extremely gross, yet for Amplifiers very low THD is promoted as measure of quality, clearly a case of straining at Gnats and swallowing the camel.

Others would argue that "better" means better in terms of how people react to the reproduced sound. On the surface this makes immense sense, in fact, it makes ABSOLUTE sense, as both recordings of music and equipment to play them back exist solely for purpose of human sensual and aesthetic enjoyment.

Thus if something measures perfect in terms of traditional measurements but fails to provide sensual and aesthetic enjoyment of music to those who listen to results, it is by ABSOLUTE definition BAD, as it is unfit for the purpose.

Where problems come in is that it seems to have so far not been possible to translate "better sound" into simple measurements that yield simple "figures of merit" so an engineer can say "if this number reads 10 it is "perfect" and can work towards making this number read as close to 10 as he can.

So one side uses false yardsticks, the other has no "true yardsticks" of their own. In fact, many attempts have been made over time to correct this, yet they have been met mostly with shrugs and with derision from some quarters. The result is that very little real progress has been made, mostly because the nature of the problems is insufficiently clear and the very real problems are by far too rarely accepted as real problems.

For example, back in the mid-90's I proposed an approach to a speaker which would be able to correct for the speakers various flaws (including HD, IMD, Power Compression and others) by "pre-distorting" the signals fed to each driver, even including the room (this was on the bass list), with an outline of the needed calibration routines, required DSP funtions etc. There was zero interest, as distortion and compression in speakers where just not perceived as valid problem that needs solving and a decade and a halve later there is still no interest.

So, to make any progress it is necessary to abolish the blinders many in audio engineering are wearing, to abandon the blind faith into "revealed truth" that cannot be altered and to stop attacking, in order to silence, anything and anyone who seems to question this "revealed truth", which is often enough to say "almost completely" not backed by empirical studies into efficacy.

However, as long as it remains easier to deny the truth and ostracise, ridicule and wherever possible censor dissenting views and ideas, this will simply not happen and things will continue as they are. A stalemate on a battlefield of mediocracy, long abandoned by the more able and competent generals and officers for fields offering more glory and pay. In fact it is much like Stanislaw Lem's description of the state of affairs in the (fictional) sciences of the (fictional) planet Solaris in his (science fiction) book of the same name, despite the absolutely stunning revelations that the planet yet had to yield...

When I was younger, less wise and more foolish, this state of affairs may have induced me to try to illustrate the situation and to try to stir up this dreary bog of sullen prejudice and muddy inertia which some call "common sense".

Good thing I'm not, or I could really get into trouble with the powers that be, innit?

Ciao T

[john curl]

Well stated, Thorsten.

[tomtt]

[Bonsai]

Originally Posted by PMA

Try audibility of noise corner (BW) modulation.

On a 100 year old recording? I would think there would be more pressing sonic issues to deal with ;-)

[simon7000]

Originally Posted by Bonsai

On a 100 year old recording? I would think there would be more pressing sonic issues to deal with ;-)

And what has improved since 100 years ago? Pavel mentioned two items. Every yardstick starts at zero even if it is not marked.

Everything you can measure is probably thousands of times better, but is the best modern reproduction a thousand times better.