ingrast said:
Walt Jung and I published it in EDN 1983. We showed how to easily measure better than 140dB down with ordinary equipment. No fancy oscillators or analysers. When I last mentioned it here someone immediately said "nulling techniques have been discredited long ago". 😕
Unfortunately the null you would get many high end preamps would be poor and tell you nothing. I wonder how the Zanden preamps would show?
scott wurcer said:
Scott, is this article available online? Is it similar to the technique on Renardson's page?
syn08, your mention of being an audio engineer in the Toronto market in which I worked through the Nineties steered me to your profile and website. I'm having trouble reconciling the strong skepticism towards low harmonics shown here with the discussion of the design targets and consequent listening results for the (impressive piece of engineering BTW) PGP amp on your site. You appear to imply the -120dB residuals (0.0001%) are responsible for the most 'unforgiving' amp your listening panel has ever heard.
rdf said:
Huh?
I'm a humble hobbyst in audio, happening to have some serious electrical engineering education and scientific background. And, I guess, a lot of common sense. I have never worked (thanks god!) in the audio industry and have no plans to, now and in the future. I do not make a penny out of this and everything I do is available for free (for non commercial use, yeah, like I could control that 😀) on my web site.
Anything else is :bs: and I would really appreciate if you could quote me otherwise. The PGP amp was a project only intended to show that such performances are possible today by using standard commercial components and technologies. Nobody ever claimed or objectively proved that this amp sounds better than any amp with decent measured performance. The listening test were just that: subjective testing. Unfortunately, all the subjects were aware of the measured performances before listening, and I am pretty sure this biased the results.
The next amp (YAP, construction is ready, to be published on my web site ASAP) has, in various incarnations, way less ambitios targets - and way less complexity. It still has some impressive features which, again, will be all available for free and in details on my web site. For now, I can tell... there is absolutely no difference in the sound between PGP and YAP.
scott wurcer said:
What I mean is there are more tools available now for postprocessing the null signal.
It should be very interesting for example to record the null residual under different source material - synmphonic orchestra comes to mind here, being distinctively muddled at the slightest amplifier weakness - and look to correlate whatever comes out with the input itself.
This is no easy task and I wish I had the time (and smarts 😉 ) to tackle it. I bet the house some insight will come through. This is on the basis of source material which, when listened through such or such gear is qualified as "sublime" or whatever adjective the reviewer fancies.
Rodolfo
andy_c said:
We just used an in-amp the same way. Amazing, a master's thesis based on that. He also has a page on capacitor DA worth looking at.
To clarify, some people consider linear frequency response and phase errors "distortion". I just think that leads to confusion, I prefer to think of distortion (when not qualified) as responses at different frequencies than the stimulus.
I once had a high end dealer claim the "John Curl Sound" was just a deliberate +-.2dB frequency response abberation in his RIAA.
It is enough to distinguish berween linear (amplitude, phase, timeresponse - but no new harmonics created) and non-linear (harmonic) distortion. Then we have non-harmonic non-linear distortion.
syn08 said:
Again, no intent to slag what by all appearances is very impressive work but an objective reading of the listening impressions here http://www.synaesthesia.ca/listening.html suggests otherwise.
john curl said:
I just pulled the cover off my Parasound HCA1000A, whose amp circuitry John designed. There are two TDK 50k pots at the input jacks, John told me a while back that he didn't really want input pots, but at least Parasound did use TDK pots. I find them useful on occasion when I demo another power amp, to equalize the volume levels. Otherwise I leave them full on. If I wasn't writing for audioXpress I would probably bypass them, but for now I will leave them and I don't really find them objectionable.
PMA said:
As I have said before: in this view, a low pass filter has distortions. Which, to me, doesn't make any sense.
Why would anyone believe the reports those who hear a difference over those who don't ?
Surely they are equally suspect.
Let us be fair.
Surely they are equally suspect.
Let us be fair.
the concept of linear and nonlinear distortions is well accepted in communications engineering.
Sorry english is not my native language. In german the meaning of 'Verzerrung' is just like that.
regards
yes it has, it changes the input signal, it is deforming the shape for example - isn't that the meaning of distortion?
Sorry english is not my native language. In german the meaning of 'Verzerrung' is just like that.
regards
fredex said:
Such differences in opinions happen when people listen to reproductions of recordings trying to compare consciously with what they remember. During live concerts when microphones, mixer, amplifiers and speakers are involved, anyone hear the difference.
Speaking of records, for me the best criterion is when people's subconscious mind react on sounds like water, helicopter, roaring beast, or when somebody asked, "Who played piano in your house?"
There is a big difference between "Your system sounds real" and jumping turning around when water suddenly start running. Or, when John was thinking he hears a real violin in the next room.
I am a lowly EE not a psycho-acoustician,yet when I look at these test data and test set-ups I see -gasp- electronic reproduction mechanisms prevalent in the evaluation or test chain(s).PMA said:
Do these people use the very best known audio gear for their tests, or is it routinely assumed that the ear not the gear provides the limits (on masking for example) of performance?
In my own equipment at home it sure seems that as I travel up the quality(price) curve (warning cliche) things I never heard before are revealed. 🙂
I am not saying they are wrong, but merely warning that my professional EE training included many built in assumptions and conclusions whose absolute accuracy or completeness my own experiences now cause me to question.
In this case dispersion and transmission on lossy media are included. Yes, pulse distortion is a common concept there.
Capacitors can have DA and not have harmonic distortion and possibly visa-versa, as long as that's clear it doesn't matter I suppose.
Capacitors can have DA and not have harmonic distortion and possibly visa-versa, as long as that's clear it doesn't matter I suppose.
Juergen Knoop said:
The lowpass filter then visibly "distorts" a pulse but not a sine wave. I think this is down to semantics and not particularly productive.
Juergen Knoop said:
I understand the concept of (e.g.) group delay, as the rate of change in the slope of the phase vs. frequency curve. If an amplifier has zero group delay distortion, which means the phase-frequency characteristic is linear, then all frequencies within a certain bandwidth are transmitted through the amplifier in the same period of time.
While the impact of linear distortions is becoming more apparent as e.g. telecoms move to higher orders of modulation (e.g. from QPSK to 8PSK) I still fail to understand the impact of such in the audio world. I'm also afraid that, in the audio world, designing for zero (or a minimum of) linear distortions conflicts with other design goals (which are, most likely, more audible).
maybe semantics don't clear things up, but 'distortion' is well defined and does include all changes in frequency and time that happen to a signal or group of except changes in level of course.
regards
regards
- Status
- Not open for further replies.