I should note that, for me personally the amuzing part was that the writer used the paper to reinforce his view despite that didn't appear to be its conclusion (even JA didn't agree with his premise). It just goes to prove that some people will use anything to try and reinforce their beliefs even if the facts don't stack up. I also noted they have linked to other research that seems to indicate that people are able to tell higher bit rates from 44k1 but only on tones!
But I did not intend any offense.
No matter what you post online, someone somewhere will be offended to some degree or another. You can't please everyone. IOW, state what you need to state.
When you speak of by-the-ear audio designers, you are displaying your own cheeky ignorance.
Jam won 2 Product of the Year and 4 other design awards for one piece of design work done to a large extent by ear. The difference between you and him in that regard is that he is an expert at it you are a cheeky ignoramus. That should bother you based on your own stated standards, but doubtful it will. Even you are welcome to visit Auburn and see for yourself.
Funny, circa 1974 we did a buy resale on a comparator where the former test engineer measured the offset with a standard op-amp test loop on an old Teradyne J263. Well the latest mask set oscillated and having no control over that I just soldered a 5uF cap to the output and lengthened the test delay. The test cable/fixture capacitance became just into the right range.
I don't think Jam is an example of that. However someone who takes hours to paint little dots on a perfectly reasonable speaker cone because it "sounds better", well.......
Matt,
Jam is absolutely not any of the other descriptions Syn08 used in that sentence. He does make certain design choices by ear though.
By the way, it is no different than me saying NJM7805 sounds better for a particular load circuit than a particular LDO that was tried. It sounded better, sorry if Syn08 doesn't like it, its just what happened. I didn't say noise was as low or voltage was as accurate with 7805, those things didn't seem to be what mattered the most at the time. Nothing new with some audio circuits behaving that way. Its doesn't mean that stupidity or malfeasance has to be involved. Completely unwarranted for someone to start yelling about BS.
Same if someone or a few people hear effects from 100kHz roll off in a preamp. Doesn't mean they are stupid, lying, ignorant fools, does it?
Jam is absolutely not any of the other descriptions Syn08 used in that sentence. He does make certain design choices by ear though.
By the way, it is no different than me saying NJM7805 sounds better for a particular load circuit than a particular LDO that was tried. It sounded better, sorry if Syn08 doesn't like it, its just what happened. I didn't say noise was as low or voltage was as accurate with 7805, those things didn't seem to be what mattered the most at the time. Nothing new with some audio circuits behaving that way. Its doesn't mean that stupidity or malfeasance has to be involved. Completely unwarranted for someone to start yelling about BS.
Same if someone or a few people hear effects from 100kHz roll off in a preamp. Doesn't mean they are stupid, lying, ignorant fools, does it?
Hi O
Very nice post.
BTW: didn't we discuss that pole-zero cancellation technique circuit many years ago at another place?
The place shut down before we finished the discussion, but it was an interesting brainstormer.
(BTW. Dadod was a "member" at that place, he did't contribute much, but asked for help all the time.)
I did research further on my H-bridge amp with diff. "VAS" and diamond OPS.
By injecting an error current into the ops diamond, I could get very low THD without pushing the ULGF and most important the THD increases propotional to the output level, there is no sign of higher "relative" distortion at low levels or cross over distortion, even if it's biased in class AB.
Stein
Question.
Do you not think that it may have been parasitic oscillations in the OPS that caused these problems rather than loop instability aka Bode instability.
Asking another way, how were you certain it was loop instability rather than parasitic?
....oh no another food fight brewing. I don't want to get into a middle of a squabble.
syn08 actually raised some good points but even if I do not agree with some of his conclusions sometimes there is something to be learned from what he has said.
To be fair there are several approaches to good audio design.
My approach is to choose your topology, build it. My main tools were a generator and scope. Once it was working listening tests were conducted until a satisfactory sonic result was obtained.
Only then the Audio Precision then came into play...............
You may have a different point of view.
I mentioned to Richard that the best distortion numbers did not mean the best sound (which he does not agree with) but this was confirmed by multiple independent listeners. Strangely Richard declined to listen to my design or compare it to another unit he had on hand. He only wanted to comment on measured distortion numbers.
Don't get me wrong the AP is a fabulous device and can give you great insight into your design but to put it bluntly if you live by the meter you will die by the meter............Instrumentation can get you only so far and good design needs a balance of measurement and careful listening.
The audio graveyard is full of many great measuring designs, remember Cherry's effort a decade or so ago (which most reviewers were enamored with for a short time), where is it now...yet the humble Dynaco Stereo 70 is as relevant today as ever was.
I put it to you....Rolex or Timex.🙂 (the Timex keeps better time)
Jam
syn08 actually raised some good points but even if I do not agree with some of his conclusions sometimes there is something to be learned from what he has said.
To be fair there are several approaches to good audio design.
My approach is to choose your topology, build it. My main tools were a generator and scope. Once it was working listening tests were conducted until a satisfactory sonic result was obtained.
Only then the Audio Precision then came into play...............
You may have a different point of view.
I mentioned to Richard that the best distortion numbers did not mean the best sound (which he does not agree with) but this was confirmed by multiple independent listeners. Strangely Richard declined to listen to my design or compare it to another unit he had on hand. He only wanted to comment on measured distortion numbers.
Don't get me wrong the AP is a fabulous device and can give you great insight into your design but to put it bluntly if you live by the meter you will die by the meter............Instrumentation can get you only so far and good design needs a balance of measurement and careful listening.
The audio graveyard is full of many great measuring designs, remember Cherry's effort a decade or so ago (which most reviewers were enamored with for a short time), where is it now...yet the humble Dynaco Stereo 70 is as relevant today as ever was.
I put it to you....Rolex or Timex.🙂 (the Timex keeps better time)
Jam
Last edited:
^ It would be interesting to make correlations between measurements and preferences though? I recall you agreeing with me that a change is often perceived as an improvement.
Great example. When it comes to audio "product of the year", I can afford to not give a flying **** about. Show me the "radiotelescope of the year" or the "MRI equipment of the year" and we can talk about. Your example directly shows the crap that brought the High End Audio industry where it is today: pretty much a joke about nothing.
Now that you mentioned it, I think we did (or started). For the rest, I don't really remember... Come up with the schematic and we can discuss, but please, I don't think this thread is appropriate. This is mostly the famous and wannabe famous audio designers turf.
For the "OPS parasitic oscillation" Bode is still in charge; but nevertheless, I got your question. It was indeed the global loop, not any local feedback loop (like the source followers). That is in general easy to identify by the oscillation frequency; not an absolute rule, but at least for the solid state audio realm, global loops oscillate around the global ULGF, local loops oscillate around the local ULGF. A mosfet OPS can oscillate locally at many 10's of MHz and the almost universal cure is to adjust the gate stoppers (either resistor or ferrite bead). That was not the case with this amplifier.
Cost of developing an MRI is many millions if not 10s or 100s of millions. Last medical linac Varian developed cost up in the 100s of millions. Lots of PhDs working for a long time to understand things at a level humans are able to engineer.
Humans at a fine grain level are much more complicated to understand than any human engineered product so far. No doubt if we spent billions of dollars to understand how to measure sound quality as humans perceive it then there would be no more arguments about it here. You would be telling me there is no way relying on an AP could possibly do it.
Since there is no money for it and probably never will be, the best audio designers will have to learn how to listen and measure, not just one or the other.
Humans at a fine grain level are much more complicated to understand than any human engineered product so far. No doubt if we spent billions of dollars to understand how to measure sound quality as humans perceive it then there would be no more arguments about it here. You would be telling me there is no way relying on an AP could possibly do it.
Since there is no money for it and probably never will be, the best audio designers will have to learn how to listen and measure, not just one or the other.
Last edited:
Scott,
Yes it would but unfortunately there are too many variables to make it a precise science.
You are so correct change is almost always perceived as improvement (a common audiophile mistake). You have to repeat the test multiple times and then some over a period of days.
Jam
Yes it would but unfortunately there are too many variables to make it a precise science.
You are so correct change is almost always perceived as improvement (a common audiophile mistake). You have to repeat the test multiple times and then some over a period of days.
Jam
Yes there are, and here's the shtick: define "good audio design". You define it as a commercial successful project for which, agreed, usually no technical performance is really needed, but only a good story to tell, even if it's full of it, what really matters is people buying into.
I couldn't care less about such a metric, and not being into the business I can afford to support the natural criteria of accuracy (or "gain with wire" performance metric, if you prefer). I am fully aware that such performance is by no means a key to commercial success; after all the amazingly engineered Halcro, to the late Bruno Putzey's Hypex NCore incredible achievements in Class D, never achieved a "cult status" like for example the Blowhard that is in the title of this thread.
Last edited:
Wrong again.
Measurement gives approximately right results, just not exactly right results in most cases.
A truly good design sounds good to a lot of people and usually measures pretty well too.
Eh? As far as I was aware he was still alive and kicking and developing newer and better class D designs?
- Status
- Not open for further replies.