It could also be the otherway around, the expansive nature of the input stage/VAS compensates other compressive elements. For sure I believe that we all agree that one great difference from live to reproduced is the limited and limitation of the dynamic range. Reproduced sounds dull compared to the vitality of live. So when a CFA sounds more live and lifelike than VFA. how can we then dismiss it...??? Well you can if you look isolated at things but how can we do that when we listen through a system..??
Lets make a simple-stupid example.... I like them... analogies....
Two amp designs have thd/IM below .001%. One uses no gnfb and the other uses a lot of gnfb. Can they sound different? Why? What test? Something other than thd/Im is at work.
Lets just say one of the other parameters -- Zout is very different and would interact with load Z (speaker) to give a varying freq response that is very noticable in listening.
OK. So, the early question was... what could make a CFA sound different from a VFA? I know that there are many T & M which show other differences besides THD/IM.... maybe it is one of them. Now we know when we normalize the two (CFA/VFA)... we still dont know if they sound different for some other reason than THD/IM.
I know the assumptions over the years have given way to new discoveries.... such as, the sound of CD's..... E.Mitner showed us that jitter of very small periods really could be heard and devised tests to show the reason.
Fast Forward many years.... looking at the math of PCM digital systems reveal that inter-sample peaks are common in commercial releases...and when they reach an interpolation process, it produces distortion products that are non-harmonic and non-musical. The new, for one, BenchMark DAC2 addresses this issue we didnt know much earlier what caused the sound but was heard.
As time goes onward the questions get answered. Here we have assumptions which are probably not true or the threshold assumptions are not true and/or we havent figured out what test to make.
THx-RNMarsh
Two amp designs have thd/IM below .001%. One uses no gnfb and the other uses a lot of gnfb. Can they sound different? Why? What test? Something other than thd/Im is at work.
Lets just say one of the other parameters -- Zout is very different and would interact with load Z (speaker) to give a varying freq response that is very noticable in listening.
OK. So, the early question was... what could make a CFA sound different from a VFA? I know that there are many T & M which show other differences besides THD/IM.... maybe it is one of them. Now we know when we normalize the two (CFA/VFA)... we still dont know if they sound different for some other reason than THD/IM.
I know the assumptions over the years have given way to new discoveries.... such as, the sound of CD's..... E.Mitner showed us that jitter of very small periods really could be heard and devised tests to show the reason.
Fast Forward many years.... looking at the math of PCM digital systems reveal that inter-sample peaks are common in commercial releases...and when they reach an interpolation process, it produces distortion products that are non-harmonic and non-musical. The new, for one, BenchMark DAC2 addresses this issue we didnt know much earlier what caused the sound but was heard.
As time goes onward the questions get answered. Here we have assumptions which are probably not true or the threshold assumptions are not true and/or we havent figured out what test to make.
THx-RNMarsh
We must open mind with humble attitude to discover a new thing. If we think our knowledge is prefect than we can not find or believe a new thing... we play as God
Last edited:
I disagree on "simple", but I fully agreed on "stupid".
We must live in parallel realities. In the one that I live in, audio distortion products are always harmonic. Non-harmonic distortion products violate the causality principle, unless you can prove that an audio device can be a time-variant system.
Last edited:
haha funny man. lets pretend we dont understand what non-harmonic means in the Context of music listening. That way we can also pretend to not understand the message. Unfortunately, inter-sample overs cause clipping in most interpolators -- like jitter was assumed to be a non-issue until it was reduced and heard the improvement.
-Thx RNMarsh
Last edited:
It may look funny to you, but I am as serious as a heart attack. With all due respect, you are talking audio gibberish.
Mr. Marsh, have you got a link to Ed Mitner's work on jitter .. especially any listening tests?
Strange how some people have an obsession about "peeking", as if it explains 'everything' - unfortunately, it doesn't
... a very large part of the answer is being fastidious, something that doesn't come naturally to many people - a solution is to encapsulate the necessary "fastidiousness" in a black box, which can't be tampered with, and which is immensely robust to being carelessly or casually treated, or how it's included into the scenario ...
Member
Joined 2009
Paid Member
Reminds me of my Botchup amp, where I started off like a JLH Class A CFA, added a mirror to the current source... almost turned into a diamond. Unfortunately I didn't get around to building it as the heatsinks were too small.
That's the generic problem with most audio - I've just used a ranking of 100 to represent that SQ level in another thread, and virtually all setups fall below that level. It's possible to reach that 100, and go well beyond it - it's just hard to do, the situation is as simple as that.
The key factor is system integrity, the elimination of all weaknesses that mitigate against that level of quality - something that the majority can't, or won't, conceive has significant bearing on the situation ...
Strange how some people have an obsession about "peeking", as if it explains 'everything' - unfortunately, it doesn't
Based on your contributions to another thread, I learned that any attempt of bringing rationality in the discussion will miserably fail when facing the endless fuzzy stream of empty words.
Peeking replaces what you can hear with what you think you can hear.
as far as I can find actual professional publication controlled listening test thresholds for jitter audibility with RedBook CD is 10s-100s of ns
the sub ns numbers tossed around are all exteme estimates of jitter of 0dB fs output at 10-20 kHz that reduce S/N enough to reduce ENOB to <16 bits - not actual demonstrations of audibility with even special test signals , much less music recordings
however realizing jitter is dubiously audible is no reason to use products with multiple ns jitter since the capability to better is cheap with competent engineering - and at higher levels it does interfere with our ppm IMD meaurements of real hardware
the sub ns numbers tossed around are all exteme estimates of jitter of 0dB fs output at 10-20 kHz that reduce S/N enough to reduce ENOB to <16 bits - not actual demonstrations of audibility with even special test signals , much less music recordings
however realizing jitter is dubiously audible is no reason to use products with multiple ns jitter since the capability to better is cheap with competent engineering - and at higher levels it does interfere with our ppm IMD meaurements of real hardware
Last edited:
And a good morning to you too ...
Unfortunately, "fuzzy" words are often the best tool for describing subjective experiences - because the 'scientists' can't be bothered investigating these sorts of areas more thoroughly than they've done so far, we're left with using 'vague' terminology to describe what's relevant.
Mental focusing, not "peeking", is the key mechanism that changes what you subjectively perceive when the 'reality' changes not one whit. Easy experiment to do: listen to just reasonable reproduction, and think that it "sounds bad" - keep thinking this way, and your mind will quickly find the off qualities in the sound, before long it will become intolerable - then, very deliberately, change your thinking patterns, put on a "happy face" - still listening to the same music, mind you - in no time at all the positive features will be so apparent, and it will be quite pleasurable to listen to. All very deliberate, all under conscious control - yet nothing externally has altered ...
Last edited:
Just saw an article about a week ago, not sure where I saw it, but the basic conclusion was that the mind actually processes information at a much higher speed than was agree upon before. So even our understanding of how we process information is still changing and many surprises are still in store. So those who base things like what level of jitter is audible may be off by a factor of more.
I understand what Richard was saying about non harmonic content creeping into the sound chain. Who is to say that there are not anomalies in a digital data stream that would show up as non harmonic relationships to a musical signal?
ps. All test equipment should come with the screws loctite'd so they cannot be opened or the circuit should be encapsulated so you have no idea what is inside. Yes this could screw with the thermal management but at least it would remove the visual aspects of looking under the hood.
I understand what Richard was saying about non harmonic content creeping into the sound chain. Who is to say that there are not anomalies in a digital data stream that would show up as non harmonic relationships to a musical signal?
ps. All test equipment should come with the screws loctite'd so they cannot be opened or the circuit should be encapsulated so you have no idea what is inside. Yes this could screw with the thermal management but at least it would remove the visual aspects of looking under the hood.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.