The problem with adjusting feedback is that you also have to adjust loop compensation. If you fail to do this then all you have is variable tone control linked to variable gain, and you don't know which one you are listening to. Hence most tests of 'variable feedback' are fatally flawed. In any case, you seem to be asking me to discover what level of feedback I prefer; this is not a valid question for me.Destroyer OS said:
Distortion varies with signal level and frequency (not to mention temperature etc.). You would need fast adaptive DSP to keep the amplifier distortion complementary to the speaker distortion. Might not go down too well with the SET brigade?indra1 said:
Simple cancellation only works approximately for second order, and that is generally agreed to be the most benign distortion. So you can partly remove the least objectionable distortion, while probably increasing all others, and it only works with certain pairs of amp and speaker. Is that progress?
Irrelevant. You are comparing small amplitude high frequency pressure waves with large amplitude low frequency mass movement and trying to draw valid conclusions? Now who was talking about "insufficient understanding"?
The apparent contradiction can be partially resolved by the knowledge that I prefer recordings which have the least messing about by the recording engineer. Think 'live concert relay by the BBC in the early 1980s': simple miking, some manual gain compression (unfortunately necessary for FM, less so for CD). The aim is to preserve the actual acoustic event, not modify it.mmerrill99 said:
In any case, the fact that a recording engineer may have damaged the simple preservation of an event is not an argument for adding further damage in the amplifier at home.
"More realistic" means more like the real thing - which is what I have been saying all along. You believe that can be achieved by piling up distortions all along the audio chain; I believe it is best achieved by reducing distortions in each link as far as reasonably possible.
In most cases with complex chips the circuit (and PCB layout) may be taken from the datasheet. If the maker says says use a X9Y dielectric cap marked as 0.1uF then he will have already taken account of the voltage coefficient. In any case, decoupler values do not usually matter that much.RNMarsh said:
One mass-market company has an audio designer (or production engineer?) who cuts corners (or cuts costs), or doesn't read datasheets. Exactly what should we deduce from that?john curl said:
A coupling cap merely has to provide sufficiently low impedance across the audio spectrum to pass the signal, block DC and limit distortion (by limiting cap signal level). Adding a bypass could make things worse, not better.RNMarsh said:
Why is THD only mentioned by people who don't believe in accuracy?indra1 said:
Re. extinguishing the candle flame:
In both cases (blowing vs. inhaling), finite-amplitude waves will be produced, starting at the mouth. These waves and the air motion thus created have different features and interact differently with the candle flame.
Blowing creates first a compression wave and a continuous bulk air flow that follows it, both propagating towards the flame; and it is the bulk flow, i.e. the jet emanating from the mouth, that extinguishes the flame.
Inhaling first creates a rarefaction wave propagating towards the flame and a bulk flow towards the person inhaling, the flow rate of which is dependent upon the pressure difference across the flame. Since this pressure difference is lower than the one in the case of blowing (no jet flow is created by inhaling), the resulting bulk flow velocity is lower, making it thus harder to extinguish the flame.
If the flame were confined in a pipe, both methods would give similar results, as bulk flow of almost the same intensity would be created in both cases.
Regards,
Braca
In both cases (blowing vs. inhaling), finite-amplitude waves will be produced, starting at the mouth. These waves and the air motion thus created have different features and interact differently with the candle flame.
Blowing creates first a compression wave and a continuous bulk air flow that follows it, both propagating towards the flame; and it is the bulk flow, i.e. the jet emanating from the mouth, that extinguishes the flame.
Inhaling first creates a rarefaction wave propagating towards the flame and a bulk flow towards the person inhaling, the flow rate of which is dependent upon the pressure difference across the flame. Since this pressure difference is lower than the one in the case of blowing (no jet flow is created by inhaling), the resulting bulk flow velocity is lower, making it thus harder to extinguish the flame.
If the flame were confined in a pipe, both methods would give similar results, as bulk flow of almost the same intensity would be created in both cases.
Regards,
Braca
Thank you. You have been most helpful. Nothing mystical is ever implied and "good sound" is subjective experience, it is the unknown that puts me on overdrive. I apologize if my posts irritate anybody, I never meant to cause any disharmony.
First I apologize if my posts also irritate you, never meant to do so.
Agreed for attempts to have perfect null. I have no wish for more distortion and perfectly happy with partial null at listening level, at least better than nothing. Any other reason for me to abandon further efforts on partial null?
Any relevance you find important to bring this matter to the discussion? I know some people that may fit the description but never seek any advice from them. I have difficulties understanding the terms being used.
I answered why I found current understanding implied in the word "accuracy" offensive in response to your question. Rewording your response I get "As current technique yields limited success, laughable progress to gain on further efforts". Do I get what you say correctly? I hope that it does not then follows that I should also join in and offensively ridicule everybody with different ideas on this matter.
I accepted that your statement may hold enough truth. I never profess understanding, knowledge or drew any conclusion on the matter. Scott Wurcer gave enough for me later on. Anything more you need?
Yes, I am not a believer in accuracy manufacturers publish, I measure to be sure. I learned the painful way not to believe in claimed accuracy, I even question the definition of accuracy used to describe amplifiers. However I doubt that I have enough relevant knowledge on the matter you asked for a meaningful answer, perhaps old timer professionals like John Curl, Nelson Pass, Howard Hoyt or Richard Marsh could tell you.
Why is "whole chain THD" rarely mentioned? 🙂
Last edited:
Yes, it is exceedingly rare. The advantage is to get the whole system distortion and the system grounding, ac power and the interfacing issues into the picture. All together, this is what we listen to.... not one piece measured by itself.
THx-RNMarsh
Last edited:
No problem to measure everything connected together. The debate is useless, as always. Wasting time.
Don't give me that reductionist ******** excuse. A change in tone doesn't even matter for understanding what more and less feedback does to the sound. The changes are consistent (regardless of tone) with every reviewer of high end audio gear's comments. They are consistent with purchasing patterns. What's inconsistent is almost everyone's conception of what the **** is actually happening, that can read an O-scope. You think because you can measure the distortion you get the whole picture, but it just isn't true. You need to hear at the speaker level what is going on, to understand the custom base; and reviewers.
Now I've heard better and worse amps with more and less feedback, so they are not all equal, but that does not change the absolute consistency of everything but electronics people's thoughts on the subject. It seems like Pass is the only one that gets it most of the time...
Here's an easy way to do it. Go listen to a CH M1 amp, you can on the fly change the amount of global. It'll explain to you real fast what you need to know. Do it on a 3 way speaker.
I take a slightly different approach. Get the IM and THD down to incredibly low levels at any freq or power level or load Z and in doing so, get SR up high enough to accomplish that goal..... and the feedback.. amount used... doesnt even enter the picture any more.
As you know, from your many CAD design iterations..... There are inherently linear circuits of very low distortion without any FB at all. But most of those topologies are derived from the CFA designs.
A good book on how to design very low distortion amplifier circuits which are inherently linear can now also be found in Audio Power Amplifiers by Arto Kolinummi
THx-RNMarsh
Last edited:
Thank you Braca, very kind of you.
In my highly unscientific way, I used flame blowing to illustrate that at driver to air interface, there is a finite difference between a wavefront 1 formed by a forward moving diaphragm and a wavefront 2 formed by a backward moving diaphragm caused by the compressible property of air. I suspected that this difference contribute to the lower limit of achievable distortion, especially on lower frequency. So I suspect that further optimization of diaphragm type of transducer will soon hit a brick wall if it is not there already.
Somehow this difference also seems to coincide with the perceived added spaciousness when small amount of leftover 2H with "negative phase" was present (that reduce the strength of wavefront 1 and reduce the strength of wavefront 2).
However, remarks by both Scott Wurcer and Pavel Macura told me to drop this line of inquiry due to accepted common scientific notion of non significance. I simply do not have enough understanding and vocabulary to hold intelligent discussion.
Thank you Mr. Marsh.
It's all too convenient to say (and makes a pretty story) that "the measurements don't represent the sound", but it's not been demonstrated that people *are* in fact able to differentiate similar measuring equipment. That has all the caveats of double blinding/etc so you're not "listening" to a circuit topology but the actual electronics in front of you.
I don't care about sighted tests, other than it makes me happy when someone's happy with the stuff they've got in front of them (even more so if I helped them get there). This has nothing to do with the audio, and everything to do with being a social being.
I don't trust reviewers as far as I can throw them. Nor have I perfected the art of throwing myself, but if I could, I'd try. 😉 If you need to get the word out about a product you're selling, then yes, reviewers are a key part of that, and you have to take the good with the bad.
And how do you know what that is?
Again how do you know you are hearing what they intended?
Sorry, I don't get where you are getting this from "run off into a field gleefully prancing as you think you have won"? I'm simply pointing to what I see as logical inconsistencies when I see them
OK, then you are in disagreement with Floyd & other experts in the field of psychoacoustics. I believe you are misunderstanding what I said
Sure, that is what Richard said. How many of us actually do this or have the necessary equipment to perform such a process?
Sure, some systems are good & some not
Of course not but it also doesn't mean that I shouldn't try to accurately represent my viewpoint.
OK, I believe you & am happy that you find the sound from your TV convincing enough for your dog 😎
OK, so the original acoustic event is not the yardstick against which you measure the fidelity of the playback system (you weren't at this event) - instead, it's some level of modification of this event that is your yardstick?
I'm not sure if you know exactly what modifications have been done & how these sounded to the recording engineer & whether you are hearing what was intended or not?
I didn't say that - I was stating that our playback systems are not capturing the full soundscape when compared to the original acoustic event (I'm not talking about mic freq response, etc) - so working with this limited signal data capture & trying to create an illusion of realism is bound to have many issues inherent in & as a result of this whole setup - it might be that some of these issues could be improved by the introduction of some harmonic distortions - distortions which we expect to hear when exposed to real world sound.
I'm being somewhat of a devil's advocate here but just following logic
After all, we are happy to introduce other distortions, dither in digital audio to deal with inherent issues in that process - to randomize signal related noise. Why? Because our auditory perception is more sensitive to correlated noise than random noise. So we introduce this distortion mechanism because of our auditory perception mechanism - if it wasn't sensitive to correlated noise, we wouldn't bother with dither
Sure, but I'm saying realism is judged by auditory processing & I believed you were defining it by measurements & accuracy. Maybe I'm wrong? Wouldn't be the first time
I just read the intro on Amazon. Its reference section contains an exhaustive list of 291 publications; probably everything ever published about audio amplifiers! This list could be useful.
Last edited:
Yes, reduce the feedback until the speaker impedance effects the FR and distortion of the whole system, "synergy".
Hi mmerrill99,
Right now I'm listening to a particularly nasty sounding little amplifier on the bench. A Dynaco SCA-35 with a set of new tubes. My normal bench system is far more accurate and doesn't make me want to leave the room. Using the same CD source, so they are on the same footing. Same speakers too.
So yes, both me and my dog are happy with the better systems. When we had a bunch of Poodles, they would come in and listen to the main system even when it was rather loud. THe same animals have left the room when a less good sounding system was playing. It appears that even the dog has his limits.
-Chris
If your system is accurate, you accept that what you hear is what was intended. If it isn't accurate, the question is pointless as you know you will not hear what was intended - ever.
Same answer as above. If you can identify the brand of piano played, I would suggest to you that the system is pretty accurate.
If you decide to use different definitions than the rest of us, well the bulk of us anyway, we cannot hope to communicate. You cannot even argue a logical inconsistency if you select your definitions to suit the argument you make.
There is no psychoacoustics involved here at all. This has to do with listening to a recorded soundtrack where you can enjoy it, or not. I guess you could suggest that the cover art can affect your mood, but we aren't suggesting this is about mood. Is the system accurate or not? That is all.
Not my problem. It doesn't take much to be able to find someone who is capable of setting this up. It might cost money, but for someone who has this as a burning question I would suggest that it would be money well spent. If you use the same mic in the live experience as the recorded one (and why wouldn't you?). For example, you could record a person talking or singing, keys rattled and whatever else you wish for, the play back the recorded media to see how closely it is matched. Of course this will depend on the quality of the recording chain and microphone. But, it's better than nothing and will at least let you know if there might be problems.
Right now I'm listening to a particularly nasty sounding little amplifier on the bench. A Dynaco SCA-35 with a set of new tubes. My normal bench system is far more accurate and doesn't make me want to leave the room. Using the same CD source, so they are on the same footing. Same speakers too.
Agree. Within reason of course.
Dogs can have rather better high frequency response than humans, and they are used to hearing live sound as that is what they pay attention to. If the system reliably fools the dogs, it probably resolves fairly well. The same dogs ignore these sounds coming from less good reproduction systems, for example the TV speakers. My wife listens to the TV speakers these days because she can no longer hear stereo. She finds the sound system reproduces some frequency ranges that she finds annoying. Damaged hearing. She enjoyed the big systems before as little as 2 months ago.
So yes, both me and my dog are happy with the better systems. When we had a bunch of Poodles, they would come in and listen to the main system even when it was rather loud. THe same animals have left the room when a less good sounding system was playing. It appears that even the dog has his limits.
-Chris
Wow, this is a very good example of circular logic.
I asked how you judged what the recording engineer intended (in order to judge the realism of your system) & you answered as above 😕
And my answer is same as above
Exactly what definitions are you talking about?
I'm afraid you don't make sense
I didn't say it was your problem - I simply asked have you done this - have you? Do you follow this process every time you audition a new device in your playback chain?
I also wonder & doubt if we can setup a suitably dynamic acoustic soundscape in our listening room to adequately exercise our playback system adequately - girl with guitar doesn't cut it, IMO
I think you are too invested in the idea of frequency response as being the criteria to judge a system by?
- Status
- Not open for further replies.