To be fair, one can also set up a servo that adds nothing audible or measurable compared with a coupling cap. Of course, that DOES require engineering rather than voodoo incantations and vague handwaves.
Yes, but it's really hard actually. Many people don't realise that a servo, even if it is supposed to work only at DC and VLF, must also have very good HF performance. It's an integrator, and if it doesn't integrate the mid- and high frequencies well, you'll find mid- and high frequency residual, often distorted, at the servo output being directly inserted into the servo input node.
Jan
oh but i do all the time listen to music all the time......
i just don't go around forums telling people this and that....
if i have to strain hard to listen for differences in sound with different caps,
where is the fun there? and if i go around forums telling people this or that cap is better,
what do i gain from it? do i win a lollipop?
i use caps the way i see fit....no more no less...
Indeed, that's why many rely on voodoo and handwaving rather than engineering.
The other workaround is to filter the output of the servo, either explicitly or indirectly.
oh but i do all the time listen to music all the time......
i just don't go around forums telling people this and that....
if i have to strain hard to listen for differences in sound with different caps,
where is the fun there? and if i go around forums telling people this or that cap is better,
what do i gain from it? do i win a lollipop?
i use caps the way i see fit....no more no less...
Looks like you've outgrown the forum
Looks like you've outgrown the forum
i am a busy guy building my own transformers used in my amps...
i simply do not have the luxury of changing caps nor dabbling with cables....
i was educated in the engineering sciences so that most of the things
that so called audiophiles get involved in, i do not have the appetite, the energy, time and most importantly money....
i have been like this ever since....
and thank's to a lot of guys like SY and you and a lot others, i gained more knowledge here than anywhere else....
Interestingly, one of the tests I do with capacitors (only works for electrolytics) is to drive them with a square wave through a moderate value resistor, say 2k, and look at the residual across the cap. With the parts proportioned correctly, it answers the question of, "how good an integrator is it?" Since the drive voltage is high (say 15V from my func gen) the caps sees essentially constant current.
The results vary widely, with higher voltage low esr parts giving an almost perfect triangle wave, and low voltage (icky 6.3V), ultra miniature and other parts showing offsets at the switch points due to esr and even obvious non-linearity. Tantalums are especially bad for non-linearity.
It doesn't tell me much that I can't get in other ways, but it's an easy test for anybody that can come up with a square wave and a scope.
If Al caps are so bad, how does M-Audio get away with using them in the output path of their sound cards like the M-192?
Granted, film caps are better, but I fail to see how anyone can hear .0006%thd.
M-AUDIO - Audiophile 192 - High-Definition 4-In/4-Out Audio Card with Digital I/O and MIDI
Granted, film caps are better, but I fail to see how anyone can hear .0006%thd.
M-AUDIO - Audiophile 192 - High-Definition 4-In/4-Out Audio Card with Digital I/O and MIDI
Marce,
To go rather OT, here it may surprise you that many of the audio greats never listened to their products until after the design was completed; particularly after the 70s, when knowledge and instruments were capable of modelling hearing to the extent that whatever was audible, could by then be quantified as such.
Leak, Walker, Williamson, Self and many others (Doug is a member here and can correct me). I am unfamiliar with the Eastern scene, but I have never read about the following conversation taking place, e.g: "There's a little brightness around 4 - 5kHz. Shall we - let me see - decrease R39 somewhat; what do you say, Jack? .... Oh! You think it is rather lacking there! .... (momentary frown) ... Then we must increase the value .....
OH! You reckon we should change the mains cord ....
Not contradicting you, Marce, but I did about the same ... must have done most things right over decades, if I have never had disagreement over the acceptability of my own amps.
Of course, commercial viability would be entirely dependant on the client's hearing taste. And with the variation in that .... For whose taste must the poor designer design?
I am intrigued and pleased that some just trust everything down to engineering, I would have thought a listen at the end of any design that involves audio reproduction would have been at least a cursory part of the design cycle.
I don't get the bit about the resistor conversation though! Wot are you trying to say?
The field I have been quite heavily involved in over the last 6-7 years does extensive listening tests, though not the sort where people sit around in a r4oom sharing anecdotes. These are modified rhyme and similar tests involving as wide a selection of victims as possible, all having a hearing test beforehand, mainly to ensure spoken words are easily understandable, but also to determine other factors.
Last edited:
Hi Marce,
(A little OT; beg that the OP will allow)
It is not just trusting engineering. Audio design is (or should be) based on practical knowledge/evidence of what is audible, otherwise it is simply physics. It has often been stated that audio design is based on hundreds of documented hearing tests using many hundreds of listeners, over decades. And surely this represents a better picture that the opinion of one batch of listeners however 'trained', in a room at one time?
It is therefore not a cursory part of the design cycle, it is an essential basic part of audio design - otherwise, as asked before, for whose specific taste must the designer design? Surely the documented results of many listening tests are not inapplicable simply because they were conducted in the past? If that is not accepted then this discussion falls apart. To repeat, any engineering of any value must be and has always been based on reality.
I tried to illustrate what might take place should someone indicate that something was amiss in a final product. What would the designer do? Would that imaginary conversation take place, involving mostly thumb-suck - or would he go back to audio engineering as above to work out how to address the perceived problem?
(A little OT; beg that the OP will allow)
It is not just trusting engineering. Audio design is (or should be) based on practical knowledge/evidence of what is audible, otherwise it is simply physics. It has often been stated that audio design is based on hundreds of documented hearing tests using many hundreds of listeners, over decades. And surely this represents a better picture that the opinion of one batch of listeners however 'trained', in a room at one time?
It is therefore not a cursory part of the design cycle, it is an essential basic part of audio design - otherwise, as asked before, for whose specific taste must the designer design? Surely the documented results of many listening tests are not inapplicable simply because they were conducted in the past? If that is not accepted then this discussion falls apart. To repeat, any engineering of any value must be and has always been based on reality.
I tried to illustrate what might take place should someone indicate that something was amiss in a final product. What would the designer do? Would that imaginary conversation take place, involving mostly thumb-suck - or would he go back to audio engineering as above to work out how to address the perceived problem?
What is an IT?
And how does it avoid using A DC blocking capacitor?
Using a DC servo does not avoid using a DC blocking capacitor.
The range of adjustment of a DC servo is some proportion of one supply rail.
Most power amps have supply rails that far exceed the supply rails of the DC servo circuit.
A DC blocking capacitor could see tens or hundreds of volts on the far side. There is absolutely no hope that a DC servo can insert sufficient current to bring this back to very near Audio Ground level.
Last edited:
Well - er - yes, Tinitus!
Over these threads one detected a lot of resistance to using a capacitor, with occasional strong feelings surfacing. Not to discombobulate the thread at this point, but I have a few questions regarding this fixation with "the best cap is no cap", which have never been answered satisfactorily. Perhaps soon after this one can start a committed thread. It is unlikely to alter entrenched perceptions, but might hopefully encourage newbies to first consider electronic basics realistically before joining the "ban-the-cap" fraternity.
But perhaps peace for now .....
Over these threads one detected a lot of resistance to using a capacitor, with occasional strong feelings surfacing. Not to discombobulate the thread at this point, but I have a few questions regarding this fixation with "the best cap is no cap", which have never been answered satisfactorily. Perhaps soon after this one can start a committed thread. It is unlikely to alter entrenched perceptions, but might hopefully encourage newbies to first consider electronic basics realistically before joining the "ban-the-cap" fraternity.
But perhaps peace for now .....
Ideally this might appear to be so , considering there are no dc offset voltages to deal with.
In practice I have come across situations where a capacitor did sound better than a direct connection and vice versa . DC offset voltages were very low. So without going into reasons , it might be a good idea to try with and without cap to find out which sounds better in 'that particular system'.
Some low cost industrial polyprop caps can sound as good as some very expensive boutique capacitors. Try all types that you can afford to get . There is never only a single solution.
I've used some polystyrene caps that sound very good. Bottom line is to experiment. You might find a surprisingly inexpensive solution.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.