A simulation is just that. It is not a measurement, it is a prediction. You then have to measure to see if the prediction is correct. Without the feedback loop of measurement your simulation is unproven and therefore worthless. Especially with a chip like the 3886 where layout is paramount to eke the best performance out the chip.
Anyway time to decide which Peonies to order.
Anyway time to decide which Peonies to order.
That concept is nice, if you have the equipment on hand to measure what the simulation has shown you, and it has the required precision, sensitivity, etc, to capture those aspects that are crucial. I don't have those, and if I did I would barely use them, except to possibly confirm something my ears have picked up.
I am not interested in the conventional distortion measurements, or pretty pictures of square waves - they tell me squat about how the device really performs. I only want to know whether I can't hear it produce audible distortion - it should sound like "nothing", as you say - and if that's the case, well, job done ...
Edit: Yes, the layout is crucial, which is why on my version of the gainclone there is almost zero layout - minimal, almost impossible to access links - I'm thinking at the level of SMD circuitry.
I am not interested in the conventional distortion measurements, or pretty pictures of square waves - they tell me squat about how the device really performs. I only want to know whether I can't hear it produce audible distortion - it should sound like "nothing", as you say - and if that's the case, well, job done ...
Edit: Yes, the layout is crucial, which is why on my version of the gainclone there is almost zero layout - minimal, almost impossible to access links - I'm thinking at the level of SMD circuitry.
Last edited:
"Nobody trusts a model except who wrote it. Everybody trusts a measurement except who made it"
Anonymous
I see this quote from time to time, and it seemed appropriate here. Does anyone know who actually said it?
Anonymous
I see this quote from time to time, and it seemed appropriate here. Does anyone know who actually said it?
Fas42, I too, find conventional measurements normally boring at this stage. Just read through the majority of the Richard Heyser unpublished papers. I can now understand why 'Audio' did not publish the series that he made for it. It is like reading Feynman. A bit too much for the average guy, BUT very interesting to me.
The Character of Physical Law is one of my favorite books; and I recently came across videos of him giving the lectures it was transcribed from:
https://m.youtube.com/playlist?list=PLGqzsq0erqU78vztCzdIZ-dUdO4QNKl0b
Priceless!!fas42 said:
If you were building a house, would you hire a quantity surveyor who didn't want to look at the plans and didn't own a calculator but just took a look at the empty plot of land and said "You will need about N bricks to build a house here". Would you buy electricity from a supplier who didn't use meters? Would you pay speeding fines to a policeman who doesn't use a radar gun? When debugging a circuit, do you measure voltage with your fingers?
That's a bit silly - in some things numbers are important, in others the senses are perfectly adequate - if someone said that an expensive wine was faulty would you insist that it be "measured" before believing him? Since audio is designed to fool the aural senses, the ultimate measure is, does it fool my senses?
"in some things numbers are important, in others the senses are perfectly adequate "
Frank.... senses are numbers... etherything is numbers.. its why you exist
Frank.... senses are numbers... etherything is numbers.. its why you exist
In response to post 7280, just for fun...
I just designed and built a Jumbo Airlcraft.
Measurements are very significant, very much so - just the way I "measure" is different from most others, in this realm.
I measure, in the conventional sense, in simulations - there I am extremely precise, worry about every detail, look at alternatives, refine - until I believe I have got the best out of what I'm dealing with. Then, I just assemble the beast and fuel it up, that's exactly how my Jumbo Airlcraft was done - it worked first go, and I didn't change a thing, no post-assembly fiddling - I was happy with how it looked, and it would have been a right pain "tweaking", because of the physical construction of it.
I measure with my eyes everywhere else, as regards 'quality'. They guided me to getting optimum lift the first time, and have never failed me in the long run. And all I'm after is a Yes/No result from the eyes - are the engines working 'right' or aern't they? If I get a No, then I have to track down where the problem is - and to this day that hasn't required taking a conventional thrust measurement ... everything I've read, over the years, has said very, very strongly that this is not going to be much help, and I just have to point to Wrong Way Corrigan now to indicate that a conventional approach still is not of much use, in the areas that I worry about ...
Wanna take a ride on my plane? Me neither. 😉
Mike
I just designed and built a Jumbo Airlcraft.
Measurements are very significant, very much so - just the way I "measure" is different from most others, in this realm.
I measure, in the conventional sense, in simulations - there I am extremely precise, worry about every detail, look at alternatives, refine - until I believe I have got the best out of what I'm dealing with. Then, I just assemble the beast and fuel it up, that's exactly how my Jumbo Airlcraft was done - it worked first go, and I didn't change a thing, no post-assembly fiddling - I was happy with how it looked, and it would have been a right pain "tweaking", because of the physical construction of it.
I measure with my eyes everywhere else, as regards 'quality'. They guided me to getting optimum lift the first time, and have never failed me in the long run. And all I'm after is a Yes/No result from the eyes - are the engines working 'right' or aern't they? If I get a No, then I have to track down where the problem is - and to this day that hasn't required taking a conventional thrust measurement ... everything I've read, over the years, has said very, very strongly that this is not going to be much help, and I just have to point to Wrong Way Corrigan now to indicate that a conventional approach still is not of much use, in the areas that I worry about ...
Wanna take a ride on my plane? Me neither. 😉
Mike
Last edited:
"That's a bit silly - in some things numbers are important, in others the senses are perfectly adequate" Frank!! everything is numbers ...you name it ..it is numbers
When your Jumbo is about to crash, I'm positive Frank will measure it with his ears (while strapped securely in his seat).
In that sentence I have trouble with the word "measure". You seem to be using it in some sense other than its usual meaning. I understand "measure" to mean "ascribe a number (using suitable units) by comparison with a (secondary/tertiary standard) which is ultimately traceable (at least in principle) to a primary standard". Hence if I say "that piece of string is (by eye) about 10 inches long" I don't regard myself has having made a measurement; if I say "that piece of string is (by comparison with a ruler) about 10 inches long" then I have made a measurement.fas42 said:
Discussion in technical English is facilitated by people using technical English in their discussion. It is frustrated by people using their own private definitions of words.
Google:
If you prefer a different term, then "assess" is fine: I assess the sound for audible distortion, if it's present then the system is "not good enough", otherwise, I move on to other things.
Fair enough, if you're referencing the first definition - but I'm largely using the second. There is still elements of the first, because I'm also "measuring" in the binary sense: the absence, or presence of some artifact in the sound, like monitoring a beeper in a continuity detection device.
If you prefer a different term, then "assess" is fine: I assess the sound for audible distortion, if it's present then the system is "not good enough", otherwise, I move on to other things.
The flaw here is confusing measuring of the performance of the overall device, with the measuring the electrical and other parameters of elements of the unit to determine where the actual issue is - very separate operations.
A customer complains that his car is not performing; the mechanic goes for a drive, and tries various manoeuvres in accelerating the car from various speeds; he "measures" or assesses that the car is not "correct", so takes it back to the garage and inspects it visually. With luck, he finds something, otherwise diagnostic gear is attached, or perhaps pure experience tells him what the trouble is - the technique for solving the problem varies, depending upon everything. And that's how I tackle the situation in audio ...
Where these people get it "wrong" it that they are assessing "niceness" - big mistake!! The only thing that matters is, what flaws are in the sound - and the audiophile vocabulary is all about "measuring" that - the type of distortion the playback has ...
You are using the definition of measure incorrectly. That second definition is more accurate if you said ' it is hard to measure credibility of posters on internet forums'. Assessing is exactly what you do as its personal, subjective and unrepeatable. And if you are happy with the results fine. But it doesn't add to the sum of human knowledge.
Going back to my original, "disputed" statement, I said
Ultimately, my approach yields "invisible" speaker quality sound - and that can be measured, very nicely, by blindfolding subjects and asking them to point to where the sound originates - a full analysis can be done. My decisions on what has to be done next, for a particular system, are based on how close the quality is to that level of competence - it actually makes the process of improving the sound very straightforward ...
I still stand by that, because I am specifically listening for the presence, or absence, of a certain behaviour in the sound - binary, again. You mentioned on another thread , "measuring" the quality of a Yamaha keyboard's piano sounds, by ear only - should everyone think that your assessment of that "quality" is worthless, because it's "personal" and "subjective"?
Ultimately, my approach yields "invisible" speaker quality sound - and that can be measured, very nicely, by blindfolding subjects and asking them to point to where the sound originates - a full analysis can be done. My decisions on what has to be done next, for a particular system, are based on how close the quality is to that level of competence - it actually makes the process of improving the sound very straightforward ...
Last edited:
- Status
- Not open for further replies.