Correct, mostly. While 'pleasing' doesn't say anything whether the reproduction is accurate or not, 'euphonic' generally means that the sound has been changed between source and reproduced, so by definition that is less accurate.
If it measure better it is by definition better, more accurate and more transparent.
That doesn't mean it is preferred or pleasing of course.
The red thread here is that you and others here fail to understand the difference between preference (personal) or being pleased by the sound, and the accuracy of reproduction.
Jan
As soon as you put more than 1 op-amp in series, you will have your local feedback.
Jan, I'm a little more worried that the amount of CCS inside the package seems not numerous enough to make the provided datasheet sample schema effective or even relevant. Well, that happens most of the time. However, it goes even worse for me when it doesn't go that way, because I probably tried everything else first.
For example, with the higher current op-amps, those are looking very poor in comparison to Class D offerings, which differ by more practically useful datasheet examples, even though they usually do omit a couple of parts just for sport, or possibly tradition. Not that the output quality achievable is much different, but that the labor to achieve it is the difference. So far as unit amplifiers, usually op-amps, both really old and really new datasheets have useful examples, and almost everything in-between, doesn't.
With the high current Class AB op-amps, the inexact documentation can be easily fixed by using far lower power voltage, then everything in the documentation except output power works perfectly well. Or much higher resistor values (same as running a simple LED on higher voltage) and significantly higher gain, can be use to get everything except the published distortion figures. It is possible to find a perfectly nice middle ground; but, that does happen to be laborious, as well as unmentioned, usually.
On a somewhat different topic, regardless of whether I may be slightly off or not, I'm not actually crazy enough to use a made for amplifying design as a buffer--that usage of op-amps doesn't work adequately. That's another datasheet fault whereby the advertised applications are more vast than what the part could do well.
Wouldn't the matter be quite different if the documentation mentioned a task that the part does extremely well and provided a thorough example of that?
Edit: This comment is now public domain. Use it however you like.
Last edited:
Getting predistortion right is difficult even when you are trying very hard and have full knowledge of the distortion you are attempting to undo. To get predistortion right by pure accident? No. This is pure vinyl myth.bear said:
Back in the late 80s had a talk from a Natsemi rep who offered a bottle of whiskey to anyone who could find a data sheet without an error in it. Things will be better now, but it still happens.
Ref your comments on 'poor in comparison to class D' would you care to give some examples as not sure what the application is.
Ref your comments on 'poor in comparison to class D' would you care to give some examples as not sure what the application is.
Jan, quite self evident.
Except for one thing, how ought one define "accurate"?
IF the test for accuracy is "closest sounding to the original source" then that is quite different than "least changed from (the potentially flawed) recorded version".
Ought not the played back version sound closest to the actual voice, and closest to the actual violin? (aka, recognized by the brain as closest sounding to "real") Most folks do not have a means of direct comparison, but some do.
It's clear that in engineering this issue generally does not come up, engineering is pretty much focused on what we can call "technical accuracy" based on equations/simulation and then measurement, which of course is exactly what its role ought to be.
Everyone understands (I would think).
Yet, LP at its best (if you have ever heard such a thing) does have a certain "quality" that is unique. LP at its average or worst is a mess. Don't confuse these.
You say "by accident". Yet all sorts of physical processes occur naturally by "accident" - it's more an inherent property both of the material itself and the method.
How about tape? Tape sounds like tape, even when there is a lot of not so great solid state before and after, the "tape" quality to some extent is present - and not just when its hit hard into saturation...
...in general I think those who are "married" to engineering as the sole criterion are missing or ignoring the basic truth that the mind and ear is pre-programmed to interpret/decode/understand/recognize sound that occurs naturally. The mind is always attempting to frame what we hear in those terms - not in some other new way. So, regardless of what engineering criterion is applied the sound that is actually most "accurate" is the one that is easiest, fastest and simplest for the mind to interpret/decode/understand/recognize. Anything that delays or otherwise reduces the speed and ability to do this, constitutes de facto less accuracy.
Now the ideal would be to have technical/engineering accuracy converge with mind/ear accuracy. The discussion about opamps is because some seem to discern something going on - EVEN IF IT IS DUE TO IMPROPER IMPLEMENTATION - when opamps are used (in some cases, and not limited to opamps, but we're talking opamps) that reduces the mind/ear accuracy.
(if it is due to improper implementation, then that needs to be discovered, documented and shown beyond reasonable doubt that with proper implementation whatever detected "problems' may be there, have gone away...)
_-_-
Quite simple. As an audio designer, I have no idea how the original is supposed to sound. None of us have, even those that were at the live event, if it is longer than a few hunderd millisecs ago.
So the best we can do for accurate, is take the signal and reproduce it as faithfully as possible. And that can only be veryfied by measurements.
Now, you may say, yeah but I feel it should have a bit more presence, or a few dB more below 200Hz, or cut slightly above 10kHz due to the live room you're in, or whatever, to approach what you *think* it should sound.
That's OK, and I say, go for it.
But don't come to me saying that I should use an amp that cuts slightly above 10kHz because it sounds sooo much better. I really hate it when someone tries to force his personal preference on the rest of the world, on the grounds that he/she likes it, and therefor it must be the best.
Jan
Last edited:
If the purpose is "accurate reproduction of what the mastering engineer heard" then you can just buy the same setup (amp/speakers) as him. Problem solved.
If the purpose is "accurate reproduction of what the mastering engineer heard" then you can just buy the same setup (amp/speakers) as him. Problem solved.
silly wabbit.
Actually one of my standard tests is to listen to the person speaking or performing through the sound reinforcement system, then turning the amplification on and off to check for changes.
Now as SY has already shown it took six opamps in series before he could distinguish between that and a straight wire. Six not five! Now the harmonic distortion of the combination should have been well below the perceptible limits. So what was it that was detectable? Or why did one opamp make a difference?
Jan, the first part of your post is something to consider.
There is no reason whatsoever that you or anyone else can not record for themselves live sound, and therefore "know" what it sounded like. Today it is completely trivial to do so.
By way of a suggestion, it would be most useful for anyone who engages in the "art of hifi" to do so, especially for those who rely mostly on engineering.
Doing this, even though it is not likely to give terribly great sounding results at the outset will make clear aspects of this process of capture--->encoding--->storage---decoding--->reconstruction more evident.
Unanticipated benefits might be forthcoming, worst case no harm is done. 😀
Saying that nobody can know what is in a given recording, I think is a mistake, since one can certainly create a baseline for comparision for themselves, and so draw some conclusions as to what has been done to and with commercial recordings. Which may in turn shed light on what is up with playback.
_-_-
There is no reason whatsoever that you or anyone else can not record for themselves live sound, and therefore "know" what it sounded like. Today it is completely trivial to do so.
By way of a suggestion, it would be most useful for anyone who engages in the "art of hifi" to do so, especially for those who rely mostly on engineering.
Doing this, even though it is not likely to give terribly great sounding results at the outset will make clear aspects of this process of capture--->encoding--->storage---decoding--->reconstruction more evident.
Unanticipated benefits might be forthcoming, worst case no harm is done. 😀
Saying that nobody can know what is in a given recording, I think is a mistake, since one can certainly create a baseline for comparision for themselves, and so draw some conclusions as to what has been done to and with commercial recordings. Which may in turn shed light on what is up with playback.
_-_-
Bear, the whole contention starts with the numerologists claim that the recording studios have some secret formula that they use in order to get the recording "right" down to the last bit on the CD. And that by playing back that CD with just that bit wrong, or with anything higher than 0 THD etc will produce much grief and loss of enjoyment.
To which I will only say what they always use to say: prove it.
I have never seen anyone say that, I have said in public if you like .75W SET amps and ultra-sensitive full range horns fine by me. Don't say it's not an "effects" box.
Two coincident pairs can record the same performance from 6 rows or 20 rows. I suspect taking out the time delay the recorded files will be very different. If op-amp rolling can switch given one, to the other, prove it.
I have never heard that claim from anyone. Have you ever been in a real recording session? It's all done by ear. If you get ten top engineers to record the same thing, (even some thing simple like a solo guitar) you will get 10 different sounding recordings, and all will sound good. So when you playback one of those recordings how do you know what it's meant to sound like, your guessing and so your also guessing at the by ear tweaks you do to your system. If you use one of the other recordings of the same thing to judge your system your tweaks will be diffrent. So go on chasing your tails, sounds like you've got nothing better to do.
And passing your music thru a LATHE is more acurate than a good opamp? (Not to mention all the other steps that degrade the sound when making vinyls) LOL. Hard to believe anyone's ears if this is what they think they hear.
Subjectivist delude them selves into thinking they know what a recorded whatever sounds like. They then use this delusion to tweak there systems. And as said hundreds of times before, go for it, just don't tell us it's more acurate.
- Status
- Not open for further replies.