John's evaluating the AD825 by a meaningless number, so why not?
From an engineering standpoint, you know, reality.
From an engineering standpoint, you know, reality.
Totally. I think it would be very nice to be able to A/B some of this stuff and really get to hear whats going on! I tried doing a digital simulation one time of that hyserisis curve thing and I use it as a plugin effect sometimes in the computer for mixing. Thing is - who knows how well it really matches reality? Unless you've got reliable listening experience it's all chat and optimism! Lol 🙂 (threads getting a bit trick to read now!)
Sure, I was keeping it as low as I could get away with to minimize other parasitics. I wanted this one as clean as poss just for balancing and isolation. Like you say, not for 'colour' this time but cos it's the best way I've got to send a signal from a mic down a long cable good condition! 🙂
I guess it's what works for "audio", AOL and THD don't compare to any number of more recent parts. I'm beginning to think it's important that the measured numbers need to be poor to remind us that, "they don't matter".
I'm starting to believe that beyond a certain point they don't matter indeed. Other than for bragging rights.
Back to our thread's title. I think opamps got a bad rap because they are mentally associated with the 741 and the '90s. The push for (poorly mastered) CDs, the utter junk that invaded the audio equipment market, vinyl and mag tape collectors having the rug pulled from under them, the elimination of tone controls and the beginning of the era of "matching preamp / amp / speaker / cables" in a never ending loop, etc. Very sad times.
Hi,
Sorry to jump here, but a two cents questions came to me and I can't keep it anymore :
is there any hearable advantage to use a rounded metal can aop (TO99) instead its normal dip 8 counterpart ?
Or does the passive parts around (résistors and caps) will dictate more the sound signature in relation to the pcb layout ??
(Sorry if stupid question, I try to conceptualy to undertand what can be the limit or headrom of tweaking from a good design: here in mind the Pre design gently shared by Monsieur Douglas Self à la LM4562 from TI).
To say it also differently : if any amelioration with the noise floor datas, does it sound differently or subjectivly better ? (Dunno : heat dissipation, shielving because the metal can ????)
regards
Eldam
Sorry to jump here, but a two cents questions came to me and I can't keep it anymore :
is there any hearable advantage to use a rounded metal can aop (TO99) instead its normal dip 8 counterpart ?
Or does the passive parts around (résistors and caps) will dictate more the sound signature in relation to the pcb layout ??
(Sorry if stupid question, I try to conceptualy to undertand what can be the limit or headrom of tweaking from a good design: here in mind the Pre design gently shared by Monsieur Douglas Self à la LM4562 from TI).
To say it also differently : if any amelioration with the noise floor datas, does it sound differently or subjectivly better ? (Dunno : heat dissipation, shielving because the metal can ????)
regards
Eldam
Interesting question. I'm a big believer that the total circuit design is very important. I think in practice pretty concrete things like impedance, linearity and bandwidth count for loads and depend on the whole circuit. After that, things like pcb layout giving better or worse RF immunity certainly makes a difference. Then theres the 'small stuff' - certainly a factor. Passive components don't match their theoretical ideal for example. For me the thing is, the effects can be so subtle that its very difficult to tell them apart from psychological effects. I'm not at all saying its all placebo, but I do think its a tricky area and people passionately disagree! 🙂 I believe people have done blind tests and that sort of thing but I dont know much about it. Does anyone know anyone have any good examples of that kind of thing?
There was an op amp listening test conducted earlier in this thread. It started with post #761 by Mooly, back on page 77. Under the conditions of that test, some people did hear some very small differences when different op amps were swapped in and out, while other other factors were held constant. If you want to listen for yourself, I would suggest using the wav files rather that the mp3s, and listen particularly to the cymbal hits and percussion brush strokes near the end of the song. Obviously, if you have to listen very carefully to notice something, while there may be some differences, it likely won't matter much to most people in typical listening situations. On the other hand, some people (not a majority) seem to be very sensitive to such differences and to them, it may be very important.
In other than listening situations, such as in recording, mixing, and mastering, I think many professionals would argue that a succession of small differences can add up to produce an inferior end product, if sufficient care is not taken at each stage of processing.
After which the very same professionals will happily indulge in "mastering for iTunes" or trying to beat "Death Magnetic" in loudness.
Actually, some complain to various degrees about some of that stuff, but they have to do what the customer wants if they want to have jobs. But, yes, there are some who take pride in making it louder than anybody else has figured out how to do.
Then does it matter if an opamp has 0.00003% vs. 0.00008% THD?
I mean, if you're designing a microphone preamp or recording or live equipment, sure, go for the best you can, but for playback of commercially recorded music, I guess those figures don't really matter?
And that the secret for a pleasant and musically engaging playback equipment does not necessarily lie within those dry numbers?
I do remember the 70s-80s, the amps of that time were not exactly shy with hiss and THD, yet a few of them were "getting it right" and offered us many thousands of hours of outstanding musical entertainment?
Was it 0.00003% THD opamps that spoiled it? Could it be that the more exactly you reproduce the recorded sound, the more it sounds like garbage? Because that's what they are actually recording in the studios these days?
I mean, if you're designing a microphone preamp or recording or live equipment, sure, go for the best you can, but for playback of commercially recorded music, I guess those figures don't really matter?
And that the secret for a pleasant and musically engaging playback equipment does not necessarily lie within those dry numbers?
I do remember the 70s-80s, the amps of that time were not exactly shy with hiss and THD, yet a few of them were "getting it right" and offered us many thousands of hours of outstanding musical entertainment?
Was it 0.00003% THD opamps that spoiled it? Could it be that the more exactly you reproduce the recorded sound, the more it sounds like garbage? Because that's what they are actually recording in the studios these days?
Last edited:
To hear 0.00003% difference in thd I think is impossible. but if You can why not take part that measures better, little here little there and You have 0.01% better thd on end of line,simple why not.
Something like noise, if You hear noise You know its not good, but if You dont hear any noise, why care at what level is it when You dont hear it? But still if You can pick part with lower noise why not.
Something like noise, if You hear noise You know its not good, but if You dont hear any noise, why care at what level is it when You dont hear it? But still if You can pick part with lower noise why not.
I don't know. My guess would be it might depend on the particular circuit in which they are used. I have a single channel, 3-band, parametric EQ that passes the signal through 14 Op Amps. Is there a noticeable difference in sound quality if you stack up 14 of them in series as opposed to just a single stage? Maybe. Probably wise to both measure and listen critically for a given application. If you can't pick a clear winner that way, I guess it doesn't matter in that case.
Perhaps the better question revolves around the metric. So, does "THD" and/or "IM" correlate to perceived quality? And then we go off into blind/ABX vs. not... but skipping that, might there be a better metric that could be found?? (of course if you believe that they all pretty much sound the same, then the question is not for you and is moot)
Please let me respond to this part separately. What sounds best is not always exact purity of reproduction. If it did, we would use only the best small diaphragm omni-condenser mics and .00003% preamps for everything, that is, if we don't absolutely need directional mics. And if we do need directional, we would only use something about like a Schoeps CMC-641 which is very accurate, flat in frequency response, and remains so quite well even off axis. In reality, nobody does that type of thing except maybe for something like recording a live symphony in a great sounding symphony hall. Why not always do it? Because if you do, somebody else will use every trick in the book to make their recordings sound better than yours and nobody will hire you to record anything (excepting maybe symphonies, and such).
Last edited:
If everything else is equal then you are right. Unfortunately for the DIY fan, a difficult to solder part (WSON package lol), a huge difference in price or a maverick opamp that would use any excuse to oscillate, can be sufficient motivation to not go for the slight, but inconsequential advantage.
Fully agreed.
My pet peeve are those "purists" riding high horses in a crusade against tone controls and such.
Tried to avoid joining in on this but its too interesting! indeed, music is art and there's no 'right'. It's a specific view and it's taken me time to come to it but thats my position for what its worth 🙂 Also, in the spirit of laying it out there, I think theres plenty thats not yet understood about the numbers. Dont get me wrong - I'm a tech'y guy, but its a big world. If in the computer I reverse the impulse response of a low pass filter so it rounds off the rightmost corners rather than the leftmost - impossible with a 'real' filter but imaginable and do-able all the same, the altered waveform sounds different, looks different, but measures the same by most conventional systems... Stuff's just not simple! 🙂
- Status
- Not open for further replies.