Right, that's why I said normal musical levels of ultrasonics. Artificial HF noise from a codec is another story. I understand a lot of things are possible in certain cases. But for music, especially from CDs, I'm still not seeing a reason to worry. From my other posts, you may know that I am not convinced every little distortion some people claim to hear is always properly attributable to imagination. I think people do hear some distortion sometimes that might not be expected from looking at schematics alone, and concluding that audible distortion is not possible in every case may not turn out to be right if only we could take measurements and get some real numbers from out in the field. I am also not convinced that all prior psychoacoustic testing was done as well as it could potentially be done now. We do know more about memory retention of subtle sound differences, and other factors, that might mitigate towards updated testing protocols. Time will tell, I suppose. Right now it's not a very hot area for academic research. Somebody has to come along who is interested in doing a lot of work, probably not for much glory.
Last edited:
You could probably get a similar effect with cymbals, if you were to translate their full spectrum up to 100kHz into the amp input. According to the link I mentioned, most of their energy is in the ultrasonic domain. Of course that doesn't happen in practice (it's cut off during recording, even for HiRes formats), but the question about the honesty of the amp producers still remains.
Mark, I came to the conclusion that I'm way more sensitive to tonal balance than to distortions. Also on whether the instrument I want to focus on is mixed in a clear, distinct way or buried into some amorphous mass of sound. While I can correct the first to some extent, the second is intractable. These are both non-controversial issues I hope, as they are definitely audible in the most stringent listening tests. I also hope that they will be addressed with priority by "the industry" once they get rid of "purism" and "loudness wars".
I'm wondering what this late drift is trying to accomplish? By and large, we're talking (at the amplifier) 20-20 kHz bandwidth (obviously can go lower, but minimal content down there), unless there's a digital glitch (shouldn't be) or noise shaping followed by very poor low-pass filtering on the output. Mic's rarely exceed 20 kHz effectively and regardless, there's little energy up high (much less an inability to hear it effectively).
Still, cymbal sounds do not start at stop with fast square wave envelopes. Hitting a cymbal with a drumstick and having it bounce off is not a that fast rise time of an event. So, i'm still not too worried. You may think of something yet that I'm not expecting though. It's often good to think things through thoroughly and try to cover all bases.
Regarding amplifiers, frequency response is probably extended more to limit phase shift in the audio band, say up to 20 kHz, than for the purpose of providing sustained full power use above normal audio frequencies. Manufacturers could design for that of course, but most consumers wouldn't want to pay for the added cost. What the consumers presumably want to pay for is the part about limited phase shift over they frequencies they are good at hearing. That's my take on it anyway.
Last edited:
Trying to get everybody on the same page is all, or a little closer to it. Not everybody has an advanced degree in engineering and this is a DIY forum. Don't happen to like calling people names as an alternative. Words like foolish, nonsense, and tinker don't end the debate, often they tend to aggravate it. And they don't seem to motivate the study of higher math either if that is their purpose. So, I am trying to take this forward a little at a friendly conversational level and see how it goes. Any problem with that? IMHO, we could use a little more of it.
Yes, a similar interest here. My mixes have been described as very clear, as it's what I like. It starts with song, arrangement, instrumentation choices, tracking choices, and mixing. That being said some music sounds better a little garbled up and distorted. Also, to make a record sound like a record, to bring some unified coherence to the overall sound, there is the issue of what people call "glue." Typically it means running the whole mix through some subtle processing, including adding a little (very little in most cases) euphonic distortion. Makes everything sound a little more like its supposed be all there to make a complete whole and not just a collection of timbrally and dynamically uncorrelated sounds. Often gluing is the mastering engineer's job, especially if the record includes tracks recorded at different studios and by different personnel.
Mark, I have had zero problems with your late comments, sorry you took it that way. It really seems you and I miss each other tone-wise.
My point was more the stirring/confusion about the demands of audio and a need to appreciate the effects of bandlimiting. Something you're trying to (peacefully) explain. 🙂
My point was more the stirring/confusion about the demands of audio and a need to appreciate the effects of bandlimiting. Something you're trying to (peacefully) explain. 🙂
I guess a good mechanical engineer could enlighten us on the vibration spectrum of two pieces of metal hit together. There's surely some serious deceleration happening. I am not any good at it so I will refrain from further commenting on the matter.
Coming back to the amps, I believe you are correct and the purpose of the extended BW is to show impeccable phase charts at "audio" frequencies and somewhere beyond them. However the designers bet on the fact that the CD player will take care of the BW limiting, (and were probably aware that it will screw the phase anyway). So those impeccable charts can be seen as disingenous - instilling a false sentiment of "everything will be perfect way beyond what you could possibly hear". Also I was talking about "hi end" stuff - in the $10k ballpark or more, where the customer has a legit expectancy that the amp should be able to deal with everything you can throw at it, within its spec sheet.
It would hald been more honest if the amp maker would had placed himself the 50kHz filter at the input of his piece. But hey, that would had produced "inferior" charts, wouldn't it? Can't charge 10 grand for such specs these days, can you?
Coming back to the amps, I believe you are correct and the purpose of the extended BW is to show impeccable phase charts at "audio" frequencies and somewhere beyond them. However the designers bet on the fact that the CD player will take care of the BW limiting, (and were probably aware that it will screw the phase anyway). So those impeccable charts can be seen as disingenous - instilling a false sentiment of "everything will be perfect way beyond what you could possibly hear". Also I was talking about "hi end" stuff - in the $10k ballpark or more, where the customer has a legit expectancy that the amp should be able to deal with everything you can throw at it, within its spec sheet.
It would hald been more honest if the amp maker would had placed himself the 50kHz filter at the input of his piece. But hey, that would had produced "inferior" charts, wouldn't it? Can't charge 10 grand for such specs these days, can you?
Last edited:
I have related that I have consistently encountered sonic change in amplifiers during first run up after repair and major resoldering, and that this change is related to peak signal level.
There is something curious going on, and it is related to the highest peak encountered in the music passage.
Once that instantaneous peak level has excited the amplifier stage, the amplifier/system takes a 'set' which continues despite reduction in signal level.
I have also stated in my experiments with filtered USB cable transferred files that ABAB sounds different to AABB and is a confounder to ABX testing.
It seems that signal embedded noise alters the PB system, and that the PB system is readily modifiable according to the nature of the signal embedded noise.
This I think is also reinforcement for the 'burn in' observations stated by many reviewers.
Dan.
Agreed. We're probably focusing on the wrong things, since the most important part is happening in the studio. And, as we already noted, a "disturbance" can be also seen as "useful signal", and a "distortion" as "euphonics". It heavily relies on the good (musical) taste of the studio people, and to allowing the listener to customize the presentation towards his own taste. There's no math formula conducing to an unique, universally accepted "best" solution.
FYI: I'm not deliberately ignoring this question. I haven't had the time/brain-power to give an answer.
P.S. That's a nice amount of phase margin on that fast of a part. 🙂 Wait, isn't 90 degrees idealized? No HF poles? (I first read this as *approximately* 90 degrees)
Last edited:
The drumsticks I am familiar with have plastic or wood tips. Some beaters use felt tips. I have seen a gong hit with a metal mallet on rare occasion, however. But the recording was band limited to CD quality. So, no smoking gun as it turned out.
Now you have the name. Let's develop a new gold mine. We need a story, and a theory around the name. I can provide some experimental data, with glowing big tubes. 🙂
I have some, from a colleague in Israel. Laser ray and tiny membrane do the trick, but nobody promised noise - free life!
Mark - just found this about the modal frequencies of a tuning fork. There's nothing to suggest that there's an upper limit, at least not until we hit some quantum mechanics principle.
http://www.acs.psu.edu/drussell/Publications/Rossing-Russell-Fork.pdf
http://www.acs.psu.edu/drussell/Publications/Rossing-Russell-Fork.pdf
It's common for academia to attribute something to the person that made the most lucid contribution on a subject. Barrie Gilbert takes some real venomous comments because the literature put his name on the 6 transistor multiplier cell a configuration he did not invent, but described in wonderful detail. People make comments without understanding what they are talking about all the time, so what.
It's fairly easy to overstate the "limitations" of Fourier theory. Oh the integrals are from plus to minus infinity so obviously they don't apply to anything real.🙄
Could you point to the section of this article that suggests this? I would love to think my Tibetan singing bowl is a quantum mechanical device.
Last edited:
They don't entertain the implications of extremely large n, it has no use to them. Someone versed in all physics modeling might be able to add some insight but constrained metal rods of substantial physical size oscillating at 100's of MHz is a stretch, but you knew that.
- Status
- Not open for further replies.