That's pretty much what I would expect large amounts of IMD, aliasing or jitter to do. The tone of a cymbal has a complex and delicate high frequency structure, and it doesn't take much to scramble it to mush.
It is also my biggest gripe about MP3, it throws away all of the structure and makes cymbals sound like a sort of random splashing.
Interesting your comment about MP3, IME this is clear "evidence" of faulty playback, rather than the MP3 encoding. Why I say this is that I have done the experiment many times of taking an MP3 file, encoded at a decent rate, and regenerated the WAV file, which I then upsampled to hi-res quality.
Play back the stock MP3 file, cymbals sound like random splashing; but then play the 'upgraded' version of that same file - and lo and behold, quite decent cymbal sound emerges! Just like 'magic', 😀 ...
How can this happen? From my POV, the MP3 decoding load on the processor is enough to inject higher levels of interference into the analogue side of the playback - and poor sound results. Bypass this step, by decoding the MP3 beforehand, and even better, upsample the uncompressed version, so that the DAC and following circuitry has an easier job of things - and quite significantly superior sound will result ...
Play back the stock MP3 file, cymbals sound like random splashing; but then play the 'upgraded' version of that same file - and lo and behold, quite decent cymbal sound emerges! Just like 'magic', 😀 ...
How can this happen? From my POV, the MP3 decoding load on the processor is enough to inject higher levels of interference into the analogue side of the playback - and poor sound results. Bypass this step, by decoding the MP3 beforehand, and even better, upsample the uncompressed version, so that the DAC and following circuitry has an easier job of things - and quite significantly superior sound will result ...
What do you consider to be a decent rate? I stop noticing the "splashy cymbal effect" somewhere between 256 and 320kbps.
The old style of USB DAC could be influenced somewhat by processor load. The new asynchronous ones are much better. The sampling clock comes from a quartz crystal in the DAC and the audio driver is slaved to it. So you either get jitter-free playback, or obvious nasty glitches if the CPU can't keep up.
I'm not sure whether you are confused or just trying to be confusing. Pressure variations in air obeys the same mathematics as voltage variations in a piece of wire, so Fourier is equally applicable. Note that Fourier theory was originally developed (IIRC) to calculate heat flow in solids. That is the wonder and mystery of maths: universal applicability.45 said:
Mathematics can't tell us what our ears/brains require; only careful listening tests can do that. Once we know (and we do know, to a large extent) we use physics and maths to deliver it. Note that "I added this gizmo, and the sound improved" tells us a lot about your preferences and almost nothing about sound reproduction.
For signal processing to re-insert information which has been removed requires intelligence: either human (off-line processing) or Maxwellian. However, it may be possible for processing to create an artificially synthesised spoof of some of what was removed; our ears/brain are sufficiently easy to fool that they could regard the spoof as more 'real' then reality itself. I suspect this is how naive NOS filterless DACs work.fas42 said:
The thing that turned me off listening to DAB radio in the UK was that below 192k the MP2 processing cannot cope with audience applause at the end of a live piece. The coder just collapses in a heap when presented with so many random transients and just produces a mush which sounds like a white noise generator with random volume variations.
Double blid test is not a solution and has nothing to do with correlating musical sound with the physical sound. You simply can't correlate them! It is wrong from the beginning whatever you do because you are not complying with the hypothesis, as told before. If you change the set-up by a tiny amount your perception of the sound (and even worse of the music which is not a simple sound but a complex event to which an emotional meaning is attributed) will change in a way which cannot be described by means of mathematical tools. A microphone will work in invariant condition and the superposition principle will apply. The brain will react to a change by making a different selection of the total information according to totally different criteria which are not known. It will pay attention to different details and could perceive as totally different (we could call this differential attention?). Music has its own language and interpretation, it has been created and played this way from the beginning. Also the instruments and places where music is played have been created following the same language and rules. Certainly not using electronic equipments. These are not only potentially dangerous but also not necessary at all.....
You are not a scientist, or if you really are you are not a good one, as you have just demonstrated you don' t have the basis to do science!
Microphones and all subsequent electronics do NOT capture MUSIC!
They capture, amplify, transmit and reconstitute sound waves via electronic forms.
It is our brains that finally interpret this as pleasurable music (or not!). We can certainly measure how accurately the original sound waves are reconstituted - measurements - but this has nothing to do with how anyone's brain reacts to the artistry of the performance.
They capture, amplify, transmit and reconstitute sound waves via electronic forms.
It is our brains that finally interpret this as pleasurable music (or not!). We can certainly measure how accurately the original sound waves are reconstituted - measurements - but this has nothing to do with how anyone's brain reacts to the artistry of the performance.
If your curious hypothesis is correct then it follows that we know precisely nothing about sound reproduction: we don't know (even approximately) what frequency range is needed, what acceptable levels of distortion may be, what noise level is perceptible/annoying etc. This is because the accepted values for them have been found by doing listening tests and correlating them with measured electrical parameters.45 said:
The fact that we do know these parameters, and have used them for decades to design audio, is ample evidence that your hypothesis is simply false. In fact, it is so false that it is "not even wrong"!
Yes, but the brain can't react to a change unless it perceives the change. And it can't perceive a change unless the sound waves hitting the ears changed in some way. This latter change belongs to the invariant world.
If you deny this, you are asserting one of two things:
The tools of the invariant world just aren't sensitive enough to detect tiny changes that the ear/brain can detect. This is the part of the hypothesis that I deny. Measuring equipment is extremely good nowadays. It is still pretty stupid, but it doesn't need to understand the nature of a change to detect one.
The hi-fi is communicating with the brain by some other means besides sound waves.
Elephant in the room: The hi-fi can communicate with the brain through the eyes by its shiny looks and glowing tubes. Hence blind testing. 🙂
Full disclosure: I'm actually an engineer, not a scientist. 🙂
Last edited:
Which paramenters? You have already jumped over everything that matters!!! Take a book and read carefully what the Fourier theorem says and then try to understand what INVARIANT means. It's not a curious hypotesis, I am afraid.
It is really sad instead that many people are so superficial and/or blind in comparing a human brain to a microphone!
Acoustics remains the science of physical sound not of music. As I said one can work out empirical laws to get in the right direction but it just ends there.
The bet is still open if you want to come and compare your digital source.
I think you need to understand what invriant means. If something is not invariant you just CANNOT measure it! It makes no sense by definition because you will never get a unique response!! Are you really a scientist????
Do you know what 'invariant' means? It does not equate to 'measurable', as you seem to think. Are you a scientist? I suspect not, at least not a properly educated physical scientist.45 said:
I don't do betting. As LP and CD are fairly easy to distinguish there is no useful information to be had by comparing them: the most that would arise is that someone could express a preference. Preferences tell us very little.
I'm pretty sure that's not what Invariant - Wikipedia, the free encyclopedia means... 😛
I have an issue with the logic behind this anyway. Just because something doesn't give a "unique response", doesn't mean that its presence can't be detected.
I don't think so, you are. I have said that something to be measurable must be invariant. I have never said that everything that is invariant is measurable!! There is a huge difference.
You are wrong again. Until now the preference has been just in one direction! No exceptions. The CD is not just good enough. Maybe in the future they will come up with better solutions.....maybe! Who knows?
You can go on forever discussing on nothing, having no experinece of what you would experience and have certainly never experienced in your life. The reason why I am so sure? You have never listened to a system that has been built following the concepts above. That is sure from what you write.
Last edited:
Yes you have issues. If something you measure now is "white" and in the nest two seconds, after the second measurement, is black what do you do with it???
Last edited:
Could be worse, I could be stuck listening to vinyl! 🙄
You are now down to the personal insult level, I take that as an admission that you have lost the debate.
You are now down to the personal insult level, I take that as an admission that you have lost the debate.
- Status
- Not open for further replies.