Thanks for the MP3 graph. Is the noise floor really that much higher? 😱
BTW, what software are you using for your FFT? I'd like to use the same, if I can.
I have the mutlitone file and will post it - host it later today. Will try some different versions, too. E.G. low pass filtered 3dB/octave, differents bands, etc.
Should be interesting to play with.
BTW, what software are you using for your FFT? I'd like to use the same, if I can.
I have the mutlitone file and will post it - host it later today. Will try some different versions, too. E.G. low pass filtered 3dB/octave, differents bands, etc.
Should be interesting to play with.
I still have Cooledit from when it was $50. Unfortunately it does not have great 24 bit support. Audition might have improved the high res file support but now costs real money. You might get enough functionality from the unlicensed versions of any number of other programs, but I have found the FFT features generally do not let you "get" at every detail (a complaint with Cooledit too).
For instance computing the equivalent noise BW of a bin is always a pain and I end up running mathematically computed reference files to calibrate all the time. Matlab or Mathcad give you real FFT libraries but GUI wise things are a pain.
I like a nice point, drag, click interface. I wish the FFT window in Audition had all the underlying access to the data as the waveform window.
For instance computing the equivalent noise BW of a bin is always a pain and I end up running mathematically computed reference files to calibrate all the time. Matlab or Mathcad give you real FFT libraries but GUI wise things are a pain.
I like a nice point, drag, click interface. I wish the FFT window in Audition had all the underlying access to the data as the waveform window.
BudP said:
You talk as if something has already been established???
Have you or anyone else done any null (or other) tests whilst altering the dielectrics, because any changes to the end sound are caused by changes to the electrical signal. If you can measure any change in the signal, then there is a real possibility that someone could hear it, given enough amplitude. If no one has done this then it is all just subjective experiences and so totally dependant on the individual.
With your explainations you have "jumped the gun", there is no need for any scientific explainations at this time, and they are inappropriate unless you are trying to cause subjective experiences in others.
Although I totally accept your subjective experiences as reported, when you claim that these are caused by dielectrics changing an electrical signal, you have something to prove not just report. You need some measurements. 🙂
Andre Visser said:
Why the urge to provide alternative explainations for why you hear stuff that is inaudible, don't you believe in psychoacoustics?
Scott's two graphs, a breath of fresh air, I can hear a difference, I can measure a difference. Come on cable guys show us your graphs.
SY said:
Like being "plumb", two guys putting up a fence,
"is it plumb?"
"No if anything it is bit over plumb"
panomaniac said:
It's probably different for the different flavors of mp3 algorithms. In my limited experience ogg passed tests better than mp3 in regards to transparency.
I think a lot of these use "perceptual encoding". Which basically means they limit everything to what they think are the limits of human perception. So anything under a certain frequency (20 or 19kHz for example) will be filtered. So I guess it does the same for dynamic range with that algorithm. I know if my ears are being swamped by music it is hard for me to hear past the -40dB mark even though my speakers are probably still outputting signal. But yeah that does look degraded.
Sorry I did not get a chance to upload Scott's signal to my website yet. Will do as ASAP.
Also downloaded the AP software and made a multitone there. Very easy. Nice that it will randomize the phase of each tone if you want. Frequencies, amplitudes and phase can be edited.
I thought Scott's choice of frequncies was a little better than the defaults in AP.
Yes, pretty much so!
Also downloaded the AP software and made a multitone there. Very easy. Nice that it will randomize the phase of each tone if you want. Frequencies, amplitudes and phase can be edited.
I thought Scott's choice of frequncies was a little better than the defaults in AP.
Yes, pretty much so!
scott wurcer said:
Yes Scott, you are right, unfortunately it seems like all are not masters of everything. I've speculated on that thought because it may explain the effect that some of us hear, done with the hope that the clever ones will realise that we are talking about different influences on sound than what they think.
I've read a paper written by a clever one with possible explanations of cable differences that correspond with what I hear, somehow some that claim to be more clever, discard his views and give explanations that are in conflict with my experiences. You may guess which one I believe.
SY said:
SY, surely we are talking about different 'timing'. This is like using a calendar for a stopwatch. 🙂
fredex said:
See my answer to Scott. It wasn't intended as an explanation, more like a thought that may help you realise where to listen for the differences we are talking about. Provided your system can reproduce it of course. 😉
Andre Visser said:
But Andre, wouldn't it be smarter to try first to establish that what some of us hear (or perceive) is in fact an audible difference? It doesn't seem efficient to start to speculate about possible causes for something of which we don't even know how real it is.
jd
Andre Visser said:
OK.
So you are talking about low level voltage changes. I still find it very hard to accept that you can hear a voltage change that is too small to measure, no matter how good your system is.
Key said:
That was a stock Fraunhofer constant bit rate (10yr. old) encoder. This is one interesting test for where the "lost" information is put. BTW the frequency information (timing?) is held remarkably well. This might be a little OT but it does relate to hearing perception.
Andre Visser said:
As I said before this is hard thankless work. Measuring -160dB THD on a Vishay resistor vs -145dB on a Dale does not in of itself have any relevance to audibility. There are no two (even samples of the same) item that will not have some difference even measurable ones.
Just imagine trying to compare headshell wires. Who would commit to setting up two TT's to measure <exactly> the same. Do you think for instance any two samples of a Denon 103D are immeasurably different?
Scott, I'm not sure whether you agree or disagree with me. For sure there will always be measurable differences between equipment or even components. Perhaps we only differ on what is audible or not.
janneman said:
Of course it is audible, I always listen with my ears. 😀
Jan I believe there are more than enough evidence for audibility, there are no one as blind as one that doesn't want to see, the same with hearing.
Andre Visser said:
Hearing happens *between* the ears.
Andre Visser said:
Yeah. And everybody carefully hides it 😉
jd
fredex said:
Who said it is too small to measure? There are enough measurements that show differences between cables.
I also believe that amplitude measurements alone can be quite misleading, seeing that an audio signal is quite complex, I believe that relative phase information (timing) are quite critical. Even a drum produce frequencies that go well into midrange, if we distort the timing between these different frequencies I'm very sure the end result will not sound like a real drum, let alone what it may do to sound localisation.
- Status
- Not open for further replies.