Page 2 of 3

Re: 44.1 kHz is slowly reaching its end

Posted: Mon Jan 22, 2018 6:45 pm
by Lyberta
loxstep wrote:Though I'm still not worried about it if the difference is inaudible.
You can run his script and hear for yourself.

Re: 44.1 kHz is slowly reaching its end

Posted: Mon Jan 29, 2018 2:28 am
by loxstep
Lyberta wrote:
loxstep wrote:Though I'm still not worried about it if the difference is inaudible.
You can run his script and hear for yourself.
I did run the script. Maybe I'm interpreting the results wrong. The original file and the 100th iteration of resampling sound identical to me. If I can't tell the difference after 100 times, I'm not worried about resampling something once.

I also compared the original and 100th iteration by inverting one and mixing them together in Audacity. I get a completely flat line, except for a small click in the beginning and end. Not sure how to explain that.

Here's a screenshot
Image

Re: 44.1 kHz is slowly reaching its end

Posted: Mon Jan 29, 2018 7:14 pm
by Markus
Hm, hard to tell. Probably caused by the resamplers filter or something, this would need some deeper investigation like reading/debugging the sox source. Maybe other resamplers don't do it. But that's just unreliable guessing.

Re: 44.1 kHz is slowly reaching its end

Posted: Wed Feb 14, 2018 10:00 pm
by forestandgarden
btw., I don't think that the rounding errors specific to S/R conversion ought to introduce a pattern if they play a role - when exactly they occur, will be semi-arbitrary and depend on when the source material crosses the rounding thresholds, and this could explain the appearance of a minimal amount of noise despite the absence of dithering.
Another thing to bear in mind when evaluating the results is that perfect noise adds up in a very peculiar fashion - not! A thought experiment might go to illustrate this: imagine 2 sine waves of the same freq&amplitude - depending on the phase difference you mix them with, the result could be anything from 0 (180°, cancelled out) to x2 (0°, adding up). Now between 2 white noise signals of same amplitude, the way peaks and valleys meet is absolutely random, you have to take an average of all possibilities, and between 0 and x2, that is: x1, your two noise signals are mixed, but the level hasn't changed.
I'm loving the research spirit in this thread, but winter and just one solar panel don't allow for much laptop use atm. Yet I'd be curious about one thing: a sine wave is a very gentle sample to test with, the smooth curve has no sudden changes and hence there's not much to interpolate to begin with, I'd be curious to know how a noise sample would behave, or how much the flanks of a rectangle would deteriorate :)
But don't get me wrong, I'm rather on the side of those who are not too worried about re-sampling losses (personally, I'm looking at the world as an event with a non-finite Fourier transformation and once it's digitally sampled, that's the damage done anyway), and regarding the original topic - I own a few not too old mobile devices, and the trend there is that any audio not 44.1khz is more likely to get spat out, with the possible exception of a few queer gsm&voice rec formats, so I'd dare predict that death isn't imminent yet for 44.1 .

Re: 44.1 kHz is slowly reaching its end

Posted: Sun Feb 18, 2018 12:01 pm
by CrocoDuck
My 2 cents on this interesting topic.
Markus wrote:Hm, hard to tell. Probably caused by the resamplers filter or something, this would need some deeper investigation like reading/debugging the sox source. Maybe other resamplers don't do it. But that's just unreliable guessing.
I think that the entity of the error might also depend on the resampler algorithm (so on sox version, in this case) but also on how the software was compiled, as I guess it could mean that floating point operation errors get propagated in a different way if the underlying assembler is different.

I would expect different resampler algorithm implementations to work in slightly different ways.
forestandgarden wrote:Now between 2 white noise signals of same amplitude, the way peaks and valleys meet is absolutely random, you have to take an average of all possibilities, and between 0 and x2, that is: x1, your two noise signals are mixed, but the level hasn't changed.
On a rigorous mathematical standpoint, the two time series will compound as observations of two random variables. So we will see a change in the statistics of the noise. For real random noise. We are stuck in pseudo-random world for computers, but provided that the seed is different it should be pretty much the same as real random.

https://en.wikipedia.org/wiki/Statistic ... processing

Adding two noise signals together will mean that we raise the total power of 3 dB. You can verify this with this simple Octave code:

Code: Select all

% How many samples of noise? This will be one second of noise at 96000 kHz
% sample rate:
timeSeriesLength = 96000;

% We generate two differet random noise sequences with 0 mean.
x = 2.0 * rand(timeSeriesLength, 1) - 1.0;
y = 2.0 * rand(timeSeriesLength, 1) - 1.0;

% Now we sum them:
z = x + y;

% And we calculate their RMS level: the dB level of their RMS value.
xLevel = 20.0 * log10(sqrt(mean(x.^2)));
yLevel = 20.0 * log10(sqrt(mean(y.^2)));
zLevel = 20.0 * log10(sqrt(mean(z.^2)));

% Display to the user, up to 2 digits.
disp(['Level of x signal = ' num2str(xLevel, 2) ' dB RMS (x is random noise)']);
disp(['Level of y signal = ' num2str(yLevel, 2) ' dB RMS (y is other random noise)']);
disp(['Level of z signal = ' num2str(zLevel, 2) ' dB RMS (z = x + y)']);
You can download Octave here to run the code: https://www.gnu.org/software/octave/. Most likely your distro repo has it packaged. A sample output of this code is as follows:

Code: Select all

Level of x signal = -4.8 dB RMS (x is random noise)
Level of y signal = -4.8 dB RMS (y is other random noise)
Level of z signal = -1.8 dB RMS (z = x + y)
Where we see that by adding two noise signals with 0 mean and same energy we end up with 3 dB more energy. The time series produced by the script are just like real sampled audio noise: floating point numbers between -1 and 1, with 0 mean.

In other words, adding noise signal from the same random process (same mean, same reference distribution) will produce a signal with the same mean but +3dB energy, which also means +3dB level.

When we think about loudness we should consider signal energy, which for noise is well represented by the RMS (Root Mean Squared) valued of the time series. The peaks don't usually give much information, as they happen only once, nor they are not meaningfully related to loudness perception.
42low wrote::mrgreen: I sometimes still even use 8Khz (in good quality that is). 8)
For particular sound purposes which need good quality but small files.
As long as your original signal is band limited to 4 kHz that will most likely work just as good.

Re: 44.1 kHz is slowly reaching its end

Posted: Sun Feb 18, 2018 4:22 pm
by CrocoDuck
42low wrote: My personal experience and oppinion is that 8khz (IMA ADPCM codec) is the lowest to keep recognizable sounds (like voices). Below that, with 4khz everything is going to sound like telephone.
Yes, human speech is usually band limited in 4 kHz - 5 kHz (although it can stretch to 7 kHz, but the most if it is within 4 kHz), so 8 kHz sampling rate works OK, as it limits the bandwidth of what you record up to 4 kHz.

Re: 44.1 kHz is slowly reaching its end

Posted: Tue Feb 20, 2018 2:34 am
by forestandgarden
Thanks CrocoDuck for the correction, I had a feeling something could be missing :)

Re: 44.1 kHz is slowly reaching its end

Posted: Sun Jun 17, 2018 11:22 pm
by protozone
I sure hope MP3 dies before 44.1 :lol:

I have been waiting for the MP3 format to just up and die ever since I discovered WavPack (WV), FLAC, OPUS, M4A/AAc, and OGG.

After getting the shpiel from hydrogenaud.io as well as doing my own listening tests, I am just done with MP3.
It's frustrating that most websites still go with MP3 even though we have all these other great formats.

I like that 48 is compatible with both video and audio.
All my recordings/composings I do at 48 kHz unless I use some old software or plugin that can't handle 44.1 kHz (there aren't many of those anymore!). Even MP3's can be 48 kHz which is where the extra bits count.

I like how at least YouTube is starting to use OPUS. I wish Amazon.com would switch ot OPUS.
256 MP3's for $1.50 or whatnot is not worth it. You can purchase nice FLAC's from other websites such as Addictech.com

I'm totally sold on OPUS for lossy and WAVpack/FLAC for lossless.

I have NEVER IN MY LIFE used WMA even though almost every portable media player I've ever had supported it.

Re: 44.1 kHz is slowly reaching its end

Posted: Mon Jun 18, 2018 9:17 am
by Lyberta
protozone wrote:I sure hope MP3 dies before 44.1 :lol:
Considering that its patents expired recently and it is now a free format, I expect a small surge in popularity of it.

Re: 44.1 kHz is slowly reaching its end

Posted: Mon Jun 18, 2018 6:52 pm
by protozone
Lyberta wrote:
protozone wrote:I sure hope MP3 dies before 44.1 :lol:
Considering that its patents expired recently and it is now a free format, I expect a small surge in popularity of it.
Wow, that's a good point. You're right. I suppose on account of that, it will be here for another 13 years :shock: :!: :| :roll: :cry: :oops:

Oh well, at least there are hardware players that can do FLAC and WAVpack. But I find it strange that at my local department stores, they pretty much stopped selling portable media players on site. And Sansa stopped producing it's wildly successful and popular "Clip"'s. Strange. Also, I like RockBox! 8) I tried it once, and it worked.

Formats are kinda fun, though. When you find something good. I even use TTA sometimes (TrueType Audio). It's brilliant. You could have 256 channels of 96 kHz lossless PCM in that one file, I think.

Re: 44.1 kHz is slowly reaching its end

Posted: Mon Jun 18, 2018 9:20 pm
by Michael Willis
42low wrote:
protozone wrote:I sure hope MP3 dies before 44.1 :lol:
Forget it. It will never die. Otherwise a lot off players could be thrown away.
Forget it, 8 track/floppy disks/VHS/zip drives/cassettes will never die. Otherwise a lot of software/music/movies could be thrown away.

Re: 44.1 kHz is slowly reaching its end

Posted: Tue Jun 19, 2018 8:20 am
by Lyberta
protozone wrote:Formats are kinda fun, though. When you find something good. I even use TTA sometimes (TrueType Audio). It's brilliant. You could have 256 channels of 96 kHz lossless PCM in that one file, I think.
WAV can have 65535 channels of 65535 bit at 4,294,967,295 Hz (4.3 gigahertz). The byterate would be 2.3 exabytes (2,300,000 terabytes) per second.

Re: 44.1 kHz is slowly reaching its end

Posted: Tue Jun 19, 2018 2:00 pm
by Michael Willis
42low wrote:Nothing will totally die.
Right, I was mostly being silly. Digital formats are very different from the physical media. Ultimately I think that produced music will be distributed in a different format, but players will continue to support mp3 as a legacy format.

stayin' alive stayin' alive

Posted: Tue Jun 19, 2018 8:09 pm
by protozone
Lyberta wrote:
protozone wrote:Formats are kinda fun, though. When you find something good. I even use TTA sometimes (TrueType Audio). It's brilliant. You could have 256 channels of 96 kHz lossless PCM in that one file, I think.
WAV can have 65535 channels of 65535 bit at 4,294,967,295 Hz (4.3 gigahertz). The byterate would be 2.3 exabytes (2,300,000 terabytes) per second.
Wow, that's amazing.
I didn't realise it had progressed to that level of sophistication.
It's been a while since I've read about that type of thing. Is that broadcast WAV format, or just regular WAV?

Re: stayin' alive stayin' alive

Posted: Wed Jun 20, 2018 7:01 am
by Lyberta
protozone wrote:Wow, that's amazing.
I didn't realise it had progressed to that level of sophistication.
It's been a while since I've read about that type of thing. Is that broadcast WAV format, or just regular WAV?
Actually, it was since the beginning. Number of channels and bit depth are stored as 16 bit integers and sample rate as 32 bit one. The main limitation is that the file size is stored as 32 bit integer which limits it at 4 GiB. WAV64 fixes this.