Testing audiophile claims and myths - Page 96

The CD standard = 16/44.1 = up to 22.05kHz

 

24/96 = up to 48kHz

 

16/32 = up to 16kHz.

 

In Foobar or Audacity you can resample 16/44.1 to to 16/32, then you'll hear the difference which you keep calling "hearing voices", "I can see better than a microscope" and "wasting everyones time".

 

A lot of users already know that speakers and room acoustics are big fish, they want to chase the small fish like lossless, ASIO and expensive sound-cards anyway. 

Edited by kiteki - 6/10/12 at 1:59am

Downsampling introduces all sorts of other issues. I didn't tell you to do that. I said to use an equalizer.

It's better to use a low-pass filter or digital equalizer. Otherwise you can get all sorts of audible artifacts.
Even then, I'm not surprised you can hear a difference. But that doesn't really mean it is very important. For most people the difference is going to be very subtle, and shouldn't really add that much to the overall enjoyment of the music.
It is wise to include 16kHz+ information, but it is not necessary for good fidelity.

The CD standard is supposed to be overkill by the way. They did not want any compromises, that's why they made absolutely sure to include the entire audible band, and also introduced a rather high 16 bit wordlength, even though 96dB's of dynamic range is higher than you would most likely ever need.

 

Even then there are potential issues, a typical graphic or parametric equalizer will not have a "brick wall" response when used to filter out everything above 16 kHz, and the more audible 10-12 kHz range will likely be attenuated as well. On the other hand, with a very steep roll-off, the resulting ringing at 16 kHz may be audible to some. Of course, it is also important to avoid clipping and to match levels accurately. The comparison should be done with files at identical sample rate to exclude the possibility of different DAC filtering (or even Windows software sample rate conversion if the DAC does not support 32 kHz) affecting the results.

It is only possible to know for sure if the test was done correctly if the files used are shared.

 

Did you use (in fb2k) the SoX resampler with Target samplerate: 32000, Quality: Best, Passband: 99% and minimum phase response followed by another resampler with the same settings but the target samplerate set to whatever your DAC supports, e.g. 44.1 kHz, to avoid resampling by the OS?

Edited by xnor - 6/10/12 at 4:12am

That's getting way too complicated. My original point was very simple. And he keeps misstating it to set up strawmen. I'll repeat it again.

Of the ten or so octaves that make up the full range of human hearing, the top one (10kHz to 20kHz) is the least important to the enjoyment of music. No instrument has fundamental frequencies up that high except for cymbals, and the harmonics are at a very low volume compared to the other octaves.

My suggestion was to familiarize yourself with what these frequencies actually sound like by playing with an equalizer and isolating the sound of music that exists above 15kHz. You can hear it, because you have human ears, but if you do this test, you'll find out it really isn't important because there's very little up there in music.

Correct balance in the middle of the audible range is much more important than messing around trying to optimize the stuff at the very edge of human perception.
Edited by bigshot - 6/10/12 at 10:16am

Sorry to ask a totally irrelevant question.

 

But, you guys are talking about DSPs in foobar, and I really don't want to start a new thread...so, are there any decent crossfeed DSPs in foobar? I can't find any in the supported components. A recommendation would be greatly appreciated.

 

It's more than just crossfeed but TB Isone is my favorite.  There's also bs2b and xnor's crossfeed.

Thanks a lot, I saw the bs2b one, but it did not seem to be well regarded. I also saw the xnor one, but didn't realize that it was not a component, so I was searching all over the foobar site to no avail.

 

Thanks!

I can personally recommend xnor's one. It is really quite good, but I stopped using it as I prefer RPGWizard's DSP chain. While that isn't crossfeed, I still advice you to check it out.

Thanks, guys!

This thread should be pinned in sound science

Interesting stuff here. I'm feeling pretty okay having not spent large amounts of money on gear before finding this thread. I bought some custom iem cables and I now expect to not hear any difference, but I'm open to being wrong. It seems that we hear very differently (not necessarily better) with our eyes open.

I have used them just a little bit and I haven't noticed a difference so far. I'll get around to some serious A/Bing soon.