Busting High-Resolution Audio with simple logic



The greatest tool is between your ears...


Oct. 19, 2018

I often see that logical thinking is missing from the audio press and - sometimes - even from academic papers. But logic (which basically means the capability of recognizing flawed logic in arguments) is important, because it makes life easier...

So I wrote down one of my basic problems with the hype about High-Resolution Audio... This is not a technical or deep psychoacoustical analysis, nor a musical test... It's a much simpler approach, some questions and a bit logic...

I will deal mainly with the sampling frequency and not the bit depth (word length), because this is the main issue.

First, I will compare the 44.1 kHz sampling frequency to 32 kHz and 22.05 kHz sampling frequencies and draw some conclusions. Then I will 'extrapolate' the results to higher sampling frequencies. I use a QA style, because I it makes everything more easier to understand. I assume normal hearing which means 'hearing ability of people between the age of 20-30 with good hearing up to 18 kHz'.

Is there audible difference between 44.1 kHz and 22.05 kHz sample rates with music?

Yes, of course, even with normal hearing and with with most types of music or instruments there is a big audible difference...

Is it possible to generate a synthetic test signal that reveals the differences between 44.1 kHz and 22.05 kHz?

(In other words, is there a test method that is repeatable and reproducible?)

Yes, there are so many. Band limited pulse, sawtooth, square wave etc... They sound quite different at 22.05 kHz and at 44.1 kHz sampling rate. (Synthetic test signals can be easily generated with audio editors)

Is there audible difference between 44.1 kHz and 32 kHz sample rates with music?

Yes, but the differences are subtle with normal hearing and mainly audible with cymbals, hi-hats, in the attack of plucked instruments and hardly recognizable with speech or flute....

Is it possible to generate a synthetic test signal that reveals the differences between 44.1 kHz and 32 kHz?

Yes, but the differences are very small...

Is there audible difference between 44.1 kHz and 48 kHz sample rates with music?

Some people will say yes, but I think the next question is more exciting:)

Is it possible to generate a synthetic test signal that reveals the differences between 44.1 kHz and 48 kHz?

I don't know that such signal with at least a positive documented test exists, so my answer is no...

Now, the next question would be about the difference between 44.1 kHz and 96 kHz, but the answer would be the same as in the 44.1 kHz vs. 48 kHz case. Up to the point that there is no synthetic test signal for revealing the difference between 44.1 kHz and 48 kHz, there is no too much sense in 96 kHz...

That's all, high resolution audio busted with some simple questions...

Notes: when people hear the difference between 44.1 kHz and 48 kHz audio, then they hear the result of an inaccurate sample rate conversion... Or they don't use ABX testing. In addition to that it's a much better approach to focus on the audible differences and not on the sound quality. If there is some kind of audible difference between to setup, then the quality will differ too, but the quality is always subjective. If there is no audible difference, then the quality has to be the same. So we can bypass the questions about sound quality focusing on the audible differences.

Csaba Horvath

Facebook    Google

Complete list of articles and software tools