Background information
Audio compression: facts, myths and a blind listening test
by David Lee
Audio compression: How important is the equipment?
26.1.2018
Translation: machine translated
It's clear that high-end sounds better. But do you also hear differences in audio compression better? A self-experiment.
My recent article on audio compression also included a blind test. The test subjects were only just able to identify the worst MP3 with a variable bit rate of 65 kbit/s without any problems. They were no longer able to distinguish a VBR of 130 kbit/s from the uncompressed signal.
Normal hardware was used for the test: a PC audio output and average headphones. This was intended to simulate an everyday setting. However, the question naturally arises as to whether you can hear more with better hardware. Many readers were of the opinion that this was the case.
So I decided to repeat the test with high-end hardware. However, only on myself, I won't re-submit all the test subjects. If a test subject count of 1 is too low for you: You can still download the test files yourself and listen to them on your (high-quality) hardware.
High end on my desk
Chief Editor Aurel Stevens shows me what real headphones are and drags the behemoth Stax SR-303 to my desk - along with the Stax SRM-323II amplifier that is needed to make the headphones sound good. I listen to some music generated directly from the raw data. The electrostatic headphones sound fantastic. However, because the files are played by a special audio player, I can't hear the test files or any tracks I know at this moment.
Next, I connect the audio output of the notebook to the Stax again. On the one hand for comparison and on the other hand so that I can hear the test files at all. Frustration: I can still only recognise the worst file. I realise that the others don't sound exactly the same, but I can't tell which is better and which is worse. The assumption is that the notebook's audio output is simply not delivering a good signal.
The audio output of the smartphone does not produce better results.
Off with this notebook output!
Now I copy the test files to a USB stick so that I can listen to them on my Onkyo CR-N755 stereo system. It should be able to deliver a high-quality signal. First I connect my Sennheiser headphones. I could only connect the Stax indirectly by connecting the Stax amplifier to the other amplifier.
The stereo system cannot play the worst files, which are always recognised anyway. This leaves the four higher-quality variants. In the case of classical music, one track seems somehow flat and I can hear other background noises in the quiet parts. This must be the second worst file (130 kbit/s VBR).
I can't pinpoint anything definite with jazz, but I have a vague feeling that some versions simply sound better than others. I make a note of an exact order on this shaky basis.
When it comes to rock, a file sounds unpleasant to me, so I make a note of that. I also make a note of the supposedly third-worst version out of a vague feeling. However, I don't have a concrete point of reference.
High end at last
A few days later: I have forgotten what I had previously noted down and am now repeating the test with the Stax.
I can now clearly hear how the background noise differs in one file. However, I can't clearly distinguish the rest.
Jazz: In the quieter moments, strange "flickering" noises can be heard in one file. This must be the second worst. I also hear this in another file, but less clearly.
Rock: I identify one file as "flat" sounding and generally poor. The others are barely distinguishable.
Conclusion
Finally, I allow myself a look at the solution. My tips were right. With both my Sennheiser and the Stax, I was able to correctly identify the second-worst compression. Surprisingly, I got the jazz sequence completely right with the average headphones. Without being able to say, however, what I was attributing this to. Just from the feeling.
With the electrostatic super duper headphones, I could no longer identify files correctly. But where I did hear a difference, I heard it very clearly and was absolutely certain that it wasn't my imagination or a vague feeling.
I come to the following conclusion: The audio output of my devices (smartphone, notebook) is simply not good enough to recognise medium compression. If I bypass this source, I can hear the difference. I don't even need extremely good headphones to do this, just average ones will do.
How to bypass the audio output
You can also achieve a high-quality audio signal on your notebook. To bypass the normal output, you need an audio interface. This is a sound card with its own inputs and outputs, which usually comes as an external device
.
Most audio interfaces are designed for musicians and therefore come with special drivers with short latencies. A classic audio interface for musicians:
Such short latencies are not necessary for normal music listening, the above product would be overqualified. Something like this will also do:
This small headphone amplifier is also a DAC. The abbreviation stands for digital-to-analogue converter, or D/A converter. Wikipedia calls it DAU, but for me that's more of an abbreviation for "stupidest user". In any case, the device gets the digital data via USB and converts it into analogue audio signals itself. And in a better way than the standard audio processing of the notebook does.
Of course, you can also simply use a high-quality audio player that already has everything you need. This one, for example, is both a media player and a USB DAC:
The FiiO Taishan is also suitable for TVs, games consoles and other devices with an optical or coaxial output.
David Lee
Senior Editor
David.Lee@digitecgalaxus.chMy interest in IT and writing landed me in tech journalism early on (2000). I want to know how we can use technology without being used. Outside of the office, I’m a keen musician who makes up for lacking talent with excessive enthusiasm.
42 comments
later