Hacker News Reader

Comment by retrogradeorbit | Hacker News Reader

Try repeating the experiment without the dither.

i found this http://www.audiocheck.net/audiotests_dithering.php

and honestly, i can easily tell the samples with dither because they added a LOT of white noise. the original and 8bit non-dither to me practically the same.

Listen for "graininess" in the 8 bit undithered sound as the voice gets quieter. You should be able to make that out. That's quantization distortion. The dithering removes that, but adds the noise.

This is the clearest demonstration of audio dither I've come across: http://youtu.be/h59LwyJbfzs

oh my god. that is even worse!

so explain why 'dithering' was 'masking' the results?! it makes it extremely obvious that it is the lower resolution sample!

That would allow the subject to easily tell the difference between the two samples by looking to see which had all 0's in the 4 lowest bits.

This makes no sense to me. Please explain. The difference between 24 and 16 is 8, not 4 for one.

Are you saying the "16-bit" sample files actually had noise added to the least significant 4 (8?) bits? This is not dithering. Dithering is adding noise BEFORE truncation. A dithered 16 bit rendering of 24 bit audio will only be 16 bit. An undithered 16 bit rendering of 24 bit audio will also only be 16 bit.

I think the GP is basically saying, how would you propose to run an internet survey asking anonymous audiophiles to blindly A/B (no peeking at the bits) full and truncated files? The dithering is probably the least questionable part of the methodology.

Dithering is known to increase the perceived dynamic range of music. As an example try listening to a song in foobar in 8bit output mode then 8bit dithered.

Are you also a fan of non-oversampling DACs?

I'm a fan of good sounding DACs.

By dithering you increase the perceived dynamic range. You change it from a listening test of the quality of the sound, to the character of the noise floor. Maybe this is why some "engineers" could tell. If you are trained in what to listen for you would know in these listening tests to completely ignore the majority of the sound and frequency range, and instead listen to the sound of the high frequency hiss on the noise floor (for instance in very quiet or completely silent sections. Just isolate a very quiet part and turn it up loud). This would be the tell tale sign of the dithering, thus 16 bit.

I do not agree that you increase the perceived dynamic range. You use it to change the noise to something more favorable. Instead of harmonic distortion, you get a much more uniform noise that doesn't distort the original signal. I really cannot see how that can be seen as anything other the better audio quality.

"Random numbers such as these translate to random noise (hiss) when converted to analog. The amplitude of this noise is around 1 LSB, which for 16 bit lies at about 96 dB below full scale. By using dither, ambience and decay in a musical recording can be heard down to about -115 dB, even with a 16-bit wordlength. Thus, although the quantization steps of a 16-bit word can only theoretically encode 96 dB of range, with dither, there is an audible dynamic range of up to 115 dB! The maximum signal-to-noise ratio of a dithered 16-bit recording is about 96 dB. But the dynamic range is far greater, as much as 115 dB, because we can hear music below the noise. Usually, manufacturer's spec sheets don't reflect these important specifications, often mixing up dynamic range and signal-to-noise ratio. Signal-to-noise ratio (of a linear PCM system) is the RMS level of the noise with no signal applied expressed in dB below maximum level (without getting into fancy details such as noise modulation). It should be, ideally, the level of the dither noise. Dynamic range is a subjective judgment more than a measurement--you can compare the dynamic range of two systems empirically with identical listening tests. Apply a 1 kHz tone, and see low you can make it before it is undetectable. You can actually measure the dynamic range of an A/D converter without an FFT analyzer. All you need is an accurate test tone generator and your ears, and a low-noise headphone amplifier with sufficient gain. Listen to the analog output and see when it disappears (use a real good 16 bit D/A for this test). Another important test is to attenuate music in your workstation (about 40 dB) and listen to the output of the system with headphones. Listen for ambience and reverberation; a good system will still reveal ambience, even at that low level. Also listen to the character of the noise--it's a very educating experience."

-- Bob Katz http://www.digido.com/articles-and-demos12/13-bob-katz/16-di...

I am aware of that statement. I still stand by my statement, but let my clarify a bit, I meant that its not just percieved increase - it is really there. Looking at a sine wave through a frequency analyzer, without dithering, it's hard to tell what the dynamic range is. You can turn down the amplitude until you can no longer see it on the analyzer. Apply dithering at that level and you can clearly see the signal again. You do not need to use your ears.

Excellent point. This makes the test somewhat flawed. Are we testing "performance" vs file size? Or overall performance for music with few quiet/silent sections?

Reply on news.ycombinator.com