Hi-Rez - do you find it obviously better?

Yes you are correct although quantisation noise still has frequency components at (2N+1)fs/2 (for N >= 0) which must be removed although they naturally attenuate as you go higher in frequency.

In fact, if you look very carefully, you will spot that I originally started talking about frequency aliasing and then changed to simplify things - but not entirely successfully. There is still one place where the term ‘signal images’ is used :frowning:

I tried to use the looser term ‘quantisation artifacts’ which Intended to include both aliasing and quantisation noise (which still crept in by mistake). ‘Digitisation artifacts’ would have been an even better terminology.

1 Like

Just realised what you were referring to. Mistake crept in when re-working the sentance. Now fixed.

1 Like

Interestingly, the reasoning I that I expounded above means that, for some people, there could be an advantage to upsampling redbook 44.1kS/s sources to 88.2kS/s on systems where the DAC’s analogue low pass filters are appropriately taylored to the sample rate and thus perform better on the 88.2kS/s version.

Something that I had not considered before :slight_smile:

Of course, this all assumes that the ‘holy grail’ of audio is a perfectly transparent system that does not colour the audio in any way (perfectly flat frequency response, absolutely linear phase) all the way from recording (through digitisation, mastering, analogue conversion, amplication and on to Speakers/headpones/iems). We all know that that is not true of many people (think of the ‘valve amps’ rule camp and even some solid state amps that are non-linear and introduce second order (and other) harmonics - and different peoples preference for different speakers.

If it were true, then 'everyone would be in the ‘measurements rule’ camp.

1 Like

That is the reason DACs use oversampling, to push the Nyquist limit a few octaves above 20kHz so they can use a much more relaxed analog low pass filter. At 4x oversampling for example, you have from 20kHz to 88.2kHz to drop from 0dB to -120dB or less. This is done during playback so no high resolution needed.

3 Likes

In the absence of noise shaping, quantization noise is distributed uniformly throughout the spectrum, from zero to fs/2.

Which, assuming the in-DAC upsampling is done competently, makes it even more unlikely that high sample rate sources give an obviously better result.

2 Likes

High Resolution is not obviously better. The recording and mastering is what matters. That being said, I have some recordings where the Hi-Rez version of the album sounds better to me than the CD quality version. I have the gear to play Hi-Rez files, so when I pay for downloads, I download the high resolution FLAC files if they are available. It can’t hurt.

1 Like

I don’t think I could pass a blind listening test between a standard 44.1 kHz/16-bit and a hi-res recording.

I can say though that I do feel all warm and fuzzy when I look over and see my DAC displaying 192 kHz/24-bit resolution. :blush:

7 Likes

It actually can.

24/192 Music Downloads are Very Silly Indeed (xiph.org)

192kHz considered harmful

192kHz digital music files offer no benefits. They’re not quite neutral either; practical fidelity is slightly worse. The ultrasonics are a liability during playback.

3 Likes

I tried an experiment about a decade ago. I had the same music in both CD and SACD formats (Vaughan Williams’ Pastoral Symphony and Norfolk Rhapsodies by Hickox and the LSO). I plugged in my best headphones into my Marantz SA8260, shuffled the discs with my eyes closed, inserted them blind by touch, and played the same track. The goal was to see whether I was able to guess which was the CD mix and which the SACD reproducibly with my ersatz double-blind test. I was able to, just about, mostly by listening to the timbre of violins. Other discs with similar setup (Glenn Gould’s second Goldberg Variations), I could not.

So the difference exists, but is vanishingly subtle, and negligible compared to the quality of the recording and mastering. I suspect HD mixes are less likely to suffer from dynamic range compression than the (then) more mainstream CD format.

1 Like

Did you ensure that both discs were being played at exactly the same volume level?

Very difficult to do.

2 Likes

Again, I’m no sound engineer, but as far as I know, all processing is done in high-res, including dynamic compression. Down-sampling to red book is just the last step, which should not include any “creative” DSP.

And exactly means exactly. From @Marian’s wonderful link, my emphasis:

The human ear can consciously discriminate amplitude differences of about 1dB, and experiments show subconscious awareness of amplitude differences under .2dB. Humans almost universally consider louder audio to sound better, and .2dB is enough to establish this preference. Any comparison that fails to carefully amplitude-match the choices will see the louder choice preferred, even if the amplitude difference is too small to consciously notice. Stereo salesmen have known this trick for a long time.

The professional testing standard is to match sources to within .1dB or better. This often requires use of an oscilloscope or signal analyzer. Guessing by turning the knobs until two sources sound about the same is not good enough.

3 Likes

Nothing can be concluded from this test. Audio mastering and playback level differences between SACD DSD and CD PCM are possibilities that render moot any blind test.

Even a hybrid SACD with both DSD and PCM layers would not escape the flaws in the aforementioned test. For example, some Telarc hybrid SACDs reportedly did not utilize a DSD digital decimation to PCM process for the CD layer. Instead, DSD was converted directly to analog, then reconverted to PCM via a different ADC for the CD layer.

AJ

1 Like

I left Vinyl in 2020, I don’t miss it. I have very expensive gear and was able to start with an Auralic Vega 2.0 and Leo clock, and some expensive cables. Later adding a Holo Audio Serene KTE give me another flavor, if I need it. IMHO I can see a bright and better future. Changing gear will be expensive but rewarding.

That will be the expensive cables polishing the sound up.

Joke!

3 Likes

Fascinating discussion everyone - thanks for contributing.

The point about matching volume levels is very pertinent when making comparisons but, to me, an even more confounding difficulty is the reliability of audio memory. Just how reliable is the memory at remembering the quaity standard of something listened to even a few seconds ago?

If sound A is obviously distorted compared to sound B then no problem, but I’m very sceptical about the ability of anyone to reliably distinuish subtle differences. Worse than that, even if a subtle difference is detected, which is “Better”, A or B? Of course it really does not matter - if you decide you like B better than A, end of story. There is no “best” but try persuading that to some audiophiles!!!

2 Likes

Q: What is the best audio resolution?

ChatGPT
The term “audio resolution” typically refers to the bit depth and sample rate of digital audio. Both of these factors contribute to the overall audio quality and fidelity of a recording. The “best” audio resolution depends on the specific requirements of the recording and playback scenario, as well as personal preferences.

  1. Bit Depth:
  • Common bit depths are 16-bit, 24-bit, and 32-bit. The higher the bit depth, the greater the dynamic range and potential signal-to-noise ratio.
  • For professional music production and high-quality audio, 24-bit is commonly used. It provides a wide dynamic range, allowing for more nuanced recording and editing without introducing noticeable quantization noise.
  1. Sample Rate:
  • The sample rate represents the number of samples of audio carried per second, measured in Hertz (Hz). Common sample rates include 44.1 kHz, 48 kHz, 96 kHz, and 192 kHz.
  • Higher sample rates can capture higher frequencies, but human hearing typically doesn’t extend much beyond 20 kHz. However, using higher sample rates can be beneficial during recording and editing to avoid issues like aliasing.
  • For most music production and playback scenarios, a sample rate of 44.1 kHz or 48 kHz is standard and sufficient. Higher sample rates are often used in professional studios for specific applications.

In general, the “best” audio resolution depends on factors like the type of audio content, the intended use (e.g., music production, gaming, casual listening), and the capabilities of the playback equipment. For everyday listening, most people won’t notice a significant difference between well-produced audio at 16-bit/44.1 kHz and higher resolutions.

It’s essential to consider the trade-off between audio quality and file size, especially for storage and streaming purposes. Extremely high resolutions may offer marginal improvements in certain scenarios but can result in significantly larger file sizes.

1 Like

Well your ChatGPT post only that while AI may be artificial, it is far from intelligent.

Good discussion all (and civil too :sunglasses:),

My 2¢ worth:

Like many of you, I have never noticed any consistent co-relation between higher sample rates and sonic performance.

Occasionally, I have had a suspicion that higher bit-rates might sound better, but they are too often a result of re-mastering which muddies the waters somewhat. :thinking:

2 Likes