[bigshot]

I would actually like to know benefits of 24bit recordings other than dynamic range? On consumer end, not post production applications.

Higher bit rates mean lower noise floors. 16/44.1 has a low enough noise floor that if you boosted the volume enough to hear the quietest possible sounds that are recorded in the file, the peaks would cross the line into causing hearing damage.

Higher sampling rates mean more high frequency content. 16/44.1 contains all of the frequencies that ideal human ears can hear. The frequencies in 96 and 192 are all beyond the range of human hearing.

Super audible frequencies are sometimes useful in the studio for complex sound filters. Exceptionally low noise floors can be useful in the studio for pulling up quiet sound elements in the mix. Neither of things have any usefulness for playing back commercially recorded music in the home.

There are two links in my sig file under the title is HD audio sounds the same as CD that have good information on this.

 

[FYFL]

Got it. Thanks.

 

[gregorio]

Of course I would like to think it was better than a standard bit-rate, but who knows?

Science knows, in addition to those of us who have done controlled tests.

I do remember however, 24bit HDCDs back in the day. Or whatever name they have used at the time. Hype?

Anything beyond 16bit for consumer playback is hype.

[1] Super audible frequencies are sometimes useful in the studio for complex sound filters. [2] Exceptionally low noise floors can be useful in the studio for pulling up quiet sound elements in the mix. [3] Neither of things have any usefulness for playing back commercially recorded music in the home.

1. Not that I'm aware of. A couple of things I am aware of and have used though: A. Pitchshifting, downwards obviously. B. With some modelling plugins; guitar amps/cabs, vintage synths and some vintage compressors and limiters for example, freqs well beyond human hearing cause IMD in the hearing band and is part of the "character" of many vintage models. However, for the last 15 years or so, these types of plugins locally upsample, so we don't have to actually record the file at say 96kHz sample rate.

2. Not really. The advantage of 24bit is when recording, it allows a great deal more headroom. The limiting factor is virtually always the combined noise floor of the mics, mic pre-amps and recording venue, rather than a 16bit noise floor.

3. Notwithstanding the previous 2 points, this statement is entirely correct though!

G

 

[bigshot]

I've used super audible frequencies for noise reduction. It can help an impulse noise reduction plugin discern what is noise and what is signal in record transcription. I was working with this a while back, and I remember it was a pain in the butt. As you say, technology moves on. Similar noise filters might not need higher sampling rates any more.

As for the sources of noise floors in a mix, absolutely. In the studios I work in they put a big emphasis on the quality of the mikes and pres so they can squeeze everything they can out of the tracks. But it usually ends up that the limited factor is the air conditioning!

 

[FYFL]

He touches on high resolution and high frequencies in this video. Claims that it all matters to a certain degree.

 

[bigshot]

No one is denying that higher bit and sampling rates aren't used in production. They are just unnecessary for playing back commercially recorded music in the home.

 

[FYFL]

No one is denying that higher bit and sampling rates aren't used in production. They are just unnecessary for playing back commercially recorded music in the home.

That's not what he said. He was talking about using reference recordings to compare your own work, style with. And mentioned lower fidelity files as less than optimal option due to slight tonality differences. Even tho extreme up-sampling help, audible differences were still there. I can confirm this based on his audio samples and my "neutral" tuned headphone moinitoring/listening rig.

 

[bigshot]

There are no audible tonal differences. If you can’t hear it in a controlled listening test, there can’t be a tonal difference. If a difference can be heard, it’s probably due to expectation bias.

Calibrate your system using a high data rate file. Then bounce that down to 16/44.1 and do a controlled listening test comparing them. See if you can hear the tonal difference.

 

[FYFL]

There are no audible tonal differences. If you can’t hear it in a controlled listening test, there can’t be a tonal difference. If a difference can be heard, it’s probably due to expectation bias.

Low res compressed samples shown noticeable roll off in high frequency range.

 

[bigshot]

When we’re we talking about low resolution compressed files? We’re talking about 16/44.1. If we are talking about an MP3 with a data rate below 128, then we have no argument.

 

[FYFL]

When we’re we talking about low resolution compressed files? We’re talking about 16/44.1. If we are talking about an MP3 with a data rate below 128, then we have no argument.

He was using the latter.

 

[bigshot]

Why? Can’t he properly encode a lossy file so it has a full range of frequencies?

 

[FYFL]

He was emphasizing caution to use lossy file as your reference in determining your own “style” or approach that you identify with while mastering your own work.
He didn't try to encode or manipulate file as a work project per se but a typical playback solution via DAC or similar. Something that you would use to play back a digital file that you like to use as a reference “style”
I think he upsample that file in steps to 352 or 384 (don't remember). A standard option with most DACs these days.

His point as well as my point is, that lossy files exhibit audible degradation in “sound quality”. In tonality as well as in resolution. Something that unnamed individual from other forum argued are not audible in “blind test”.... for most people.

Edit. Up-sampling was used as an example that someone could use to compensate for lossy file on their playback “rig”. It didn't do much.

 

[bigshot]

Again, I am not talking here about mastering engineers. They should be using masters in their work. I'm talking about you sitting on your couch listening to music from your collection. What recording engineers do with a recording is completely different than what consumers do. For the purposes of listening to music in the home, a high data rate lossy file is fine.

Some comments...

Above the threshold of transparency a lossy file suffers no degradation in sound quality or tonality. If it did, it wouldn't be transparent!

At higher data rates lossy files don't roll off high frequencies. All of that happens at lower data rates, not higher ones.

Upsampling does absolutely nothing to change the tonality of digital audio. All it does is add empty space to the file.

The thing that doesn't get picked up in blind tests is bias. And that's the whole point of a blind test!

Here in Sound Science we can ask for proof established by means of controlled testing. Blather about hoodoo magical unmeasurable untestable qualities works better in other forums.

 

[danadam]

Huh, I thought that the term "high resolution" was pretty established to mean something > 16/44. I don't think I've ever before heard it used to mean high bitrate of a lossy file.