Main Area and Open Discussion > General Software Discussion

Why 24-bit/192kHz music files make no sense - and may be bad for you!

<< < (12/16) > >>

IainB:
I don't know if anything will come of this, but could be relevant and it is interesting:
Is There Any Merit To Neil Young's Plan To Improve The Quality Of Digital Music?

IainB:
This is interesting:
Dolby's TrueHD 96K Upsampling To Improve Sound On Blu-Rays
SpoilerStowie101 writes in with a story about your Blu-ray audio getting better.
"The audio on most Blu-ray discs is sampled at 48kHz. Even the original movie tracks are usually only recorded at 48kHz, so once a movie migrates to disc, there isn't much that can be done. Dolby's new system upsamples that audio signal to 96kHz at the master stage prior to the Dolby TrueHD encoding, so you get lossless audio with fewer digital artifacts. The 'fewer digital artifacts' part comes from a feature of Dolby's upsampling process called de-apodizing, which corrects a prevalent digital artifact known as pre-ringing. Pre-ringing is often introduced in the capture and creation process and adds a digital harshness to the audio. The apodizing filter masks the effect of pre-ringing by placing it behind the source tone — the listener can't hear the pre-ringing because it's behind the more prevalent original signal."

Renegade:
NECROTHREAD! ARISE! <chanting to old ones that have no name... />



Seems Pono is just about here:

http://www.rollingstone.com/music/news/neil-young-plans-pono-launch-for-2014-20130904

Neil Young's music service Pono, which will provide listeners with downloads of high-resolution songs made to sound like their initial recordings, is almost ready to roll. It's set to launch in early 2014, according to a Facebook post written by Young.

"The simplest way to describe what we've accomplished is that we've liberated the music of the artist from the digital file and restored it to its original artistic quality – as it was in the studio," wrote Young. "So it has primal power."

--- End quote ---

I'm going to come out on Neil's side here.

I'm also going to say f*&( science. But I say that for scientific reasons.

The original article:

http://people.xiph.org/~xiphmont/demo/neil-young.html

The ear hears via hair cells that sit on the resonant basilar membrane in the cochlea. Each hair cell is effectively tuned to a narrow frequency band determined by its position on the membrane. Sensitivity peaks in the middle of the band and falls off to either side in a lopsided cone shape overlapping the bands of other nearby hair cells. A sound is inaudible if there are no hair cells tuned to hear it.
--- End quote ---

And THAT, ladies and gentlemen, is why they fail.

Neil is right.

Monty is wrong.

But I am getting ahead of myself there. Consider that the tl;dr.

There is no good science done in this area. There is only partial science.

The article does an excellent job of addressing a lot of issues. It explains quite a bit. But it never asks the right questions. The discussion is far too limited, and so is the science.

The entire argument that "24/192 makes no sense" centers around sampling theory and the human ear. This is not sufficient. More recent graphs of human hearing do not really get much more accurate than the original studies from the 1920's (or whenever - I'm not going to bother looking it up as it's a side issue). Sure, there are some differences in A-weighted vs. other weightings, but even when they become more accurate, weightings are regularly misused by professionals who really should know better. But... it all goes back to reliance on weightings as accurate representations of... wait for it... what levels of sound humans can identify as being heard.

I need to repeat that.

"what levels of sound humans can identify as being heard"

That's important. It's what different scales are based on and it is the sum total of the industry approach to "sound".

When you buy a microphone or good speakers, you'll see graphs that show how they perform. Blah blah blah. All the same basic science.

Now... imagine 100 people out in the jungle spread out say 30 m from each other in a line walking forward. As they walk forward, our hero, Joe, is in the center when the hair on his arms rises up and he gets a sick feeling. The only thing he "hears" is the crunch of twigs beneath his feet and the odd calling out from people near him for the little boy. Another hero, Fred, is on the far right of the line. He has no such sensation.

We backtrack just a moment in time and over to the far left. Not so much a hero, but more of a victim, Harold looks ahead and sees a large tiger. It roars. This is where Joe's hair stands up and he gets the sickly feeling.

The tiger then proceeds to eat Harold. NOM NOM NOM NOM~! Slurp! Burp! Harold was very tasty and the tiger is very happy at such a wonderful meal.

What happened there?

The tiger's roar contained a lot of energy. Harold got the full exposure to that energy. Fred received zero exposure as he was too far away. But they're not the interesting parts of the story.

Joe was too far away to receive any audible energy from the tiger's roar. However, he did receive some of that energy, which spooked him and caused the hair on the back of his neck to rise up, and for Joe to get that sickly feeling that he couldn't quite identify. That portion of the tiger's roar that Joe "felt" was around 3 Hz as higher frequencies don't travel as far as lower frequencies. If it were 40hz, we'd simply say that a DCer fell on top of Joe. :P ;D

(I could have summed that up really quickly, but the little story where Harold gets eaten was much more fun! ;D )

The current science as we have in the article completely dismisses these cases by limiting "sound" to "the audible spectrum". The fact of the matter is that sound has a greater effect than just that limited definition, and nobody is asking these questions.

Again, f*&( "science" for being so stupid. You don't get to try to talk about a scientific topic then limit the discussion to what you like and exclude all the inconvenient evidence that flies in your face. That's not science. That's hyper-focused bullshit. Now, it may be really good shiny awesome math and cool charts hyper-focused bullshit, but it's still bullshit.

Neil is onto something. He knows that there is something that we are missing. However, nobody is going to be able to articulate that in a scientific context until there is research into these areas.

I am not putting forward that I know all the answers. I am putting forward that the wrong questions or not enough questions are being asked, and that the discussion has been artificially limited in a horribly irresponsible way if you actually give a crap about evidence based science. Just because you can't explain or don't understand some evidence doesn't mean that you can exclude it, which is exactly what modern sound science does, as clearly illustrated in the article.

(Took me long enough to get around to reading that.)

Joe Hone:
I've enjoyed reading through this older thread and some of the opinions being offered. I work in the audio industry and for my purposes, 16 bit works fine. We have done all of the blind testing time allows and find that 24/192 isn't "better," but it may be "different." And as for discerning listeners claiming to hear a difference where science says there is none, we have a mic preamp in the studio that has an "air" switch, which activates frequency filters above what any human ear can hear. But those frequencies interact with frequencies that are audible, and it is interesting to hear how that switch opens up the top end of say a voice or guitar being tracked - even if "air" isn't right for that particular track. So I won't be arguing with Neil Young. But I'm also still working in 16 bit. (We cheat, because we track to tape which preserves all of the warmth you want to hear but then go into the DAW for editing and mixing.)

I work with first generation sound all day long, but it gets compressed, equalized, mixed, mastered and replicated before it arrives in your playback system. To me, it already has lost much of what makes it musical before you ever hear it, but mixing requires compression, equalization and reverb to give the instruments and voice space to be heard through your speakers. Fortunately, most of us get caught up in the composition/performance rather than worry that much about the sound, which is why mp3 is tolerable. And mp3 is vastly superior to my introduction to music - the AM transistor radio I got for my birthday in 1968.

JavaJones:
IANAAE (I Am Not An Audio Engineer), so take the following with a grain of salt. ;)

Frequencies that home speakers/headphones cannot reproduce are irrelevant without massive improvements in sound reproduction capability in the home. Given the fact that this has not happened in the last 50+ years (incremental improvements only), I don't expect it to happen in the next 50. Even if it did, the improvements would be so minor for most listeners that the cost and hassle of replacing all their equipment would not be desirable for most. By the time such a conversion was complete, it would be time to re-buy the White Album anyway, and it could be mastered in 512/4096 for all I care. In other words by the time you'll be able to actually hear the difference (due to limitations in your equipment, not necessarily your ears!), you would have likely bought the thing again anyways, i.e. there will be something better than "Pono". Buying Pono stuff now doesn't help you though.

For now and the foreseeable future, much as sub-audible frequencies may be *perceivable* and have an effect *in person*, they are not relevant for recorded music. Nor, in fact, are they relevant for *any* amplified music since there are multiple limits in place there, not least of which are the speakers, but also any live processing being done (reverb, compression, etc.). Even if your entire amplification system is analog, the speakers are still a limiting factor. As are mics that recorded it in the first place, for that matter! There is *so much compromise* throughout any music production process, whether analog or digital, that I think it's a bit silly to cling so tightly to the "purity" of reproducing the finished results with 100% accuracy. Hell, the placement of speakers in a person's room, or how old their headphones are (and thus how much wear they have been subject to, how clean and undamaged their drivers are) will likely impact the sound they perceive far more than the difference between 16 and 24 bit or 192kHz vs. 44.1kHz.

But forget all that, this is what really matters, and where real science comes into it (not the theoretical, the practical!). Multiple blind tests have been done that show that even so-called audiophiles, even self-processed "super hearers", cannot in fact hear the difference between high bitrate MP3s/AACs and original CD recordings. If that's true, how can we expect to hear difference in the even smaller (relatively speaking) quality differential between CD quality and Pono? Now you can argue theoreticals all you want, but in the end there is one great way to answer this compellingly, and that is to run blind tests with Pono, with 24/192 audio vs. 16/44.1, and let's just see what the results are. This reminds me of the Randi Foundation's million dollar prize for proof of the supernatural - so far nobody has won. :D

Until that happens, as far as I can see at this point you're going to be buying files in a proprietary format that are 6 times larger than they need to be, using more bandwidth and hard drive space than necessary (and probably paying more for the privilege too). It's wasteful and unnecessary.

Of course Xiph.org has done a far better job than me of explaining why all of this is misguided. :D

By the way Joe: http://www.youtube.com/watch?v=NGJ9Z0wOGYk

- Oshyan

Navigation

[0] Message Index

[#] Next page

[*] Previous page