Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Psychoacoustic modeling (Read 1206 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Psychoacoustic modeling

Hello, I would like to know which of the lossy audio formats has the best psychoacoustic modeling and if that means having greater transparency and quality

Re: Psychoacoustic modeling

Reply #1
Quote
if that means having greater transparency and quality.
It would mean the best quality for a given bitrate.  i.e. Good quality and high-compression.   

Quote
Hello, I would like to know which of the lossy audio formats has the best psychoacoustic modeling
I don't know...  MP3 and AAC can both often be transparent at higher bitrates and once it's transparent the quality can't improve.     The "improvement" would be lower bitrates (smaller files).     But in some cases I believe AAC beats MP3.   (I haven't done tests myself).

I don't have any complaints with any of my AC3 (Dolby) DVDs, but of course I can't A/B or ABX them.    It's lossy but some of my concert DVDs with surround sound are the most enjoyable music I own!

Some encoders are optimized for lower bitrates (smaller files) rather than sound quality.    Obviously, whatever they use for cell phones is optimized to minimize bandwidth.

Re: Psychoacoustic modeling

Reply #2
Hello, I would like to know which of the lossy audio formats has the best psychoacoustic modeling and if that means having greater transparency and quality
Psychoacoustic modeling is a feature of a particular audio encoder, not the audio format which you're encoding to. It's possible for a "dumb" encoder to generate a file of a modern format like xHE-AAC which sounds terrible. However, modern formats like xHE-AAC generally, due to better compression technology and more research during the last years, make it easier for an encoder's psychoacoustic model to operate.

So if you want the best possible "theoretical quality" (assuming your particular encoder does a decent job), then I'd always go with the newest format.

Chris
If I don't reply to your reply, it means I agree with you.


Re: Psychoacoustic modeling

Reply #4
Not wanting to sound too cynical to newcomers to this or any other field (everyone is one in almost everything) or worse: like an old geezer (I reckon I'm halfway into it) reminiscing about "how good things were back then", but I can definitely see guess where questions like these come from, thanks to some ten-fold increase in these so-called "best of lists" in the last decade and a half or so - be it about programming languages or lawn mowers. The still-standing "specialized press" have had their fair share of such clickbaits as well, with their regular "best drummer/album/movie ever" BS!

People have just gotten used to this "why reading the book/article? Let's all skim the Cliffs notes/wiki article and get done with it!" drivel! And with such, second-hand bias, instead of properly researched opinion, seems to have become not only acceptable, but even commendable!

Back to our lossy audio compression field, I recall that, at least in the late 90s, when promises as far-fetched as MP3 Pro or VQF came up, the first thing we did was downloading the (usual proprietary) tools, (so we'd already start off by taking it with pinch of salt) and then tried them there and then, before wanting to get miraculous end-all answers to fields as subjective as this one in particular or as encompassing as programming languages.

All this rather chatty pondering (sorry) makes me wonder:  What ever happened to the good olde hands-on approach? I mean, most tools are free or even open source, people can easily take THEIR OWN conclusions for themselves IF they're willing to do so!

Edit: strikethrough and some rephrasing/typo correction
Listen to the music, not the media it's on.
Wavpack hybrid -hb4.3x4

Re: Psychoacoustic modeling

Reply #5
I have read in some forums and internet sites that there are formats such as musepack that, despite being old, implement a good psychoacoustic model to encode, so the newest would not always be the best (taking into account the blind personal listening test - MultiCodec a ~ 192 VBR kbps by IgorC)

Re: Psychoacoustic modeling

Reply #6
The Musepack encoder has a very good psychoacoustic model, but from my personal experience, the MPC codec only works acceptably well at roughly 96 kbps stereo and above. If your primary interest is around 200 kbps, then probably most encoders of the last few years (except maybe some FFmpeg specific open AAC encoders, Kamedo2 may know) should perform decently. But yes, in case of doubt, do some blind listening tests, as includemeout wrote.

Which of course begs the question, how many formats is/are "AAC"?
Quite a few ;) With "modern" I meant codecs like xHE-AAC, MPEG-H Audio, and AC-4, though one may not have access to all of those.
If I don't reply to your reply, it means I agree with you.

Re: Psychoacoustic modeling

Reply #7
I’m not sure the people commenting here are fully able to. There is a difference between a psychoacoustic model and the implementation of a psychoacoustic model.

Re: Psychoacoustic modeling

Reply #8
I’m not sure the people commenting here are fully able to. There is a difference between a psychoacoustic model and the implementation of a psychoacoustic model.
Care to expand on that, please?
Listen to the music, not the media it's on.
Wavpack hybrid -hb4.3x4


Re: Psychoacoustic modeling

Reply #10
I’m not sure the people commenting here are fully able to. There is a difference between a psychoacoustic model and the implementation of a psychoacoustic model.

Yes Teacher
Please do teach us the difference.

Re: Psychoacoustic modeling

Reply #11
I’m not sure the people commenting here are fully able to. There is a difference between a psychoacoustic model and the implementation of a psychoacoustic model.
Care to expand on that, please?

What part?
Just like The Irish Man said.
Listen to the music, not the media it's on.
Wavpack hybrid -hb4.3x4

 
SimplePortal 1.0.0 RC1 © 2008-2021