[PCWorld]

Post your comments for Thomson Pumps up Volume With Tools for MP3HD Lossless Codec here

[mpheadley]

Why not just use wav files, and if your device isn't compatible, make it able to play wav files!

[mpheadley]

Ok, after re-reading, I understand now they will be compatible with existing mp3 players, but again I still say, why not use wavs at home and high quality mp3 on the go? Do you really hear the difference with headphones between mp3s and wav files now?

[oldschoolh4ck3r]

Why not encode at very high bitrates using the existing MP3 encoding? If 192 Kbps is CD quality, 320 Kbps should be just like lossless. Also, HOW ABOUT SURROUND SOUND? Dolby 5.1/6.1/7.1 and DTS should be an option. Now THAT would be innovative. There were talks of doing this by the MP3 creators years ago.

[WinTard]

Thank you for great news!

As an audiophile I welcome the fact this new MP3HD is lossless compression and will be compatible with existing MP3 decoding algorithms.

However I deplore the fact it is proprietary but understand the practical reasons behind it. And I suspect some ripoff of the open-source FLAC algorithms tweaked towards compatibility with existing MP3 technology is what this MP3HD is all about.

I already use FLAC - Free Lossless Audio Codec, for my media collection; alas it isn't implemented everywhere like in my car audio... Thus I maintain a duplicate stream of .FLAC and .MP3 files.

And even though I only encode using 320Kbps (or better) MP3, it still is lossy, meaning we lose bits of info. I can hear the difference.

FLAC results in lossless compression of .WAV files by about 50% of the original size. (a bit like a .ZIP file/stream). I wonder if this MP3HD will be similar, or achieve better lossless compression?

Regardless, I'm going to try it now.

PS: a .WAV file results when you 'rip' a standard audio CD sampled at 44.1KHz (which results in a theoretical maximum sinusoidal waveform quantization of 22.05KHz according to Nyquist's Theorem)

[beobrian]

If you use headphones in good quality, you can easily hear the difference. You will meet great loss of sound quality, with standard MP3.

[beobrian]

I am sorry, but 192 k.bit/pr. sec is no near CD quality, you can hear great loss of sound quality. Only WAV or the different lossless formats, will give you the CD sound quality. You can very easy hear and feel the difference.

[knowwi]

Can't speak to the quality of the work, but the very same site with the lossless MP3 has a surround Mp3 and a tool that can "emulate" surround sound from stereo....

[mjd420nova]

I really get a kick out of everyone tossing the term "lossless" around like it really means anything. Anytime you compress analog information so that it takes up less digital space, something gets discarded and even when de-compressed, that information is lost and cannot be recovered. When dealing with analog signals, it isn't so much about discarding certain parts of the discrete frequencies, but the way all of that signal interacts with the other parts of the signal. Multiple harmonics and sub-harmonics make up as much of the real music as any other part. It isn't so much what you hear but what you feel. I have tried every lossless compression scheme that is available to the public and found everyone to be just more hype than real preformance. And I mean not by just listening, I use frequency counters, sine, square and triangle wave signal generators and spectrum analyzers in a laboratory environment. None of them has ever returned an analog signal that is even 90 percent of the real signal as it was original produced, compressed and then de-compressed.

[WinTard]

While I agree with most of your statements mjd420nova, I will put it to you:

Do we live in a digital or analog world?

My point is at some point, molecular, atomic or sub-atomic, there is a limit where things become insignificant. Both from the miniscule, to the macroscopic scales of things. If you set your oscilloscope on 1 millisecond or 1 microsecond scale, you will note and observe a greatly different perspective of things.

To that idea, I put it to you that in terms of compression algorithms, there is a world of distinction between lossy and lossless. And this distinction is purely absolute in terms of mathematics and logic.

Computer zipping programs, regardless of the algorithms involved are usually lossless compression algorithms. This means that for every bit truncated in the compressed format representation of a bit stream, once reconstituted to its original format, every single bit of information is 100% there, bit-for-bit. However, in a lossy compression scheme of things, re-expanding the compressed bit-stream will result in discrepancies to the original bit stream.

One could argue that even in the analog world, dealing with finite gradient steps (at the sub-molecular level) ^1^, transcoding any type of information from one state to another, in the case of audio, going through a microphone membrane, to convert acoustic pressures into electrical signals, further encoded, packaged, stored, transmitted, duplicated, reconstructed, amplified, and finally ending up as acoustic pressures emanating from a loudspeaker; all these intermediate steps are 'lossy' in essence because 'distortions' will occur, regardless of the medium being in a digital or analog domain.

I give you an example why I paid attention to my speakers being able to reproduce 50 KHz within a -6dB from 1 KHz: Well musical tonality is basically composed of fundamental frequencies, with multiple harmonics which compose the 'timbre' of the music. Thus explaining why an identical frequency note on a piano, trumpet, guitar, or violin each sounds distinct. They are all producing similar fundamental frequencies, yet are distinguished by their harmonics.

Thus taking into consideration that multiple harmonics have valleys and dips, they then to either amplify (standing waves) or cancel each other, resulting in an orchestral sound, whether modern or classical, is irrelevant. But an instrument having a fundamental frequency of say 10KHz will have a second order harmonic frequency of 20KHz and third order and fourth order to infinity. The summations of these harmonics will result into fundamental frequencies within the audible range. Example. A 48 KHz and 50 KHz harmonic will result in a 2 KHz summation, which falls well within the audible range.

The ultimate speaker I know of, but alas, can't afford, is a Pioneer (yes Pioneer Electronics of Japan, the World's largest loudspeaker manufacturer) professional ribbon tweeter that can reproduce 200 KHz linearly. At $800 a piece, I need twenty of them per side, plus the high-quality amplifiers to drive them individually, (yes I am an audiophile enthusiast)... But that is one of my future goals... But that driver will reproduce a clean 20 KHz square wave? And that's only for the super-tweeter...

Back to earth, my point is that even the concept of a 'straight wire with gain' from Mark Levinson is mere lossy transcoding technology.

Sorry for the rant everyone, but this topic of High-Fidelity is extremely dear to my heart, and has been since age nine. I'm also crazy to admit I've spent well over $100K in the last twenty years or so on my home theatre. Mainly by tinkering and building things myself from parts... Imagine if I had to purchase brand-name... I just couldn't afford it.

~~~~~~~~~~
You are as young as your dreams, and as old as your doubts.
{Chinese Proverb}
-----
^1^ Well sound is air pressure waves. Air is roughly composed of 19% Oxygen, and 81% Nitrogen (when pure)... Air molecules are therefore finite. And in between the molecules is void space (in which no air pressure variation exist). Thus explaining why sound requires a medium for transmission. Radiation however, such as heat, light, or cell phones radio-transmissions do not require any medium to propagate. Anyone noticed the Sun lately? Radiation can be good or bad for you...