[doublem9876]

What do you guys think about apple releasing their own mac-specific audio encoding called MAC, just like Windows has their windows-specific WMA? MAC could stand for Macintosh Audio Compression. Agreements and Disagreements?

[Brad]

Absolutely not.

We already have some great audio codecs out there. AAC and the "Apple Lossless Codec" immediately comes to mind for compressed audio. For uncompressed audio, we have the WAV and AIFF formats.

Why bother with another format, especially when it will cater to an extremely small minority of users?

[Luca]

Brad's right... AAC already appeals to a small enough number of people since it's basically only available for QuickTime/iTunes users. Meaning all Mac users and several million (but by no means a majority of) PC users. How could such a codec possibly survive or be a good idea if it's Mac-only? It just doesn't make any sense at all.

If you're promoting this then you must also support MS's attempts at making their own stuff the "industry standard" without including Mac (and in many cases Linux) support. The thing is, MS has way more pull than Apple in these matters, so if they want WMV3 to be the standard video codec (and fuck all you Mac and Linux users, you're not good enough to play our amazing video codec), they can push for that. If they want IE6 to be the ONE and ONLY browser that works for certain sites, they can push for that too (and again, fuck all you Mac/Linux users who want open standards).

Non-cross-platform is simply a Bad Thing, period. I don't care if it's Apple or MS pushing for it, it's just bad news.

[DMBand0026]

Proprietary formats = teh bad. That's just another way for stuff to not work for some people and work for others. I just don't like the idea of it, K.I.S.S. Keep it simple stupid, we don't need another codec, and it certainly doesn't need to be proprietary.

[Franz Josef]

What they said. Very bad idea.

[NaMo4184]

In the legendary words of lamb chops."bhahaahaaaah d"

[oldmacfan]

AIFF is my fav...

[Gm7Cadd9]

how about calling audio compression for what it is...DATA OMISSION.

In audio, compression involves leveling peaks in audio to help control the dynamic range.

In data, compression implies there is some decompression going on, i.e. zipped files, or stuffit type files.

MP3s and the like (yes even Apple lossless) use data omission, as well as compression, and it sucks, MP3s and everything along those lines have a purpose and a place, they are portable, storing thousands of songs on an ipod is a great idea, but why not have a format is the purely uncompressed. They say you can hardly tell a difference between apple lossless and a CD...yet apple lossless is still somewhere around 11 times SMALLER, so what is omitted? The high end and the low end, the harmonics, the sibilance, the warmth. You say you can't hear a difference, but that's because you are used to hearing crap.

It boggles my mind how everyone is sooooo caught up in this whole YEAR OF HD crap, and yet not only do we settle for crappy audio we encourage it! I think this came about from MP3s being "free" and then we they became legit no one cared anymore because an entire generation was raised on this new technology, while I don't pirate anymore, I had my rebel care-free days, now if you expect me to pay for something I expect it to be the highest quality possible, otherwise there should be NO comparison to a CD costing so much money...I agree it sucks to buy a CD with only 1 good song, so give the option when downloading to download whichever format you choose...make it interesting and allow downloadable AC3 files for surround.

so to end my thread hijack I would like to say that I wouldn't mind a proprietary format so long as it was verbatim of what the original source was. Sure it would take up a lot more space...I am a recording engineer, and when I work in the studio on a session at 24 bits and 192KHz a 32 track session of a 4 minute song is easily 90 gigs...the bounce (also 24bit 192k) would be around 130 megs, which is huge! But then again hard drives ARE getting bigger, maybe I am crazy in wanting songs to be 130 megs, in fact saying it out-loud I know it is crazy and stupid...but next time you hear an mp3 that is 1.3 megs, remember the initial source was probably 130 megs....and those 128.7 megs had to be of some importance...so to end my long winded rant, yes make a new format, make it a higher quality, maybe not the BEST quality, but we need standards people...AM I THE ONLY ONE WHO CARES ABOUT THE RULES?!?!

...let's hear it for 1-bit DSD 2.87Mhz....anyone?

-Roy

[Luca]

Even CD-audio files aren't 130 MB per song. More like 35 MB (for a 4-minute song). I haven't yet met a single person (in person or online) who needs anything more than uncompressed CD audio to enjoy their music. I do know a couple people who insist on CDs and won't even listen to 320 kbps MP3 files, and several who are fine with MP3s if they're at the maximum possible bitrate of 320 kbps.

Personally I think it's a waste for me to use anything above 192 kbps VBR MP3, because my audio equipment isn't high-end enough to take advantage of it. That's the case with many people. Of course, one view is "garbage in, garbage out": give people crappy audio equipment (i.e. iPod earbuds) and they'll put up with 128 kbps compressed files.

[The Return of the 'nut]

Originally Posted by Luca Even CD-audio files aren't 130 MB per song. More like 35 MB (for a 4-minute song). I haven't yet met a single person (in person or online) who needs anything more than uncompressed CD audio to enjoy their music. I do know a couple people who insist on CDs and won't even listen to 320 kbps MP3 files, and several who are fine with MP3s if they're at the maximum possible bitrate of 320 kbps.

you have never heard anything above CD quality then. My dad's friend is a music producer. In the studio the difference is amazing between CD and the high resolution sources.

DVD Audio and SACD offer vast improvements over the CD.

Heck, a high end turntable (very high end) theoretically produces better sound than a CD.

And these differences are certainly audible withe the right equipment. Of course few have that equipment.

[Kickaha]

Originally Posted by Gm7Cadd9 MP3s and the like (yes even Apple lossless) use data omission,

Uh, no. Lossless means just that. Lossless = no loss of data = no data omission.

Some codecs are strictly compression, but recreate the original bit stream bit for bit. Those are lossless. Most toss out pieces that are theoretically harder for the human ear to hear anyway. Those are lossy.

Big difference.

[Brad]

Originally Posted by Gm7Cadd9 so to end my thread hijack I would like to say that I wouldn't mind a proprietary format so long as it was verbatim of what the original source was.

For a recording engineer, you don't seem to know much about DSP.

There is no such thing as "verbatim" digital audio recording. You can reduce the quantization errors to miniscule levels and prevent frequency aliasing by increasing the bits and sampling rate, as is certainly doable with current technology, but it is impossible to perfectly reproduce any analog signal in a digital form.

That said, once it's in digital form it stays exactly the same unless you use a lossy compression algorithm. Apple Lossless, FLAC, and others can be lossless, meaning that none of the original signal is lost from the compression algorithm. Using the Apple Lossless codec on an audio file is essentially like using zip or some other common "data" compression algorithm on a file. What goes in before compressing is exactly what comes out after decompressing.

We don't need a special, new format simply to reproduce audio at higher bit rates and frequencies. Audio IFF and others already support the 24-bit 192 KHz audio that you were describing. If you need compression, again, there are already several lossless compression codecs out there that will return 40-50% reduction in size.

I'm afraid I have no idea what it is you actually want that doesn't already exist.

[doublem9876]

okay, obviously nobody understood what I was asking for. All I wanted to know was whether you guys liked the name or not. I thought it was cool that I could use the three letters of mac to stand for something, I didn't want a whole discussion on audio compressions and if we need more, it was only for the purpose of the name.

[Robo]

Even the name is flawed, because Apple already uses the name ".mac" for their internet services.

Sorry...

[Wrao]

Originally Posted by doublem9876 I thought it was cool that I could use the three letters of mac to stand for something,.

Nooooooo!!!!! *falls over*

[LudwigVan]

Originally Posted by The Return of the 'nut And these differences are certainly audible withe the right equipment.

And the right ears.

I have to stop cranking the volume when I'm wearing my headphones. My handle may be "LudwigVan," but I don't want to end up deaf like Beethoven.

[Wrao]

Originally Posted by LudwigVan And the right ears.

It's mostly the right equipment if you ask me. I think even non-muso types with relatively insensitive ears will still be able to hear leagues of quality difference A/Bing something on very high end equipment and shoddy equipment.

I have very sensitive ears, but it wasn't until I upgraded my headphones recently that I have suddenly started to clearly notice differences in various levels of compression and uncompressed audio.

Of course, I don't let them ruin a song, but it can be a little bit annoying.

It's amazing what a good pair of headphones brings out in music.

[sunrain]

Originally Posted by The Return of the 'nut you have never heard anything above CD quality then. My dad's friend is a music producer. In the studio the difference is amazing between CD and the high resolution sources. DVD Audio and SACD offer vast improvements over the CD. Heck, a high end turntable (very high end) theoretically produces better sound than a CD. And these differences are certainly audible withe the right equipment. Of course few have that equipment.

I think that's why DVD Audio and SACDs haven't really taken off. Most people don't have the equipment to take advantage of the increase in fidelity. After all, most consumers think that Bose is the height of modern audio. *shudder*

I have a pretty sweet audio setup, but there aren't enough recordings (that I'm interested in, anyway) out to warrant buying a player for those high-end formats. Of course, if the next DVD (or HD-DVD, or Blue-Ray) player I buy is able to play higher fidelity formats, I might consider buying some recordings.

[Squozen]

Originally Posted by doublem9876 okay, obviously nobody understood what I was asking for. All I wanted to know was whether you guys liked the name or not. I thought it was cool that I could use the three letters of mac to stand for something, I didn't want a whole discussion on audio compressions and if we need more, it was only for the purpose of the name.

Let's just create a Moronic Applenova Conversation then. Oh, wait...

[Squozen]

Originally Posted by Gm7Cadd9 MP3s and the like (yes even Apple lossless) use data omission, as well as compression, and it sucks, MP3s and everything along those lines have a purpose and a place, they are portable, storing thousands of songs on an iPod is a great idea, but why not have a format is the purely uncompressed. They say you can hardly tell a difference between apple lossless and a CD...yet apple lossless is still somewhere around 11 times SMALLER, so what is omitted? The high end and the low end, the harmonics, the sibilance, the warmth. You say you can't hear a difference, but that's because you are used to hearing crap.

You are mistaking AAC (which is lossy) and Apple Lossless (which is lossless). Apple Lossless files usually end up between 650 and 900 kbps, which is nowhere near 11 times smaller than standard CD Audio (1411kbps).

[Franz Josef]

Originally Posted by sunrain After all, most consumers think that Bose is the height of modern audio. *shudder*

Hey, speaking as a consumer, don't dis my Bose

[chucker]

Originally Posted by The Return of the 'nut you have never heard anything above CD quality then. My dad's friend is a music producer. In the studio the difference is amazing between CD and the high resolution sources.

A German audio magazine invited 10 sound experts (producers, performers, etc.) and 10 amateurs to a "blind" sound test. They created a room with $20,000 worth of audio equipment. Pretty much the best you could get. They would take three formats: 128 kbit/s MP3, 256 kbit/s MP3 and lossless CD quality. The participants would not know in advance which of the three formats the piece they were hearing was in; they had to guess.

Before taking the test, almost all of the sound experts were convinced that they would easily be able to distinguish the CD quality audio.

The results? The experts would get far more matches on average than the non-experts -- no surprise. None of the experts would have even 90% of their guesses right. With some of the music, they were incapable of distinguishing even 128 kbit/s MP3 from the original CD audio track. Sometimes, the 256 kbit/s MP3 version, even after listening to it several times, would sound *better* than the CD version, which was explained by certain (compression-unrelated) algorithms in MP3 that make the sound more suitable for human ears.

The conclusion in the end, however, was: not only is 128 kbit/s perfectly suitable for the average ear; 256 kbit/s MP3 is certainly suitable for even fans of classical music.

Don't forget that AAC is a far superior code, taking up roughly 3/4s the bitrate for about the same quality.

So don't go telling me that DVD Audio or Super Audio CD is of any use, unless you take a similar test and aren't proven wrong.

[StevesMom]

Stupid idea. Sony tried something similar with ATRAC, and look where that got them.

[sunrain]

Originally Posted by chucker A German audio magazine invited 10 sound experts (producers, performers, etc.) and 10 amateurs to a "blind" sound test. They created a room with $20,000 worth of audio equipment. Pretty much the best you could get. They would take three formats: 128 kbit/s MP3, 256 kbit/s MP3 and lossless CD quality... ...So don't go telling me that DVD Audio or Super Audio CD is of any use, unless you take a similar test and aren't proven wrong.

Can you provide some linkage for this article/test?

You're being needlessly harsh. Be a little civil. DVD-Audio and SACD are of plenty use, even aside from any debate over discernible quality. Both formats handle multi-channel audio up through six channels, I believe. Both formats are also capable of handling a variety of bit-rates, which means that entire collections of songs/pieces could be recorded (at CD quality) onto a single disc. Imagine buying all of Beethoven's symphonies, on a single disc. Very cool. (assuming you like Beethoven )

[The Return of the 'nut]

Originally Posted by chucker A German audio magazine invited 10 sound experts (producers, performers, etc.) and 10 amateurs to a "blind" sound test. They created a room with $20,000 worth of audio equipment. Pretty much the best you could get. They would take three formats: 128 kbit/s MP3, 256 kbit/s MP3 and lossless CD quality. The participants would not know in advance which of the three formats the piece they were hearing was in; they had to guess. Before taking the test, almost all of the sound experts were convinced that they would easily be able to distinguish the CD quality audio. The results? The experts would get far more matches on average than the non-experts -- no surprise. None of the experts would have even 90% of their guesses right. With some of the music, they were incapable of distinguishing even 128 kbit/s MP3 from the original CD audio track. Sometimes, the 256 kbit/s MP3 version, even after listening to it several times, would sound *better* than the CD version, which was explained by certain (compression-unrelated) algorithms in MP3 that make the sound more suitable for human ears. The conclusion in the end, however, was: not only is 128 kbit/s perfectly suitable for the average ear; 256 kbit/s MP3 is certainly suitable for even fans of classical music. Don't forget that AAC is a far superior code, taking up roughly 3/4s the bitrate for about the same quality. So don't go telling me that DVD Audio or Super Audio CD is of any use, unless you take a similar test and aren't proven wrong.

your little story is laughable.

I don't believe it for one moment because I can easily tell the difference between those 3 formats even on low to mid range equipment and so can others. especially 128K MP3 which honestly sounds like shit and if played throuugh high end equipment sounds as if your music is being played through a waterfall or something

you also do nothing to dismiss the fact that formats with higher quality than CD can be better and that people can tell the difference.

again, I'm guessing you have never heard anything above CD quality audio, and it seems that is true.

[StevesMom]

Originally Posted by chucker They created a room with $20,000 worth of audio equipment. Pretty much the best you could get.

$20,000 buys you reasonably good prosumer stuff. Don't kid yourself this is "Pretty much the best you could get" - for that you need to spend hundreds of thousands of dollars.

[Kickaha]

'Nut, I remember the same study, but it was only anecdotal. I think it says more about the 'experts' actual aural acuity (as compared to their opinion of themselves) than it does about the sound quality. ie, I believe that such a study could happen, I've seen the same thing from snobby folks who SWORE they could tell the difference, but were frequently fooled by relatively low-quality recordings.

OTOH, I used to be able to hear *WELL* above what the established human hearing range is supposed to be, and I had a co-worker once who *could* tell between various encodings with nearly 100% accuracy... he couldn't listen to most CDs because there was a funky 'whine' that gave him a headache. I've seen audio equipment systems that were *amazing*, and the sound quality was improved immensely. In such cases, yeah, 128bit is going to sound tinny to a lot of people... but other people will *not* be able to tell the difference.

It's like my advisor - he simply cannot tell much difference between a $15 bottle of wine and a $150 bottle of wine. He tastes *a* difference, but not 10x worth. (Consequently, he's the master of the under $12 wine. ) An acquaintance of mine, however, is frightening in his ability to blindly distinguish between wines. He can nail the grape, region, year, and usually the *vinyard*.

People's senses vary wildly.

[Gm7Cadd9]

Well I certainly missed a lot since I left work yesterday.

First off, for being a recording engineer I do happen to know a thing or two about DSP. you provide a good argument Brad, everything is true, but what I am suggesting is less conversions in the chain. A/D converters are where the bulk of quality is lost, but even when it is digital and changes formats from within digital it loses quality. This is why dithering was invented. A complex way to boost the lower bits higher up. A program like Toast CAN convert a 24 bit audio stereo interleaved file into 16-bit 44.1, how you ask? it just chops off those extra 8 bits, and quantized those extra samples. how is it done from within pro tools? Using complex algorithms, in fact it's messed up, they add white noise to the signal...and blah blah blah, I am not the authority on this expert and that's not the point, LESS conversions in ANYTHING, audio, video, whatever is the ideal.

As for the German test, I find it crazy, has no credibility, nothing backing it, and even so statistics can tell a million different stories form the same numbers. HOWEVER it IS funny you should mention that story. I too have a story, MAYBE not a fact, but when I was a recording student I learned about the Great Rupert Neve, inventor of the world's best sounding Mic Pre. He said that while humans can hardly hear anything about 16k (let alone 20k) does not mean that they can't tell the difference when it isn't there. He did a similar test, and supposedly proved that while you could not hear 24k, you could hear a difference when a 24k harmonic was added to a 1k test tone. What does this mean to you and I? Well it's the reason you can tell the difference between hearing a piano in person and on a recording, even with the best of the best of the BEST gear and almost NO bottleneck you never question if you are hearing the piano in the room or on a recording. The reason is microphones cannot capture all the high end harmonics, and even if they can, they get lost somewhere, usually the mic pre, which is why Neve invented better analogue mic pres.

Along those lines, yes even an uncompressed CD file is not 130 megs, but someone mentioned it as well...SACD and DVD-A are better than CD. Keep in mind, when these artists are in the studio they are NOT recording at 16-bit 44.1. They are either recording on analog tape (if they have the patience and the budget) or on a pro tools HD rig, probably at 96k or 192k and 24 bit...at the VERY Least they will record at 48k. And what do they teach? Keep the quality as HIGH as possible until the very last bounce, then convert it, see also: less conversions. So in the studio it sounds amazing, but imagine there were no CDs, and it was straight to mp3, as much as that would suck do you think they would be tracking and mixing at mp3 quality? Of course not. the only reason these terrible formats exist is because WE are perpetuating them, it would be one thing if we didn't know how to make something sound better. And while you claim on consumer gear to not be able to tell, you haven't heard anything great then. You don't have to have an amazing setup to hear a difference, and even if a lot of it is subconscious, it's still very cool.

To return to a previous point I made, what if you did the same test with people and LCD, Plasma and DLP, and HD and component vs. composite...sure there is an argument there, some say they could, some say they couldn't and some people say it's apples and oranges. But I believe it was the great George Lucas who said something along the lines of Audio being 50% of the motion picture experience.

Wow, too many points to address, but as for Apple lossless being lossless...I don't buy it, I love everything apple does, and kudos to them for inventing the BEST mp3, leave it to apple to make something the BEST it can be, but all they did was polish a piece of crap. It is STILL lossless, it HAS to be, the file size is smaller, if that means cutting off the peaks, cutting the ultra low end and high end, it's still data omission, just probably not in the fundamental frequency range. I am NOT an expert on this, I suppose I don't even know how it works, but according to some basic scientific principles if it once was bigger and now is smaller, something got lost somewhere, usually in science the mass changes because of heat, and if matter cannot be created or destroyed then yeah, nevermind. you want a TRUE lossless format then talk about MLP..THAT is lossless. I would much rather debate DVD-A vs SACD, but when the debate is over varying mp3 formats I can't help but laugh. 1-bit DSD abandons the LPCM we have been using since the inception of digital audio and for once does not rely on a bit depth system using voltage to determine dynamic range, in 1 bit delta-sigma modulation the new technology samples at 287,000Hz well above what the human can hear, and while your speakers cannot reproduce a lot of that, the idea is to keep the source as high quality as possible up until the end, and the recording process where the initial A/D takes place is where it all counts.

Bleh, sorry I am so long winded, and jump from random point to random point, I just came into reading so much, I don't want everyone to think I am all anti-mp3, I just think we need someone swinging for for a better format. Keep mp3 for ipods and portable devices, and even computers if you so desire, but why not promote the better formats?!? If you can't win them over on quality then at the very least try to win them over with surround sound!!!

-Roy

[Kickaha]

Originally Posted by Gm7Cadd9 Wow, too many points to address, but as for Apple lossless being lossless...I don't buy it, I love everything apple does, and kudos to them for inventing the BEST mp3, leave it to apple to make something the BEST it can be, but all they did was polish a piece of crap. It is STILL lossless, it HAS to be, the file size is smaller, if that means cutting off the peaks, cutting the ultra low end and high end, it's still data omission, just probably not in the fundamental frequency range.

Welcome to information theory.

Assume a string of letters: the cat the man the plan they flew on the plane

47 characters

Convert them like so: Take substrings and cache them, then rewrite the original string:

A = the
B = plan

A cat A man A B Ay flew on A Be

Now the string is 34 characters long. Add in the 7 original characters as a lookup table, and you're still at 41 characters. You can do recursive caching and lookups like so:

A = the
B = plan
C = 'A '

Now you can rewrite it as:

Ccat Cman CB Ay flew on CBe

27 characters + 9 character lookup table = 36 characters. That's a 23.5% reduction in data size.

You've reduced the size of the data, but you can always flip the algorithm around and recreate the original string *EXACTLY*.

That's lossless compression.

Compression *MAKES* files smaller, that's *what it does*. Lossless/lossy only refers to whether the decompressed result looks exactly like the original file... not what the compressed file in the middle is.

Now do you get it? Apple Lossless = lossless. What you get in the end is *EXACTLY* what you had in the beginning. MP3/AAC/etc = lossy. What you get in the end is *not* exactly what you had in the beginning.

Lossy compression literally makes it so that the *end* result, *after* you uncompress it, is still smaller than the original because data has been thrown away. That's where the loss of sound comes from.

[Gm7Cadd9]

Thanks for the explanation, it does clear a lot up, and it makes a lot more sense. I am still a little skeptical though. I see how you can substitute, and even even anticipate as in some cases, but is it really the same? like 1k brings out a certain sound on everything, but even a boost at 1k does not make a guitar sound like a snare drum or vice versa. Along those lines, the transient/wave form drawn is very complex, if you were talking about variables and substitution I could see that with something simple like sine waves and graphing a function, it's just sooo complicated when you zoom in on a waveform it seems like there would be near infinite number of variables and few of them being interchangeable, I think I understand the idea of taking out variables and caching them, and when the file is played back they are "put back in"

And something else I read on lossless packing (in mp3 that is) that says it uses something along the lines of predictive quantization. Almost like how oversampling works, this I could see. In digital if you graph a sine wave it is not fluid and smooth, but goes in steps, the more steps the more "smooth" the line becomes. Oversampling uses math to take the distance between any two points and plot more points that don't *really* exist between them, the result is a much "smoother" waveform. If done properly oversampling makes a waveform better, but done poorly it adds noise to a signal, up-sampling and oversampling are terms that are thrown around a lot, and sometimes it's a very good thing, but more often than not, it is NOT a good thing.

So I suppose, in the end if lossless does do everything with no loss, even with the super complicated variables, then yay! BUT, along the same lines, why not use this technology for better formats? If it is as perfect as it sounds, then make AC3 files smaller, make a multi-channel SACD mix smaller, maybe then people will take to these formats and stop embracing the mp3 so much!