AAC and VQF

Posted by: Terminator

AAC and VQF - 20/02/2000 10:52

Does the empeg play aac and vqf files? Ive been reading some about these formats and plan to test them out to see how good they sound compared to mp3. Supposedly the vqf files can be compressed down to a smaller size and still sound as good as mp3s.

Term

Posted by: rob

Re: AAC and VQF - 20/02/2000 12:34

AAC is a very expensive format to licence, so we would only do so if there was great demand for it. It probably wouldn't be a free upgrade either, so it would have to wait until such time as we can distribute codecs in a modular fashion.

The issues with VQF (more properly known as TwinVQ) are more involved. The format was developed by NTT and licenced exclusively to Yamaha. The format is proprietry and Yamaha refuse to release specifications - although a few people have reverse engineered it. Given the relatively small demand we have received to support this format, it's almost certainly not worth us risking the legal minefield that would come with it. I believe there is an open VQF project underway for Linux, so maybe someone else could hack up an empeg codec, if and when we provide support for such things.

Rob


Posted by: JeepBastard

Re: AAC and VQF - 20/02/2000 13:42

i've done some extensive testing with all formats. Your better off with mp3 at 160. The amount of tools you have access to do manage your mp3s is lot greater than those for VQF and AAC. Those two formats lacked the critial mass to become commercially popular. It was helped by the fact that the companies tried to hold the formats way too close to their chest.

remember a2b? even better the company they deecided to use as their developer did a piss poor job of presenting it to the public. Telephone companies trying to play with digital music. heh


Empeg Kicks Ass
S/N 00203
http://www.iretro.com
Posted by: Terminator

Re: AAC and VQF - 20/02/2000 22:47

Have you tested the ePac format?
http://www.audioveda.com/html/about/audioveda.html
I bet that it is what is left of a2b. Lucent technologies developed it. The compression ratios look good though, and the player/ripper/encoder is in alpha.

Posted by: Terminator

Re: AAC and VQF - 20/02/2000 22:56

Will the ARM processor be able to handle these formats if and when you decide to support them though? The tradeoff of most of these new formats appears to be that you have to have more cpu horsepower to play them.

Term

Posted by: rob

Re: AAC and VQF - 21/02/2000 02:17

> Will the ARM processor be able to handle these formats if and when you
> decide to support them though?

We currently use around 30% of the available processing power to decode MP3, plot a visual, and do all the other background stuff.

Rob


Posted by: altman

Re: AAC and VQF - 21/02/2000 09:27

MP3 beat ePAC in listening tests at the last perceptual audio conference, I seem to remember. Probably at >128kbit bitrates, though.

Hugo


Posted by: mcgrant

Re: AAC and VQF - 28/04/2000 15:41

You can put me down as someone interested in the AAC format when it comes out. I'm firmly entrenched in 192kbps and above for MP3 as it is, but I'd love to save disk space with a better encoder.

But perhaps it would be better still to wait for the MPEG-4 audio format standard to settle down, and go with that. My understanding is that it is AAC + a couple of new twists, and that in theory MPEG-4 decoders will be able to play MPEG-2 AAC files.

Michael
mark 2 waiting list...



Posted by: rob

Re: AAC and VQF - 28/04/2000 17:24

The ARM processor is making huge headway in embedded audio applications and we can benefit from these development efforts. Libraries are, or will be, available for most established or forthcoming formats of any significance.

The big issue for us is whether the demand makes the licencing burden worthwhile. In the case of AAC this is a very significant amount which would probably have to be passed on to clients. Right now there isn't the demand.

We've already signed up for WMA (I'm not sure if that will make Consumer 1.0, but if not it won't be far behind) and we're keeping a close eye on future formats.

Rob


Posted by: ianken

Re: WMA (WAS RE:AAC and VQF) - 11/05/2000 13:36

I am looking forward to WMA support. I've switched to it from MP3. It compresses much faster and sounds significantly better at the same bitrate. To get good 128k MP3 files you have to use the Fraunhoffer codec on "high quality" mode and that takes for ever. I get better 128k WMA files and can rip and compress in a single pass at 8X on my system. I tried the XiNG codec via AudioCatalyst and was not impressed with the quality, too much aliasing even at 128k.

Posted by: tfabris

Re: WMA (WAS RE:AAC and VQF) - 11/05/2000 13:47

I tried the XiNG codec via AudioCatalyst and was not impressed with the quality, too much aliasing even at 128k.

Interesting... I switched from Fraunhofer to Xing because I could still hear aliasing in Fraunhofer High Quality even at 128k.

The trick is that Xing can do variable bit rate. The high-frequency aliasing and artifacts go away when you do VBR. That's the advantage of AudioCatalyst: When you do the middle-quality VBR, you get better sound than Fraunhofer at 128k-fixed, for about the same file size.

Did you test AudioCatalyst in Variable Bit Rate mode or just 128k-fixed?

Tony Fabris
Empeg #144
Posted by: bryan

Re: WMA (WAS RE:AAC and VQF) - 12/05/2000 05:51

Has anyone played with LAME recently?

Supposedly it now produces better quality than Fraunhofer.

Posted by: Bruno

Re: WMA (WAS RE:AAC and VQF) - 17/05/2000 17:32

In reply to:


Interesting... I switched from Fraunhofer to Xing because I could still hear aliasing in Fraunhofer High Quality even at 128k.

The trick is that Xing can do variable bit rate. The high-frequency aliasing and artifacts go away when you do VBR. That's the advantage of AudioCatalyst: When you do the middle-quality VBR, you get better sound than Fraunhofer at 128k-fixed, for about the same file size.


Couldn't agree with you more, Tony (although I would use Xing's encoder with another front-end such as Easy CD-DA Extractor, which allows more flexibility in file naming). Xing's encoder is super-fast (at least as fast as WMA encoding when I played with it a while back), VBR gives excellent quality to compression ratios, and the format is not proprietary. One of the best features about the Empeg is that it is based on an open platform. How sad to put another one of Microsoft's attempts to dominate a key market sector (in this case multimedia streaming servers) on such a cool piece of kit.

And the guff that Microsoft put out about the compression/quality ratios of WMA compared with MP3 was their usual marketing rubbish. I played with it quite a lot, and from what I could tell, WMA may have been a little better at the same bitrate, but it certainly didn't offer similar quality at half the bit rate as they claimed, even when listening to Microsoft's own sample files. At 64kbps, a lot of WMAs sound hollow. You would want to use at least 96kbps, and probably 128kbps to play safe. And in that case, your files would be barely any smaller than VBR MP3s at high-normal (75%) quality setting.

To anyone thinking of using WMA, I would suggest that you have a look at WMA's digital rights management capabilities, built in to their proprietary metadata structure, to get an idea of what Microsoft intend to achieve with WMA. And then think about the the flexibility of MP3, that allows options such as ID3v2 tags, that you would be sacrificing if you abandoned it for the closed solution that is WMA (admittedly, theoretically you could use ID3v2 tags with WMA, but it is unlikely that Microsoft will support this).


Cheers,

Bruno
[email protected]

Posted by: Jens

Re: WMA (WAS RE:AAC and VQF) - 10/07/2000 06:41

Sorry to join the thread late ... I've been off the forum for a while; my car order was delayed due to Ford deciding they're not making any of the 2000 models, so I have to wait until the Autumn (Fall) to get a 2001 model. I digress ...

I've been using WMA format at 128bit, ripping CDs to my new dedicated multimedia and phone server - it's hooked up to my Marantz/Mission stereo with a wireless keyboard and 53" TV for a monitor ;-) It rips at about 6x speed and on a blind WMA/CD test, I couldn't tell the difference in a home environment, so it's going to be fine for the rag-top.

I see that the empeg will later support the WMA format ... my question is will there be any track info when I import the files, or will I have to rerip them all? There's obviously something getting stored as Media Player 7 has full track, album and artist details logged.

Here's a thought ... I have cars in both the US and the UK that I'd like to empeg enable. I can get cradles for them both, sure, but what about the 6 hour flight? Can I get a battery pack? Lead acid batteries aren't allowed on the plane. ;-)

J.

Posted by: altman

Re: WMA (WAS RE:AAC and VQF) - 10/07/2000 07:40

We read the WMA header, so any info that media player gets the empeg will get too.

Hugo


Posted by: Jens

Re: WMA (WAS RE:AAC and VQF) - 10/07/2000 08:45

Splendid. Thanks!

Now I just need the empeg and the car to put it into!