i want to convert my cd's to m4a at 192 kb/s but I don't know which aac codec is the best nero, itunes, faac, dobly (http://www.dolby.com/professional/product_manufacturing/consumer.html) include with easy cd-da extractor, ...


excuse for my bad english

I'd go for Nero. 128 or 160, 192 kb/s/stereo is a waste of your harddisk space...

iTunes won the AAC listening test @ 128kbps..

yes but later versions of itunes had degraded audio quality.

Only if you can listen above 16kHz.

Hmm.. except the high frequency "issue" (which has been blown out of proportions on this board) the overall quality in QT 6.5.1 is better than the previous versions.

Give me proof.

And I don't know what you understand by "out of proportions", but we have listening test results from board members showing that it indeed degraded quality. Where are your test results showing quality didn't degrade all that much?

Does AAC have a LAME equivalent? Most people seem to use either itunes or Nero (from the sounds of things), but does anyone consider one to be "better" than the other?

If so, is there a specific verision that is preferred overall (e.g. LAME 3.90.3 rather than earlier or later variants)

No. Both iTunes and Nero offer high quality encoding, and neither seems to be obviously better than the other (quality-wise).

What about FAAC? I've heard a bit about them and they seem to be mildly popular around the place.

I don't know too much about AAC, but I'm interested in ABXing it sometime, like OGG vs LAME vs AAC vs MPC :/

Like this?

Well I haven't seen that till now. Interesting. What's Vorbis aoTuV?

I found on rjamorim's site that FAAC doesn't score to well at all. In this test:

http://www.rjamorim.com/test/aac128v2/results.html

Its surprising how well iTunes did at 128kbps CBR, even beat Nero VBR which had an average of 140kbps

magic75 said: "Now that was a surprise... Lame as good as AAC??? Anyone expected that?"

Did he expect LAME to beat AAC? Or AAC to beat LAME?

BTW, what type of encoder does Winamp 5.05 use for AAC? It has two types: Raw and MP4 available. Have they ever been tested? I thought MP4 and AAC were the same, or is MP4 AAC an AAC encoded file in an MP4 container?

Bloody confusing lossy formats and technicalities, its enough to make me stick to .WAV's

AAC to beat lame, for sure.

Remember AAC was developed to be the successor of MP3. So, you expect it to deliver better quality.

Also, the MPEG forum touts that AAC has 30% efficiency over MP3 - that is, it should sound the same at bitrates 30% smaller.

I think an explanation to what happened is that the Lame developers did a wonderful job taking the MP3 format to its limits. Lame is an excellent encoder and is already taking nearly the most of what the format can give. Most AAC encoders, on the other hand, are just past their infancy and have lots of room for improvements. It's also worth mentioning AAC is reportedly much harder to tweak than MP3.



"is MP4 AAC an AAC encoded file in an MP4 container" <- that's correct.

Winamp uses the Dolby encoder.

And no, I didn't test Winamp AAC. I kinda repent for it now, but this test's discussion became so catastrophic at one point, with people tampering polls, people getting into flame wars about what encoder to feature, and so on, that I decided it was enough and went with only the 5 encoders that were already chosen by then.

It wasn't my biggest mistake when conducing tests, though. Thankfully.

Regards;

Roberto.

Vorbis aoTuV is essentially Ogg Vorbis 1.0.1 with special tunings done by Aoyumi. Most of these tunings are now in Ogg Vorbis 1.1.



The AAC encoder in Winamp 5.05 seems to have problems with castanets.

http://www.hydrogenaudio.org/forums/index....showtopic=24285

I abxed it successful up to 224 kbps.

Thanks for clearing that up. Any thing musical these days is very dependent on the situation. There's zillions of different encoders, decoders, transcoders, cd burners and software such as Nero that can all have an effect on apparent sound quality.

Not to mention the effect different sound cards, speakers or headphones can have

I was thinking about encoding all my songs into AAC (using Winamp), but that was before I came across EAC, LAME and FLAC (as well as a million other lossy and lossless formats).

What version of LAME was used in that test? I'm currently using LAME 3.90.3 (which supposedly has a lot of backing around the audio community as being the "best"). But I think I saw a test around here somewhere that suggested LAME 3.96.1 may be better

BTW, what is a container? I read a bit about it on the FLAC sites faq, but the concept still seems weird to me.

The current idea I have is that a container is just like a second format with added stuff that allows tags.

Is MP3 natively in a "container"??



What do you mean? Castanets as in the musical instruments?

I'd say the time mp3 born is before the concept of a "container", so mp3 is just raw encoded stream with added feature such as id3tag.

Yeah. There are at least two killer samples that have castanets in it. Winamp AAC had some problems encoding them transparently, even up to 224 kbps.

3.96.1



3.96.1 wasn't as tested as 3.90.3. But I believe it's quality is probably on par with 3.90.3. I really don't believe quality degraded a lot, if it degraded. And it has the advantage of being much faster.

I use 3.96.1 for my MP3 encodings.

QuickTime 6.5.2 and iTunes 4.7 were both released today with claimed quality improvements...

Beliefs don´t count, you must provide proof that quality did not degraded a lot or a little.


(just kidding )

MPEG Layer 3 streams are most commonly contained in MPEG 1 containers. However Mp3 can also be muxed into MPEG 2, MPEG 2.5, MPEG 4 or Matroska conatiners, perhaps there are some other which I missed. Mp3 streams muxed into MPEG 2 or MPEG 2.5 containers usually still have the .mp3 extension. AFAIK every type of multimedia file (maybe every file period) has to have some sort of container.

There is no multimedia container wrapping MP3 streams. They can be considered RAW, since it's just the compressed information and frame headers, pretty much like ADIF AAC.

I don't consider frame headers a "container"

Also, if it was inside the MPEG "container", you would need an MPEG demultiplexer to playback MP3.

MPEG 1 and 2 don't have native containers. The video and audio streams are just multiplexed together, they are not wrapped inside something else.

MPEG 2.5 is just an extension to the MP3 standard for very low sampling frequencies. It doesn't make sense to speak about it as another MPEG multimedia portfolio.

Are you sure about this ?

SebastianG

That's what I have always seen in multiplexers. With several different ways to multiplex the streams: PS, TS, etc.

I guess I did not understand what you mean by "native container". AFAIK the elementary streams are somehow packetized and then muxed on a packet basis.

SebastianG

I stand corrected. Thank you for clarifying Roberto.

I think it's worth mentioning that iTunes AAC is not yet gapless unlike Nero AAC.

Although it seems Nero isn't doing it the correct way.
http://www.hydrogenaudio.org/forums/index....ndpost&p=254841

(not that there is some sort of correct way - the standard doesn't provide specifications for gapless information storage. But it seems the current way is actually dangerous and bad for decoders)

Does faac handle gapless info correctly?

It handles it the same way as Nero.

what's about http://www.aacplus.net/ ?
is better nero or it?

There are some quite full notes on AAC here.

Are those of any help? From what's said there, it seems like aacPlus is another name for HE/AAC. Presumably, anyone's who is using HE/AAC is licensing it from them. It seems to be intended for very low bitrates. But Coding Technologies claim, "delivers CD quality stereo at 48 kbps" sounds a little dubious to me. Apple only just told us:



I mean, the company is new to me, and if their technology is five times as good as Apple's wouldn't we all have heard a bit more about it?

Granted Steve Jobs may have been covering EMI's arse a little for them, since Apple already said much the same of 128 kbps AAC awhile back. (A little more money on singles (but no more on albums) may be what EMI really wants, in order to nudge customers back towards albums, and the increased bitrate may be a sweetener for the price increase rather than being strictly necessary for "equal to CD quality" for most listeners.) And after all, how long's a piece of string? Still and all ...

Coding Technologies is the inventor of Spectral Band Replication (SBR), the advanced technology found behind MP3Pro and HE-AAC. It's safe to say that they know how to code a good AAC implementation, but last year's HE-AAC listening test as well as personal experience clearly prove that the "48 kbps = CD quality" claim is nothing but a shady marketing feint. You could obviously compare it to Microsoft's "64 kbps = CD quality" running gag, introduced with the release of the WMA 8 encoder to push the format against MP3 and its common "128 kbps = CD quality" claims. Besides, if you want to test the Coding Technologies codec yourself, simply use Winamp's internal AAC encoder.

But - in the low bitrate sections Coding Technologies' claim is partially true, at least if their encoder's compared to Apple's implementation. The fact that the latter neither supports HE-AAC nor any other useful low-bitrate technologies, like Intensity and Parametric Stereo (based on statements made on these boards, personally I don't have any experience concerning Apple's AAC codec), leaves it at a clear disadvantage. A 64 kbps Coding Technologies (HE-AAC) file sounds a lot closer to CD quality than its 64 kbps iTunes (LC-AAC) counterpart, though, of course, for a trained hearing this bitrate should still be too low to achieve complete transparency. We'll know more about that after the upcoming multiformat test, where Nero's HE-AAC codec is gonna be included at the same bitrate, will be finished.

Thanks for confirming both points. What I'd guessed, but it's good to have it confirmed.

IOW, in answer to Francesco's question, he'd be getting this technology in Nero's, or anyone else's, HE-AAC.

thanks
what i think that i understood is not encode below the 128 in acc

question: which is best quality AAC encoder in 128-192 kbit/s CBR?
thanks

question:


but is 128k acc really the cd quality? is true?

Based on the previous listening test of 128 kbps, i think we can safely say that a lot of people will have difficulty distinguishing between 128 kbps (MP3, AAC, OGG etc) and a CD.

I guess a lot of things have changed since 2004 - is nero now pretty much the standard as far as AAC encoders go?

Thanks.

i can hear the difference between 128mp3 & cd
i'm talking abut 128 acc & cd

Only youself can tell. AAC and MP3 compression algos cause different artifacts. You might be sensible (or just trained) to one kind and not to the other. Thus YOU might be better off with it (or not). Just test it.

If your question implied that AAC at 128kbps was artifact safe: No it is not.

I'm pretty sure the last listening test showed iTunes and Nero approximately equal at 128kbps (with iTunes having a SLIGHT edge up on Nero), but I could be misremembering.

I think you are right, the iTunes AAC encoder has a slight edge over Nero's. It was really small though. I don't know the actual quality score but if the iTunes AAC encoder scored a 4.4 then Nero's scored a 4.3 (not the actual numbers but I believe the two encoders were that close to each other).

I personally trust the Nero AAC encoder more than the iTunes AAC encoder simply because Nero's seems to act more like Lame in that, when encoding in VBR, you choose a quality level rather than a bitrate. I am sure Apple implemented their VBR scheme for a reason (to give people a known file size/bitrate instead of quality) but I much prefer picking a quality level over having a known bitrate/file size.

For the people that were not paying attention to the original posting date (like me).

This thread was started in 2004 and the postings on this thread before Francesco's April 7, 2007, question is now about 2.5 years old so there have been some changes with various AAC encoder quaity since this original topic started in 2004.

is there a test about that ?
about the best aac??
and the best setting to have the cd quality in acc?

I'm on the Mac (sometimes Linux) and so don't use Nero, but I think I'd use it if I were on Windows for that very reason.

For example, take old mono recordings of something - say, Andrés Segovia playing guitar. If you encode those with the suggested default in LAME you'll get a bitrate of around 100kbps. Now if I were to set the iTunes/Core Audio encoder to 160kbps VBR, that would be great for most stuff, but with material like this it won't drop below 160kbps (or not by much). It seems to me that would be throwing bits away when the information simply isn't there on the CD, owing to the recording limitations of the time.

Likewise, there are tracks that LAME will encode at over 200kbps on the same setting. It seems to me that where LAME would do that, then there's probably a warrant for it and on those occasions a setting of 192kbps VBR in iTunes would be better.

Now I suppose one could take a lot of time assessing the likely needs of particular material and ABXing sample tracks from each album at various different bitrates in the iTunes encoder. But I'd sooner just let the encoder take care of it. I take it that with Nero I could simply set a quality of 0.5, and it would take care of that in accordance with the, perhaps widely varying, needs of different material.

True that. iTunes' "VBR" mode functions like ABR. This is not as efficient because it sticks to a specified average bitrate, instead of using how many (or how little) bits are needed to maintain a consistent quality level.

It's better than CBR, though.

I plan to rip my entire Mozart collect to AAC. As the project is rather big. I want to make sure I will use the right tool for it.

I spend sometime to search this forum and the net. I found there are a few AAC encoders for Mac.

1. iTune, Use Core Audio
2. MAX sbooth.org/Max It works pretty well, easy and use Core Audio too.
3. XLD http://tmkk.hp.infoseek.co.jp/xld/index_e.html I can't really get it work on my Mac. But some user find it better than MAX.

I am not sure which one is the better encoder. I plan to do some tests later, but not for now.

Do someone has an answer for this?

thanks ahead.

Vic

I may be wrong, but I don't believe that XLD will encode from a CD. I can't find anything to indicate that it will in its menus or on its website.

What you can do with it is to "split" a single lossless file with imbedded cuesheet into multiple tracks and at the same time transcode those tracks into another format (such as AAC). It uses the Core Audio encoder for that.

So, AFAICT, you can't use XLD to encode AAC unless you first use some other means (e.g. Max or a Windows program) to encode multiple tracks to a single lossless file.



I think you can save yourself some tests. Max would be the obvious choice if you wanted to encode to FLAC or MP3; but you want to use AAC, and, as you already said, both iTunes and Max use the Core Audio one. There's no determining which is the better encoder because both use the same encoder.

So the choice of which of the two to use for ripping/encoding comes down to other factors than the encoder.

First off, I suppose there is the ripper. Speed matters a lot to some people. Max is slow when using its default ripper, but quite fast when using CDParanoia. The ripper in iTunes is even faster. CDParanoia is supposed to be good at error correction, but unless you've got very scratchy CDs I don't suppose that will be a problem.

There is also the matter of what online database the program gets its information from. Max, a very nice program in other ways and what I use myself, unfortunately uses MusicBrainz, which is not a good database, and particularly poor for classical music. For example, MusicBrainz will probably return the artist field as "Mozart" in all cases and leave the composer field blank. By contrast, iTunes CDDB is usually quite good and only needs a little cleaning up.

Other than that it's really down to whether you prefer the interface of the one program or the other.

As far as I know XLD uses the same Core Audio encoder that is standard on OS X *. If there are any differences in what people prefer it should be features of both apps. Do you have more info on which user finds it better and for what reason?

* It would be a real hassle to use a different AAC encoder. You'd have to get it from somewhere, possibly with license issues and incorporate and distribute it in your app. All this while one of the worlds best LC-AAC encoders is already present and free to use on any OS X installation.