It seems the marketing employees at Ahead have gone a little overboard with the ink lately. For more information take a look at IBC2003 Report (http://doc.hydrogenaudio.org/Members/JohnV/IBC%202003/IBC2003/view). It's JohnV's report on the IBC 2003 exhibition. Scroll done through Ahead's flyer to the heading "Audio Codec Features". Once you've read the whole section it becomes apparent that some liberties have been taken whilst writing a description for HE-AAC and AAC. For example:
Support for High Efficiency AAC in MPEG audio. This offers the highest possible audio quality at bitrates as low as 64k, 48k, 32kb/s, or even 5:1 ability at 128 kb/s.
HE-AAC 128kb/s? Last time I checked (just 10 seconds ago) Nero can only encode in HE-AAC at a maximum of 80kb/s. So this statement appears to be a flat out lie, even from an objective point of view.
In fact CD quality stereo at 48kb/s
Wow, did I miss a thread here or something? This is the first news I've heard of these CD quality claims at 48kb/s. Even HE-AAC is not equivalent to CD quality at 48kb/s and rjamorim's listening test, rjamorim's listening test (http://audio.samharris.us/test/64test/results.html), proves this even at a higher bitrate of 64kb/s.
Transparent quality at 128kb/s. MPEG4-AAC provides transparent audio quality at 128kb/s according to all professional and formal listening test conducted so far.
That certainly is a lot to take in. I mean transparent audio quality at 128kb/s? Not according to this test, rjamorim's 128kb/s AAC listening test (http://audio.samharris.us/test/aac128test/results.html), carried out by rjamorim. In fact if Nero AAC provides transparent audio at 128kb/s then how can QuickTime AAC rank higher than it in the test? I guess there is a level of audio quality beyond transperancy? Or maybe it's just the fact that Nero AAC isn't transparent at 128kb/s and Aheah is spreading propaganda.
Quality of MP3 at 50% the space. With High Efficiency AAC in MPEG audio, you will get the same quality of a MP3 at 50% the space
Once again another incorrect statement as rjamorim's listening test (see the first link above) proves. 64kb/s HE-AAC failed to equal the same quality of LAME MP3 at 128kb/s. I'm beginning to see a pattern of lies forming here and I'm becoming very tired of refuting them .
I think Ahead should stand up and take some responsibility for their misleading marketing campaign surrounding HE-AAC and AAC. There certainly needs to be some corrections made so that people are not misinformed about the quality of Ahead's products. I agree that He-AAC and AAC do provide superb audio quality and I use their product a lot however it is important they do not exaggerate the audio quality that can be achieved, expecially when they are trying to substantiate their false claims with objective data. The last thing Hydrogenaudio community needs is for AAC and HE-AAC myths to start circulating around the internet and have to debunk them. Must I remind everyone about the 128kb/s MP3=CD quality myth. Such myths can only hurt and hinder the audio community. Maybe Ivan Dimkovic can explain Ahead's statements or have a word to the marketing people at Ahead?
P.S. And people thought Xiph's marketing claims for a "patent free" codec were outlandish!...sheesh
I too am tired of companies blatantly lying to the general public simply because they won't know any difference. While I thoroughly respect the AAC codec and the work behind it, I do not have any respect for a company making false claims about their product, regardless of how good it is. Instead of trying to fit in and beat Microsoft and other competition at their own game of bitrate marketing b.s., I think Nero should be the better company, acknowledge the true tested ability of their codec, and simply list the inconsistancies and lies of the competition. You know, set the record straight. This would bring them major props in the audio world, and would educate those uninformed in these matters.
The bitching about Ahead's marketing should be directly comparable to Microsoft's BS marketing about 64==CD-quality. The general lesson is that a codec only sounds as good as you think it does. In other words, people truly up on audio compression are going to find the right answers regardless. Those that don't, well, do we feel responsible correcting idiotic points of view?
HE-AAC 128kb/s? Last time I checked (just 10 seconds ago) Nero can only encode in HE-AAC at a maximum of 80kb/s. So this statement appears to be a flat out lie, even from an objective point of view.
If you spent more time reading instead of flaming one company, you would notice
5.1 in the text - which is SIX channel audio encoded at 128 kbps. Sometimes it is just hard to read, right? So, for people with cheaper tickets:
HE AAC is available for 5.1 (SIX) channel encoding at 128 kbpsOf course HE AAC is not available for two channel audio @128 kbps. There would be no point.
Regarding 48 kbps claim, sorry for making you upset, but this is translated claim of the principal inventors of the algorithm who own patents, not by Ahead Software (or, God Forbid, me). CD quality is a subjective measure, not exact. The only thing I can be sure is that HE AAC beats WMA, Ogg, mp3Pro or any other codec with a large improvement @48 kbps - some of them already are marketed as "CD quality" at 48 kbps - so this is a hard case for marketing dev. Since there is no defined measure of "CD Quality" it is completely pointless to argue about that - that's just how marketing works and marketing is unavoidable in a real-capitalist-world, sorry.
Regarding transparency @128 - latest Ahead AAC codec is vastly improved in some problem samples compared to the one used in the test (it will be shipped with the new web release). It can certainly win subjective ITU scale >4.0 ranking which
in fact is a ITU border of transparency (please read ITU-R BS.1116 and relevant documents) - like with CD quality, it is also a subjective measure - for me, no AAC codec is transparent @128 kbps, but for >80% of people it certainly is. In fact, for most people - both QT and Nero are transparent @128 kbps.
If we would respect HA measures for "CD Quality" no codec would be able to have that claim attached - even MPC has some problem samples at its own highest settings - so, we must define some worldwide measure, like ITU-R impairment scale and border level of 4.0, like I already explained.
Sigh. I don't see why Ahead has to repeat the claims of the principal inventors of the algorithm, seemingly without consideration of how viable that claim is. If the inventor was deaf and claimed it's transparent at 8kb/s, would that be thrown into the marketing machine?
On the other hand, HE AAC @ 64 could well be equal to MP3 @ 128. It was reasonably close to LAME, so how would it compare to Xing which is one of the most wide-spread encoders?
To me it looks like HE AAC is better than WMA, so the ludicrous claims of Microsoft also have to be bettered, so neither can claim to be the bigger man at this stage.
edit:
If you spent more time reading instead of flaming one company, you would notice 5.1 in the text
No, you would notice the 5:1 in the text. Which makes you think of compression ratios, 16/44/mono to 128k is ~5:1, which is what I first thought when seeing that.
MPEG4-AAC provides transparent audio quality at 128kb/s according to all professional and formal listening test conducted so far.
That implies that all professional and formal listening tests conducted so far have deemed AAC to be transparent at 128kb/s? Was Roberto's test not professional and formal, or not considered?
Well, all relevant AAC codecs in Roberto's test scored >4.0 on ITU scale, which puts them into indistinguishable from the original ITU range. Term "transparent" is not mathematically defined, and the closest thing I would tie it to is that ITU scale.
Also, similar results were obtained in official MPEG listening tests.
I am no expert in marketing & sales, so I am definitely not going to comment on "CD quality @48 kbps" label attached on many codecs. The only thing I could say is that HE AAC scored considerably better than any other competitor @48 kbps on EBU listening tests. Since "CD Quality" is not defined by any math measure it is meaningless to discuss about it - everyone has its own preference, for some people it might well be 48 kbps.
I can claim that HE AAC @48 kbps performs significantly better than MP3 @96 kbps- that is something I can definitely measure.
Sorry for the "5:1" - I am not sure who made this mistake, I hope I clarified it a bit.
It seems the marketing employees at Ahead have gone a little overboard with the ink lately. For more information take a look at IBC2003 Report (http://doc.hydrogenaudio.org/Members/JohnV/IBC%202003/IBC2003/view). It's JohnV's report on the IBC 2003 exhibition. Scroll done through Ahead's flyer to the heading "Audio Codec Features". Once you've read the whole section it becomes apparent that some liberties have been taken whilst writing a description for HE-AAC and AAC. For example:
Well, imo it's quite natural that brochures try to give as good picture as possible, but who doesn't know it's advertizement talk? To be honest, I didn't even read the brochure's audio section so carefully, my report is anyway much more concentrading on the video side, and the brochure gives quickly some information of its features.
Support for High Efficiency AAC in MPEG audio. This offers the highest possible audio quality at bitrates as low as 64k, 48k, 32kb/s, or even 5:1 ability at 128 kb/s.
HE-AAC 128kb/s? Last time I checked (just 10 seconds ago) Nero can only encode in HE-AAC at a maximum of 80kb/s. So this statement appears to be a flat out lie, even from an objective point of view.
I'm not very enthusiastic to defend an advertizement brochure, but it does say "5.1" referring to 5.1 audio. Futhermore you might also want to check your own checkings because Nero HE-AAC maximum is currently 96kbps, not 80kbps (the highest preset is only 80kbps though, but you can set it manually to 96kbps).
In fact CD quality stereo at 48kb/s
Wow, did I miss a thread here or something? This is the first news I've heard of these CD quality claims at 48kb/s. Even HE-AAC is not equivalent to CD quality at 48kb/s and rjamorim's listening test, rjamorim's listening test (http://audio.samharris.us/test/64test/results.html), proves this even at a higher bitrate of 64kb/s.
This is pretty old news.. MS claims CD quality at 64kbps, if you want to indicate to ignorant masses that you have something better, you gotta use the same advertizement talk. Anyway, my report is not meant for newbies who can't understand the difference of advertizement brochure and real world. Of course I could have taken only some of the sections from the brochure, but I thought it would be interesting for people to see the actual Nero Digital brochure.
Quality of MP3 at 50% the space. With High Efficiency AAC in MPEG audio, you will get the same quality of a MP3 at 50% the space
Once again another incorrect statement as rjamorim's listening test (see the first link above) proves. 64kb/s HE-AAC failed to equal the same quality of LAME MP3 at 128kb/s. I'm beginning to see a pattern of lies forming here and I'm becoming very tired of refuting them .
Well, this claim is, depending on the case not so wrong. Sure, it's not better than Lame 128 ABR, but there are many MP3 implementations which infact are worse quality at 128kbps than Nero AAC-HE 64. I don't like these kind of comparisons either, but they are part of the advertizement nowadays..
The last thing Hydrogenaudio community needs is for AAC and HE-AAC myths to start circulating around the internet and have to debunk them.
That's why group blind listening test results are very much at present here. I don't think that any rumors start to circulate because of some scanned advertizement brochure!?! And frankly any rumor like that can pretty much only spread among those who believe that 64kbps WMA is CD quality, so I don't think there's much harm done anyways.. Everybody can see what it is, and should take it as what it is.
P.S. And people thought Xiph's marketing claims for a "patent free" codec were outlandish!...sheesh
Well, the claims are by nature very different. In Ahead's case it follows the marketing trends, and probably refers to the fact that audio quality is always very subjective experience. Of course I strongly disagree with claims like "64kbps is CD quality, 48kbps is CD quality", (In fact older members remeber even back from the r3mix days that I defended a view that any lossy audio can't be CD quality), but there's not much that can be done to change the marketing strategy of companies.
... The only thing I could say is that HE AAC scored considerably better than any other competitor @48 kbps on EBU listening tests...
Are the results of those test available somewhere?
Edit: Do you mean this (http://www.ebu.ch/trev_291-dietz.pdf?display=EN)?
Some of us here know, and hear very well that CD-Quality is not possible at 48 kbps, because we have been artifact-trained and have high quality equipment like the Audiophile Revo, Senn HD(?)-580 cans or Klipsch Promedia speakers. However, for the rest (the so-called masses), they're listening to their music on rudimentary hardware. They wouldn't hear a difference, especially with HE-AAC being such an, well, efficient codec.
I detest marketing too, but hey, it isn't deception per se, as justified above. I would be contented with wma if I didn't know it was a format that took away user rights unknowingly, I would be contented with Musicmatch Jukebox if I hadn't stumbled on www.r3mix.com, I would be contented with -r3mix if I hadn't stumbled onto Ha.org, and I would be contented with --aps if I hadn't read up on Musepack.
I detest marketing too, but hey, it isn't deception per se, as justified above. I would be contented with wma if I didn't know it was a format that took away user rights unknowingly, I would be contented with Musicmatch Jukebox if I hadn't stumbled on www.r3mix.com, I would be contented with -r3mix if I hadn't stumbled onto Ha.org, and I would be contented with --aps if I hadn't read up on Musepack.
ignorance is bliss
or is it?
@ JohnV
Thanks for your comments and I just wanted to clarify that I was in no way attacking your report. My intention was to focus on Ahead's brochure.
@ Ivan
Mac is right in pointing out the brochure stated 5:1 compression ratios at 128kb/s and this is what I was refering to. Your correction about 5.1 audio does make more sense though. It does always pay to read the brochure before you post doesn't it Ivan ?
I could swear that I read that before it went out to print Now... it seems that I must increase my caffeine dosage...
In fact CD quality stereo at 48kb/s
All I can say is that this is just marketing BS. It does not help much to say that MS did a similar thing.
Well, all relevant AAC codecs in Roberto's test scored >4.0 on ITU scale, which puts them into indistinguishable from the original ITU range. Term "transparent" is not mathematically defined, and the closest thing I would tie it to is that ITU scale.
For anyone who hasn't read it, one set of tests say the following:
10.7. Statistical indistinguishability
“Is the performance of the coding of AAC codecs at the test bitrate distinguishable from the original signal?” In general, from Figure 5, we see that the performance of the AAC codecs is statistically distinguishable from the original signal. However, for certain items, the codecs give indistinguishable performance. The AAC Main 128 codec is indistinguishable from the original for 8 of 10 items, the AAC Main 96 for 3 items, the AAC LC 128 for 8 of the 10 items, AAC LC 96 for 4 items, and AAC SSR 128 for 8 items. For comparison, MPEG-1 Layer II was statistically indistinguishable for 4 items, and MPEG-1 Layer III for 3 items.
10.8. EBU “Indistinguishable quality”
“Is the performance of AAC codecs at the tested bitrate achieving ‘indistinguishable quality’ in the EBU definition of that phrase?”
A detailed description of this criterion and the statistical tests required to analyse it are found in [1]. The following chart is in the same format as the chart in section 6.9 of that document. Rather than calculate a “cut-off” as described there, we are directly comparing the confidence intervals as recommended by the EBU definitions. The data for these comparisons is found in Annex 7.
Codec Items failing Ratio (if needed)
AAC Main 128 4 0.9528
9 0.9448
AAC Main 96 0,1,4,8,9
AAC LC 128 1 0.8931
2 0.8501
AAC LC 96 0,1,2,3,4,6
AAC SSR 128 1 0.8436
2 0.8420
MP2 192 1,2,3,4,7
MP3 128 0,1,2,6,8,9
Thus, we see that the AAC Main 128 and AAC LC 128 codecs provide ‘indistinguishable quality’ in the EBU sense of the phrase . The AAC SSR 128 codec fails to meet this criterion by a margin of less than 1% relative to the decision criterion.
10.9. Most critical material
“Is the following requirement of ITU-R Recommendation BS.1115 fulfilled? ‘For emission, the most critical material for the codecs must be such that the degradation may be perceptible but not annoying (grade 4)’”
It is difficult to know exactly what statistical criterion to use in evaluation of this question. The following table shows the mean and lower bound of the confidence interval of the rating score of the most critical (i.e., lowest-rated) test item for each codec, extracted from the table in Annex 7.
Codec Mean Lower bound
AAC Main 128 4.4227 4.1051
AAC Main 96 3.4818 3.0077
AAC LC 128 3.8500 3.4409
AAC LC 96 3.1409 2.5833
AAC SSR 128 3.7409 3.2518
MP2 192 2.3182 1.9423
MP3 128 1.6318 1.3105
It is clear that the AAC Main 128 codec meets this criteria, since it is statistically unlikely that for any of the critical items, the true rating is as bad as “perceptible but not annoying”. For the cases of AAC LC 96, MP2, and MP3, in each case the confidence interval contains “slightly annoying”. The other three cases (AAC Main 96, AAC LC 128, and AAC SSR 128) do not contain “slightly annoying” in the confidence interval, but each has a mean lower than “perceptible but not annoying”.
(from http://www.mp3-tech.org/ (http://www.mp3-tech.org/) programmers docs "Report on the MPEG-2 AAC Stereo Verification Tests" w2006.zip)
It's interesting that there are three quality thresholds, going upwards: "good enough for emission", then "indistinguishable quality" (both of which are fudging the issue),
then "Statistical indistinguishability", which is the kind of thing we aim for here, which no codec can manage on all samples.
It's even more interesting that mp2 (used in most of the world except the USA in digital broadcasting) doesn't meet
any of these EBU requirements, and yet is used widely throughout the EBU.
This does seem to excuse the praise for 128kbps AAC. But 48kbps = CD quality? Surely "proven in independent tests to be better than
anything else currently available at this bitrate" would be more truthful, and more impressive. Afterall, people simply don't believe "CD quality" labels on codecs anymore, but they might look at that claim a little more seriously.
Cheers,
David.
In fact CD quality stereo at 48kb/s
All I can say is that this is just marketing BS. It does not help much to say that MS did a similar thing.
I agree. To me the terms "CD-Quality", "Transparency" and "Lossless" all have concise definitions that should be adhered to, marketing or not. Actually I don't acknowledge "marketing boo-boo" as a valid excuse in this case since Xiph is under complete attack right now because of their own marketing-speak concerning Vorbis. Either this is a no-no as much as Xiph's predicament, or Xiph is as "in the clear" as Ahead is minimalizing the accountability of their "48kbps CD-quality" claim. (Except that from what I've seen
more people have proven even the higher 64kbps to be *not* CD-quality than have proven Vorbis to be not patent-free...but I won't get started on that road. Just a point of comparison.)
If you're talking about what "sounds just as good as CD", then the proper term is "transparency", and it's entirely subjective. 48kbps may be transparent to one person, where another needs 64 and another needs 256 and another (with remarkable hearing) may only find lossless to be "transparent". If they said "48kbps is
transparent", that would be fine. They'd be stretching the truth implying it's tranparent for *everyone*, but at least they'd be using the right word.
"CD-Quality" refers to bit-equivalency IMO, and CD-quality is only found with a lossless encoding. It's is like being pregnant...either you are or you're not, and it's not a subjective term. (And the fact that some CDs are released now with transcoded lossy formats is another matter...CDs that are not CD-quality. ) A (measurable) determining factor is the ability to transcode from anything "CD-quality" to any lossless format and remain lossless from source audio. 48kbps to FLAC fails this measurement.
These are very important terms to use properly if those "in the know" about digital audio encoding are ever to educate the rest of the world and try to bridge the "knowledge gulf". Because the other route is to exploit the "rest of the world" by telling them FUD like "48kbps is CD quality". If you don't know any better then at worst you're uninformed. If you *do* know better then at worst you're a liar, and in turn using lies to exploit people for profits. Shame, shame...
(These statements are not accusations against any specific individuals, but rather statements of philosophy.)
I'd have to agree with your post ScorLibran, the term "CD quality" is refering IMHO to a bit identical copy of the CD and this can only be achieved with lossless encoding. "Transperancy" would have been a more accurate term to use despite it still being aplicable to only a small group of listeners in the case of 48kb/s.
I don't see what all the fuss about "CD quality" is. It begs the question of "What is quality?"
The term was first used in the PR world when claiming that 128kbps MP3 - be it Xing, FhG or LAME - was CD quality. Just like WMA @ 64 and HE-AAC @ 48 isn't transperant, neither is MP3 @ 128. You're expecting the vocabulary of one profession, public relations, to be the same as that of others, such manufacturers, broadcasters or your own. Do as you always do; use your common sense and never trust anyone praising their own product, until you or someone you trust has verified it. Why should Nero be treated any differently?
Many companies contradict their own marketing if you look hard enough. For instance, on their PR pages, Apple claim that 128kbps AAC is The Best. On their developer pages, the claim is that where MP3 requires at least 128kbps/channel to be indistinguishable, AAC requires 96kbps/channel. Which claim do you find more likely? And in which case is Apple most interested in being close to the truth?
Nero A2C is a structure of AAC (mp4) among others. Most of those others are free, but Ahead try to make money out of all new they put in their product packages. It was not a long time ago, when mp3pro was some "new". And Ahead tried to make money with it's great (en) / (de)codec whilst better one's were found here and there... After a few months it was free and staid as free.
It was just MPEG-3 layer 2 or 2.5, which were before that adobted by makers of LAME and others. All said that 64kbps encoding is the same as 128kbps MPEG-1 layer III (and 'Stereo Quality' of course!). and 1/2 - 2/3) os size!
Ppl can claim that mp3 @ 128 and mp3pro @ 64 gives You a Stereo quality. The same involves HE-AAC @ 64 (or 48k), But I have ears - Just taking off all the DFX's and other "sound enhancements" and listen. After that I put the original CD to my (good) player and listen - whoa - whatta difference. The lower and higher frequencies are back and the midway is not like drinking soap anymore!
At the beginning of CD-Ripping at after mid 90's there was Fraunhofer Institute's MPEG1-Layer III and all said that 96k is 'Stereo-Q' and all above is luxus (with that poor and slow kinky codec!!!???). Then came Xing (97) and put all upside down - a good and FAST codec, which was changeable, presetable and took the best out of each CPU Instruction sets...
(and then Gogo - as fast as Xing, Used Different CPU 'sets' , but was poor as a begger in ste street at quality...)
And Now we live the (let's say) fourth generation time in the audio codecs evolution. Nero's AAC has been hacked, ppl can use it for no charge ("thanks" to the "Evil Hackes") as they did before that to the MP3Pro-Codec...
Still my experience is that a good audio compression is beginning at :
1. 192k with MPEG-3 layer I or variable 160-224 (Lame 3.93.1 or Xing)
2. 96k with MPEG-3 Layer 2.5 (Lame 3.93.1 or Xing)
3. 128k with WMA profile (seems to be highest...) (MS)
4. 128k-160k AAC-HE or variable Extreme::High (as in Nero-Presets)
5. Ogg-Vorbis VBR / ABR 160-192k (Poikosoft Power-Encoder etc..)
6. Monkey's Audio at Normal Compression (Lossles but takes up all HD-Space :-)
7. and all the lossles new audio compressors for the Hi-Fi men...!
I can do with the first one (192k Basic MP3 with Good Q=2 at Lame ripping), because new HD's are not too expensice anymore
With 140 bucks U get 150Gigs new HD (7200rpm 8M Cache, No-Noize Mechanism etc...And about 10 albums fit to a single 80min CD (50 Cent each about...)
Filling four of those HD's U get 0.6TB Space and with $35 U get controller for CD-DVD-RW's with four extra IDE-Ports...
So, I rather listen good music as HI-FI as possible. These new techiques just mess You around. Just now is a good time to test all possible and wait for the final 4th generation codec(s)...
P.S And welcome all to my PC - This is my first message at hydrogen's - I have mostly been hanging out with FileSharing-Forums etc Music related before...
(Just tell me if I talk too much - I promise to read more than write ;-) ...
1. 192k with MPEG-3 layer I or variable 160-224 (Lame 3.93.1 or Xing)
2. 96k with MPEG-3 Layer 2.5 (Lame 3.93.1 or Xing)
I think you are confused here.
There exists:
MPEG Layer I, Layer II and Layer III. These are different encoding methods, Layer III being the well-known 'MP3' and the other ones dating back mostly to Video CD and similar things.
MPEG 1, 2 and 2.5. These are, IIRC, bitstream/header formats, and don't relate to the actual quality of the audio part.
Not Confused @ all. That was just my (poor) finnish to english writing with too fast, without thinking, just one part of the story. Of cource I'm familiar with the MPEG-2 Codec used mostly at bitrate 224k when making (S)VCD's - that is just a thing we have to accept : much (not so lossles) audio at VCD takes the file a lot bigger - just hoping for the days (at near future), when U can use MP3 (or AAC) and DivX 5.x at VCD's (AND play them with [now quite expensive] home players)...
MPEG-3 Layers are (so far) I, II and 2.5.... (and layers 2, 2.5 are often referred as MP3Pro...)
~~~ Woops, now I'm talking to 'Vorbis developer' - Just You speak - I'll Listen ~~~~
So sorry my mistake (not a leak of knowledge...) ;_)
Of cource I'm familiar with the MPEG-2 Codec used mostly at bitrate 224k
MPEG-1 Layer II
MPEG-3 Layers are (so far) I, II and 2.5.... (and layers 2, 2.5 are often referred as MP3Pro...)
MPEG-1 Layer III is 32khz/44.1khz/48khz
MPEG-2 Layer III is 16khz/22.05khz/24khz
MPEG-2.5 Layer III is 8khz/11.025khz/16khz
MP3Pro uses MPEG-2 Layer III with SBR, and there exists no Layer 2.5.
Yes, Still one mistake - the 2.5 at MP3Pro - I have had to remember that one...
Still writing too fast concidering my thoughts being at same level ;-)
Yes, I Know the frequencies well - for example layer MPEG-1: 44100,
MPEG-2: 22050, MPEG-2.5: 11025 (in hz's)....
Still MP3Pro at 22050 hz has it's points - it's not just dividing the frequency - that should sound awful. I'm Not an expert in that technology in a "bit-level"...
Sorry once my "colons and semis"
BTW : Lately I've been trying Ogg quite a lot and it seems good (and fast to encode with good quality [ABR & VBR]). Used Poikosoft's Power Encoder and #1 CD-Ripper for that.
In what way do You see Ogg evolving at near future. Are You gonna make new "layers" or put the file in half to get the same Q as the AAC-PPL do (try...)?
The one and only sad point is that all the nowadays "home theaters" and portable players accept only MP3-1 (at layer 3) as compressed music and some accepts MS-WMA...
So this is all-the-time-evolwing world, heaven knows what we got after 1 year...?
In what way do You see Ogg evolving at near future. Are You gonna make new "layers" or put the file in half to get the same Q as the AAC-PPL do (try...)?
Although I wrote several Vorbis-related utilities and code, I'm not the main developer not inventor of Vorbis, so I cannot answer what kind of design changes are planned for the future. But I do know quality will be improved in the next version by simply fixing errors and tuning the current Vorbis 1.0 better...
"In addition to Real's own codecs, the popular open audio format Ogg Vorbis is being integrated into the Helix DNA client, according to Jack Moffitt, CEO of Xiph.org." - BetaNews
So Good point for these new codecs and apps using these : Availibility to get off of the Horrible Windows to say Linux (no Wine anymore), Apple etc...
And just when a project is "Open-Sourced", Every (talent) ppl can put their spoons to the soap and the most talent man or group chooses and puts all together...
As LAME (Ain't Mpeg Encoder) with GNU-Origin
And AAC has many good groups - like the Faac-Project (Still a bit DOS-Like, but good front-ends are there to join the formats and their goodies...
And Ogg-Vorbis, completely open (with sources), patent Free, professional, mimetype new official since feb. this Y. Waiting for v1.1...
MP4 or AAC - Which one wins the mimetype competition? MP4 has been "adopted" to different Audio / Video purposes (solely by real and apple - streaming audio and video - envivio TV etc) and still is used as a variation for AAC (I like mp4 more as a file-ext, but it's confusing...)
Still my winamp's aac-plugin (again from poikosoft and free) plays well exts .aac and mp4, but in case of aac it tell me to "convert" the header and put it in mp4-format, after that I can use tagging for those...(mp4's real or not)
So I beg You an explanation for mp4 and mp4 - I think that I understant the point partly and the "mime-formats" are this-year-fresh, but I still think I have missed a point(s)?
That's why I'm here - hopely mostly asking than explaining or playing mr. perfect ;-) (Me - Never That !!!)
As LAME (Ain't Mpeg Encoder) with GNU-Origin
LAME (http://lame.sourceforge.net/rationale.html) is not a part of the GNU project; it was independently developed.
And Ogg-Vorbis, completely open (with sources), patent Free, professional, mimetype new official since feb. this Y. Waiting for v1.1...
Except for the patent part both MP3 and AAC are as open as Ogg Vorbis, with open source implementations. You mentioned some yourself, actually.
MP4 or AAC - Which one wins the mimetype competition? MP4 has been "adopted" to different Audio / Video purposes (solely by real and apple - streaming audio and video - envivio TV etc) and still is used as a variation for AAC (I like mp4 more as a file-ext, but it's confusing...)
The MP4 (and M4A) and AAC do not denote the same format. MP4 is a container format for AAC and MPEG-4 video just like Ogg is a container format for Ogg Vorbis, Ogg Theora & friends. M4A is often used for audio only mp4's. Raw AAC should be avoided - it has many drawbacks compared to MP4.
Azo-999: I think that you should check www.mp3-tech.org for some introduction to mpeg audio, mp3 and mp3pro.
And Ogg-Vorbis, completely open (with sources), patent Free, professional, mimetype new official since feb. this Y. Waiting for v1.1...
Except for the patent part both MP3 and AAC are as open as Ogg Vorbis, with open source implementations. You mentioned some yourself, actually.
Well, it didn't take long till we faced this issue regarding Vorbis' "patent freeness" again..
Azo-999: you might want to read this (http://www.hydrogenaudio.org/forums/index.php?showtopic=13531).
About That Vorbis Patent(ness) I read the article in general reviews (don't remember anymore the address - I do not bookmark everything)
Yes - I did read the comments - thanks!
My dear friend mr. Darwin told me to ask : Which one came first - The Ogg or the Vorbis? That's all folks about patents and GNU's. It's not a relevant point for the future of mankind !
And I came here mostly to learn about all-audio-video-related things, NOT to be a genius at all. And I have found GOOD links, suggestions, howto's etc so far...
I work as a freelancer Systems Analyst / Programmer and can read well VB, Delphi, C, C++ (& CygWin which I use to make sources from SourceForge etc to win32 executables) and some others. The first two I use, but not as making audio related apps (except a few frontends & helper apps), merely doing with different databases, SQL, ADO/RDO, Btrieve and some Web-projects (XML, Java (+Script) etc). And Forms to use them...
So be nice to me and accept my stupidness (and the habit, that I cannot be mouth or fingers shut among You experts).
I will also go to other sections here and simply thow up some Q's according my musical needs (which will after all relate to codecs, quality and apps I use, because music itself is so near my heart, and also the various technologies around them (Just quite new at my interests) - As at the age of 42 I'm still not an "oldie", who'll say : "It has always worked, what more do I need")
So I wish You a VERY good morning, day, evening or night depending wherever You happen to live ....
I'm GMT+2 18:40 PM - This forum's machines must be at U.S. west coast or ...? because it's 7:40 AM there and You are just wakie up...
ignorance is bliss
or is it?
Way back i nthe day, when I was learning to encode in MP3, I used to encode at 128. Then, I learned to hear the artifacting, so I switched to 160, re-encoding all my files. Then I heard a few artifacts even there, so I switched to 192, figuring that should finish it, seeing as how all the release groups released at 192CBR. But I heard a few artifacts there, even. So, not knowing how to better that, I gave up, disheartened with MP3.
I was sure to check on LAME development, and one day I noticed something called --alt-preset showing up in the logs. As best I could tell from them, (and I was a bit confused) it allowed specific VBR encoding for certain genres of music. I searched for documentation, so I could try it out, but found nothing.
Then, suddenly one day, I decided to try Vorbis, stumbled on HA, found out about musepack, and here I am today.
The point is that I was very unhappy that I couldn't find information on something better than MP3, while being confident that something
had to exist and that I couldn't possibly be the only person on the planet that could hear the loss.
As an aside, at one point, I thought I could hear the quality 44.1kHz 16bit sound had, artifact-wise. I know now I couldn't, and I was always skeptical about it.
Lossy format is Lossy - Whatever the format or codec is when talking about packing original wave to 1:8 - 1:15 at size (Even the original Wav can be lossy, if using bad rippers etc..)
So at our "Digital Age" the original music (at concerts, live, studios...) is always analog and we hear every point of it as serious of sinus wave-forms mixed together, the quality in our head is depending of the condition of our ears, our age and the way we orient to the source. This is real music, live.
Then we go to studios with quality analog tapes or digital samples taken with HQ microphones and ADC-Converted into some HD. And we begin to mix them, join end enhance more and more using computers nowadays. This is the start of getting to lossy formats or could we say "no more original" sound. We say that it's even better than the original after all studio kicks, but that's not true : it only sounds better, but it has come to artificial format which is after all screwing then pressed (burned) to CD's, DVD's etc.
Then the artificial music reaches our homes, we put the CD to a player : after first made ADC at studios it now goes through transistors, condensators, resistors, diodes and finally DAC (Digital to Analog Converter), vise versa at 16 to 24 bits samples generally - so we hear "analog" music with samplerates from 44100khz to 192000khz and even more if having a Goooood Stereos (or soundcard at PC)...
Now - we want to keep the music at our PC's or CD's as much compressed as possible to get our collection not to fill whole living room (and share it with others with KaZaA, FileShare, DC++ etc with reasonable speeds - so the size matters there). Now most of us have transferred ourselves to ADSL or T1/T3 and BW's 512kbps to 10 megs or over...and still we want more and more smaller files with "not so lossy" formats like MP4 (AAC). The Q is rising after all : WHY?
Can anyone answer me thruthfully, where we need all those new techniques. Is it just a trend of this time when all the other things too are evolving so fast and ppl like to study, research, invent all new and try to "push the limits" to "atom level"?
For example five Years ago a 80 gigs HD was SCSI-interfaced and costed BUCKS - now we get slim noiseless small ATA166 HD's let's say 250Gigs with $200 and after a year God knows how much and how cheap...We are in kinda merrygoround with all this and that. We buy something and after a year (or half) it's old and others say : "U still using those caveman-products - lookie what I got..."
To my point : I understand that MP3 and Napster were founded at last century and with 28k modems ppl shared 64k to 96k MP3's "near transparent (then)" and the smaller the piece the faster You got it! (10-30 min per title). Nowadays this is not the point anymore. I know that many still have 56k modems, because lack of money or just the point that they live at countryside ("Outside Nutbush City Limits") or just a place where the technology is still unreachable...
My final point : I have after all studies decided to make VBR MP3's (from 160k to 256k) with best possible encoders and best possible quality and ripping speed makes me nothing : I can get 28x to 40x realtime MP3-RipEncode with Xing or Gogo, but I use L.A.M.E (which is very flexible, thanks to You makers of it..) with best Q-level 1 ABR ot VBR (which makes LF's and HF's sound better). The realtime ripping goes with 5x to 10x and I can wait (and have two PC's...)
Still I am NOT saying NO to AAC (MP4) - it has it's points for example when speaking of 5.1 chan and AC3 mixing / Splitting / rendering to DVD-Video etc...
Just have to make some "twice a work" and that's when the lossy is coming more lossy!
I'm not an APE - so I'm not using it (Except at my Chinese Horoscope...)
Can anyone answer me thruthfully, where we need all those new techniques. Is it just a trend of this time when all the other things too are evolving so fast and ppl like to study, research, invent all new and try to "push the limits" to "atom level"?
Easy - this had been discussed dozens of times on HA :
Bandwidth is $, one MB is $ and reducing this, reduces costs and enables broader range of service to be transmitted.
Sure, you can store PCM wave files on future 1 TB portable players, but it does not remove the networking and streaming problem, which is the core reasons why most of the coding solutions were invented.
What if you want HDTV resolution with 5+1 channel hi-fi audio to be streamed to your computer over a cheap internet connection? Would you use uncompressed stream?
Ok, let's calculate...
1920x1280x60 FPS RGB = 421 MB/s = 3368 Mbit/s
5.1 audio @ 48 kHz = 4.39 Mbit/s
I guess you are not going to see this in your home anytime soon, and the cost for the provider alone to stream audio only with 4.39 Mb/s is way too big.
HE AAC and MPEG-4 AVC could down this to ~2-3 Mb/s peak which is acceptable and cost efficient.
Yes, Not Waves (My Goddesss) - as my conclusion I told my compromize would be a good MP3 now and a good say AAC after 1 Y, when the technology around it has evolved enough. Just waiting a little - still testing among others...
Steaming audio? We all do not need it and I think that streaming Video is taking much more Space outta the lines (if we are kidding with 48100 khz audio broadcasts - aren't we?)
And the bottlenecks are changing all the time - now it's Net, tomorrow it could be something else. As when transferring ourselves to neural networks, only the cables at our houses are "bottlenecks" then.
So after all the transfer-speeds rises all the time, but we seem to need smaller and smaller files (more lossy than 192k VBR MP3 [or 96k MP3Pro]) . Odd?
I think that part of this after all is a trend. For example ppl are working faster and harder and on the edges of their capasities all the time (mostly mentally). We are asked to do alone what 3 to 5 men did before. Burnouts, alcoholism, drugs....
Now 1 "man-year" produces more and more. Neural networks and PC's at near future - no more transmission limits - new howses are build with cables, satellites etc to accept the future...TB's HD's, GB's transfer speeds... that's only a start of our future and it comes soon with this logarhitmic speed (who want's to go to moon anymore - "anotherone bites the dust"...).
We are digitalizing more and more (I got yesterday two long portscan-attempts from the "Godollo University of Agricultural Sciences"....Where are we going my friends?
My final point : I have after all studies decided to make VBR MP3's (from 160k to 256k) with...... best possible quality...
... with best Q-level 1 ABR ot VBR...
AFAIK, the settings that produce LAME mp3s with the highest possible quality are the --alt-presets, and IIRC, they use a q-setting of 2.
Yes, Not Waves (My Goddesss) - as my conclusion I told my compromize would be a good MP3 now and a good say AAC after 1 Y, when the technology around it has evolved enough. Just waiting a little - still testing among others...
Why has MPEG adopted AAC, is MP3 "enough" - no, why?
- MP3 does not support multichannel (MPEG-2 Layer III does, but in a very crippled way)
- MP3 does not support all industry needed sampling rates (64, 96 kHz...)
- MP3 has algorithmic drawbacks (M/S issue, sfb 21 issue, itd...)
- It is possible to achieve MP3 quality at less space with advanced algorithms
- Framing of MP3 is not that good, AAC even support 960-sample framing to fit better in some telco equipment
Steaming audio? We all do not need it and I think that streaming Video is taking much more Space outta the lines (if we are kidding with 48100 khz audio broadcasts - aren't we?)
Well, the industry is progressing - maybe in 5 years there will be nice 192-320 Kbps 48 kHz 5.1 broadcasts for your car audio?
Or you would like to see HDTV movie in a plane, or interactive TV channel where you could be able to select what kind of programming you want.. this will not be possible without advanced compression and multimedia technologies, like MPEG-4, MPEG-7, MPEG-21...
And the bottlenecks are changing all the time - now it's Net, tomorrow it could be something else. As when transferring ourselves to neural networks, only the cables at our houses are "bottlenecks" then.
"Neural Networks" ? - I thought neural network is something else not a transport medium
So after all the transfer-speeds rises all the time, but we seem to need smaller and smaller files (more lossy than 192k VBR MP3 [or 96k MP3Pro]) . Odd?
No.
Not everybody has the golden ears, or the necessary high quality playback hardware, or the HD space, to only need high bitrate encoding. Low bitrate codecs play their role in saving HD space for those who need it.
Now 1 "man-year" produces more and more. Neural networks and PC's at near future - no more transmission limits - new howses are build with cables, satellites etc to accept the future...TB's HD's, GB's transfer speeds...
There will be observable bottlenecks in the years to come, and as long things are not instantaneous, there will always be ever-increasing speeds.
As for your philosophical musings, I appreciate them, especially about the part about working like mad for practically nothing... but they are off-topic. I suggest that you start your musings in the Off-Topic subforum, please.
AFAIK, the settings that produce LAME mp3s with the highest possible quality are the --alt-presets, and IIRC, they use a q-setting of 2.
Yes - that was a "Press Error" - I ment preset 2, but pressed 1 - that's what I use....
I'm still a bit studying IIRC (with my ears - they are not so well as they used to be as I was under 40...). But I like to listen good music (with my 6.1 SB Audigy II Platinum with big and good Speakers), not just collecting it!
Btw : thkx 4 correcting my "visual error"
Btw2 : Nice to get good feedback with story made in 10 mins for just relaxing my fingers. I had no means to hurt anyone's thoughts, because we all have at least one - and it's different than those others. Philosophic? Yes - off-topic, but always interesting - no matter where it starts - why not at AAC-Thread?
Or you would like to see HDTV movie in a plane, or interactive TV channel where you could be able to select what kind of programming you want.. this will not be possible without advanced compression and multimedia technologies, like MPEG-4, MPEG-7, MPEG-21...
I do not fly a lot - ok I was not blaming new compression methods, just rising some critisism and Q's around them - MPEG-21 -lmao....
(P.S Do I smile too much ...? Just wondering by myself...)
We buy something and after a year (or half) it's old and others say : "U still using those caveman-products - lookie what I got..."
LOL!!!
btw, have you seen my Hi-Fi?
(http://www.david.robinson.org/pics/medgram.gif) (http://www.david.robinson.org/pics/gramophone1.jpg)
Seriously, if you look at the trend in digital radio and TV: bandwidth is finite, but people want to put more and more content through it. So, better codecs are very useful.
Trust me - I know - I live in a country where digital radio = MPEG 1 layer 2, and the average bitrate has been pushed down to 128kbps to fit more stations in. (and the content is transcoded from MPEG 1 layer 2 based servers!)
If you're in the USA, you can enjoy DirectTV (is that what it's called?) where the bitrates are low, and many channels are supposedly using 3/4 or 1/2 NTSC resolution, just to fit them all in.
So better codecs would be very useful. Though in broadcast, the MPEG-1 and -2 hardware is so common that it will take a long time for it to be superseded - and MPEG-4 might be out of date by then.
Cheers,
David.
Can anyone answer me thruthfully, where we need all those new techniques. Is it just a trend of this time when all the other things too are evolving so fast and ppl like to study, research, invent all new and try to "push the limits" to "atom level"?
Because MP3 is far from perfect:
http://www.mp3-tech.org/content/?Mp3%20Limitations (http://www.mp3-tech.org/content/?Mp3%20Limitations)
Because MP3 is far from perfect:
- Joint stereo limitations - I do not use Joint Stereo!
- Too limited maximum frame size - I can do with that technical issue and bit reservoir...
- Inoptimal window sizes - Corrected at AAC, but not much for my ears....
- Scalefactor band 21 problem - sfb21 inflating - could be a bit better
- Hybrid transform scheme - some noise that can not be totally removed. plain
MDCT Corrects this and has not to be "backwards compatible" with layer II.
- Unwanted Echo with some combinations - I can hear it only if having some DSP on (say DFX for wa2, bmbn for wa3 etc...)
- Mixed blocks limitation - ISO standards, ISO standards - Not my ear....
- Missing :
1. PNS (perceptual noise substitution)
2. SBR (spectral band replication)
3. TNS (temporal noise shaping)
4. LTP (long term prediction)
Q: Does Ogg Vorbis or AAC have them all and can I hear it somehow (I mean in a real world - not at tests on white papers)?
Nothing bad guys - just thowing back A's (and my Q's)...
- Joint stereo limitations - I do not use Joint Stereo!
I had a feeling you'd say that when you talked about quality mode in Lame.
Have you read the FAQs?
- Too limited maximum frame size - I can do with that technical issue and bit reservoir...
But, even if you don't care about compression ratio, this point is important. It means it's impossible to encode some sounds properly with mp3.
- Hybrid transform scheme - some noise that can not be totally removed. plain
MDCT Corrects this and has not to be "backwards compatible" with layer II.
mp3 has the filter bank. If you don't use the filter bank, it's not mp3, therefore it's a new codec. A totally incompatible one, but one step on the way to AAC.
Rather than asking why we should be interested in new codecs, why don't you consider: Why should we use old, imperfect, inefficient ones?
Cheers,
David.
Rather than asking why we should be interested in new codecs, why don't you consider: Why should we use old, imperfect, inefficient ones?
Who told U that You should? Me? No - I was just talking about my habbits now and in the meanwhile waiting the AAC to evolve a bit more - It's not a ready codec by anyhow - but at a very interesting level evolving to a one very conciderable at near future!
I'll wait eagerly Ahead's new version and in the meanwhile test FAAC/FAAD combination and the new frontends which make them "Windows-Like"....
There's so much even at sourceforge.net to test and the free nero plugins at neroplugins.cd-rw.org etc....
I was just talking about my habbits now and in the meanwhile waiting the AAC to evolve a bit more - It's not a ready codec by anyhow - but at a very interesting level evolving to a one very conciderable at near future!
I'll wait eagerly Ahead's new version and in the meanwhile test FAAC/FAAD combination and the new frontends which make them "Windows-Like"....
What do you mean AAC is not ready codec? There are many different AAC encoders and many of them seems to be ready. AAC itself has been ready for some time now.
In latest Nero there is new AAC encoder, but how long are you going to wait? And for why? Are you telling us that if encoder has room to improve, it's not ready? Do you think that Lame is ready?
- Joint stereo limitations - I do not use Joint Stereo!
You should Joint stereo is integrated in order to get rid of some stereo unmasking problems, and not just to reduce the bit rate as many unexperienced people think.
- Too limited maximum frame size - I can do with that technical issue and bit reservoir...
No you can't - limitation is here to stay, and there is no way to remove it without breaking the MP3 standard. Because of this limitation, some files will never be transparent.
- Inoptimal window sizes - Corrected at AAC, but not much for my ears....
Inoptimal window size generates 2 problems:
1. Some samples can't be transparent at all (fatboy, for example)
2. Tonal and steady samples are coded with more bits than needed
- Scalefactor band 21 problem - sfb21 inflating - could be a bit better
Also a problem that is impossible to solve without breaking the MP3 spec, or heavy overcoding the stream.
- Hybrid transform scheme - some noise that can not be totally removed. plain
MDCT Corrects this and has not to be "backwards compatible" with layer II.
Also impossible to solve in MP3... PQF filterbank is useless and just introducees additional noise - removing PQF and setting the filterbank to the proper size leads you towards AAC
- Unwanted Echo with some combinations - I can hear it only if having some DSP on (say DFX for wa2, bmbn for wa3 etc...)
I can hear unwanted echo without DSP plug-ins, and many others can do the same as well...
- Mixed blocks limitation - ISO standards, ISO standards - Not my ear....
Not yours.. but ...
- Missing :
1. PNS (perceptual noise substitution)
2. SBR (spectral band replication)
3. TNS (temporal noise shaping)
4. LTP (long term prediction)
Q: Does Ogg Vorbis or AAC have them all and can I hear it somehow (I mean in a real world - not at tests on white papers)?
AAC has PNS, SBR, TNS and LTP - Nero AAC has PNS, SBR and TNS ... they make obvious difference in their relevant application areas.
Who told U that You should? Me? No - I was just talking about my habbits now and in the meanwhile waiting the AAC to evolve a bit more - It's not a ready codec by anyhow - but at a very interesting level evolving to a one very conciderable at near future!
AAC has 4 mature implementations: Nero, Fraunhofer, Dolby and QuickTime - all of them are able to beat MP3 at the same bit rate.
What do you mean AAC is not ready codec?
In this world there are so many "ready" things - name one. Nothing comes to mind....?
And as I told I'm NOT just waiting, testing all the time and just now installing Foobar v0.7 and AAC for it (I know, that one of the winamp guru went to "foo" and made the "bar" and it has become quite a player nowadays). Just the DFX's missing, but there are many choices for that as seen at http://www.foobar2000.org/components.html (http://www.foobar2000.org/components.html) .
So - @ first there's a spec, then comes codec and after that various player-makers etc adopts the codec (or buids an own release as many seems to do all the time - invent the wheel again and again). as in this quote :
AAC has 4 mature implementations: Nero, Fraunhofer, Dolby and QuickTime
and we could set a few "lonely riders" among them trying to make it "Moneyless" (as the hackers have done to Nero AAC.DLL up to 6.0.0.15 and will well continue...)
Just as you said :
There are many different AAC encoders and many of them seems to be ready
, that means competition and variousness and by that way not ready at all...(when will PC be ready? Or Windows - our GatesWarez whole-life beta since 1991 (or as MS-DOS since 1982 [or false said 1985 at winver.exe] )...
Nothing is ready in this world, my friend and that is just the thing that makes life interesting. If all should be - what should we do - use the "steem-radios" to listen LM, AM broadcasts with amplitude modulation.
OK - off-topic - AAC is well - I hope all the good future to it (as for L.A.M.E, which aint mpeg encoder)
And for Ogg ((Un)patent(less)), which did not get any more points than MP3 (after all this IS AAC thread...)
Just as you said : There are many different AAC encoders and many of them seems to be ready
, that means competition and variousness and by that way not ready at all...(when will PC be ready? Or Windows - our GatesWarez whole-life beta since 1991 (or as MS-DOS since 1982 [or false said 1985 at winver.exe] )...
Nothing is ready in this world, my friend and that is just the thing that makes life interesting. If all should be - what should we do - use the "steem-radios" to listen LM, AM broadcasts with amplitude modulation.
Hunh?
So the only things in this world that are "ready" are things which have reached perfection-of-concept and cannot as such be improved upon? OK...the only things in the world that are "ready" are dirt and rocks. Nothing else is ready, by that standard.
The fact that AAC has multiple implementations means that it is not only "ready" but well evolved. A codec begins with only one implementation as a "proof-of-concept". Once it's concept is proven, then it's ported as desired (and as allowed).
And if there's competition, that means it's not "ready"? That's a completely invalid implication. Different implementations of AAC are not actually "competition" because the different implementations
compliment each other. E.g., if compatiblity issues of my platform restrict me from using QuickTime, then I have three other choices for AAC encoding. Either way, I'd still be using AAC, right?
Ivan has already explained the "readiness" of AAC concerning features and capabilities. As for sound quality, AAC is not only "ready", it's in
first place at a particular bitrate in a recent formal listening test (http://www.hydrogenaudio.org/forums/index.php?showtopic=13464).
Edit: tpyos...and clarification...
- Joint stereo limitations - I do not use Joint Stereo!
If you don't use Joint Stereo on your mp3s at high bitrates, you are
not encoding them with the highest quality possible. I suggest you take the --alt-presets as is, and don't bother adding -m s behind them.
Quality of MP3 at 50% the space. With High Efficiency AAC in MPEG audio, you will get the same quality of a MP3 at 50% the space
Once again another incorrect statement as rjamorim's listening test (see the first link above) proves. 64kb/s HE-AAC failed to equal the same quality of LAME MP3 at 128kb/s. I'm beginning to see a pattern of lies forming here and I'm becoming very tired of refuting them .
how true but you talk about 64kb .... go and use 96 the compare it to 192MP3
http://www.warp2search.net/forums/viewtopic.php?t=2614 (http://www.warp2search.net/forums/viewtopic.php?t=2614)
and I dunno which nero thingy u use but if you use the package linked above I can choose 300+ kB which is more then enough for ANY possible sound out there ...