Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Why Continue MP3 Development Given AAC? (Read 59860 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Why Continue MP3 Development Given AAC?

Reply #50
I have a question who really has performed ABX test to see if there was an improvement for LAME 3.97/3.98/3.99 for the same bitrate.

AFAIK each new version boosts more bitrate on difficult parts. That makes uncomparable 3.97 -V5 vs  3.98 -V 5. 
3.99 -V0 is better than 3.98.4/3.97 -V0 because it has higher bitrate.  It's ok as V0 targets to highest quality close to max 320 kbps.

But the question is: were there  any substantial improvements while keeping the same bitrate?

I remeber Guruboolez has compared 3.98 and 3.97 http://www.hydrogenaudio.org/forums/index....st&p=460922

To refresh the minds 3.97 beta 2 was realesed in 2005. 6 years ago.

And also some japanese guy http://d.hatena.ne.jp/kamedo2/20111214/1323849399

I really like very much LAME project but I can't see clear improvements without increasing bitrate (since 3.97).

Why Continue MP3 Development Given AAC?

Reply #51
I know of no devices that won't play AAC.
The audio unit in my car will not play AAC. It will, however, play MP3.
lossyWAV -q X -a 4 -s h -A --feedback 2 --limit 15848 --scale 0.5 | FLAC -5 -e -p -b 512 -P=4096 -S- (having set foobar to output 24-bit PCM; scaling by 0.5 gives the ANS headroom to work)

Why Continue MP3 Development Given AAC?

Reply #52
I know of no devices that won't play AAC.
The audio unit in my car will not play AAC. It will, however, play MP3.


I suspect your head unit is not of recent manufacture(?) If it was, (I believe) not supporting AAC (like music sold in the iTunes music store) would be a glaring flaw and possible commercial suicide.

(I should have worded that differently. If you notice, I did acknowledge the existence of players that don't support AAC in my second paragraph. I meant I have no experience with such players. This is why I followed the sentence you quoted with those player in which I have experienced.)
The Loudness War is over. Now it's a hopeless occupation.

Why Continue MP3 Development Given AAC?

Reply #53
.....
I really like very much LAME project but I can't see clear improvements without increasing bitrate (since 3.97).

Sure there was improvement.
Trumpet for instance was real bad when using VBR up to 3.97. 3.98 improved significantly on it. Same goes for the 'sandpaper noise' issue. Just to talk about problem samples I care about. IIRC eig improved too.
Improvements like these don't necessarily bring average bitrate up. AFAIK average bitrate for the various -V levels did not significantly vary from version to version before 3.99. For 3.99 it's by design, but even then for -V levels around -V3/-V2 average bitrate is very close to that of 3.98.
Of course individual birate of specific tracks can vary significantly reflecting changes in the psy model.
lame3995o -Q1.7 --lowpass 17

Why Continue MP3 Development Given AAC?

Reply #54
I know of no devices that won't play AAC.

I know plenty. Neither of my DVD players, which are 5-10 years old, play AAC. (One of them plays OGG Vorbis, tho, which is kinda interesting.)
My Nokia phone does play AAC, but not gaplessly. The program I use for playing music gaplessly on it doesn't support AAC.
Many PMPs I see don't play AAC, even new ones. For example the Phillips MP3 players and MP4 players only play MP3, WAV, WMA, FLAC, APE, according to the specs.

I don't have any hard numbers, but I'd say quite a significant number of devices currently in existence don't support AAC, but pretty much all of them support MP3. You have to take into account devices released some years ago as well, they are still relevant.
On sound quality.. I only did some limited testing myself.. while I think AAC does sound better at 128kbps (which makes it very attractive for some uses), I'm not sure there's an audible advantage over MP3 at 200kbps (or V2 bitrates), which is kinda the standard bitrate used for music nowadays.

Why Continue MP3 Development Given AAC?

Reply #55
Many PMPs I see don't play AAC, even new ones. For example the Phillips MP3 players and MP4 players only play MP3, WAV, WMA, FLAC, APE, according to the specs.


Sheesh, how could Philips brand a PMP as MP4, if it doesn't play AAC (it does play MP4 vids)? That verges on misleading advertising.

On MP3 development, my sense from reading the tests is that while LAME hasn't improved much in the last few years, other MP3 codecs have caught up a lot of ground. Specifically, IIRC, the iTunes MP3 codec used to do badly, but in the last comparative test did well.

Why Continue MP3 Development Given AAC?

Reply #56
Sure there was improvement.

Not without rising the bitrate.


Trumpet for instance was real bad when using VBR up to 3.97. 3.98 improved significantly on it. Same goes for the 'sandpaper noise' issue. Just to talk about problem samples I care about. IIRC eig improved too.

Trumpet (sample14 from last public test):
3.97 V0 VBR NEW -  226 kbps
3.98.4 V0 - 258 kbps
3.99.3 - 287 kbps


eig sample:
3.98.4 V0 - 249 kbps.
3.99 V0 - 295 kbps

Average bitrate of V0 has gone way up for 3.99 on my set of albums.

Quote
Improvements like these don't necessarily bring average bitrate up.

http://www.hydrogenaudio.org/forums/index....st&p=772362
The bitrate has gone up and I failed to find at least one post where You or any other person have submit results for 3.98.4 -V0 and 3.99 -V0.x (like -V 0.25 - 0.5 or so) for the same average bitrate on large collection of albums.

Until now I have found only the results (>15 samples) of Japanese guy. And conclusion is pretty clear: NO improvement for V2 (LAME 3.99 vs 3.98.4).

Why Continue MP3 Development Given AAC?

Reply #57
Well, there has also been a very nasty mess in the patent arena for AAC that could yet dwarf the mess for MP3 issues.

It is interesting, a 1992 paper by Johnston and Ferierra (sp?) in the ICASSP has, effectively, the guts of AAC (MPEG-2 kind) inside of it, except for the issue of TNS (which is indeed a valuable addition) and the non-uniform quantizer, which is ok for low rates, but a major, disastrous inefficiency for higher rates.

Funny how long it's been, almost 20 years.
-----
J. D. (jj) Johnston

Why Continue MP3 Development Given AAC?

Reply #58
Sure there was improvement.

Not without rising the bitrate.
....

I was talking about quality improvements and average bitrate of 3.98 vs. 3.97 at this point. Does your experience differ from mine?

As for 3.99 I think I was the first to give an average bitrate table for various -V levels and for 3.99 and 3.98. As I wrote above 3.99 behavior is diffeent here by design.
lame3995o -Q1.7 --lowpass 17

 

Why Continue MP3 Development Given AAC?

Reply #59
Can You provide the results where You have compared different version of LAME with setting those provide the equal bitrate on quite/enough big set of albums?

Because we're talking about some of your findings/post/topic (?)  and I can't find them.

Thank You.

Why Continue MP3 Development Given AAC?

Reply #60
Cell phones in Japan used to support AAC in 3gp/3g2 format, but not MP3. Probably there were none supporting pure MP3... I don't know. They were selling heavily DRM-ed music for phones since 2002 or so, at very high price. Apparently they needed some container format for that purpose.
Panasonic Blu-ray recorder (DIGA) and Nintendo DS-i also comes in mind as such.
In Japan, AAC is the codec of digital broadcasting. Therefore, every video device must support AAC anyway.

Having said that, MP3 seems still much more popular in Japan, too ... at least as a codec for music.

Why Continue MP3 Development Given AAC?

Reply #61
I know of no devices that won't play AAC.
The audio unit in my car will not play AAC. It will, however, play MP3.


I suspect your head unit is not of recent manufacture(?) If it was, (I believe) not supporting AAC (like music sold in the iTunes music store) would be a glaring flaw and possible commercial suicide.


My new (2012) car will play only MP3 and WMA.
And I don't think they are going out of business just yet.

Why Continue MP3 Development Given AAC?

Reply #62
My new (2012) car will play only MP3 and WMA.
And I don't think they are going out of business just yet.


Can you connect an iPod to your car's stereo?
The Loudness War is over. Now it's a hopeless occupation.

Why Continue MP3 Development Given AAC?

Reply #63
My new (2012) car will play only MP3 and WMA.
And I don't think they are going out of business just yet.


2009 Toyota.  MP3 / WMA.

Quote from: Engelsstaub link=msg=0 date=
Can you connect an iPod to your car's stereo?
only through the line-in jack or via an optional > $100 adapter/
Creature of habit.


Why Continue MP3 Development Given AAC?

Reply #65
I was only wondering if it was possible. If it supports the iPod then it would have AAC playback capabilities and I wouldn't understand why it wasn't possible with the USB stick.

Based on replies from you and Soap, I understand that some manufacturers of car stereos don't support AAC. That's just stupid (IMO) when the majority of legally-purchased music is obtained through iTunes. It's even stupider (IMO) that they seem to think more people would be concerned about WMA playback.

As for me, I just use AAC for my lossy-encoding needs. I've nothing against MP3...it has some clear advantages. I just like the uniformity in my iTunes library and have to use it when encoding video as well.

Every device I own (and that's quite a few if I count kids and phones) supports it. I have an older Alpine stereo in my car. It supports AAC and has an iPod interface. Perhaps if my circumstances and usage were different, I'd be using MP3 instead. It would certainly be "just as good" as I only encode at "iTunes Plus" bitrates or 160 Kbps for video.
The Loudness War is over. Now it's a hopeless occupation.

Why Continue MP3 Development Given AAC?

Reply #66
That's just stupid (IMO) when the majority of legally-purchased music is obtained through iTunes. It's even stupider (IMO) that they seem to think more people would be concerned about WMA playback.


Most people who purchase music from iTunes use ipods.  Since you have to pay to include AAC support in hardware devices, a lot of companies simply don't bother. 

Every device I own (and that's quite a few if I count kids and phones) supports it. I have an older Alpine stereo in my car. It supports AAC and has an iPod interface. Perhaps if my circumstances and usage were different, I'd be using MP3 instead. It would certainly be "just as good" as I only encode at "iTunes Plus" bitrates or 160 Kbps for video.


Yes, AAC is fine if you stick with the Apple ecosystem.


Why Continue MP3 Development Given AAC?

Reply #68
Yes, AAC is fine if you stick with the Apple ecosystem.


I thought one of the things established in this thread was that AAC is NOT an Apple proprietary format. But obviously, in some eyes, it is for ever tainted by being the format chosen for iTunes.
(Don't bother, I know you didn't SAY that, and I know MP3 is more universal, but it would be nice to keep the AAC = Apple = DRM = teh ebbil set of associations out of it.)

Why Continue MP3 Development Given AAC?

Reply #69
Pretending Apple never came into play, you don't think DRM is fair game in this discussion?

After all, it was presented as a reason why some devices don't support mp3.

Why Continue MP3 Development Given AAC?

Reply #70
and have to use it when encoding video as well.

http://en.wikipedia.org/wiki/Blu-ray#Audio

AAC is not mentioned there, so what devices are you talking about? I presume something from Apple? Furthermore you mentioned Handbrake earlier which uses the outdated FAAC encoder (the PC version at least) and probably is worse than LAME.

Why Continue MP3 Development Given AAC?

Reply #71
...so what devices are you talking about? I presume something from Apple? Furthermore you mentioned Handbrake earlier which uses the outdated FAAC encoder (the PC version at least) and probably is worse than LAME.


Me, earlier in this same thread: "An interesting side-note: every popular device I know of, for playing ripped videos (using Handbrake, etc.,) requires the audio to be mixed down in the AAC format. This includes the XBox 360, Playstation 3, and others. I would hate to have encoded my rips with MP3 audio and then try to find something other than VLC that will play it."

Core Audio/Mac OS...not FAAC. None of the devices I listed or implied were made by Apple except the iPod. XBox is Microsoft (so is my daughter's Zune HD, which will not play a video encoded with MP3 audio either.) PS3 is Sony.
The Loudness War is over. Now it's a hopeless occupation.


Why Continue MP3 Development Given AAC?

Reply #73
I mean the comparison of different versions of LAME in quality terms with setting those provide the equal bitrate on quite/enough big set of albums.

My question still stands: were there any substantial improvements of LAME  while keeping the same bitrate in last 5 years (3.97/3.98/3.99)?
Please, direct link to the proof.

Why Continue MP3 Development Given AAC?

Reply #74
When coupling with H264, usually AAC is used (since it is the defacto, standard MPEG4 audio codec).
Youtube videos are currently H264+AAC or WebM(VP8+Vorbis). Everybody uses H264+AAC for video. Therefore, it's a must-have in the recent video devices.
In fact, Adobe flash was using H263+MP3 in the past. However, since they have adopted H264+AAC, it has become popular choice very rapidly. It's probably because H264 is efficient enough to deprecate older video codecs.

Of course AAC itself has good points from technical point of view. It has multichannel support, and efficient at lower bitrate. Therefore, much suited for video streaming, where the bandwidth is limited.
However, (personally) I regard the adoption of AAC in the video world is mainly a by-product of H264.