Skip to main content

Topic: HE-AAC v.1 & v.2 comparison (Read 78064 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • Garf
  • [*][*][*][*][*]
  • Developer (Donating)
HE-AAC v.1 & v.2 comparison
Reply #25
Quote
Qs: So at the current state of development, what would be the use of HE-AAC v2 for an audiophile, and if so what bitrate would be transparent?
[a href="index.php?act=findpost&pid=325239"][{POST_SNAPBACK}][/a]


HE-AAC (and HE-AAC v2) are essentially useless for "transparent" encodes.

  • nyaochi
  • [*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #26
Quote
is there any technical reason for this exclusion of 44.1?
[a href="index.php?act=findpost&pid=325251"][{POST_SNAPBACK}][/a]

I don't think so. I presume that the source code is only for reference purpose. I found the code for 44.1kHz in many places. But the mapping table to determine encoding parameters for given sample rate, stereo mode, and bitrate does not have entries for 44.1kHz. We can easily add 44.1kHz support by copy-and-pasting entries for 48kHz.
  • Last Edit: 05 September, 2005, 04:26:20 PM by nyaochi

  • nyaochi
  • [*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #27
And I found an interesting sample that exposes a clear flaw of 48kbps (SBR+PS).

"Wish I" in album "Finally Woken" by Jem.

original
aacPlus v2 48kbps (SBR+PS)
aacPlus v2 48kbps (SBR)

I was astonished to listen to the SBR+PS sample. It eliminates the sub melody (played by the classic guitar?) from the right channel too much that it's like a different session. I don't complain of other artifacts for this bitrate, but it's not acceptable for me that an encoder changes the volue valance so drastically. SBR without PS sample does not suffer from this problem. We may have to be familar with the original sample to notice this kind of problems.

  • bond
  • [*][*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #28
Quote
Quote
is there any technical reason for this exclusion of 44.1?
[a href="index.php?act=findpost&pid=325251"][{POST_SNAPBACK}][/a]

I don't think so. I presume that the source code is only for reference purpose. I found the code for 44.1kHz in many places. But the mapping table to determine encoding parameters for given sample rate, stereo mode, and bitrate does not have entries for 44.1kHz. We can easily add 44.1kHz support by copy-and-pasting entries for 48kHz.
[a href="index.php?act=findpost&pid=325255"][{POST_SNAPBACK}][/a]
sounds evil, i guess this also answers the question whether this reference code is the same as the commercial one sold by coding tech 
I know, that I know nothing (Socrates)

  • Loke
  • [*][*]
HE-AAC v.1 & v.2 comparison
Reply #29
hmm...
In the AAC+ v1 & v2 encoder from Coding Technologies isn't the LC-AAC part of it made by fraunhofer?

I seem to remember this was the case with their mp3PRO-encoder, when it came 4 years ago. It was a normal lowbitrate mp3-stream, with additional information for the high frequencies (SBR).
The lowrate mp3 part was made with a fraunhoferencoder, while SBR-part was made by CT.

Is this true?

Now winamp's got two AAC encoders; one by Dolby, and now one made by CT.

  • Enig123
  • [*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #30
The HE-AAC encoder included in Helix can encode to bitrate up to 128kb. I wonder if HE will help in such bitrate.

Also it was reported somewhere in doom9's forum Helix's LC-AAC encoder performes very well @ bitrate of 192 kb. Would someone test it at such high bitrate?

  • Digisurfer
  • [*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #31
This was an exciting test, thanks Guru! I'm really looking forward to seeing how Nero's soon to be released aacPlusv2 encoder stacks up after seeing this test. It's been a while since I first started taking an interest in HE-AAC, but a year later it's starting to look like things are finally beginning to heat up. Now to get all the audio player and tagger developers on the same page as far as support goes.

Edit: Would be cool if mp3pro was thrown into with aacPlusv2 tests if there is another one in the future.
  • Last Edit: 05 September, 2005, 11:04:06 PM by Digisurfer

HE-AAC v.1 & v.2 comparison
Reply #32
Quote
The HE-AAC encoder included in Helix can encode to bitrate up to 128kb. I wonder if HE will help in such bitrate.

Also it was reported somewhere in doom9's forum Helix's LC-AAC encoder performes very well @ bitrate of 192 kb. Would someone test it at such high bitrate?
[a href="index.php?act=findpost&pid=325298"][{POST_SNAPBACK}][/a]


Are the Helix LC and HE AAC encoders freely available for download?
If yes, could someone provide a download link for the latest version?

And another question: does anyone know of an AAC-capable mobile
phone that can play back HE AAC streams? I'm planning to buy the
SonyEricsson W800 Walkman phone and would like to know if HE AAC
is supported.

Thanks
Wanna buy a monkey?

  • bond
  • [*][*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #33
Quote
Are the Helix LC and HE AAC encoders freely available for download?
If yes, could someone provide a download link for the latest version?

https://helix-producer.helixcommunity.org/downloads.htm
I know, that I know nothing (Socrates)

HE-AAC v.1 & v.2 comparison
Reply #34
Should I download the Command Line Application or the SDK?

I can't find the version Guruboolez used (v.11).
  • Last Edit: 06 September, 2005, 05:04:46 AM by yourtallness
Wanna buy a monkey?

  • dimzon
  • [*][*][*]
  • Banned
HE-AAC v.1 & v.2 comparison
Reply #35
Quote
Should I download the Command Line Application or the SDK?

I can't find the version Guruboolez used (v.11).
[a href="index.php?act=findpost&pid=325348"][{POST_SNAPBACK}][/a]

looking for it too...

  • kurtnoise
  • [*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #36
Quote
Should I download the Command Line Application or the SDK?

I can't find the version Guruboolez used (v.11).
[{POST_SNAPBACK}][/a]


[a href="https://helixcommunity.org/beula/download/]https://helixcommunity.org/beula/download/[/url] . register needed but it's free.. 

CLI is enough but keep in mind that the AAC stream is embedded in the rm container. However, Guru uses an other way...  The AAC streams are directly embedded in mp4 file.

  • dimzon
  • [*][*][*]
  • Banned
HE-AAC v.1 & v.2 comparison
Reply #37
Quote
Quote
Should I download the Command Line Application or the SDK?

I can't find the version Guruboolez used (v.11).
[{POST_SNAPBACK}][/a]


[a href="https://helixcommunity.org/beula/download/]https://helixcommunity.org/beula/download/[/url] . register needed but it's free.. 

CLI is enough but keep in mind that the AAC stream is embedded in the rm container. However, Guru uses an other way...  The AAC streams are directly embedded in mp4 file.
[a href="index.php?act=findpost&pid=325353"][{POST_SNAPBACK}][/a]

Can anybody provide command line sample?

  • bond
  • [*][*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #38
i think the helix encoder can output mp4 too, not sure
I know, that I know nothing (Socrates)

  • Enig123
  • [*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #39
The following information is originally from doom9's forum. I just forget the exact url. In this way make Helix Producer can output aac stream in mp4 container directly.

Code: [Select]
Codec Bitrate Label                 4cc flavor  Compatibility *)
48 Kbps Stereo - RA 10 with aacPlus racp  3     RealPlayer 10
32 Kbps Stereo - RA 10 with aacPlus racp  4     RealPlayer 10

64 Kbps Stereo - RA 10 with aacPlus racp  0     RealOne
96 Kbps Stereo - RA 10 with aacPlus racp  1     RealOne
128Kbps Stereo - RA 10 with aacPlus racp  2     RealOne


Keng!
11th June 2005, 07:45

From your RealPlayer folder <C:\Program Files\Real\RealPlayer\producer>, copy both [mp4wrtr.dll] in <plugins> folder & [hxfilewriter.dll] in <tools> folder to Helix DNA Producer directory, [mp4wrtr.dll] in <plugins> folder & [hxfilewriter.dll] in <tools> folder.

Now create an audience file: audio.rpad

<?xml version="1.0"?>
<audience>
<streams>
<audioStream>
<pluginName type="string">rn-audiocodec-realaudio</pluginName>
<codecName type="string">raac</codecName>
<codecFlavor type="uint">1</codecFlavor>
<streamContext type="bag">
<presentationType type="string">audio-only</presentationType>
<audioMode type="string">music</audioMode>
</streamContext>
</audioStream>
</streams>
</audience>

Then launch producer with following commandline:
    Producer.exe -i CDImage.wav -o CDImage.m4a -ad audio -dv -dt

Also work for .mp4 extension!

This is useful if you rip your AudioCD with Foobar2000 or CDex!
The commandline for Foorbar2000 should like this:
-i %s -o %d -ad audio -dv -dt
Remember to use mp4/m4a in the extension box!
So, no more mka -> aac -> mp4!
  • Last Edit: 06 September, 2005, 07:31:19 AM by Enig123

  • Assault
  • [*]
HE-AAC v.1 & v.2 comparison
Reply #40
@ Enig123

I've just tried encoding a wav file to an he-aac file but I only get an lc-aac file with 64kbps and a sampling rate of 32khz with your audience file (raac flavor 0) . Do you know the codec type and flavor I have to use to get an he-aac file? I tried racp flavor 0 but without success. Thanks for your help!

EDIT: Problem solved. I found racp.dll on https://helixcommunity.org/beula/download/ and now it works fine with racp flavor 0.
  • Last Edit: 06 September, 2005, 12:10:10 PM by Assault

  • JohnV
  • [*][*][*][*][*]
  • Developer
HE-AAC v.1 & v.2 comparison
Reply #41
Quote
I read somewhere that Nero 7 is announced for October. I guess that the new generation of Nero Digital HE-AAC will be released with the new Nero package and I’m sure that Ivan had built a competitive encoder.

Yeah, a bit bad time for us (or for you) for this test. The Nero Digital audio encoder you are testing here is now very much outdated.
During the last weeks, AAC-HE and PS have received very significant quality updates.
When we release the new one pretty soon, remember to redo this test.
  • Last Edit: 06 September, 2005, 02:49:51 PM by JohnV
Juha Laaksonheimo

  • rjamorim
  • [*][*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #42
Awesome test, as always, my friend.

You seem to be getting a taste for debunking age-old prejudices
Get up-to-date binaries of Lame, AAC, Vorbis and much more at RareWares:
http://www.rarewares.org

  • optimus
  • [*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #43
I've just integrated Helix Encoder support (as well as Winamp AAC+ encoder) in my MEnc, with all info I got in this thread, check it out here.



Download it here.
This includes:
MEnc executable
MPlayer (latest CVS version)
Helix Producer (stripped for audio encoding only and added aacplus audience files)
Winamp AAC+V2 encoder
Lame MP3 encoder (latest CVS version)
  • Last Edit: 07 September, 2005, 10:36:46 AM by optimus
MediaCoder the transcoder
http://mediacoder.sourceforge.net

  • dand
  • [*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #44
Quote
Quote
Quote
- is the 3gpp reference the same or a different encoder than the one coding technologies sells?

Probably with minor differences.
[a href="index.php?act=findpost&pid=325196"][{POST_SNAPBACK}][/a]


Some experimenting will quickly tell they're "not so minor" IMHO...
[a href="index.php?act=findpost&pid=325198"][{POST_SNAPBACK}][/a]

Garf, what are the main differences (spare us a time...)?

  • Oki
  • [*][*]
HE-AAC v.1 & v.2 comparison
Reply #45
Quote
- is the 3gpp reference the same or a different encoder than the one coding technologies sells?
Quote
Probably with minor differences.[a href="index.php?act=findpost&pid=325196"][{POST_SNAPBACK}][/a]
Quote
Some experimenting will quickly tell they're "not so minor" IMHO...
[a href="index.php?act=findpost&pid=325198"][{POST_SNAPBACK}][/a]
Quote
Garf, what are the main differences (spare us a time...)?[a href="index.php?act=findpost&pid=326576"][{POST_SNAPBACK}][/a]

The main differences should be speed optimizations and more options in the commercial one. I do not hear big differences in quality.

Regards,
Oki
  • Last Edit: 13 September, 2005, 04:09:45 AM by Oki

  • SirGrey
  • [*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #46
Interesting...
Nero lost to CT this time. Life is full of surprises 
Thanks a lot for your efforts, Guruboolez !!!

  • moozooh
  • [*][*][*][*]
HE-AAC v.1 & v.2 comparison
Reply #47
Cause Nero is simply outdated. We'll see the real winner when the long-awaited new version comes out.
Infrasonic Quartet + Sennheiser HD650 + Microlab Solo 2 mk3. 

  • Garf
  • [*][*][*][*][*]
  • Developer (Donating)
HE-AAC v.1 & v.2 comparison
Reply #48
Quote
Quote
- is the 3gpp reference the same or a different encoder than the one coding technologies sells?
Quote
Probably with minor differences.[a href="index.php?act=findpost&pid=325196"][{POST_SNAPBACK}][/a]
Quote
Some experimenting will quickly tell they're "not so minor" IMHO...
[a href="index.php?act=findpost&pid=325198"][{POST_SNAPBACK}][/a]
Quote
Garf, what are the main differences (spare us a time...)?[a href="index.php?act=findpost&pid=326576"][{POST_SNAPBACK}][/a]

The main differences should be speed optimizations and more options in the commercial one. I do not hear big differences in quality.

Regards,
Oki
[a href="index.php?act=findpost&pid=326577"][{POST_SNAPBACK}][/a]


I actually did mean quality. Can you try some PS problem clips? I might have had an older/newer encoder so sorry if I messed up.

  • Oki
  • [*][*]
HE-AAC v.1 & v.2 comparison
Reply #49
Quote
I actually did mean quality. Can you try some PS problem clips? I might have had an older/newer encoder so sorry if I messed up.[a href="index.php?act=findpost&pid=326773"][{POST_SNAPBACK}][/a]
If you make those clips available to us I am sure we could test them. I am not the best artifact hunter over here but i can help anyway. Guruboolez has proven habilities for this and it can be interesting to compare the CT's reference 3GPP rel6 audio codec against their commercial one.

Regards,
Oki