IPB

Welcome Guest ( Log In | Register )

> Upload forum rules

- No over 30 sec clips of copyrighted music. Cite properly and never more than necessary for the discussion.


- No copyrighted software without permission.


- Click here for complete Hydrogenaudio Terms of Service

6 Pages V   1 2 3 > »   
Reply to this topicStart new topic
Winamp FhG AAC encoder with gapless fix, Updates from the 5.62 release
benski
post Jul 4 2011, 18:51
Post #1


Winamp Developer


Group: Developer
Posts: 670
Joined: 17-July 05
From: Brooklyn, NY
Member No.: 23375



Steve from the AAC listening test topic found a bug related to gapless information in m4a files encoded with Winamp/FhG's new AAC encoder.

Updates from the 5.62 release
* Fixed a bug that was preventing gapless playback information from being written to the encoded file in some circumstances
* Fixed compatibility issue that was preventing HE-AAC files from being used with iTunes
* Fixed a bug that was causing the ADTS encoder to fail with the default settings

To install, just unzip this to your Winamp plugins folder. It should overwrite an existing file (unless you are using an older version of Winamp).

This will be released with 5.63 (or 5.621 depending on what version number we decide upon)
Attached File(s)
Attached File  enc_fhgaac_v1_01.zip ( 434.58K ) Number of downloads: 1004
 
Go to the top of the page
+Quote Post
lvqcl
post Jul 7 2011, 15:41
Post #2





Group: Developer
Posts: 3328
Joined: 2-December 07
Member No.: 49183



I noticed that this encoder produces 88.2 kHz .m4a files when the input is 44.1 and bitrate>=320kpbs. 48kHz remains untouched.

Is it a bug or a feature?
Go to the top of the page
+Quote Post
benski
post Jul 7 2011, 17:20
Post #3


Winamp Developer


Group: Developer
Posts: 670
Joined: 17-July 05
From: Brooklyn, NY
Member No.: 23375



QUOTE (lvqcl @ Jul 7 2011, 10:41) *
I noticed that this encoder produces 88.2 kHz .m4a files when the input is 44.1 and bitrate>=320kpbs. 48kHz remains untouched.

Is it a bug or a feature?


A little bit of both. One issue is that the configuration dialog currently has no idea what the input samplerate / channel count is going to be. And the encoder must be trying really hard to honor the bitrate request. I'll do a bit more checking on it - is it all bitrates >320kbps, even 324kbps? I'll check myself also.
Go to the top of the page
+Quote Post
lvqcl
post Jul 7 2011, 18:08
Post #4





Group: Developer
Posts: 3328
Joined: 2-December 07
Member No.: 49183



QUOTE (benski @ Jul 7 2011, 20:20) *
is it all bitrates >320kbps, even 324kbps? I'll check myself also.


All bitrates >= 320. Even 320 kbps. (2 channels, 16 bit).
Go to the top of the page
+Quote Post
Alex B
post Jul 7 2011, 23:14
Post #5





Group: Members
Posts: 1303
Joined: 14-September 05
From: Helsinki, Finland
Member No.: 24472



By chance I noticed a serious encoding artifact (I'd actually describe it as a bug) when I tested this encoder with some gapless albums. The album is Transatlantic's The Whirlwind. I noticed that the track #4 (A Man Can Feel) has a suspicious track peak value of 2.46 after encoding it using the FhG AAC VBR preset 3. The track has a loud artifact in the left channel at about 2 min 1 s.

I tried to create a 30 s sample in which the artifact would occur at 25 s, but the cut source file did not produce the same artifact. Apparently only the complete track triggers it.

I verified the problem with two different computers.

I created samples of the reference file and the decoded file (decoded to float and cut). I reduced the volume levels by -8 dB to prevent clipping before I saved the samples in the 16-bit integer format. The artifact occurs at about 5 s.

The reference sample: Attached File  A_Man_Can_Feel__reference.flac ( 754.23K ) Number of downloads: 237

The decoded sample: Attached File  A_Man_Can_Feel__decoded_aac.flac ( 689.35K ) Number of downloads: 246

Here is how the decoded sample looks in Audition:






EDIT

benski and C.R.Helmrich, I have posted a PM to you.

This post has been edited by Alex B: Jul 7 2011, 23:38


--------------------
http://listening-tests.freetzi.com
Go to the top of the page
+Quote Post
lvqcl
post Jul 8 2011, 04:22
Post #6





Group: Developer
Posts: 3328
Joined: 2-December 07
Member No.: 49183



QUOTE (Alex B @ Jul 8 2011, 02:14) *
I tried to create a 30 s sample in which the artifact would occur at 25 s, but the cut source file did not produce the same artifact. Apparently only the complete track triggers it.


Maybe removing first N*1024 samples will help?
Go to the top of the page
+Quote Post
Alex B
post Jul 8 2011, 12:46
Post #7





Group: Members
Posts: 1303
Joined: 14-September 05
From: Helsinki, Finland
Member No.: 24472



So you think that the AAC frame border position may have an effect on this. That would be a logical explanation because I don't think a VBR encoder would take into account the past audio content from tens of seconds or minutes before the current frame.

This post has been edited by Alex B: Jul 8 2011, 12:57


--------------------
http://listening-tests.freetzi.com
Go to the top of the page
+Quote Post
Alex B
post Jul 8 2011, 14:18
Post #8





Group: Members
Posts: 1303
Joined: 14-September 05
From: Helsinki, Finland
Member No.: 24472



That did the trick. I removed N*1024 samples from the beginning. Here is a ~15 s sample that produces the same artifact when encoded (using Winamp/FhG AAC v.1.01, VBR preset 3). The problem occurs at about 12.6 s. It is clearly audible when replay gain is set to prevent clipping.

Attached File  A_Man_Can_Feel.flac ( 1.73MB ) Number of downloads: 245


This post has been edited by Alex B: Jul 8 2011, 14:26


--------------------
http://listening-tests.freetzi.com
Go to the top of the page
+Quote Post
MrSinatra
post Jul 13 2011, 02:31
Post #9





Group: Members
Posts: 39
Joined: 11-June 10
Member No.: 81386



Benski,

i wish you were posting this at the winamp forums!
Go to the top of the page
+Quote Post
C.R.Helmrich
post Aug 23 2011, 11:59
Post #10





Group: Developer
Posts: 686
Joined: 6-December 08
From: Erlangen Germany
Member No.: 64012



QUOTE (Alex B @ Jul 8 2011, 00:14) *
By chance I noticed a serious encoding artifact (I'd actually describe it as a bug) when I tested this encoder with some gapless albums.

Thanks a lot, Alex, for this report! This is a very rare bug which has been fixed in a newer release of Fraunhofer's AAC encoder (version 03.02.03) a few weeks ago and should reach Winamp soon.

QUOTE (lvqcl @ Jul 7 2011, 16:41) *
I noticed that this encoder produces 88.2 kHz .m4a files when the input is 44.1 and bitrate>=320kpbs. 48kHz remains untouched.

Is it a bug or a feature?

Thanks for reporting! It's a bug (because it was trivial to solve) and is also fixed in the new version.

Please, anyone, if you find such configuration problems or artifacts, report them here on HA along with the input file, just like Alex did, so we can reproduce and hopefully fix them.

Chris

This post has been edited by C.R.Helmrich: Aug 23 2011, 12:08


--------------------
If I don't reply to your reply, it means I agree with you.
Go to the top of the page
+Quote Post
b66pak
post Aug 23 2011, 18:45
Post #11





Group: Members
Posts: 58
Joined: 2-February 10
Member No.: 77800



@C.R.Helmrich

hi, do you plan to add adts support for VBR mode too?
_
Go to the top of the page
+Quote Post
benski
post Aug 23 2011, 19:37
Post #12


Winamp Developer


Group: Developer
Posts: 670
Joined: 17-July 05
From: Brooklyn, NY
Member No.: 23375



QUOTE (b66pak @ Aug 23 2011, 13:45) *
@C.R.Helmrich

hi, do you plan to add adts support for VBR mode too?
_


The encoder supports VBR encoding in ADTS format. We (Winamp) have chosen not to implement it as there is no defined seek-table for ADTS files.
Go to the top of the page
+Quote Post
b66pak
post Aug 23 2011, 19:57
Post #13





Group: Members
Posts: 58
Joined: 2-February 10
Member No.: 77800



i hope you will reconsider...(may be as a silent option - for a CLI wrapper only!)...this would be very handy (with adts STDOUT support too! - may be as a silent option - for a CLI wrapper only!) for pipeing VBR/CBR aac from an audio decoder to an audio-video muxer...
_

sorry for OT...
_

This post has been edited by b66pak: Aug 23 2011, 19:58
Go to the top of the page
+Quote Post
punkrockdude
post Oct 13 2011, 16:16
Post #14





Group: Members
Posts: 244
Joined: 21-February 05
Member No.: 20022



I am looking forward to a fix for the upsampling issue. Regards.
Go to the top of the page
+Quote Post
punkrockdude
post Oct 13 2011, 17:58
Post #15





Group: Members
Posts: 244
Joined: 21-February 05
Member No.: 20022



Are you considering higher bitrate VBR presets? I use Q 0.9 with Nero which result in ~375 kbps with modern music and around ~250 kbps with older less hifi music. You (Fhg) have constant bitrate up to 448 kbps for us who want some reassurance headroom when it comes to the bitrate but a VBR mode that uses the highest bitrates and results in maybe 320-375 kbps would be much appreciated. Regards.
Go to the top of the page
+Quote Post
C.R.Helmrich
post Oct 13 2011, 20:21
Post #16





Group: Developer
Posts: 686
Joined: 6-December 08
From: Erlangen Germany
Member No.: 64012



QUOTE (punkrockdude @ Oct 13 2011, 18:58) *
Are you considering higher bitrate VBR presets? I use Q 0.9 with Nero which result in ~375 kbps with modern music and around ~250 kbps with older less hifi music. You (Fhg) have constant bitrate up to 448 kbps for us who want some reassurance headroom ...

I assume you like the idea of headroom in order to avoid non-transparent encodings. Actually I tuned FhG's encoder for precisely that, using also the most "killer" items I could find here on HA. At the highest VBR mode, any item should be transparent. If it is not, please notify me. Until then, I see little point in adding a very-high-rate VBR mode.

In any case, you'd have to ask benski since he decides (and has to implement) which modes are/will be available in the Winamp GUI.

Chris


--------------------
If I don't reply to your reply, it means I agree with you.
Go to the top of the page
+Quote Post
Alex B
post Oct 14 2011, 09:10
Post #17





Group: Members
Posts: 1303
Joined: 14-September 05
From: Helsinki, Finland
Member No.: 24472



Am I correct to assume that the v. 03.02.03 is still not available?

Edit: Winamp 5.621 is still the latest build and its encoder version is 03.02.02. (at least in my installer, downloaded July 14)

This post has been edited by Alex B: Oct 14 2011, 09:43


--------------------
http://listening-tests.freetzi.com
Go to the top of the page
+Quote Post
Anakunda
post Oct 26 2011, 13:13
Post #18





Group: Members
Posts: 450
Joined: 24-November 08
Member No.: 63072



Great! The only downside is I'm not using Winamp. Is therre a facet for foobar2k, or a plugin for f2k?

This post has been edited by Anakunda: Oct 26 2011, 13:13
Go to the top of the page
+Quote Post
benski
post Oct 26 2011, 14:09
Post #19


Winamp Developer


Group: Developer
Posts: 670
Joined: 17-July 05
From: Brooklyn, NY
Member No.: 23375



QUOTE (Alex B @ Oct 14 2011, 04:10) *
Am I correct to assume that the v. 03.02.03 is still not available?

Edit: Winamp 5.621 is still the latest build and its encoder version is 03.02.02. (at least in my installer, downloaded July 14)


Winamp 5.622 with FhG AAC 3.2.3 just got released moments ago.

This post has been edited by benski: Oct 26 2011, 14:49
Go to the top of the page
+Quote Post
lvqcl
post Oct 26 2011, 15:16
Post #20





Group: Developer
Posts: 3328
Joined: 2-December 07
Member No.: 49183



Thanks for the info!
Go to the top of the page
+Quote Post
b66pak
post Oct 27 2011, 17:28
Post #21





Group: Members
Posts: 58
Joined: 2-February 10
Member No.: 77800



@benski can you update your attachement from the first post?
_
Go to the top of the page
+Quote Post
IgorC
post Oct 28 2011, 14:37
Post #22





Group: Members
Posts: 1534
Joined: 3-January 05
From: ARG/RUS
Member No.: 18803



Apparently new version 3.2.3 has not only bug fixes but also quality tunings.
Go to the top of the page
+Quote Post
Gainless
post Oct 28 2011, 14:49
Post #23





Group: Members
Posts: 169
Joined: 28-October 11
Member No.: 94764




QUOTE
I assume you like the idea of headroom in order to avoid non-transparent encodings. Actually I tuned FhG's encoder for precisely that, using also the most "killer" items I could find here on HA. At the highest VBR mode, any item should be transparent. If it is not, please notify me. Until then, I see little point in adding a very-high-rate VBR mode.

In any case, you'd have to ask benski since he decides (and has to implement) which modes are/will be available in the Winamp GUI.

Chris


Depends on what you call "transparent".
The encoder is indeed very efficient and every detail is clear at the highest VBR mode, but the sound is still kinda "compact" and a bit unnatural compared to the lossless files.
320 kbps would be perhaps a bit overkill, but what speaks against a mode "6" at around 256 kbps?

Edit:
Sorry for the wrong quote I answered to at first!

This post has been edited by Gainless: Oct 28 2011, 15:28
Go to the top of the page
+Quote Post
b66pak
post Oct 28 2011, 17:47
Post #24





Group: Members
Posts: 58
Joined: 2-February 10
Member No.: 77800



can anybody help with the new .dll?
_
Go to the top of the page
+Quote Post
A_Man_Eating_Duc...
post Oct 28 2011, 19:36
Post #25





Group: Members
Posts: 925
Joined: 21-December 01
From: New Zealand
Member No.: 705



Download the latest Winamp package.


--------------------
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.
Go to the top of the page
+Quote Post

6 Pages V   1 2 3 > » 
Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 26th July 2014 - 04:39