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: Original File Length (OFL) for mp3, mp3Pro, aac... (Read 11421 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Original File Length (OFL) for mp3, mp3Pro, aac...

On AES 114th held in Amsterdam,  Thomson Multimedia researchers presented a way to signal exact codec delay and file length in a backwards compatible fashion - in current mp3, mp3pro, aac, etc...  coding algorithms.  It is known as OFL - or Original File Length

Algorithm deals with non-exact file length and delays  introduced by modern frequency-domain coding systems.  Data for OFL is written in ancilliary data so it is completely backwards compatible - decoder without OFL will just ignore the OFL data.

OFL will allow gapless playback with OFL compatible decoders.

AES paper number is:  5830


PS: They are mentioning LAME in the paper

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #1
Hopefully LAME and Foobar2000, Nero AAC and FAAD will add the true gapless playback as soon as possible.
Juha Laaksonheimo

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #2
Are you telling us that they published a paper just for this simple issue?

Well, at least we will have a uniform way to store delay.

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #3
Quote
Are you telling us that they published a paper just for this simple issue?


Well... maybe it is student work, I am not sure 

What is interesting is that they used three MP3 encoders in the table:  FhG, FhG FastEnc and LAME 3.91  and showed their appropriate delays - inclusion of LAME in Thomson paper really surprised me,  a good news indeed!

Quote
Well, at least we will have a uniform way to store delay.


I really hope they will publish this OFL specs (at least for MP3 and AAC)

There was also another good Dolby paper regarding new codec with variable frame length adjusted in a manner that it perfectly matches the video frame rate - but I guess this is for the new forum section (Scientific / Engineering discussion)

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #4
Quote
Hopefully LAME and Foobar2000, Nero AAC and FAAD will add the true gapless playback as soon as possible.

Not just gapless - this also removes codec delay (silence at the beginning of the track)  because  encoder/decoder pairs from one company are usually adjusted to match their delay only -  for example,  FhG's AAC implementation has different delay than Ahead's AAC implementation, etc... 

With OFL decoder will automatically be aware of encoder delay, without a need to "guess" or not to do anything.

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #5
Quote
OFL will allow gapless playback with OFL compatible decoders.


When can we expect to see these?  I assume we will see OFL on harware-based players and that it is possible to be added as a firmware upgrade?

This is a VERY good idea to standardize this. How is it implemented?  Through tagging of the files?  Will old files be able to be scanned and made compatible?  Will this be a built-in feature in encoders?
WARNING:  Changing of advanced parameters might degrade sound quality.  Modify them only if you are expirienced in audio compression!

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #6
I do not have the paper yet.
But it seems to me that :
*this is stored in ancillary data (in a similar way to Lame tag)
*any player could be upgraded to this
*it should be possible most of the time to update already existing files, but the question is how to know the delay of previously encoded files

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #7
Quote
*it should be possible most of the time to update already existing files, but the question is how to know the delay of previously encoded files


Yeah, it all goes back to "which encoder was used", etc.  This is more complex than say ReplayGain, in that this has to be much more specific.  Even if we can't perfectly convert old files to OFL, it will still be worth it, regardless.
WARNING:  Changing of advanced parameters might degrade sound quality.  Modify them only if you are expirienced in audio compression!

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #8
Does anyone know if progress is being made on OFL, or should I say, the implementation of it?  It is looking more and more like this is just what mp4 and aac need.  BTW, is anyone here at hydrogenaudio a member of the AES?
WARNING:  Changing of advanced parameters might degrade sound quality.  Modify them only if you are expirienced in audio compression!

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #9
Quote
Does anyone know if progress is being made on OFL, or should I say, the implementation of it?  It is looking more and more like this is just what mp4 and aac need.  BTW, is anyone here at hydrogenaudio a member of the AES?

Well.. I was pushing gapless to Nero AAC/FAAD2, and Menno got it even working pretty soon, but there were problems with other components of Nero (plugin usage related), so it didn't make it.
Maybe someday in the future..
Juha Laaksonheimo

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #10
Quote
Maybe someday in the future..

Depressing... 

Well, hopefully we'll see it in the near future... 
WARNING:  Changing of advanced parameters might degrade sound quality.  Modify them only if you are expirienced in audio compression!

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #11
Is this the method that was finally implemented into Nero?
WARNING:  Changing of advanced parameters might degrade sound quality.  Modify them only if you are expirienced in audio compression!

Original File Length (OFL) for mp3, mp3Pro, aac...

Reply #12
Quote
Is this the method that was finally implemented into Nero?

No, Nero makes use of the data fields already available in the MP4 file format. Nero doesn't change anything in the actual AAC data, only writes different frame durations for the first and last frames.

Menno