hi,
Has anyone here ever tried ISO AAC encoder? How do you like the quality of LC@128kbps?
I used to think its high complexity of computing would to some extent guarantee an average quality but it turns out to be quite poor (faraway behind) against other commercial ones, e.g. NERO.
Is this true? I am not sure whether I have made mistakes in building ISO binary. Anyone understand how that huge gap comes up?
Thanks.
Full Version: ISO AAC encoder
The ISO encoder is only built to demonstrate how to produce ISO-compliant output. This says nothing about quality, speed or efficiency. It probably uses a very basic psymodel and its various parameters have not been tuned by listening tests.
Thanks. I am very curious in an optimum psy model implementation. Is there any quality optimized psy model? Any reference code, implementation or paper? I know that NERO is a leading product in this field but it is not open source product.
The ISO encoder is only built to demonstrate how to produce ISO-compliant output. This says nothing about quality, speed or efficiency. It probably uses a very basic psymodel and its various parameters have not been tuned by listening tests.
QUOTE(eofor @ May 8 2008, 17:51) 
The ISO encoder is only built to demonstrate how to produce ISO-compliant output. This says nothing about quality, speed or efficiency. It probably uses a very basic psymodel and its various parameters have not been tuned by listening tests.
QUOTE(andyquan @ May 8 2008, 15:47)
Thanks. I am very curious in an optimum psy model implementation. Is there any quality optimized psy model? Any reference code, implementation or paper? I know that NERO is a leading product in this field but it is not open source product.
An LGPL AAC Encoder is being worked on as part of the Google Summer of Code for the FFmpeg project. It is already functional and its author is now working on psychoacoustic models.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.