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: AAC Conformance Encoder / Decoder (Read 4692 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

AAC Conformance Encoder / Decoder

I noticed that in the ISO refsoftware, it is using type 32 bits FLOAT instead of 64 bits DOUBLE for most cases.

I was wondering if I should be using 32 bits FLOAT in all cases? Is the 32 bits FLOAT sufficient for floating point calculations such as those in the MDCT/ IMDCT, predictors?

Currently I am using 64 bits DOUBLE in all cases.. and I wonder if this would effect the conformance tests?

AAC Conformance Encoder / Decoder

Reply #1 –
32 bit float should be enough for 16-bit precision - we ported AAC Encoder from 64-bit double to 32-bit float in 2002, and the resulting encoder provided bit-to-bit identical streams like 64-bit version.

Decoder works in 32-bit float natively, and passes all conformance tests set by ISO.
 

AAC Conformance Encoder / Decoder

Reply #2 –
Quote from: wkwai on
I noticed that in the ISO refsoftware, it is using type 32 bits FLOAT instead of 64 bits DOUBLE for most cases.

I was wondering if I should be using 32 bits FLOAT in all cases? Is the 32 bits FLOAT sufficient for floating point calculations such as those in the MDCT/ IMDCT, predictors?

Currently I am using 64 bits DOUBLE in all cases.. and I wonder if this would effect the conformance tests?

 

Could any of you comment on the compliance of the FIXED point version??