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: Can Opus encode speech below 6 kbit/s like Codec 2? (Read 4579 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Can Opus encode speech below 6 kbit/s like Codec 2?

Just curious. Some modes of Codec 2 seem very simple, seemingly enough for CELT mode to replicate. Basically, one needs to be able to generate white noise (possible), periodic signals (sort of possible by filtering white noise using CELT's pitch postfilter) and to adjust signal power per band (of course possible).

Am I missing something?

Re: Can Opus encode speech below 6 kbit/s like Codec 2?

Reply #1
As for the available options:

Opusenc --bitrate N.nnn
Set target bitrate in kbit/sec (6-256 per channel)

And the front page for the official doc reads:
Sampling rates from 8 to 48 kHz
Bit-rates from 6 kb/s to 510 kb/s


But digging further into the API you can find that:
#define OPUS_SET_BITRATE(x)    
Configures the bitrate in the encoder.
Rates from 500 to 512000 bits per second are meaningful


So maybe by using the library directly you can test bit-rates as low as 0.5 kbps?

I'm not sure if such a low value is later capped to a minimum elsewhere.

Re: Can Opus encode speech below 6 kbit/s like Codec 2?

Reply #2
The codec specification says 6 kbps. However, the libopus encoder clamps the value between 300 kbps per channel and 500 bps. opusenc itself accepts values down to 500 bps, although the corresponding error message points out that 6 kbps is the "meaningful" lower limit.

 

Re: Can Opus encode speech below 6 kbit/s like Codec 2?

Reply #3
I was discussing this not too long ago, and the consensus is, that Opus creates somewhat meaningful codecs at 10kb/s and above, so I'm not sure these endeavours even make sense.