Re: Exhale CBR and low bitrate support
Reply #5 – 2025-02-19 20:13:24
Exhale is currently the only practical way to get USAC encodings with low sampling rates. For Mac, GNU/Linux or BSD users like me Exhale is the only free encoder, but I think also for Windows users, who can buy the encoder for a small fee thanks to Poikosoft. This does not mean we should be stubborn. In the past I tried Exhale at every mono/stereo combination for every bitrate and every sample, today the conversation has been removed. The results were not consistent, then Exhale improved in five presets. In the meantime Opus, which I have always criticized (today I have to give a lot of satisfaction to Jean-Marc Valin and to those who have suffered my criticisms), has improved a lot and honestly if you do not want a paid MPEG-D USAC encoder, I recommend you try it, for speech only it is fantastic, you will get much better results than Exhale. If instead you need to speak with a musical background that does not mask the voice, e.g. -32 LUFS for content normalized to -23 LUFS, as is done for broadcast in Europe, you will be better off using the paid xHE-AAC. Exhale It can be used at higher bitrates for all types of content, but at low bitrates it is no match. Think about it because if you thought the speech with Exhale at low bitrate was good, you will have to retrain your hearing (in the sense of recalibrating, training again) to pick up the artifacts, which are many. From the command line:opusenc --speech --bitrate 24 input.wav output.opus