Hi [UTF-8?]ÐÐ°ÐºÑ [UTF-8?]ÐаÑпов <[email protected]>
(Sorry no name in plain text) I think we are comparing apples to oranges. As I see it, Codec2 voice compression is about determining what a human voice is doing and digitising that. And, to as low a data bitrate as practical in carrying good intelligibility. Digging a voice out of a noisy environment is not the main purpose. And as I see it, the 1200/2400 modes are interim steps towards the lower bitrate modes. Alan VK2ZIW On Sun, 12 Jan 2020 07:01:01 +1100, David Rowe wrote > Hi [UTF-8?]ÐакÑ, > > Nice work on your tests! OK so this is what I hear: > > 1/ On the noise free English samples, codec 2 at 1200/2400 is > slightly worse than MELP at the same rates. This is consistent with > other tests > (e.g. academic papers using the two codecs as references). > > 2/ With noisy speech the MELP samples appear to have removed the > interfering noise - the noise is suppressed in the output samples, > and not faithfully reproduced. This suggests the MELP > implementation you have used also has a pre-processing step to > remove noise. > > So it's not quite an A/B comparison - the core MELP codec is being > tested on audio that has had the noise suppressed. > > With noisy input speech, the Codec 2 samples are distorted and > unpleasant to listen too, but still (as you suggest) intelligible. > > For noise suppression in Codec 2 applications we use the Speex noise > suppressor (codec2/misc/speexnoisesup.c for a command line version). > > 3/ It's difficult for me to evaluate the Russian samples, as I don't > speak the language, but thanks for the feedback. Yes, Codec 2 was > developed on just English samples. > > 4/ You are correct Codec 2 1200/2400 hasn't been actively developed > in over 5 years, recent efforts have been at lower bit rates, and LPCNet > > (at 1733 bit/s). In particular to support digital voice systems for > HF radio. > > Cheers, > David > > On 11/1/20 11:08 pm, [UTF-8?]ÐÐ°ÐºÑ [UTF-8?]ÐаÑпов wrote: > > Sorry for sending HTML message, I didn't expect it to mess up the archive. > > > > Plaintext copy of previous message follows: > > > > ------ > > > > Hello! > > > > Recently I had tested Codec2 on audio samples on both English and Russian languages, using MELPe as a reference. Moreover, I used three noise modes: no noise at all, gaussian noise and voice noise (e.g. when you are talking in a crowded room). As I don't have a source code for MELPe 600, I had tested only 1200 and 2400 bitrates. Some of MELPe tests were repeated in a environment where encoder and decoder processes were separated to ensure that no audio information is leaked through internal buffers and structures. > > > > To my disappointment, however, Codec2 performed much worse than MELPe, producing yet still intelligible (in some cases), but heavily distorted result. I made page with original and encoded samples: https://m-k.mx/static/codec2/. > > > > I am surprised why Codec2 performs so bad even on high bitrates where MELPe produces almost identical audio without any audible artifacts. Codec2 also degrades significantly on non-English samples (VQ codebooks are fine-tuned for English?). While degradation at noisy samples is expected (especially in voice noise), Codec2 sometimes produces very distorted and hardly intelligible results, while MELPe exhibit only slight degradation. > > > > So my questions are: > > > > 1. Are my experiments correct? Maybe I'm missing something important, like equalizer in front of Codec2 (--eq option appears to have no impact on the quality). > > 2. Are high-bitrate modes like 2400 and 3200 still up-to-date? I have seen a lot of effort put into 700C mode, but sometimes high-bitrate modes are useful as well. > > > > > > _______________________________________________ > > Freetel-codec2 mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 > > > > _______________________________________________ > Freetel-codec2 mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 --------------------------------------------------- Alan Beard OpenWebMail 2.53 _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
