similar to: The inband FEC option within Opus

The -inbandfec option means that the encoder is *allowed* to use FEC, not that it has to. The final decision is based on the bitrate and the percentage of packet loss (and a few minor factors). If you're not seeing a difference, it means that given the options you're using, the encoder thinks actually using FEC would result in worse quality. Cheers, Jean-Marc On 08/05/17 07:11 PM,

I didn't expect this, is it normal? When you use inbandfec, and have packet losses, you end up with more audio samples than you started with. With short frame (2.5), the FEC isn't supposed to do anything but it shouldn't do that should it? and 20mS frame, should do 'something' but? add audio? I'm using opus_demo for this. Here's the steps (I have my test audio in

Hello, i'm a research student working on everyday sounds classification. i need to draw a dendrogram to show how the classification is done, but while i never used R before, i guess that a help from someone would be great :) does any one of you did something like that before? Thank you -- Ayoub

Hello, I'm interested in taking this task for this year's round of the Google Summer of Code. I'm a Computer Science undergraduate student at the University of Milan-Bicocca, I have 5+ years of experience with C/C++ but I'm quite new to the LLVM project. I'd like to have further information about the task requirements and whether it's still available. Thank you in advance.

Hello, i'm trying to draw a dendrogram to show the clustering hierarchy. i'm working on 2 classes but each class contains more than 3000 instances. i'm using the hclust function to do that! but i can't read anything from this figure. i would like to color each class with its own color so at least i can extract some information from the clustering. does any one of you did

Hi, I am looking at the conformance test using opus_compare and the test vectors and I am a bit confused about FEC. Is there a way to validate that a decoder processes LBRR frames correctly? Thanks, Alexis

Hi, When running opus_demo with FEC enabled, there is a bug related to the length of FEC frames. The FEC frames will be decoded with a length of max_frame_size and the resulting output file will be much longer than the input file.

Hi, I would like to monitor FEC usage in order to include it in RTCP EX or calculate MOS estimation, etc. However the Opus codec library does not seem to expose such information. "Was LBRR found and used or was it PLC ?" I saw in WebRTC that they are using a technique to parse the "frame header" WebRtcOpus_PacketHasFec() It this something that is supported ? What would you

Hi guys,

At this bitrate the encoder likely decides that it's better to put all the bits in the normal packet than use FEC. When you enable FEC it steals a lot of bits from the non-FEC content. Also, the use of FEC depends on the reported percentage of packet loss. The more loss there is, the lower the threshold for enabling FEC. Overall, the encoder attempt to make the best decision on a

having a hard time communicating on IRC, thank you gmaxwell, very informative. anyway, we were discussing the proper implementation of FEC on the decoder side. well, encoder side is just a boolean thing so that's alright. i gave an example where the receiver lost 5 rtp packets, 1 2 3 4 and 5 during which, we call opus_decode with a null pointer and fec=0 for every packet lost. now, when it

Hi all, Newbie to the group. Just started using Opus as part of a WebRTC project and amazed by the versatility of the codec. Great stuff!!!! I have been trying to understand the performance of Opus inband FEC in the presence of bursty loss. Although I do not have exact characterization of the loss profile, we are seeing issues over WiFi. RTCP reports about 33% loss, but I am guessing a lot of it

What do you mean by "implement"? You're just using the Opus built-in PLC (passing NULL), right? The PLC generally attempts to find periodicity and replicate it. I guess if your signal isn't periodic it can lead to a repetition that isn't great. It's something that could probably be improved in the PLC. Cheers, Jean-Marc On 08/25/2015 01:21 PM, Scott Boekweg wrote:

Hi, I changed the VAD sensitivity by varying VAD_SNR_FACTOR_Q16 and now it shows the difference in silence frames. Thanks Suresh On 4 October 2016 at 17:57, Suresh Thiriveedi <sthiriveedi at gmail.com> wrote: > Hi, > > When we pass around 9K samples of only ambient noise (no voice), the > encoder which is enabled FEC+DTX is detecting only some 140 frames as > non-voice

Hi Jon, On 27/01/17 08:31 AM, Jon Lederman wrote: > We are trying to use Opus in a VoIP environment for sending stereo > audio. We have noticed a phenomenon where when FEC is enabled and > packet_loss_percentage>0, that there is a mixing of audio from the > left channel into the right channel and vice versa. That is, rather > than hearing each channel in its pristine form as

Hi, I'm trying to use the FEC feature. I have a service which does the encoding with OPUS_SET_INBAND_FEC(1) and OPUS_SET_PACKET_LOSS_PERC(20) with 10ms packets. I'm not clear on the decoding process though. 1- When a packet is lost, do I need to call decode with fec=1 ONLY or do I need to call decode with fec=0 after as well? 2- How do I know up front the size of the pcm that I send to

Hi, I noticed that in some scenarios, Opus 1.2.1 produces better quality than 1.3.1 does. In the use case here, I'm enabling FEC and "transcode" signals from telephony networks (PCMU, 8kHz sampling) to VoIP (48kHz here). In this case, Opus always produced some leakage/ringing above 4kHz but for 1.3.1, these artifacts became worse. The small script below can be used to demonstrate

Hi, I noticed that in some scenarios, Opus 1.2.1 produces better quality than 1.3.1 does. In the use case here, I'm enabling FEC and "transcode" signals from telephony networks (PCMU, 8kHz sampling) to VoIP (48kHz here). In this case, Opus always produced some leakage/ringing above 4kHz but for 1.3.1, these artifacts became worse. The small script below can be used to demonstrate

In a fork of seanbright's opus patch for 13 there are further patches for Forward Error Correction and Package Loss Concealment, both of which ought to very useful in voip: https://github.com/traud/asterisk-opus Anybody used these patches ? Puzzled why they weren't committed to the main patch. sean

Hi, When we pass around 9K samples of only ambient noise (no voice), the encoder which is enabled FEC+DTX is detecting only some 140 frames as non-voice (returning only TOC, no frame content). We were expecting all or more to be identified as non-voice. Our idea was to check how the decoder re-generates the original ambient noise during the silence duration (when we feed NULL to decoder) when