Does Anyone Know If There Is A Problem With The Chrome Browser And Asterisk Cmp2k Video
Using asterisk 16.1.1.
I’m setting up a test using the cmp2k (Cyber Mega Phone 2K Ultimate Dynamic Edition).
I have noticed Chrome 72 had some issues with video streams. I just upgraded to Chrome 73 and see they still have some issues. If I have 2 calls in a confbridge with video set to none. I then set the video source to a Chrome browser and the Remote Video shown to both calls from Firefox and Chrome do not update. However, if I set the video source to the Firefox browser, my Remote Video is accurate in both Firefox and Chrome. I confirmed that asterisk is indicating the video source changed by looking at the AMI BridgeVideoSourceUpdate event.
When I use the Firefox (65.0.2) browser I can set either call to be the video source and the Remote Video updates accordingly.
Is this caused by Chrome’s video sent to asterisk being some format which asterisk can’t use in the confbridge? The endpoints for both Firefox and Chrome calls are set to disallow = all allow = opus,ulaw,vp9,vp8,h264
Both calls use the same bridge and user settings.
[3]
type = bridge language = en internal_sample_rate = 0
mixing_interval = 20
record_file_append = no max_members = 10
video_mode = none
[4]
type = user admin = no marked = no startmuted = no music_on_hold_when_empty = no quiet = yes wait_marked = no end_marked = no dsp_drop_silence = yes dsp_silence_threshold = 2500
dsp_talking_threshold = 160
denoise = no jitterbuffer = yes talk_detection_events = yes dtmf_passthrough = no announce_user_count = no announce_join_leave = no announce_user_count_all = no announce_only_user = no send_events = no echo_events = no announce_join_leave_review = no
Have a great day!
Dan
4 thoughts on - Does Anyone Know If There Is A Problem With The Chrome Browser And Asterisk Cmp2k Video
The way I understand it is that video stream has changed, so the browser needs some way to know that. Otherwise the decoder thinks it’s invalid data and drops it. In these cases either Asterisk needs to issue a renegotiation
(currently not supported), or the codec needs to contain video stream information in their payload. I spoke with Joshua Colp about this some as he’s had some dealings with this and he had the following to say:
“Some codecs (such as VP8/V9) embed information about the video stream within their payload. Asterisk does not currently rewrite this information, so when a stream change occurs in Asterisk using the selective source functionality this can cause the receiving side (the browser) to drop the payload as it sees it as not being part of the existing stream. Different browsers can behave differently, such as resetting the video decoder to handle the new stream. Rewriting this information in the video payload is not currently supported.”
So you are probably seeing it work or not in Chrome vs Firefox due to browser, and codec support of such occurrences.
…. ….
VGhhbmsgeW91IEtldmluLg0KDQpBbnkgaWRlYSBpZiBHb29nbGUgZGV2ZWxvcGVycyBtYXkgZGVj aWRlIHRvIHJlc2V0IHRoZSBkZWNvZGVyLCBwZXIgSm9zaHVh4oCZcyBleHBlcmllbmNlPw0KT3Ig cGVyaGFwcyBhc3RlcmlzayBkZXZlbG9wZXJzIHdvdWxkIGNvbnNpZGVyIGV2ZW50dWFsbHkgYWRk IHRoZSByZXdyaXRlIHN1cHBvcnQgaW4gYXN0ZXJpc2s/DQoNCkFsd2F5cyBmdW4gZGVhbGluZyB3
aXRoIG5ld2VyIHRlY2hub2xvZ3kgYXMgaXQgZ29lcyB0aHJvdWdoIHNldmVyYWwgcmV2aXNpb25z Lg0KDQpIYXZlIGEgZ3JlYXQgZGF5IQ0KDQpEYW4NCg0KRnJvbTogYXN0ZXJpc2stdXNlcnMgPGFz dGVyaXNrLXVzZXJzLWJvdW5jZXNAbGlzdHMuZGlnaXVtLmNvbT4gT24gQmVoYWxmIE9mIEtldmlu IEhhcndlbGwNClNlbnQ6IFRodXJzZGF5LCBNYXJjaCAxNCwgMjAxOSA0OjUwIFBNDQpUbzogQXN0
ZXJpc2sgVXNlcnMgTWFpbGluZyBMaXN0IC0gTm9uLUNvbW1lcmNpYWwgRGlzY3Vzc2lvbiA8YXN0
ZXJpc2stdXNlcnNAbGlzdHMuZGlnaXVtLmNvbT4NClN1YmplY3Q6IFJlOiBbYXN0ZXJpc2stdXNl cnNdIERvZXMgYW55b25lIGtub3cgaWYgdGhlcmUgaXMgYSBwcm9ibGVtIHdpdGggdGhlIENocm9t ZSBicm93c2VyIGFuZCBhc3RlcmlzayBjbXAyayB2aWRlbw0KDQpPbiBXZWQsIE1hciAxMywgMjAx OSBhdCAxMDoxNSBBTSBEYW4gQ3JvcHAgPGRhbkBhbXRlbGNvLmNvbTxtYWlsdG86ZGFuQGFtdGVs Y28uY29tPj4gd3JvdGU6DQpVc2luZyBhc3RlcmlzayAxNi4xLjEuDQoNCknigJltIHNldHRpbmcg dXAgYSB0ZXN0IHVzaW5nIHRoZSBjbXAyayAoQ3liZXIgTWVnYSBQaG9uZSAySyBVbHRpbWF0ZSBE
eW5hbWljIEVkaXRpb24pLg0KDQpJIGhhdmUgbm90aWNlZCBDaHJvbWUgNzIgaGFkIHNvbWUgaXNz dWVzIHdpdGggdmlkZW8gc3RyZWFtcy4gIEkganVzdCB1cGdyYWRlZCB0byBDaHJvbWUgNzMgYW5k IHNlZSB0aGV5IHN0aWxsIGhhdmUgc29tZSBpc3N1ZXMuICBJZiBJIGhhdmUgMiBjYWxscyBpbiBh IGNvbmZicmlkZ2Ugd2l0aCB2aWRlbyBzZXQgdG8gbm9uZS4gIEkgdGhlbiBzZXQgdGhlIHZpZGVv IHNvdXJjZSB0byBhIENocm9tZSBicm93c2VyIGFuZCB0aGUgUmVtb3RlIFZpZGVvIHNob3duIHRv IGJvdGggY2FsbHMgZnJvbSBGaXJlZm94IGFuZCBDaHJvbWUgZG8gbm90IHVwZGF0ZS4gIEhvd2V2
ZXIsIGlmIEkgc2V0IHRoZSB2aWRlbyBzb3VyY2UgdG8gdGhlIEZpcmVmb3ggYnJvd3NlciwgbXkg UmVtb3RlIFZpZGVvIGlzIGFjY3VyYXRlIGluIGJvdGggRmlyZWZveCBhbmQgQ2hyb21lLg0KSSBj b25maXJtZWQgdGhhdCBhc3RlcmlzayBpcyBpbmRpY2F0aW5nIHRoZSB2aWRlbyBzb3VyY2UgY2hh bmdlZCBieSBsb29raW5nIGF0IHRoZSBBTUkgQnJpZGdlVmlkZW9Tb3VyY2VVcGRhdGUgZXZlbnQu DQoNCldoZW4gSSB1c2UgdGhlIEZpcmVmb3ggKDY1LjAuMikgYnJvd3NlciBJIGNhbiBzZXQgZWl0
aGVyIGNhbGwgdG8gYmUgdGhlIHZpZGVvIHNvdXJjZSBhbmQgdGhlIFJlbW90ZSBWaWRlbyB1cGRh dGVzIGFjY29yZGluZ2x5Lg0KDQpJcyB0aGlzIGNhdXNlZCBieSBDaHJvbWXigJlzIHZpZGVvIHNl bnQgdG8gYXN0ZXJpc2sgYmVpbmcgc29tZSBmb3JtYXQgd2hpY2ggYXN0ZXJpc2sgY2Fu4oCZdCB1
c2UgaW4gdGhlIGNvbmZicmlkZ2U/DQoNClRoZSB3YXkgSSB1bmRlcnN0YW5kIGl0IGlzIHRoYXQg dmlkZW8gc3RyZWFtIGhhcyBjaGFuZ2VkLCBzbyB0aGUgYnJvd3NlciBuZWVkcyBzb21lIHdheSB0
byBrbm93IHRoYXQuIE90aGVyd2lzZSB0aGUgZGVjb2RlciB0aGlua3MgaXQncyBpbnZhbGlkIGRh dGEgYW5kIGRyb3BzIGl0LiBJbiB0aGVzZSBjYXNlcyBlaXRoZXIgQXN0ZXJpc2sgbmVlZHMgdG8g aXNzdWUgYSByZW5lZ290aWF0aW9uIChjdXJyZW50bHkgbm90IHN1cHBvcnRlZCksIG9yIHRoZSBj b2RlYyBuZWVkcyB0byBjb250YWluIHZpZGVvIHN0cmVhbSBpbmZvcm1hdGlvbiBpbiB0aGVpciBw YXlsb2FkLiBJIHNwb2tlIHdpdGggSm9zaHVhIENvbHAgYWJvdXQgdGhpcyBzb21lIGFzIGhlJ3Mg aGFkIHNvbWUgZGVhbGluZ3Mgd2l0aCB0aGlzIGFuZCBoZSBoYWQgdGhlIGZvbGxvd2luZyB0byBz YXk6DQoNCiJTb21lIGNvZGVjcyAoc3VjaCBhcyBWUDgvVjkpIGVtYmVkIGluZm9ybWF0aW9uIGFi b3V0IHRoZSB2aWRlbyBzdHJlYW0gd2l0aGluIHRoZWlyIHBheWxvYWQuIEFzdGVyaXNrIGRvZXMg bm90IGN1cnJlbnRseSByZXdyaXRlIHRoaXMgaW5mb3JtYXRpb24sIHNvIHdoZW4gYSBzdHJlYW0g Y2hhbmdlIG9jY3VycyBpbiBBc3RlcmlzayB1c2luZyB0aGUgc2VsZWN0aXZlIHNvdXJjZSBmdW5j dGlvbmFsaXR5IHRoaXMgY2FuIGNhdXNlIHRoZSByZWNlaXZpbmcgc2lkZSAodGhlIGJyb3dzZXIp IHRvIGRyb3AgdGhlIHBheWxvYWQgYXMgaXQgc2VlcyBpdCBhcyBub3QgYmVpbmcgcGFydCBvZiB0
aGUgZXhpc3Rpbmcgc3RyZWFtLiBEaWZmZXJlbnQgYnJvd3NlcnMgY2FuIGJlaGF2ZSBkaWZmZXJl bnRseSwgc3VjaCBhcyByZXNldHRpbmcgdGhlIHZpZGVvIGRlY29kZXIgdG8gaGFuZGxlIHRoZSBu ZXcgc3RyZWFtLiBSZXdyaXRpbmcgdGhpcyBpbmZvcm1hdGlvbiBpbiB0aGUgdmlkZW8gcGF5bG9h ZCBpcyBub3QgY3VycmVudGx5IHN1cHBvcnRlZC4iDQoNClNvIHlvdSBhcmUgcHJvYmFibHkgc2Vl aW5nIGl0IHdvcmsgb3Igbm90IGluIENocm9tZSB2cyBGaXJlZm94IGR1ZSB0byBicm93c2VyLCBh bmQgY29kZWMgc3VwcG9ydCBvZiBzdWNoIG9jY3VycmVuY2VzLg0KDQogICAuLi4uIDxzbmlwIGNv bmZpZz4gLi4uLg0KDQotLQ0KS2V2aW4gSGFyd2VsbA0KRGlnaXVtIC0gQSBTYW5nb21hIENvbXBh bnkgfCBTZW5pb3IgU29mdHdhcmUgRGV2ZWxvcGVyDQo0NDUgSmFuIERhdmlzIERyaXZlIE5XIC0g SHVudHN2aWxsZSwgQUwgMzU4MDYgLSBVUw0KQ2hlY2sgdXMgb3V0IGF0OiBodHRwczovL2RpZ2l1
bS5jb20gJiBodHRwczovL2FzdGVyaXNrLm9yZw0K
It’s hard to really know the way things will go in the end, it’s always evolving.
It’s certainly something that could be done, I just don’t know of anyone working on it right now or if anyone plans to.
—
Joshua C. Colp Digium – A Sangoma Company | Senior Software Developer
445 Jan Davis Drive NW – Huntsville, AL 35806 – US
Check us out at: http://www.digium.com & http://www.asterisk.org
—
Thank you Joshua
—–Original Message—