Commit Graph

3599 Commits

Author SHA1 Message Date
Saxon 947f818bc6 stream/mts: modified error in FindPid in the case that we can't a packet with given pid so that we show pid 2019-02-27 16:36:59 +10:30
Alan Noble 8a13bb7007 Temporary fix: Make ringBufferSize size larger for RTMP to work over flaky networks. 2019-02-27 05:53:49 +00:00
Saxon b669d2645b Merge branch 'master' of http://bitbucket.org/ausocean/av 2019-02-26 09:07:21 +10:30
Saxon 6143baed0b rtmp: doing conn.close log message after we actually check that there is still a connection, otherwise we might have a nil logger and cause a panic 2019-02-26 09:06:25 +10:30
kortschak e64bae84d1 Merged in test/vet-failure (pull request #152)
revid: fix build

Approved-by: Alan Noble <anoble@gmail.com>
2019-02-25 08:58:57 +00:00
Saxon 796bc12e09 Merge branch 'master' into improve-defaulting 2019-02-25 18:43:16 +10:30
Dan Kortschak 8240fea9ab revid: fix build 2019-02-25 18:39:01 +10:30
Saxon 7d3d8c3374 revid: change ring buffer size depending on output type 2019-02-25 18:36:39 +10:30
Alan Noble 4919c972c6 Merged in revid-log-debug-message-upon-ring-timeout (pull request #151)
Only log a debug message when ring.ErrTimeout occurs, not a warning.

Approved-by: Saxon Milton <saxon.milton@gmail.com>
Approved-by: kortschak <dan@kortschak.io>
Approved-by: Alan Noble <anoble@gmail.com>
2019-02-25 07:58:36 +00:00
Saxon 2818f308df revid: defaulting to Mpegts packetization if output is Http or Rtp and Defaulting to Flv packetization if output is Rtmp 2019-02-25 18:26:20 +10:30
Alan Noble a90b35b319 Only log a debug message when ring.ErrTimeout occurs, not a warning. 2019-02-25 07:14:30 +00:00
Saxon Milton c247e63bd0 Merged in default-mode-crash-fix (pull request #150)
revid-cli: initialising config.Output slice when no outputs are defined so that a default output can still be set

Approved-by: Alan Noble <anoble@gmail.com>
2019-02-25 06:42:38 +00:00
Alan Noble 95303b4e71 Set default log level to Info (to match the comment). 2019-02-25 06:31:12 +00:00
Saxon 5d5af5601f revid: not closing destinations on call to revid.Reset() anymore 2019-02-25 17:00:31 +10:30
Saxon 8340143622 rtmp: not logging conn close message unless we do have a connection 2019-02-25 16:56:37 +10:30
Saxon 7f2d53e778 revid: now checking for Packetization var from netreceiver 2019-02-25 16:29:44 +10:30
Saxon b7d1bd17a7 revid: initialising ring buffer in different location 2019-02-25 16:03:28 +10:30
Saxon be6e168774 revid: fixed quantization and bitrate logic in config.Validate() 2019-02-25 15:56:51 +10:30
Saxon cc4d683ba6 revid: calling reset inside revid.Update once config is updated 2019-02-25 15:47:43 +10:30
Saxon 5a2d89ef8f revid: using only one output when we're getting config from netreceiver 2019-02-25 15:31:42 +10:30
Saxon e118a639d4 revid: outputting config when we update 2019-02-25 15:28:19 +10:30
Trek H d95ce7efe6 ADPCM: fixed conversions and calculations for block variables 2019-02-25 14:43:26 +10:30
Saxon 37850e8350 revid-cli: initialising config.Output slice when no outputs are defined so that a default output can still be set 2019-02-25 12:59:57 +10:30
Trek H 936677f747 ADPCM: added descriptions to file headers and author to readme 2019-02-25 11:55:13 +10:30
Saxon Milton 855dfbae01 Merged in netsender-run-in-revid-cli (pull request #141)
cmd/revid-cli: replaced use of send() with netsender.Run() and implemented readPin func.

Approved-by: kortschak <dan@kortschak.io>
Approved-by: Alan Noble <anoble@gmail.com>
2019-02-24 23:17:17 +00:00
Alan Noble 471cd56c54 Merged in rtmp-dial-error-close (pull request #149)
Removed incorrect Close upon Dial error.

Approved-by: kortschak <dan@kortschak.io>
Approved-by: Alan Noble <anoble@gmail.com>
2019-02-24 23:15:27 +00:00
Alan Noble 01d85af133 Removed incorrect Close upon Dial error. 2019-02-22 20:45:13 +00:00
Saxon Milton 9b57306aed Merged in revert-encoder-timing (pull request #148)
stream: reverted timing changes from most recent PR
2019-02-17 03:08:00 +00:00
saxon c58b207419 stream: reverted timing changes from most recent PR 2019-02-17 13:35:55 +10:30
saxon 9e9e349cbf stream: reverted some changes I made to tick() in both mts encoder and rtp encoder 2019-02-17 13:27:01 +10:30
saxon f1625d27f5 revid: fixed destination slice bug 2019-02-17 13:11:17 +10:30
saxon ca0a008c59 revid: correct cc logic is discontinuity.go 2019-02-17 12:50:33 +10:30
saxon 819c9a784c revid: mtsSender_test.go passing segmenting and discontinuity tests 2019-02-17 05:32:44 +10:30
saxon 3f3d587eeb revid: mtsSender test for segmenting actually working now 2019-02-17 03:35:59 +10:30
saxon b3b8c6bb44 stream/mts: checking data is also good in mts segment test 2019-02-17 00:31:30 +10:30
saxon a5cb1c5abb stream/mts: made modifications such that the segment test is passing 2019-02-16 23:52:40 +10:30
saxon 3356457c71 revid: not wrapping t.Errorf 2019-02-16 16:43:15 +10:30
saxon bb091f5961 revid: wrote test for mtsSender_test.go to see if the mtsSender is segmenting properly 2019-02-16 16:33:39 +10:30
saxon 4ddf87d63d stream/mts/discontinuity.go: fixed expectedCC logic 2019-02-16 14:38:10 +10:30
saxon 361f5edbc9 stream/mts: wrote sender interface and a minimalHttpSender type 2019-02-16 14:26:51 +10:30
saxon 34daa45b46 Merge branch 'master' into mts-chunking 2019-02-16 09:17:43 +10:30
Saxon Milton 725b40b114 Merged in fix-encoder-timing (pull request #147)
stream/mts & stream/rtp: fixed timing calculations
2019-02-15 22:46:30 +00:00
saxon da5fdccaf3 stream/mts & stream/rtp: fixed timing calculations
Currently time calculations in the mts encoder are based on the premise that each mpegts packet is a 'frame', and in the rtp encoder that
each encoding of multiple packets into one rtp packet is a 'frame', these are both false. Mpegts encoding and rtp encoding should do
timestamp calculations using a reasonable approximation of time at encoding/send.
2019-02-16 09:09:54 +10:30
saxon 281aa47fd5 stream/mts: fixing discontinuities that could be caused by ringbuffer 2019-02-16 01:40:35 +10:30
saxon bea0000340 stream/mts: creating ring.Chunk so that we can use another loadSender 2019-02-15 23:55:51 +10:30
saxon b6bf41b17d stream/mts: made mtsSender more general
The mtsSender can now use any loadSender to send fixed and segmented mts data to.
2019-02-15 23:35:45 +10:30
saxon 31d36577b1 stream/mts: ausOceanSender => mtsSender 2019-02-15 23:17:13 +10:30
saxon 4044368837 stream/mts: fixed commenting in discontinuity.go 2019-02-15 22:24:07 +10:30
saxon 3aa94887eb revid/senders.go: commenting of ausOceanSender 2019-02-15 14:33:18 +10:30
saxon 6964ac513e stream/mts: adding discontinuity.go which contains discontinuity repairer 2019-02-15 13:49:37 +10:30