4 ms·
Interestingly, the video feed gets corrupted around 11th minute, and neither YouTube under Linux/Firefox, nor VLC, nor mpv can play it: the picture freezes, and
by nine_k 6y ago
Interestingly, the video feed gets corrupted around 11th minute, and neither YouTube under Linux/Firefox, nor VLC, nor mpv can play it: the picture freezes, and error diagnostics are printed.
Youtube player under Windows / Firefox does play it through, forcing its way past some garbage frames with noise and rainbow artifacts.
- floatingatoll 6y agoIf you got a snapshot of that corrupted file, I know at least the VLC and Firefox projects would appreciate the bug report, since "freeze + error" is often a short step away from "buffer overflow" — and simply alone the fact that you captured a bytestream under real world circumstances that crashed their players, and have that bytestream for the bug report, is always of maximum interest with such things.
- nine_k 6y agoA great idea; will do.
- Denvercoder9 6y agoWhich video feed? The YouTube stream linked in the submission plays fine under all three players you mentioned for me. SpaceX did have issues with freezing picture around that mark, but that's all on their end (note the clock keeps ticking).
- bostonsre 6y agoYea, would guess transmission issues from the rocket to the ground and the clock is added on the ground somewhere. I wouldn't expect the clouds to interfere if the transmission went up to space then back down (would expect them to do something like that at some point). But maybe microwave from ground to rocket could have had clouds interfere with the signal?
- walrus01 6y agoDownloading it via youtube-dl for the highest quality version also results in a file that at first glance has many freezes in the video stream, when played in VLC. But I think there's nothing wrong with the youtube stream. The freezes were in the downlink from SN15 to the ground. If you go to 9:37 in the video, you can see that the overlaid text timestamp in the spacex stream continues to increment (3:06, 3:07, 3:08, etc). It actually is a video that plays normally. Their ground based equipment that was adding the framing and T+ overlay was working normally and sending out a normal stream, it was just receiving frozen video on its input side. Meaning that they were feeding normally formatted video to youtube at that time. What i got from youtube-dl looks like the following: https://pastebin.com/1JvVY6s4 https://pastebin.com/1JvVY6s4
- nine_k 6y agoThe one I've downloaded is a FullHD h264 stream. It freezes at 11:41, with VLC reporting a few lines of errors: [h264 @ 0x7f4d35907640] get_buffer() failed [h264 @ 0x7f4d35907640] thread_get_buffer() failed [h264 @ 0x7f4d35907640] decode_slice_header error [h264 @ 0x7f4d35907640] no frame! It of course freezes completely, with the overlaid clock and all. VLC version 3.0.13 Vetinari (3.0.12.1-158-g6977abc430), must be pretty fresh.
- walrus01 6y agoNot what I'm seeing here with a version of VLC that's slightly older than yours. At 11:41 in the video, the video feed from starship to ground does freeze, but video continues normally with the overlaid time clock counting seconds. VLC media player 3.0.12 Vetinari (revision 3.0.12-1-0-gd147bb5e7e) I used the default youtube-dl options which downloaded the video as VP9 codec (See above link to pastebin for ffmpeg's info for the file), not H264, so how it's handling a hiccup in the stream may be part of that. available formats show as follows format code extension resolution note 249 webm audio only tiny 51k , opus @ 50k (48000Hz), 5.20MiB 250 webm audio only tiny 66k , opus @ 70k (48000Hz), 5.98MiB 140 m4a audio only tiny 130k , m4a_dash container, mp4a.40.2@128k (44100Hz), 14.90MiB 251 webm audio only tiny 153k , opus @160k (48000Hz), 11.40MiB 278 webm 256x144 144p 73k , webm container, vp9, 30fps, video only, 3.60MiB 160 mp4 256x144 144p 78k , avc1.4d400c, 30fps, video only, 3.57MiB 242 webm 426x240 240p 98k , vp9, 30fps, video only, 4.31MiB 243 webm 640x360 360p 154k , vp9, 30fps, video only, 6.98MiB 133 mp4 426x240 240p 165k , avc1.4d4015, 30fps, video only, 7.32MiB 244 webm 854x480 480p 244k , vp9, 30fps, video only, 10.95MiB 134 mp4 640x360 360p 288k , avc1.4d401e, 30fps, video only, 13.78MiB 247 webm 1280x720 720p 465k , vp9, 30fps, video only, 23.04MiB 135 mp4 854x480 480p 545k , avc1.4d401f, 30fps, video only, 21.75MiB 248 webm 1920x1080 1080p 788k , vp9, 30fps, video only, 26.61MiB 136 mp4 1280x720 720p 1004k , avc1.64001f, 30fps, video only, 39.90MiB 137 mp4 1920x1080 1080p 1985k , avc1.640028, 30fps, video only, 79.93MiB 271 webm 2560x1440 1440p 3509k , vp9, 30fps, video only, 147.72MiB 313 webm 3840x2160 2160p 6214k , vp9, 30fps, video only, 342.33MiB 18 mp4 640x360 360p 210k , avc1.42001E, 30fps, mp4a.40.2@ 96k (44100Hz), 24.25MiB 22 mp4 1280x720 720p 475k , avc1.64001F, 30fps, mp4a.40.2@192k (44100Hz) (best)
- bostonsre 6y agoHow do they get the video stream from the rocket to the ground? Maybe the clouds caused interference?
- bodhiandphysics 6y agoRocket exhaust is highly charged and can cause significant interference! It’s easy to loose a com link
- bostonsre 6y agoInteresting.. how wide of an impact area would the exhaust have? Would it only impact signals traveling directly through it or would there be some bubble of interference around it?
- bodhiandphysics 6y agoIonized gas escapes from the exhaust plume and causes interference in general... the actual aerodynamics are... complicated!