[00:16:01 CEST] <Guster_> hi there [00:16:19 CEST] <Guster_> can anyone "new" help me plese with ffmpeg [00:23:27 CEST] <Draz> Padding on video appears grey: Applying pad filter to video stream gives non-black padding. http://pastebin.com/vuHEVSGZ [00:24:16 CEST] <Draz> I've tried a few things, including applying format filters and outputting with a different codec, but no success. [00:24:29 CEST] <Draz> Help appreciate. [00:24:41 CEST] <Draz> *appreciated. [00:25:10 CEST] <Draz> Guster_, "When asking a question be precise, detailed...". What's your issue? [00:25:46 CEST] <Guster_> cant unicast with ffmpeg to ffserver [00:26:20 CEST] <Guster_> there is log from ffmpeg [00:26:21 CEST] <Guster_> http://pastebin.com/fXw9qw6V [00:26:44 CEST] <Guster_> at ffserver just Fri Apr 22 00:12:30 2016 127.0.0.1 - - [POST] "/feed1.avi HTTP/1.1" 200 1282 [04:03:52 CEST] <anadon> People kicking around? [04:28:26 CEST] <DHE> anadon: standard procedure is to ask your question and wait [04:36:43 CEST] <anadon> DHE: I know, but some channels stay dead for a very long time. I just ping once in awhile. [04:38:30 CEST] <DHE> on the other hand it's late in north america, night in europe, and you've wasted valuable time playing IRC tag already. [04:39:15 CEST] <anadon> DHE: Don't heap on a mistake. [04:40:13 CEST] <DHE> sorry, just trying to make a point [05:37:40 CEST] <wyatt8740> Is there a way to make ffmpeg print ONLY info? (meaning I don't want it to print warnings or errors, but only print the information about encoding that you'd get if it were going smoothly) [05:38:54 CEST] <wyatt8740> I get spammed by distracting 'Past duration [number here] too large' when combining two streams (video and audio devices) into a single capture [05:39:53 CEST] <anadon> Could you grep through the output [05:40:07 CEST] <wyatt8740> not while preserving colors, since I'd have to redirect stderr [05:40:12 CEST] <wyatt8740> info seems to print to stderr too [05:40:33 CEST] <wyatt8740> and it behaves differently by not using colors if you put it through a pipe [05:47:34 CEST] <wyatt8740> also, grepping seems to prevent the progress from being displayed [05:47:40 CEST] <wyatt8740> probably due to buffering or something [05:48:31 CEST] <anadon> I'd wait for someone more expert than me (I'm not at all) [05:48:51 CEST] <wyatt8740> fiddling with socat but I never was able to make this work for anything else so my hopes are low [05:49:20 CEST] <wyatt8740> (once tried to log all communication between my real RS-232 terminal and my PC with socat, it failed miserably) [05:49:32 CEST] <anadon> That's dedication [05:50:33 CEST] <wyatt8740> the terminal is really cool and I'm trying to make it work as well as possible :\ [05:50:44 CEST] <wyatt8740> the termcap definitions seem to be missing a few things [05:51:08 CEST] <wyatt8740> also I was able to discover and report a bug in GNU Nano where syntax highlighting caused the entire screen to redraw with every keystroke [05:51:31 CEST] <wyatt8740> it happened regardless of terminal but in modern terminal emulators it was so fast no one noticed :| [05:52:17 CEST] <anadon> wyatt8740: Sounds l=more like you should be bugging #gnu or something. [05:52:19 CEST] <wyatt8740> http://i.imgur.com/DKBqQ2b.jpg [05:52:27 CEST] <wyatt8740> no, this is a ffmpeg problem [05:52:30 CEST] <wyatt8740> I got the nano bug fixed [05:52:31 CEST] <wyatt8740> :) [05:52:49 CEST] <wyatt8740> and this ffmpeg problem is not 1985 terminal related [05:53:05 CEST] <anadon> nethack? [05:53:16 CEST] <wyatt8740> that's rogue [05:53:22 CEST] <wyatt8740> I do have nethack here too though [05:53:33 CEST] <anadon> On a 1985 machine? [05:53:49 CEST] <wyatt8740> no, lol, that's a terminal [05:54:09 CEST] <wyatt8740> my i5-3350p desktop PC has rogue and nethack on it :\ [05:54:27 CEST] <wyatt8740> also a serial port [05:56:05 CEST] <wyatt8740> anyway, if anyone can help me to make ffmpeg not print errors or warnings but still print info messages, please speak up :) [06:35:35 CEST] <furq> wyatt8740: -v quiet -stats [06:39:54 CEST] <furq> also i'm disappointed that someone with a 3161 on his desk isn't using irssi [06:54:20 CEST] <wyatt8740> furq: nice, you recognized the model [06:54:23 CEST] <wyatt8740> did you use one? [06:59:08 CEST] <wyatt8740> also when I find a pic I'll show you - I do use irssi on it [06:59:14 CEST] <wyatt8740> just not in that pic :\ [07:06:11 CEST] <wyatt8740> durandal_1707: I forget, isn't your name taken from Roland's sword? (Roland the knight of charlemagne) [07:06:36 CEST] <furq> yeah i definitely didn't type "ibm terminal" into google images and then find another picture of your desk [07:06:48 CEST] <furq> that's not how i know the model number [07:09:29 CEST] <wyatt8740> furq: oh, okay. But there are about a million and a half IBM terminals [07:09:41 CEST] <wyatt8740> and I would've been too lazy :\ [07:09:56 CEST] <furq> well your desk is on the second row so it worked out great [07:10:06 CEST] <wyatt8740> lol [07:10:09 CEST] <wyatt8740> is it now? [07:10:31 CEST] <wyatt8740> so it is [07:10:36 CEST] <wyatt8740> with the amber screen I need to fix [07:13:48 CEST] <wyatt8740> most of their offerings [07:13:54 CEST] <wyatt8740> *don't even work on normal PCs [07:13:58 CEST] <wyatt8740> you need to have a IBM mainframe [08:17:55 CEST] <wyatt8740> I really think ffmpeg is one of the most awesome free/OSS projects out there. I'd say it along with the linux kernel and maybe QEMU (...and ffmpeg and qemu are both things that were started by Fabrice Bellard. Wow). [09:17:16 CEST] <thebombzen> does anyone know if the drawtext filter supports WOFF2? [09:28:35 CEST] <xbloodhound> I have been trying to add two separate animations as png watermarks using something like this -filter_complex "[0:v][1:v][2:v] overlay=0:0" but I am getting error that says too many inputs specified for overlay filter etc. so my question is it possible to add two overlays at all or am I doing something wrong? [09:29:53 CEST] <xbloodhound> I only know -filter_complex "[0:v][1:v] overlay=0:0" works at least [09:32:00 CEST] <furq> xbloodhound: [0:v][1:v]overlay=0:0[tmp];[tmp][2:v]overlay=0:0[out] [09:43:39 CEST] <xbloodhound> ugh that worked thx [09:43:58 CEST] <xbloodhound> been messing around for a while [09:59:25 CEST] <termos> is there a way to reduce the blocking time of avio_open? it's taking 500ms each call to open rtmp:// sockets, it's too long when opening 20+ of them [11:04:54 CEST] <Eiken_> when i apply a cube lut generated in nuke the output looks exactly the same [11:05:21 CEST] <Eiken_> -vf lut3d=/mnt/home/erjo/testlut.cube [11:05:24 CEST] <Eiken_> is this incorrect? [11:50:54 CEST] <durandal_1707> wyatt8740: nope [11:58:43 CEST] <aspiers> hi all, any ideas why ffmpeg 3.0.1 would record audio perfectly via pulseaudio, but when I add -f x11grab for screencasting, the audio is truncated at the last ~3s of every capture ? [11:59:02 CEST] <aspiers> and BTW a/v sync is perfect in this scenario [11:59:52 CEST] <aspiers> I used: ffmpeg -thread_queue_size 512 -f pulse -i default -ac 1 -f x11grab -video_size 1280x720 -framerate 30 -i :0.0+0+0 -y test.mp4 [13:41:03 CEST] <GusterChe> hi there [13:41:23 CEST] <GusterChe> me again like last night :D [13:41:41 CEST] <GusterChe> same problem with my ffmpeg and ffserver :D [13:42:37 CEST] <GusterChe> anyone with experience to pvt? :D [13:54:32 CEST] <GusterChe> there are all outputs from ffserver and ffmpeg with ffserver.conf http://pastebin.com/1KxwWQDZ if soneone know what i did wrong please help [14:01:05 CEST] <GusterChe> is anyone online? xD [14:13:17 CEST] <GusterChe> hi oldcode [14:14:36 CEST] <GusterChe> hi fracting :D [14:15:07 CEST] <BtbN> Don't use ffserver if you can't figure stuff out on your own. [14:27:08 CEST] <GusterChe> there are all outputs from ffserver and ffmpeg with ffserver.conf http://pastebin.com/1KxwWQDZ if soneone know what i did wrong please help [15:26:00 CEST] Last message repeated 1 time(s). [15:29:35 CEST] <Joeboy> Hi. I'm trying to convert some stereo files to mono, but the resulting files seem to get converted from s24le to s16le. Is there a way I can make ffmpeg not do that? [15:29:38 CEST] <Joeboy> I'm doing: [15:30:05 CEST] <Joeboy> ffmpeg -y -i input.wav -map_channel 0.0.0 junk.wav -map_channel 0.0.1 output.wav [15:36:17 CEST] <furq> -c:a pcm_s24le [15:36:36 CEST] <Joeboy> Actually -acodec copy seems to work [15:36:37 CEST] <furq> or maybe -c:a copy will work but i don't think it'll make any difference [15:36:50 CEST] <furq> yeah that'll do it [15:36:56 CEST] <Joeboy> Sorry, should have experimented more before asking [15:38:03 CEST] <furq> if you don't specify then ffmpeg will pick whatever output codec it thinks is best for the output container [15:38:14 CEST] <furq> which for wav is pcm_s16le [15:38:39 CEST] <GusterChe> furq: can u help me with my problem? [15:39:08 CEST] <furq> i can tell you to not use ffserver because it's basically abandoned and it sucks [15:39:10 CEST] <furq> does that help [15:39:38 CEST] <GusterChe> yes xD [15:39:52 CEST] <GusterChe> and, i have one question more if u have some time :D [15:40:38 CEST] <furq> well that depends on how long it would take to answer the question [15:40:41 CEST] <GusterChe> if i want to use nginx and rtmp module. do u can give me guildlines? :) [15:40:55 CEST] <furq> https://gist.github.com/fur-q/d7028f51c38f7d0bb56e [15:40:56 CEST] <furq> how's that [15:41:28 CEST] <furq> you can ignore the lua stuff if you don't care about auth [15:43:37 CEST] <GusterChe> and how to set unicast to server with ffmpeg? [15:44:09 CEST] <furq> it's rtmp, it only does unicast [15:56:37 CEST] <GusterChe> furq: thanks, this works fine for now [15:56:53 CEST] <GusterChe> i go to test it for another things [15:57:05 CEST] <GusterChe> thanks a lot and cya! [17:56:39 CEST] <Joeboy> Ok, I asked a question earlier and thought I'd got it sorted but it turns out I hadn't. So I'm going to ask it again. [17:57:07 CEST] <Joeboy> I'm trying to convert an s24le stereo file into two s24le mono files. [17:57:58 CEST] <Joeboy> (actually lots of files) [17:58:17 CEST] <Joeboy> I got as far as: ffmpeg -y -i $f -map_channel 0.0.0 junk.wav -acodec copy -map_channel 0.0.1 "gve-london-audio/oktava/$f" [17:58:48 CEST] <Joeboy> But that gives me stereo files, for some reason. [17:59:06 CEST] <Joeboy> without the -acodec copy, I get s16le mono files. [17:59:51 CEST] <Joeboy> I just want to do a raw copy of one track from the stereo file into a mono file. [18:01:02 CEST] <Joeboy> ffmpeg (at least my version) doesn't seem to have any 24 bit output options [18:09:24 CEST] <durandal_1707> Joeboy: -codec:a pcm_s24le [18:13:08 CEST] <Joeboy> durandal_1707: Thanks! [20:27:12 CEST] <johnnny22-afk> i have a mpegts that starts at some pts, is there a way to re-encode that pts/dts, so that it starts at 0 ? [20:28:23 CEST] <c_14> ffmpeg should do that by default [20:49:25 CEST] <johnnny22-afk> m'ok, but I have a 39h mpegts, which has a rollover around 13h or so from the beginning. ffprobe seems to tell me that the duration is about 13h long. After re-encoding it, ffprobe still tells me that the duration is about 13h. I was hoping to make the duration be ~26h or so, being that max duration of a mpegts without rollovers.. Should I be adjusting something else ? [20:49:55 CEST] <johnnny22-afk> maybe it's the DTS value that should be changed ? [20:52:49 CEST] <c_14> ffprobe doesn't actually go through the entire video to find the duration, it usually either uses a value stored in the container [20:56:17 CEST] <aspiers> hi all, any ideas why ffmpeg 3.0.1 would record audio perfectly via pulseaudio, but when I add -f x11grab for screencasting, the audio is truncated at the last ~3s of every capture ? [20:56:22 CEST] <aspiers> I used: ffmpeg -thread_queue_size 512 -f pulse -i default -ac 1 -f x11grab -video_size 1280x720 -framerate 30 -i :0.0+0+0 -y test.mp4 [20:57:09 CEST] <c_14> aspiers: is this a regression? [20:57:39 CEST] <aspiers> c_14: since the last time I tried it, yes - but that was at least 6-12 months ago [20:58:05 CEST] <aspiers> c_14: A/V sync is still flawless, however [20:59:37 CEST] <c_14> Can you test with a recent git build? If it still doesn't work with that, open a ticket on trac and mention that it's a regression (preferably mention which release it last worked with) [21:00:09 CEST] <aspiers> c_14: I already downloaded a static build from git, but it was missing support for pulseaudio :-/ [21:00:16 CEST] <aspiers> and I don't have time to compile one unfortunately [21:01:16 CEST] <c_14> You can make a bugreport and mention that you only tested with 3.0.1, but the first thing you're going to be asked is if you can still reproduce with a build from git [21:01:44 CEST] <aspiers> hah [21:02:26 CEST] <aspiers> yeah, that's fair - although it would only take someone who already has a git build ~20s to copy/paste my command and see if it's the same for them ... [21:02:40 CEST] Action: c_14 doesn't have pulse [21:02:42 CEST] <c_14> therefore can't test [21:02:45 CEST] <aspiers> ok [21:04:30 CEST] <johnnny22-afk> c_14: aah ok ,right. In my case it's a .ts file, so I'm not sure where it takes that duration value. [21:04:53 CEST] <johnnny22-afk> c_14: is there a way to make it check the whole file to compute the duration of a particular .ts file ? [21:06:27 CEST] <johnnny22-afk> maybe -formatprobesize to the whole size of the file ? [21:07:19 CEST] <c_14> There is -count_packets and -count_frames, but I don't think either is used to get a duration, you can use ffmpeg -i input -f null /dev/null and the last line of output will contain the total duration [21:08:42 CEST] <johnnny22-afk> yeah, i guess that would be one way :) [21:09:08 CEST] <johnnny22-afk> adding -c copy for speed : P [21:11:17 CEST] <CoffeeFluxx> Heyo! I ran a command to mux some h264 video and aac audio into mp4, but somehow it's creating an empty srt track http://pastebin.com/raw/q8T2Ja4K [21:11:56 CEST] <CoffeeFluxx> Any idea what I'm doing wrong? Neither of the files I'm muxing in has subs, so I'm a bit confused where that came from [21:12:14 CEST] <c_14> CoffeeFluxx: the m4a presumably had a subtitle trac [21:12:25 CEST] <c_14> can you pastebin the complete console output of the first command? [21:12:32 CEST] <c_14> also, it's not srt. It's mov_text [21:12:32 CEST] <CoffeeFluxx> sure [21:13:40 CEST] <CoffeeFluxx> http://pastebin.com/raw/ccnWubmm [21:15:02 CEST] <c_14> hmm, weird [21:15:14 CEST] <c_14> you should be able to get rid of it with -sn though [21:16:38 CEST] <CoffeeFluxx> yeah, i checked both the input files and it wasn't there [21:19:53 CEST] <CoffeeFluxx> it's still there, even after adding -sn [21:20:08 CEST] <CoffeeFluxx> lemme pastebin [21:20:24 CEST] <c_14> add -map_metadata -0 [21:22:01 CEST] <CoffeeFluxx> http://pastebin.com/raw/ytfy4Nth [21:25:16 CEST] <CoffeeFluxx> added the -map_metadata -0, but it's still there [21:26:41 CEST] <c_14> eeh, honestly no clue. Try updating to a newer build? Does the mov_text stream always show up or only with those two files? [21:28:19 CEST] <CoffeeFluxx> no clue, normally i'm just muxing to mkv so i use a different program [21:31:59 CEST] <c_14> Does a subtitle track show up when muxing those 2 files to mkv instead of mp4? [21:35:45 CEST] <CoffeeFluxx> conversion fails [21:35:45 CEST] <CoffeeFluxx> http://pastebin.com/raw/fSYti06j [21:36:16 CEST] <c_14> well, that's fun [21:36:20 CEST] <c_14> try mpegts? [21:37:54 CEST] <CoffeeFluxx> same error [21:40:13 CEST] <c_14> try muxing to mkv with the older build (if you still have that one) [21:42:21 CEST] <CoffeeFluxx> nope :\ [21:42:23 CEST] <CoffeeFluxx> lemme grab one [21:44:17 CEST] <CoffeeFluxx> hmm i can't find older windows builds [21:45:17 CEST] <kepstin> the issue with your most recent command is that the m4a file has a piece of cover art that's being encoded as video [21:45:31 CEST] <kepstin> you'll want to use some -map parameters to select specific streams [21:45:42 CEST] <kepstin> e.g. -map 0:a -map 1:v for your most recent example [21:45:51 CEST] <kepstin> that'll also solve the subtitle issue [21:46:55 CEST] <CoffeeFluxx> i thik i tried that earlier and it didn't get rid of the subs [21:47:03 CEST] <CoffeeFluxx> the video is definitely correctly in the output [21:47:07 CEST] <CoffeeFluxx> i'll try it again, though [21:50:14 CEST] <CoffeeFluxx> yeah the subs are still there [21:54:09 CEST] <c_14> It's probably a bug. I'd open an issue on trac, though if there's no good way to reproduce the issue... If you could upload the files (or parts thereof) to reproduce that would probably help [21:54:24 CEST] <CoffeeFluxx> huh, re-encoded the audio with -map 0:a and it appeared there too [21:57:18 CEST] <CoffeeFluxx> i'll file a report, though unfortunately i can't upload the files :\ [21:57:48 CEST] <CoffeeFluxx> i'll see if i can repro by encoding some other stuff the same way, though [22:19:57 CEST] <spirou> if I want to let ffmpeg convert 5.1 sound in a video to normal stereo, would it be very complicated parameters? [22:20:40 CEST] <JEEB> if input audio channels are mapped in a normal way and in a single track - it should be as simple as -ac 2 after input [22:21:11 CEST] <JEEB> it only gets complicated if you have to start doing channel mapping or multiple tracks into channels of a single track [22:21:13 CEST] <spirou> since I have to let it reencode the sound anyway to make the TV accept it, I guess I could tell it to so stereo [22:22:57 CEST] <spirou> (I have a mp4, that the tv didn't like, and when I had ffmpeg put it in a mkv instead (with -c copy) there was no sound, so I assume the tv don't like that kind of sound, or that kind of sound in a mkv) [22:23:02 CEST] <spirou> "Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz, 5.1, fltp, 447 kb/s (default)" [22:24:38 CEST] <spirou> JEEB: ok so like this then? ffmpeg -i in.mp4 -ac 2 -vcodec copy out.mkv [22:24:54 CEST] <JEEB> add -c:a aac -b:a 192k or so [22:25:20 CEST] <JEEB> if it complains about experimentality then update FFmpeg or add -strict experimental which enables "experimental" features [22:26:08 CEST] <JEEB> and you might want to use -c:v copy for video in the short way as well in later commands :) [22:27:15 CEST] <spirou> aha, the "-c:v copy" would be the new way of say "-vcodec copy" [22:27:22 CEST] <JEEB> yup [22:30:09 CEST] <spirou> yeah I had to write -strict experimental after the input file. I assume it is same thing as -strict -2 as it suggested [22:30:31 CEST] <spirou> btw, what do "[matroska @ 0xa79c120] Codec for stream 0 does not use global headers but container format requires global headers" means? [22:31:41 CEST] <JEEB> spirou: yeah - except the text-based value is generally more understandable than "-2" which is the internal value :) [22:31:57 CEST] <JEEB> also you should generally update your FFmpeg [22:32:08 CEST] <JEEB> the aac encoder was vastly improved recently [22:32:14 CEST] <JEEB> and the experimental flag was removed [22:32:36 CEST] <JEEB> among other improvements [22:35:12 CEST] <spirou> ah yeah I'm still on "7:2.8.3+git1~trusty1", I haven't updated for some time... the newest would be "7:3.0.0+git1~trusty1" it says here [22:37:01 CEST] <spirou> I'll should remember to let it do its updates before doing anything I want to store permanently [22:39:33 CEST] <epifanio> Hi [22:39:50 CEST] <spirou> Ideally I should try to make some ffprobe-using script that could tell if the TV would be OK with a file... [22:39:55 CEST] <spirou> hello epifanio [22:41:33 CEST] <epifanio> Im dealing with 2 windows applications (for which I do not have access to the src code), the first one takes as input a directory with png and generates a .avi as output. The second application takes the .avi (previously generated) as input, and try to make a mosaic from each frame [22:41:56 CEST] <epifanio> i want replace the first application with some open source code, using ffmpeg [22:42:32 CEST] <CoffeeFluxx> submitted a bug report - any suggestions in the meantime on how to work around the issue? [22:42:53 CEST] <epifanio> i do know how to generate a video using a sequence of pngs with ffmpg, but my output is not compatible wiht the input video needed by the second application [22:43:42 CEST] <epifanio> so, I tried to run ffmpeg -i goodoutput.avi on the file generated by the first application, this is the output https://gist.github.com/e7ef38889217aee0fa1e1aefceeaf2a8 [22:44:39 CEST] <epifanio> can you suggest me the correct syntax/codec i need to use with ffmpeg to generate a video of the same type as the one described by the ffmpeg -i output? [22:44:45 CEST] <kepstin> epifanio: should be as simple as setting the codec to match, "-c:v rawvideo -pix_fmt bgr24" [23:13:34 CEST] <epifanio> kepstin: Thank you! Im trying with: !cat *.png | ffmpeg -framerate 5 -f image2pipe -c:v png -i - -c:v rawvideo -pix_fmt bgr24 test3.avi [23:14:26 CEST] <epifanio> ive a video output, but the framerate is not accepted this the log: https://gist.github.com/31592fd73aa4569a73dd73a30422e31a [23:15:34 CEST] <epifanio> debug [23:16:19 CEST] <epifanio> the app which triys to open the video goes in segfault [23:17:58 CEST] <epifanio> this the log of ffmpg -i on the output video i used: https://gist.github.com/74270e975ae79348c44c63db6467cd3c [23:17:58 CEST] <epifanio> [23:18:15 CEST] <epifanio> the fps are set correctly [23:20:51 CEST] <epifanio> ill try again ffmpeg -i on both files and post here to double check their format match [23:24:03 CEST] <epifanio> the output is https://gist.github.com/epifanio/f9ec086b2f946a56583cf46f1de47fcf left.avi is the video which works as input for the app Im using, test3.avi is the one i generated with ffmpeg [23:26:24 CEST] <kepstin> epifanio: I don't see any notable differences. [23:26:46 CEST] <kepstin> if the application that tries to open it segfaults, well, that's a pretty bad application bug... [23:27:01 CEST] <kepstin> that said, you might want to try making the height/width a multiple of 8 or 16. [23:27:12 CEST] <epifanio> it is a vey bad app which they are forcing me to use .. :( [23:27:26 CEST] <kepstin> at this point, you're just debugging that application :/ [23:29:44 CEST] <epifanio> they just dont care, they give me an app which work (please let me use this term & I know) that takes a directory with some png and make a video out of it & my problem is that i have 70.000 images and i need to group them in subfolders of 20 images each .. then make videos for each subfolder [23:30:31 CEST] <epifanio> so i wanted to script my way to generate those videos, on my unix machine .. rhater than klick who knows how many times on a windows application gui [23:31:22 CEST] <epifanio> i saw a little difference in the metadata: encoder : Lavf57.25.100 [23:31:41 CEST] <epifanio> and the size: 330x330 vs 1360x1024 [23:33:47 CEST] <kepstin> yeah, I suspect the size is the main issue [23:36:24 CEST] <epifanio> how can i force the output to match this size? which is actually the original size of the png, a sample is: http://epinux.com/hbc_img/selection/left/201503.20150619.181140656.204627.pn... [00:00:00 CEST] --- Sat Apr 23 2016