[FFmpeg-user] output length not right

Good morning all, I have here two similar commands, which output two different length of the resulting mov. Is a fairly complex command, so I might be doing something wrong along the line... The first command generates a .mov (h264 )from a sequence of png and mix in a wav (stereo) audio, the outp

Re: [FFmpeg-user] How to turn PNG with alpha channel into MOV with alpha channel?

Hi Nemo, have you been trying this ? http://stackoverflow.com/questions/644684/turn-image-sequence-into-video-with-transparency Cheers. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] single frame duration with concat

Hi Sorry for going back to this, but there's a behaviour I'm not quite getting. if I set the filter as : [in]trim=start_frame=0:end_frame=1[out] it doesn't insert anything at the beginning. if I set the filter as : [in]trim=start_frame=0:end_frame=2[out] It does insert two frames at the beginnin

Re: [FFmpeg-user] drawtext metadata

Thanks Moritz, me neither. But that doesn't mean I can't try (later). What would be great is not having to rely on a fixed set of metadata, but able to pull (as text) any available in the source images. Ta! L. ___ ffmpeg-user mailing list ffmpeg-user@ffm

Re: [FFmpeg-user] drawtext metadata

No i did not. Here it is . hope it helps. http://share.efestolab.uk/pydio/data/public/694f85 L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] drawtext metadata

tried with file.creator but still doesn't pick up anything. Should I be opening a ticket for this ? Thanks L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] drawtext metadata

doesn't look like. efesto@hdd-XPS-15-9530 A010C009_101029_L1O3 $ ffmpeg -i A010C009_101029_L1O3.0284.dpx ffmpeg version N-72356-g8aa9853 Copyright (c) 2000-2015 the FFmpeg developers built with gcc 4.8 (Ubuntu 4.8.2-19ubuntu1) configuration: --prefix=/home/efesto/ffmpeg_build --pkg-config-flag

[FFmpeg-user] drawtext metadata

Gm all. I'm trying to burn in on a clip some informations present on the metadata the a source dpx sequence. I can see a lot of metadata using djv: Time = 2012:08:16:14:52:55:BST Creator = Nuke Source Offset = 0 0 Source Center = 960 540 Source Size = 1920 1080 Source Border = 0 0 0 0 Source

Re: [FFmpeg-user] How to change FPS without re-render

can you try setting -r before and after the input ? L. 2015-06-03 18:55 GMT+02:00 : > I tried: > > ffmpeg -i in.mov -r 24000/1001 out.mov > > which is creating the right fps but encoding it to h264. Then I tried: > > ffmpeg -i in.mov -r 24000/1001 -vcodec copy out.mov > > which is not fixing the

Re: [FFmpeg-user] EXR to DPX Error

No black frame with the exr sequence. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] EXR to DPX Error

Just noticed I picked up the dpx rather than the exr. I'll be doing another test, but still a black frame. L. 2015-06-03 19:52 GMT+02:00 lorenzo angeli : > While keep testing I hit a weird thing though. > > If converting to dpx I found no problem , as well as generating a simple

Re: [FFmpeg-user] EXR to DPX Error

While keep testing I hit a weird thing though. If converting to dpx I found no problem , as well as generating a simple H264 out of them. with a filter_complex * chain the frame 10 comes out black without any reason I can explain. Here a tar.gz with the output mov and the report: http://share.efe

Re: [FFmpeg-user] EXR to DPX Error

Hi Malte, just downloaded and tested. Although I can see the message '[exr @ 0x305a660] Found more than one compression attribute.' , the dpx sequence has been correctly generated. The only thing I did was renaming the sequence to r..exr here is the output: ffmpeg -probesize 500 -analyzedur

Re: [FFmpeg-user] EXR to DPX Error

Hi Malte, I tried download the sequence but didn't let me. For the error you are having : "Found more than one compression attribute." I do have it as well, while converting EXR to something, but apart the issue with PIZ compression, never found it creating issue on the output. L. ___

[FFmpeg-user] curve expression

Gm all , one quick question, is there any way to use expressions in a curve filter ? from the docs I can see stated : *> To avoid some filtergraph syntax conflicts, each key points list need to be defined using the following syntax: x0/y0 x1/y1 x2/y2 * so I guess not, but I though better doub

Re: [FFmpeg-user] EXR to DPX Error

Hi Sanders, out of curiosity , what compression do you have set in the EXR? I found that with PIZ compression , ffmpeg often (if not always) fails . There was a bug reported some time ago but I think that has been fixed already : https://trac.ffmpeg.org/ticket/4386 Hope it helps. L. ___

Re: [FFmpeg-user] Alexa input dpx

Still on the same subject, do you know if there's any way to convert this : float from_func_AlexaV3LogC(float v) { return v > 0.1496582f ? std::pow(10.f,(v - 0.385537f) / 0.2471896f) * 0.18f - 0.00937677f : ( v / 0.9661776f - 0.04378604) * 0.18f - 0.00937677f; } To a curve filter (I guess is the

Re: [FFmpeg-user] single frame duration with concat

Good point, thanks. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] single frame duration with concat

Thanks Lou, trim works like a charm. Cheers. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] Alexa input dpx

While looking for a solution I found this : https://github.com/MrKepzie/Natron/blob/workshop/Engine/Lut.cpp (http://mozilla.org/MPL/2.0, hopefully is compatible with ffmpeg) Natron is a sort of drop in place for Nuke (Foundry) and allowes the basic color transformations on input I do require (as w

[FFmpeg-user] single frame duration with concat

Hi all, I'm trying to concatenate a single image before an image sequence. My aim is to retain the first image for just one frame, and not one second (as I'm doing now). I've been trying all sorts of things, setting -r 25/1 -t 1 -loop 1 but none of these combination seems to be working. Does have

Re: [FFmpeg-user] should I be opening a feature request ?

I thought I did but it seems I didn't . Thanks again. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

[FFmpeg-user] should I be opening a feature request ?

Hi, I popped over this feature request : https://trac.ffmpeg.org/ticket/1949#comment:12 which is about showing the current frame. I've checked around and it seems there's no way to format it though. (eg : 0010 rather than 10.00) Should I be creating a new feature request, or re opening that t

Re: [FFmpeg-user] no such filter

same thought here. thanks for pointing me over the null filter. Cheers! L. --- Efesto Lab LTD, Chancery Station House, 31-33 High Holborn, London WC1V 6AX

Re: [FFmpeg-user] no such filter

Thanks Moritz, sorted out using copy, so I'm sure the data doesn't get touched, and the filter chain gets recognized. thanks for the explanation, out of curiosity, do you think is better use 'null' or 'copy' for doing nothing ? (kind of a weird question I know) Cheers. L. _

Re: [FFmpeg-user] no such filter

ok kind of sorted out, having 'copy' as the only filter*, but if not defined, shoudn't be implied that I just want to copy the input to the output ? *[0:v]copy[clip] Ta. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/l

[FFmpeg-user] no such filter

Hi all, I'm trying to understand why this command doesn't work. I think I'm guessing why (there are no filter set ), but I'm wondering whether is correct that is failing. The streams in and out are (or at least I think are) correctly defined. This doesn't work: -filter_complex "[0:v][clip];[1:v][

Re: [FFmpeg-user] Alexa input dpx

Lut wise I have generated a set from OCIO default configuration with all the various permutations (alexa to log, log to lin etc...) , but I can't find any good combination to apply on the source (alexa) dpx to make it appear right. L. ___ ffmpeg-user mail

Re: [FFmpeg-user] Alexa input dpx

Ok found all I needed about haldclut. Thanks for the help again. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] Alexa input dpx

Thanks to all, Yes the ability to support log dpx would be really appreciated as well. Where can I find more informations about haldcut filter, I tried googling but not much popped up. Thanks everyone for the help ! L. ___ ffmpeg-user mailing list ffmpeg-

Re: [FFmpeg-user] Alexa input dpx

Cheers Carl, here is the complete command with output debug* : http://pastebin.com/wn7j6Bsv The problem is that the input sequence of dpx has an AlexaV3LogC color space, and I'm applying a lut which expect to have that type of color space as input. In Nuke (which is the tool I'm using to check th

[FFmpeg-user] Alexa input dpx

Hi All, No command line here, just an overall question about color management. I have a sequence of DPX which are supposed to be in ALEXAv3LogC, how can I tell ffmpeg to use that color space to read the images in ? Or , what color space ffmpeg assume is by default a sequence of dpx (linear I guess

[FFmpeg-user] keeping one frame for 3

Hi , I'm trying to have one still frame and an image sequence clip joined with concat. It seems to be working fine, but I'm not sure how to make the still image lasting 3 frames. at the moment I'm using : ffmpeg -loop 1 -t 1 -i -i .%4d. --filter_complex "[in][out]" . out.mov What's the c

Re: [FFmpeg-user] Text on multiple lines

Ok I'll try again and let you know. Thanks. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] Text on multiple lines

Sorry did inserted again the previous reply. Damn mobile. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] Text on multiple lines

I'm not sure how to break a line (being just text ), rather than with \r or \n. I've been trying all sort of thing but without success. Wouldn't be possible to have the draw text able to understand these common delimiters? On 18 May 2015 11:43, "Carl Eugen Hoyos"

Re: [FFmpeg-user] Text on multiple lines

Hi Carl, In a way yes(works pressing enter in the command line) . I need to find a way to do it programmatically from python. Ta. L. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] Text on multiple lines

Sorry , no way then to tell ffmpeg to break the line somehow ? Cheers. L. -- --- Efesto Lab LTD, Chancery Station House, 31-33 High Holborn, London WC1V 6AX

Re: [FFmpeg-user] Text on multiple lines

Thanks Carl, got it now, hitting enter seems to be working, but I need to find a good way to make this working from a generated command line. Is there any character I can actually use as replacement for \n to make ffmpeg understand there's a carriage return ? Cheers. L. ___

Re: [FFmpeg-user] Text on multiple lines

color=black@0.5:text='two>lines'[out]" out.mov [/code] but the text still remains on the same line. Thanks. L. Ps. -v trace was just badly placed. 2015-05-16 10:40 GMT+02:00 Carl Eugen Hoyos : > lorenzo angeli efestolab.uk> writes: > > > -v trace > > W

Re: [FFmpeg-user] Text on multiple lines

7% 3.3% 0.4% direct: 0.9% skip:51.2% L0:31.1% L1:67.7% BI: 1.2% [libx264 @ 0x3ee3540] 8x8 transform intra:82.3% inter:89.2% [libx264 @ 0x3ee3540] coded y,u,v intra: 50.0% 1.1% 0.3% inter: 20.2% 0.2% 0.0% [libx264 @ 0x3ee3540] i16 v,h,dc,p: 12% 40% 5% 43% [libx264 @ 0x3ee3540] i8 v,h,dc,ddl,

[FFmpeg-user] Text on multiple lines

Hi all, I found this old thread on StackOverflow: http://stackoverflow.com/questions/8213865/ffmpeg-drawtext-over-multiple-lines on the subject of having text split on multiple lines. I tried to replace \n with \f or other suggested, but I've got no luck. Is there any way I can simply take a text

Re: [FFmpeg-user] help with file paths under windows

THanks Moritz, I'll be having a second look at it. Cheers! L. 2015-05-15 13:12 GMT+02:00 Moritz Barsnick : > On Thu, May 14, 2015 at 17:39:46 +0200, lorenzo angeli wrote: > > Now, this seems all to be working fine under linux, but on winows, I > > started finding a bunch of i

[FFmpeg-user] help with file paths under windows

Hi All, first of thanks in advance for anyone which will be able to help me out. I'm developing a command line tool for simplifying the generation of review material , which includes a slate (a first image with notes) and then the proper clip with text, masks , fonts, luts etc . The final comm