[HEVC] x265.EXE: mingw builds

[vhelp]

i had a lot of trouble with different parts of the tools code, like the console piping, (had to rebuild from scratch) and then many hours with encorporating a progress bar, (couldn't get each of the elements parsed correctly) and a few other code parts. but those parts are now working, as you can see in the upgraded version below. i want to add one more feature, maybe two, before i upload something. it is not a pretty looking gui but works for me.

[Marchand]

You should change your avatar to Dungeon Master .... this is amazing .... in a few hours and you've got ..... My congratulations .....I'll wait for your wonderful work. Thank Dungeon Master ...

[selo3000]

i had a lot of trouble with different parts of the tools code, like the console piping, (had to rebuild from scratch) and then many hours with encorporating a progress bar, (couldn't get each of the elements parsed correctly) and a few other code parts. but those parts are now working, as you can see in the upgraded version below. i want to add one more feature, maybe two, before i upload something. it is not a pretty looking gui but works for me.

"gui for video encoding"


program for end-user download link?

[El Heggunte]

NEW build:= x265_0.3+437-fc12faa1672e.7z

Slightly less slow than before, apparently

[DarrellS]

no stdin support still. it is at the very bottom of the consortium's todo list though the race is on for someone, anyone, to add that support. maybe possbile through another language via .dll creativity. don't know how to call .dll functions from c/c++/c# from other languages though.

but i can't get the frame number from avs2yuv for some reason
right now i plan to extract mediainfo->vid length(source vid for avs)->frame count->filesize

then use ffmpeg to convert to .y4m video, then you read "its" studin/output data with your front-end gui tool:

ffmpeg -i video.avs -an -f yuv4mpegpipe -pix_fmt yuv420p -y video.y4m

i always add -y because i sometimes get a message to overwrite or not. this way no interuption.

nothing is overkill as long as is the only objective or objective(s) any/all work

you may be able to "pipe" to x265 w/out intermedium file in this way.
edit: piping does not work even though some website say you can, but i tried many variations and nothing works. guess we are still stuck with yuv or y4m intermediates.

I guess it should be no surprise... There is not one single MPEG command line encoder that accepts stdin input in the whole world. Well, ffmpeg . They'll accept avs scripts or ini files but no "stdin" or "-" input. There are a lot of popular command line encoders that do not support stdin. A lot of geniuses out there developing encoding software but hardly any of them think that stdin is important. Usually our only hope is that ffmpeg will implement those encoders but they seem afraid to include a lot of popular encoders. Hopefully they aren't afraid to implement the x265 encoder and are just waiting for a finished product so they can implement both the encoder and decoder.

[El Heggunte]

.............

I guess it should be no surprise... There is not one single MPEG command line encoder that accepts stdin input in the whole world. Well, ffmpeg . They'll accept avs scripts or ini files but no "stdin" or "-" input.

Yes, there exists --- mpeg2enc:

[C:\]
=>mpeg2enc
**ERROR: [mpeg2enc] Output file name (-o option) is required!
--verbose|-v num
    Level of verbosity. 0 = quiet, 1 = normal 2 = verbose/debug
--format|-f fmt
    Encoding profile
    [0 = Generic MPEG1, 1 = standard VCD, 2 = user VCD,
     3 = Generic MPEG2, 4 = standard SVCD, 5 = user SVCD,
     6 = VCD Stills sequences, 7 = SVCD Stills sequences, 8|9 = DVD,
     10 = ATSC 480i, 11 = ATSC 480p, 12 = ATSC 720p, 13 = ATSC 1080i]
--display-hsize|-x [32..16383]
   Set the the display-horizontal-size hint in MPEG-2 output to
   something other than the encoded image width
--display-vsize|-y [32..16383]
   Set the the display-vertical-size hint in MPEG-2 output to
   something other than the encoded image height
--aspect|-a num
    Set displayed image aspect ratio image (default: 2 = 4:3)
    [1 = 1:1, 2 = 4:3, 3 = 16:9, 4 = 2.21:1]
--frame-rate|-F num
    Set playback frame rate of encoded video
    (default: frame rate of input stream)
    0 = Display frame rate code table
--video-bitrate|-b num
    Set Bitrate / peak bitrate of compressed video in KBit/sec
    (Peak bitrate if a target bitrate and/or quantisation floor is set
    (default: 1152.0 for VCD, 2500.0 for SVCD, 7500.0 for DVD)
--target-video-bitrate|-t
   Set target bitrate for entire video stream in KBit/sec
--nonvideo-bitrate|-B num
    Non-video data bitrate to assume for sequence splitting
    calculations (see also --sequence-length).
--quantisation|-q num
    Image data quantisation factor [1..31] (1 is best quality, no default)
    When quantisation is set variable bit-rate encoding is activated and
    the --bitrate value sets an *upper-bound* video data-rate
--ratecontroller|-A [0..1] (default:0)
    Specify ratecontrol alorithm
--output|-o pathname
    Pathname of output file or fifo (REQUIRED!!!)
--target-still-size|-T size
    Size in KB of VCD stills
--interlace-mode|-I num
    Sets MPEG 2 motion estimation and encoding modes:
    0 = Progressive (non-interlaced)(Movies)
    1 = Interlaced source material (video)
    2 = Interlaced source material, per-field-encoding (video)
--motion-search-radius|-r num
    Motion compensation search radius [0..32] (default 16)
--reduction-4x4|-4 num
    Reduction factor for 4x4 subsampled candidate motion estimates
    [1..4] [1 = max quality, 4 = max. speed] (default: 2)
--reduction-2x2|-2 num
    Reduction factor for 2x2 subsampled candidate motion estimates
    [1..4] [1 = max quality, 4 = max. speed] (default: 3)
--min-gop-size|-g num
    Minimum size Group-of-Pictures (default depends on selected format)
--max-gop-size|-G num
    Maximum size Group-of-Pictures (default depends on selected format)
    If min-gop is less than max-gop, mpeg2enc attempts to place GOP
    boundaries to coincide with scene changes
--closed-gop|-c
    All Group-of-Pictures are closed.  Useful for authoring multi-angle DVD
--force-b-b-p|-P
    Preserve two B frames between I/P frames when placing GOP boundaries
--quantisation-reduction|-Q num
    Max. quantisation reduction for highly active blocks
    [0.0 .. 4.0] (default: 0.0)
--quant-reduction-max-var|-X num
    Luma variance below which quantisation boost (-Q) is used
    [0.0 .. 2500.0](default: 0.0)
--video-buffer|-V num
    Target decoders video buffer size in KB (default 46)
--video-norm|-n n|p|s
    Tag output to suit playback in specified video norm
    (n = NTSC, p = PAL, s = SECAM) (default: PAL)
--sequence-length|-S num
    Place a sequence boundary in the video stream so they occur every
    num Mbytes once the video is multiplexed with audio etc.
    N.b. --non-video-bitrate is used to the bitrate of the other
    data that will be multiplexed with this video stream
--3-2-pulldown|-p
    Generate header flags for 3-2 pull down of 24fps movie material
--intra_dc_prec|-D [8..11]
    Set number of bits precision for DC (base colour) of blocks in MPEG-2
--reduce-hf|-N num
    [0.0..2.0] Reduce hf resolution (increase quantization) by num (default: 0.0)
--keep-hf|-H
    Maximise high-frequency resolution - useful for high quality sources
    and/or high bit-rates)
--sequence-header-every-gop|-s
    Include a sequence header every GOP if the selected format doesn't
    do so by default.
--no-dummy-svcd-SOF|-d
    Do not generate dummy SVCD scan-data for the ISO CD image
    generator "vcdimager" to fill in.
--playback-field-order|-z b|t
    Force setting of playback field order to bottom or top first
--multi-thread|-M num
    Activate multi-threading to optimise throughput on a system with num CPU's
    [0..32], 0=no multithreading, (default: 0)
--correct-svcd-hds|-C
    Force SVCD horizontal_display_size to be 480 - standards say 540 or 720
    But many DVD/SVCD players screw up with these values.
--no-constraints
    Deactivate constraints for maximum video resolution and sample rate.
    Could expose bugs in the software at very high resolutions!
--no-altscan-mpeg2
    Deactivate the use of the alternate block pattern for MPEG-2.  This is
    A work-around for a Bug in an obscure hardware decoder.
--dualprime-mpeg2
    Turn ON use of dual-prime motion compensation. Default is OFF unless this option is used
--custom-quant-matrices|-K kvcd|tmpgenc|default|hi-res|file=inputfile|help
    Request custom or userspecified (from a file) quantization matrices
--unit-coeff-elim|-E num
    Skip picture blocks which appear to carry little information
    because they code to only unit coefficients. The number specifies
    how aggresively this should be done. A negative value means DC
    coefficients are included.  Reasonable values -40 to 40
--b-per-refframe| -R 0|1|2
    The number of B frames to generate between each I/P frame
--cbr|-u
    For MPEG-2 force the use of (suboptimal) ConstantBitRate (CBR) encoding
--chapters X[,Y[,...]]
    Specifies which frames should be chapter points (first frame is 0)
    Chapter points are I frames on closed GOP's.
--help|-?
    Print this lot out!

[C:\]
=>

[vhelp]

due to the way the newest commit of x265 fc12faa updates the progress string, the last part no longers says [100.0%] complete. so my app no longer functioned properly..progressbar would show less then 100% completed. i had to write a workaround which i don't like because i wonder if that change in the fc12faa effected the final output of those frames. i wonder if other apps had this problem and had to be updated also. maybe ozok should give his gui a test run (i can't test on my pc bc it doesn't run on my xp bc mp4box fails at the muxing and crashes) and report back if his doesn't have this issue. or, maybe its just my implementation of piping in my app. my app is not quite ready yet, thanks to most time spent debugging this issue.

[DarrellS]

.............

I guess it should be no surprise... There is not one single MPEG command line encoder that accepts stdin input in the whole world. Well, ffmpeg . They'll accept avs scripts or ini files but no "stdin" or "-" input.

Yes, there exists --- mpeg2enc:

[C:\]
=>mpeg2enc
**ERROR: [mpeg2enc] Output file name (-o option) is required!
--verbose|-v num
    Level of verbosity. 0 = quiet, 1 = normal 2 = verbose/debug
--format|-f fmt
    Encoding profile
    [0 = Generic MPEG1, 1 = standard VCD, 2 = user VCD,
     3 = Generic MPEG2, 4 = standard SVCD, 5 = user SVCD,
     6 = VCD Stills sequences, 7 = SVCD Stills sequences, 8|9 = DVD,
     10 = ATSC 480i, 11 = ATSC 480p, 12 = ATSC 720p, 13 = ATSC 1080i]
--display-hsize|-x [32..16383]
   Set the the display-horizontal-size hint in MPEG-2 output to
   something other than the encoded image width
--display-vsize|-y [32..16383]
   Set the the display-vertical-size hint in MPEG-2 output to
   something other than the encoded image height
--aspect|-a num
    Set displayed image aspect ratio image (default: 2 = 4:3)
    [1 = 1:1, 2 = 4:3, 3 = 16:9, 4 = 2.21:1]
--frame-rate|-F num
    Set playback frame rate of encoded video
    (default: frame rate of input stream)
    0 = Display frame rate code table
--video-bitrate|-b num
    Set Bitrate / peak bitrate of compressed video in KBit/sec
    (Peak bitrate if a target bitrate and/or quantisation floor is set
    (default: 1152.0 for VCD, 2500.0 for SVCD, 7500.0 for DVD)
--target-video-bitrate|-t
   Set target bitrate for entire video stream in KBit/sec
--nonvideo-bitrate|-B num
    Non-video data bitrate to assume for sequence splitting
    calculations (see also --sequence-length).
--quantisation|-q num
    Image data quantisation factor [1..31] (1 is best quality, no default)
    When quantisation is set variable bit-rate encoding is activated and
    the --bitrate value sets an *upper-bound* video data-rate
--ratecontroller|-A [0..1] (default:0)
    Specify ratecontrol alorithm
--output|-o pathname
    Pathname of output file or fifo (REQUIRED!!!)
--target-still-size|-T size
    Size in KB of VCD stills
--interlace-mode|-I num
    Sets MPEG 2 motion estimation and encoding modes:
    0 = Progressive (non-interlaced)(Movies)
    1 = Interlaced source material (video)
    2 = Interlaced source material, per-field-encoding (video)
--motion-search-radius|-r num
    Motion compensation search radius [0..32] (default 16)
--reduction-4x4|-4 num
    Reduction factor for 4x4 subsampled candidate motion estimates
    [1..4] [1 = max quality, 4 = max. speed] (default: 2)
--reduction-2x2|-2 num
    Reduction factor for 2x2 subsampled candidate motion estimates
    [1..4] [1 = max quality, 4 = max. speed] (default: 3)
--min-gop-size|-g num
    Minimum size Group-of-Pictures (default depends on selected format)
--max-gop-size|-G num
    Maximum size Group-of-Pictures (default depends on selected format)
    If min-gop is less than max-gop, mpeg2enc attempts to place GOP
    boundaries to coincide with scene changes
--closed-gop|-c
    All Group-of-Pictures are closed.  Useful for authoring multi-angle DVD
--force-b-b-p|-P
    Preserve two B frames between I/P frames when placing GOP boundaries
--quantisation-reduction|-Q num
    Max. quantisation reduction for highly active blocks
    [0.0 .. 4.0] (default: 0.0)
--quant-reduction-max-var|-X num
    Luma variance below which quantisation boost (-Q) is used
    [0.0 .. 2500.0](default: 0.0)
--video-buffer|-V num
    Target decoders video buffer size in KB (default 46)
--video-norm|-n n|p|s
    Tag output to suit playback in specified video norm
    (n = NTSC, p = PAL, s = SECAM) (default: PAL)
--sequence-length|-S num
    Place a sequence boundary in the video stream so they occur every
    num Mbytes once the video is multiplexed with audio etc.
    N.b. --non-video-bitrate is used to the bitrate of the other
    data that will be multiplexed with this video stream
--3-2-pulldown|-p
    Generate header flags for 3-2 pull down of 24fps movie material
--intra_dc_prec|-D [8..11]
    Set number of bits precision for DC (base colour) of blocks in MPEG-2
--reduce-hf|-N num
    [0.0..2.0] Reduce hf resolution (increase quantization) by num (default: 0.0)
--keep-hf|-H
    Maximise high-frequency resolution - useful for high quality sources
    and/or high bit-rates)
--sequence-header-every-gop|-s
    Include a sequence header every GOP if the selected format doesn't
    do so by default.
--no-dummy-svcd-SOF|-d
    Do not generate dummy SVCD scan-data for the ISO CD image
    generator "vcdimager" to fill in.
--playback-field-order|-z b|t
    Force setting of playback field order to bottom or top first
--multi-thread|-M num
    Activate multi-threading to optimise throughput on a system with num CPU's
    [0..32], 0=no multithreading, (default: 0)
--correct-svcd-hds|-C
    Force SVCD horizontal_display_size to be 480 - standards say 540 or 720
    But many DVD/SVCD players screw up with these values.
--no-constraints
    Deactivate constraints for maximum video resolution and sample rate.
    Could expose bugs in the software at very high resolutions!
--no-altscan-mpeg2
    Deactivate the use of the alternate block pattern for MPEG-2.  This is
    A work-around for a Bug in an obscure hardware decoder.
--dualprime-mpeg2
    Turn ON use of dual-prime motion compensation. Default is OFF unless this option is used
--custom-quant-matrices|-K kvcd|tmpgenc|default|hi-res|file=inputfile|help
    Request custom or userspecified (from a file) quantization matrices
--unit-coeff-elim|-E num
    Skip picture blocks which appear to carry little information
    because they code to only unit coefficients. The number specifies
    how aggresively this should be done. A negative value means DC
    coefficients are included.  Reasonable values -40 to 40
--b-per-refframe| -R 0|1|2
    The number of B frames to generate between each I/P frame
--cbr|-u
    For MPEG-2 force the use of (suboptimal) ConstantBitRate (CBR) encoding
--chapters X[,Y[,...]]
    Specifies which frames should be chapter points (first frame is 0)
    Chapter points are I frames on closed GOP's.
--help|-?
    Print this lot out!

[C:\]
=>

I forgot about that one. It comes with mjpegtools, right? I heard it was very slow. Is the quality better than ffmpeg?

[El Heggunte]

.......................

I forgot about that one.

It comes with mjpegtools, right?

Yes$S

I heard it was very slow. Is the quality better than ffmpeg?

"Very-slow" is a point-of-view IMHO yes, according to a comparison test I ran three or four years ago, its quality is somewhat better than ffmpeg's MPEG-2. One good thing: mpeg2enc can read the custom quantization matrix from an external file One bad thing: its bitrate control still is far from perfection Also, and according to the manpages, "if constant quantizer =< 3", random artifacts may happen
So I'd recommend mpeg2enc only for creating "generic" m2v streams and mpg files,
but if you intend to use it for DVD or Blu-Ray authoring, then «play at your own risk»

[DarrellS]

I can't get it to work. There aren't any input options at all. The generic command doesn't list any errors but it refuses to run. If I list height, width or framerate it says the commands are no good. Not sure what I changed on one attempt and it said it couldn't find the imagemagic header. What? If it doesn't accept an input file then I guess it isn't going to see a header. It has a boat load of commands but it's usage consists of...

[mpeg2enc.exe] Output file name (-o option) is required!

Not sure what you're supposed to do with that. I assume it relies on another execute file to open the video and the mpeg2enc.exe file to encode it. That doesn't do me any good in Virtualdub.

I did find some settings that are supposed to make the ffmpeg encoder better. I'm testing now to see the difference.

[DarrellS]

.......................

I forgot about that one.

It comes with mjpegtools, right?

Yes$S

I heard it was very slow. Is the quality better than ffmpeg?

"Very-slow" is a point-of-view IMHO yes, according to a comparison test I ran three or four years ago, its quality is somewhat better than ffmpeg's MPEG-2. One good thing: mpeg2enc can read the custom quantization matrix from an external file One bad thing: its bitrate control still is far from perfection Also, and according to the manpages, "if constant quantizer =< 3", random artifacts may happen
So I'd recommend mpeg2enc only for creating "generic" m2v streams and mpg files,
but if you intend to use it for DVD or Blu-Ray authoring, then «play at your own risk»

I'm not really interested in encoding mpeg2 for myself but seems a few others are and I'd like to give them another option with the external encoder feature in Virtualdub. Seems the recommended encoders are HCenc and QuEnc and the only option in Virtualdub for those is frameserving. Shame neither of them accept "-" as input.

[El Heggunte]

Well, I only know mpeg2enc does work when fed by avs2yuv. Also, mpeg2enc expects stdin only, therefore specifying input options is unnecessary. I don't use VirtualDub 1.10, so I have no idea of what's going wrong between it and mpeg2enc.

[El Heggunte]

...
i wonder if that change in the fc12faa effected the final output of those frames.
....

Over here, no dropped nor corrupted frames thus far.

To check for dropped frames, I use a very-short encode (25 or 30 frames only), mux to MKV, and open this latter in VirtualDub. In spite of the "MISSING CODEC" warning, the framecount indicator in the seek bar still works anyway.

[ozok]

I've just updated my GUI with a new few fatures. You can get it from here: https://bitbucket.org/ozok/x265gui/downloads
Changes:

-Added: x265 options
-Added: CPU and RAM usages
-Added: Option to see information about source file
-Added: Taskbar progressbar for Windows 7
-Added: Logging
-Fixed: GUI didn't terminate properly if one of the backends wasn't found
-Updated: Mp4box to 0.5.1-DEV-rev4707

There are probably many bugs, so please report them if you find one or two.

[selo3000]

I've just updated my GUI with a new few fatures. You can get it from here: https://bitbucket.org/ozok/x265gui/downloads
Changes:

-Added: x265 options
-Added: CPU and RAM usages
-Added: Option to see information about source file
-Added: Taskbar progressbar for Windows 7
-Added: Logging
-Fixed: GUI didn't terminate properly if one of the backends wasn't found
-Updated: Mp4box to 0.5.1-DEV-rev4707

There are probably many bugs, so please report them if you find one or two.

@Özok


gui for hevc player ?


KILL BILL - The Whole Bloody Affair Hevc-AAC 720p [Marchand].mkv not playing?

[ozok]

No, it's for x265.exe.

[vhelp]

regarding x265.exe commit fc12faa

there seems to be a leak in this build. at first, i thought it was my gui and i test earlier versions, same problem, but i hadn't noticed it in other x265.exe encoder builds that i encoded longer clips with. but in one of my encodes a few minutes ago, my netbook (500mb ram, 1.3g total virt ram) with windows xp home sp3 popped up a 'virtual mem' error and had to increase its size. taskmgr.exe is reporting that slow stair-step of death. i started the encoding job at 250mb of available ram. now its at 940mb, and my netbook gets progressively slower due to the now constant accessing of the hdd swapping in/out ram processing.

see example image below.

ozok, can you test your gui with taskmgr open under performance tab ?

[ozok]

You are right vhelp, after a point it reaches 2GB ram usage and stops working. AFAIK it is the limit a 32bit app can use. Here is my ss from process explorer: https://forum.videohelp.com/images/imgfiles/t8OlMns.png

[vhelp]

my organization blocks your url can't see. but thanks for reporting it.

can you also test older x265 builds ? i didn't notice this problem before. and i've done a lot more 'longer' encodes without problems. don't know what changed in this (fc12faa) though. maybe revert back or maybe we (window xp users) are out of luck moving forward, which would end my participation here

[ozok]

I don't think problem was there with previous build supplied here. Because I was using an older build before you asked me to do a test with fc12faa. I use Win7 so I don't think it's the fault of OS. I remember one of older builds had a problem like this, but I'm not sure if it was related to a possible memory leak.

Edit: I also noticed that there is a significant file size difference between older build and fc12faa.
x265 [info]: HEVC encoder version 0.3+238-88e0c10bf47b = 5.25 mb
x265 [info]: HEVC encoder version 0.3+437-fc12faa1672e = 4.15 mb

[El Heggunte]

Well, I think the x265 mailing list does need to receive the feedback from more people

http://mailman.videolan.org/listinfo/x265-devel

For the record, Vid01 is there already, and me too of course ---
--- following the advice from Min Chen himself

[El Heggunte]

........

Edit: I also noticed that there is a significant file size difference between older build and fc12faa.
x265 [info]: HEVC encoder version 0.3+238-88e0c10bf47b = 5.25 mb
x265 [info]: HEVC encoder version 0.3+437-fc12faa1672e = 4.15 mb

There has been a lot of "code clean-up" lately, maybe it has something to do with that Anyway,

the binaries compiled with GCC 4.7.3 have always been bigger than the binaries compiled with GCC 4.8.1,

whereas the .EXEs built with Visual Studio are always the smallest ones

[Brazil]

I've just updated my GUI with a new few fatures.
-Updated: Mp4box to 0.5.1-DEV-rev4707

There are probably many bugs, so please report them if you find one or two.

Thanks for this but it seems that you have extracted MP4Box from the GPAC framework so it can't work as is. You also need the following files for MP4Box to work:
js.dll or js32.dll depending on the version being used
libgpac.dll
msvcr100.dll

Or compile your own static build without the need of MSVC 2010, like this one:
https://drive.google.com/folderview?id=0BwV03nn6LPd9ODRlU3h1SURqNGM&usp=sharing

[mindwin]

El Heggunte

The binaries compiled with GCC 4.7.3 have always been bigger than the binaries compiled with GCC 4.8.1.

Whereas the .EXEs built with Visual Studio are always the smallest ones

Is there any differences except the binary size between GCC and MSVS builds?
Perfomance etc.

[El Heggunte]

Between GCC 4.7.3 and GCC 4.8.1: no.

I don't use the MSVC builds, so I don't know.

[vhelp]

is anyone sure that when run under w7 system, there is no mem leak ?

. windows xp home, sp3 .
i test the last version/build, x265.exe (0.3+238-88e0c10bf47b) and it does not have a mem leak problem. it is the usual 70mb usage, my taskmgr.exe shows a flat line report, no slowup in pc performance, etc., exactly what you want. lets hope that the next version doesn't have this mem leak problem.

[Jacobr]

there seems to be a leak in this build. at first, i thought it was my gui and i test earlier versions, same problem, but i hadn't noticed it in other x265.exe encoder builds that i encoded longer clips with. but in one of my encodes a few minutes ago, my netbook (500mb ram, 1.3g total virt ram) with windows xp home sp3 popped up a 'virtual mem' error and had to increase its size. taskmgr.exe is reporting that slow stair-step of death. i started the encoding job at 250mb of available ram. now its at 940mb, and my netbook gets progressively slower due to the now constant accessing of the hdd swapping in/out ram processing.

i also had that leak problem and apart from that i think i experienced random crash at the beginning of encode.
went back to previous one

found out why length was wrong in my mkv encodes, but now i've got different issues with gui. i want to find out when does the mux start&end exactly, so i check filesize periodically, but it returns 0 until file is ready. it still 'kind-of' works but it looks unstable.
any of you here knows how could you check size on growing file, it can be winapi or some function called from system dll, i can use these things from autoit3

[ozok]

@Brazil, I'm uploading a package with dlls you said. I got gpac installed so I missed that.
@vhelp, there is a memory leak under win7 in my system.

[El Heggunte]

lets hope that the next version doesn't have this mem leak problem.

"Let's hope "?

¿Bugs being fixed without being properly-reported first? I doubt that.

http://mailman.videolan.org/listinfo/x265-devel

smtp://steve_àtt_borho_dótt_org

[vhelp]

@ozok, ty. i am making similar additions to my gui. also, your gui does not complete one of its process at the mp4box stage and closes down bc of it. not sure it was reported to you. the error, 'unable to locate component, libgpac.dll'.

so w7 also suffers this mem leak, maybe it just takes a bit longer to notice. anywa, that' good news, in a twisted kind a way bc at least its not intentional on xp