On Monday, May 2, 2022, Andrea paz <[email protected]> wrote:

Tried the build with megapatch_12. Everything OK: it compiles without errors. I enclose cin5_ffmpeg4.log. There are many "warning+error" concerning Latex and the html manual and many "warning" concerning c++. I don't know if they are important or not; everything works fine for me. TrueHD loading and playback works fine; I tried it with 2 examples containing video and 22 audio tracks. I have not tested the encoder. Otherwise, I wouldn't know what to test since I don't have termux/bsd/bluray. You tell me what to test.

usual operations.. or even just try to do something you never tried before... my idea for truehd encoder test was to make .thd soundtrack from known pcm (uncompressed) source, then load/save as pcm again.. two pcm files must be exactly (crc sum) the same - patches aimed at fixing thd last frame not written correctly..

Then I tried a build with --without-thirdparty (also removing dv; libdpx and firewire), to get ffmpeg5 working. Everything works fine. As I noticed before, putting opencv plugins is not possible: CinGG doesn't start. Incompatibility with ffmpeg5?

I do not think they are related much? I mean, opencv can link with libav* libs, but they are c, not c++. you still can't buld even with static opencv3 tar source from cingg's website? I totally forgot about binary opencv plugins.. guess they should be built with roughly same c++ compiler as main Cin, and linked to same openexr (hm...) can you provide failed startup log for binary opencv plugins and failed build log for opencv3 static?

I attach cin5_ffmpeg5.log, if you want to make comparisons. There are some warnings but not the same as I had with internal ffmpeg4.

Andrea -- Oops, you do have to be "root" unless you do the workaround as described in the manual but this is just a simple one time test to see if you have the same problem that I do. On Mon, May 2, 2022 at 9:36 AM Phyllis Smith <[email protected]> wrote:

*Andrea,* since you have the megapatches all in, could you run the binary and test "truehd"? These are the steps:

load file File-> BD Render Profile - use down arrow to choose *Truehd* and take rest of defaults Ignore Warning about root because you do not need to actually create bluray media It will automatically stop CinGG and you will be at a terminal window See if you get the same error message that I get or whatever you get:

+ mount -t udf -o loop /tmp/bd_20220502-092612/bd.udfs /tmp/bd_20220502-092612/udfs + bdwrite /tmp/bd_20220502-092612/udfs /tmp/bd_20220502-092612/bd.m2ts *unknown bluray audio format 0 ch*

On Mon, May 2, 2022 at 5:17 AM Andrea paz via Cin < [email protected]> wrote:

...

Tried the build with megapatch_12. Everything OK: it compiles without errors. I enclose cin5_ffmpeg4.log. There are many "warning+error" concerning Latex and the html manual and many "warning" concerning c++. I don't know if they are important or not; everything works fine for me. TrueHD loading and playback works fine; I tried it with 2 examples containing video and 22 audio tracks. I have not tested the encoder. Otherwise, I wouldn't know what to test since I don't have termux/bsd/bluray. You tell me what to test.

Then I tried a build with --without-thirdparty (also removing dv; libdpx and firewire), to get ffmpeg5 working. Everything works fine. As I noticed before, putting opencv plugins is not possible: CinGG doesn't start. Incompatibility with ffmpeg5? I attach cin5_ffmpeg5.log, if you want to make comparisons. There are some warnings but not the same as I had with internal ffmpeg4. -- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin

Andrea, you will have to install 'mkudffs' if you don't mind. I looked on the internet, and it appears to be available for Arch. Following Phyllis' instructions (BD render) I render up to 100% and

then I have these errors (from root):

+ mkudffs -b 2048 /tmp/bd_20220502-182538/bd.udfs 26524 /tmp/bd_20220502-182538/bd.sh: line 14: mkudffs: command not found

On Monday, May 2, 2022, Andrea paz via Cin <[email protected]> wrote:

@andrew I didn't quite understand your truehd audio example. I took 2 pcm s16 files I loaded them in CinGG and burned them as truehd: I have no errors and the rendering I loaded it again in CinGG with no errors. I don't think this is what you wanted.

can you notice sha256sum of pcm files, then after roundtrip via truehd, try to save thd file as opened by cingg as s16 pcm, with another name. make sha256sum of this file... they both should be the same (unless libavcode/libavformat insert some metadata there? look with ffprobe/mediainfo)

Following Phyllis' instructions (BD render) I render up to 100% and then I have these errors (from root):

Render::render_single: Session finished. ** rendered 917 frames in 14.584 secs, 62.877 fps ++ dirname /tmp/bd_20220502-182538/bd.sh + sdir=/tmp/bd_20220502-182538 ++ cd /tmp/bd_20220502-182538 ++ pwd + dir=/tmp/bd_20220502-182538 + PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/lib/ jvm/default/bin:/opt/rocm-5.1.1/bin:/opt/rocm-5.1.1/hip/bin: /usr/bin/site_perl:/usr/bin/vendor_perl:/usr/bin/core_ perl:/home/paz/cinelerra5/cinelerra-5.1/bin + mkdir -p /tmp/bd_20220502-182538/udfs ++ du -cb /tmp/bd_20220502-182538/bd.m2ts ++ tail -1 ++ sed -e 's/[ ].*//' + sz=45932544 + blks=26524 + rm -f /tmp/bd_20220502-182538/bd.udfs + '[' -f /tmp/bd_20220502-182538/bd.meta ']' + mkudffs -b 2048 /tmp/bd_20220502-182538/bd.udfs 26524 /tmp/bd_20220502-182538/bd.sh: line 14: mkudffs: command not found

OpenCV: I compiled with opencv from the CinGG website. The build did not stop, but the result is that CinGG does not have the opencv plugins. See attachment.

I can't give any errors when I try to start CinGG with ffmpeg5 and the opencv binaries. The terminal reports nothing at all.

@Phyllis Smith Here what I get after installing udftools (which contains mkudffs): Render::render_single: Session finished. ** rendered 917 frames in 14.627 secs, 62.692 fps ++ dirname /tmp/bd_20220502-192458/bd.sh + sdir=/tmp/bd_20220502-192458 ++ cd /tmp/bd_20220502-192458 ++ pwd + dir=/tmp/bd_20220502-192458 + PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/lib/jvm/default/bin:/opt/rocm-5.1.1/bin:/opt/rocm-5.1.1/hip/bin:/usr/bin/site_perl:/usr/bin/vendor_perl:/usr/bin/core_perl:/home/paz/cinelerra5/cinelerra-5.1/bin + mkdir -p /tmp/bd_20220502-192458/udfs ++ du -cb /tmp/bd_20220502-192458/bd.m2ts ++ tail -1 ++ sed -e 's/[ ].*//' + sz=45950976 + blks=26533 + rm -f /tmp/bd_20220502-192458/bd.udfs + '[' -f /tmp/bd_20220502-192458/bd.meta ']' + mkudffs -b 2048 /tmp/bd_20220502-192458/bd.udfs 26533 filename=/tmp/bd_20220502-192458/bd.udfs label=LinuxUDF uuid=62701408d8215c0e blocksize=2048 blocks=26533 udfrev=2.01 start=0, blocks=16, type=ERASE start=16, blocks=4, type=VRS start=20, blocks=76, type=ERASE start=96, blocks=16, type=MVDS start=112, blocks=16, type=ERASE start=128, blocks=4, type=LVID start=132, blocks=124, type=ERASE start=256, blocks=1, type=ANCHOR start=257, blocks=26016, type=PSPACE start=26273, blocks=3, type=ERASE start=26276, blocks=1, type=ANCHOR start=26277, blocks=91, type=ERASE start=26368, blocks=16, type=RVDS start=26384, blocks=148, type=ERASE start=26532, blocks=1, type=ANCHOR + mount -t udf -o loop /tmp/bd_20220502-192458/bd.udfs /tmp/bd_20220502-192458/udfs + bdwrite /tmp/bd_20220502-192458/udfs /tmp/bd_20220502-192458/bd.m2ts + umount /tmp/bd_20220502-192458/udfs + echo To burn bluray, load writable media and run: To burn bluray, load writable media and run: + echo for WORM: growisofs -dvd-compat -Z /dev/bd=/tmp/bd_20220502-192458/bd.udfs for WORM: growisofs -dvd-compat -Z /dev/bd=/tmp/bd_20220502-192458/bd.udfs + echo for RW: dd if=/tmp/bd_20220502-192458/bd.udfs of=/dev/bd bs=2048000 for RW: dd if=/tmp/bd_20220502-192458/bd.udfs of=/dev/bd bs=2048000 + kill 350045 Terminato I get a working bd.m2ts file.

can you confirm it contain trueHD audio (run mediainfo)?

ffprobe: Stream #0:1[0x1100](ita): Audio: pcm_bluray ([128][0][0][0] / 0x0080), 48000 Hz, stereo, s16, 1536 kb/s mediainfo: Audio ID : 4352 (0x1100) Menu ID : 1 (0x1) Format : PCM Format settings : Big / Signed Muxing mode : Blu-ray Codec ID : 128 Duration : 38 s 240 ms Bit rate mode : Constant Bit rate : 1 536 kb/s Channel(s) : 2 channels Channel layout : L R Sampling rate : 48.0 kHz Bit depth : 16 bits Stream size : 7.00 MiB (16%) Language : Italian

I added to megapile_12 the patch 0001-..... I uncommented the line with "truehd" in bdcreate.C I built CinGG without problems. I started CinGG from root. In BD Render the profile "truehd" appeared. The render progresses up to 100% then there is a crash. At the terminal I have: Render::render_single: Session finished. ** rendered 917 frames in 14.092 secs, 65.072 fps ++ dirname /tmp/bd_20220502-210520/bd.sh + sdir=/tmp/bd_20220502-210520 ++ cd /tmp/bd_20220502-210520 ++ pwd + dir=/tmp/bd_20220502-210520 + PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/lib/jvm/default/bin:/opt/rocm-5.1.1/bin:/opt/rocm-5.1.1/hip/bin:/usr/bin/site_perl:/usr/bin/vendor_perl:/usr/bin/core_perl:/home/paz/cinelerra5/cinelerra-5.1/bin + mkdir -p /tmp/bd_20220502-210520/udfs ++ du -cb /tmp/bd_20220502-210520/bd.m2ts ++ tail -1 ++ sed -e 's/[ ].*//' + sz=46442496 + blks=26773 + rm -f /tmp/bd_20220502-210520/bd.udfs + '[' -f /tmp/bd_20220502-210520/bd.meta ']' + mkudffs -b 2048 /tmp/bd_20220502-210520/bd.udfs 26773 filename=/tmp/bd_20220502-210520/bd.udfs label=LinuxUDF uuid=62702b9e8438e328 blocksize=2048 blocks=26773 udfrev=2.01 start=0, blocks=16, type=ERASE start=16, blocks=4, type=VRS start=20, blocks=76, type=ERASE start=96, blocks=16, type=MVDS start=112, blocks=16, type=ERASE start=128, blocks=4, type=LVID start=132, blocks=124, type=ERASE start=256, blocks=1, type=ANCHOR start=257, blocks=26256, type=PSPACE start=26513, blocks=3, type=ERASE start=26516, blocks=1, type=ANCHOR start=26517, blocks=75, type=ERASE start=26592, blocks=16, type=RVDS start=26608, blocks=164, type=ERASE start=26772, blocks=1, type=ANCHOR + mount -t udf -o loop /tmp/bd_20220502-210520/bd.udfs /tmp/bd_20220502-210520/udfs + bdwrite /tmp/bd_20220502-210520/udfs /tmp/bd_20220502-210520/bd.m2ts unknown bluray audio format 0 ch Despite the CinGG crash, I get a bd.m2ts file. Ffprobe indicates: Stream #0:1[0x1100](ita): Audio: truehd ([131][0][0][0] / 0x0083), 48000 Hz, stereo, s32 (24 bit) Stream #0:2[0x1100]: Audio: ac3 ([131][0][0][0] / 0x0083), 0 channels, fltp

On Monday, May 2, 2022, Phyllis Smith <[email protected]> wrote:

Andrew, what other software has to be installed to get Truehd to work? Andrea does not get the error I do so there must be something additional that has to be installed. We need to know for documentation.

i tried with tsmuxer, but I guess bdwrite need some more logic to deal with it too... without bdwrite we do not have burnable bluray disk image, just stream alone (some modern players play those anyway..)

On Mon, May 2, 2022 at 12:42 PM Andrew Randrianasulu < [email protected]> wrote:

...

well, this is not exactly what we hoped for.. in this case.

sorry TWO issues going on

1) tsmuxer choke on ffmpeg's lpcm as encoded without ONE MORE PATCH..

2) truehd encoder in ffmoeg may trip bdwrite...

for truehd encoder you need to either uncomment line in bdcreate.C {with truehd profile, in /* */ pair..} or manually choose audio when batch render window come up...

On Monday, May 2, 2022, Andrea paz <[email protected]> wrote:

...

...

can you confirm it contain trueHD audio (run mediainfo)?

ffprobe: Stream #0:1[0x1100](ita): Audio: pcm_bluray ([128][0][0][0] / 0x0080), 48000 Hz, stereo, s16, 1536 kb/s

mediainfo: Audio ID : 4352 (0x1100) Menu ID : 1 (0x1) Format : PCM Format settings : Big / Signed Muxing mode : Blu-ray Codec ID : 128 Duration : 38 s 240 ms Bit rate mode : Constant Bit rate : 1 536 kb/s Channel(s) : 2 channels Channel layout : L R Sampling rate : 48.0 kHz Bit depth : 16 bits Stream size : 7.00 MiB (16%) Language : Italian

On Monday, May 2, 2022, Phyllis Smith <[email protected]> wrote:

We can NOT make LV2 a requirement because the plugins are too unreliable - just like we can not make OpenCV a requirement. When patching the LV2 patches from the pile - which is the following: 0052-Defend-against-missed-lv2-headers.patch 0056-Attempt-at-fixing-lv2-detection.patch 0062-dynamic-lv2.patch It will not compile, so I think these patches will have to be redone. Here are the error messages I immediately get:

I think this is due to lv2 set to auto.... so, you try to make static build as always and it fails to build / use lv2 and friends?

checking libusb headers... yes checking for lv2 availability... Package lilv-0 was not found in the pkg-config search path. Perhaps you should add the directory containing `lilv-0.pc' to the PKG_CONFIG_PATH environment variable Package 'lilv-0', required by 'virtual:world', not found Package lilv-0 was not found in the pkg-config search path. Perhaps you should add the directory containing `lilv-0.pc' to the PKG_CONFIG_PATH environment variable Package 'lilv-0', required by 'virtual:world', not found checking for lilv/lilv.h... no checking lilv headers... no Package serd-0 was not found in the pkg-config search path. ... and so on for all of them

On Sun, May 1, 2022 at 2:49 PM Andrew Randrianasulu < [email protected]> wrote:

...

ls megapile_12/ 0002-Improve-truehd-decoder-encoder-from-ffmpeg.git.patch

== I think you can skip it until you find any info about your players on supporting or not pure truehd tracks..

0004-mem-and-resource-leaks-in-indexfile-indextate-cppche.patch 0005-cppcheck-in-cursor.C.patch 0006-cppcheck-in-resourcepixmap.C.patch

== I hope those are correct... but not really programmer!

0007-Add-objrem-target-for-thirdparty-Makefile-removes-on.patch

== as before - personal build trimmer

0008-Freebsd-13-conditional-include-in-exportedl.C.patch 0009-fileexr-forward-port-openexr-3.patch 0010-filegif-addition-of-nearly-all-quntize.c-from-giflib.patch 0011-unsigned-long-cast-for-freebsd-in-bctrace.C.patch 0012-dirent64-and-readdir64-aliases-for-freebsd-in-guicas.patch 0013-freebsd-includes-in-plugins-titler.patch 0014-freebsd-defines-in-guicast-bcresources.C.patch 0015-freebsd-in-indexfile.C.patch 0016-freebsd-in-file.C.patch 0017-freebsd-in-dvdcreate.C.patch 0018-pkg-config-based-dynamic-openexr-link-fix-for-freebs.patch 0019-disable-frei0r-and-libvmaf-for-freebsd-13-dynamic-ff.patch 0020-update-blds-bsd.bld-for-freebsd-13.patch 0021-Attempt-at-adding-freebsd-detect-in-plugins-Makefile.patch 0022-Attempt-at-freebsd-netbsd-detection-in-cinelerra-Mak.patch

== most of freebsd

0023-experimental-switch-python-in-guicast-to-python3-unc.patch

== this one not strictly needed but 0057 depend on it.. can squash both into one

0024-Add-android-only-hacked-Dynamic-makefile-libs.patch 0025-Make-cin-compile-with-ffmpeg-5-unconditionally.patch 0026-Dynamic-termux-configure.patch 0027-Hopefully-fixed-dynamic-liba52.patch 0028-Cinelerra-Makefile-dynamic-section-fix-for-non-andro.patch 0029-Hopefully-fix-dyn-liba52-on-termux.patch 0030-Dynamic-termux-libzmpeg.patch

== dynamic compile on termux part 1

0031-Patches-for-attempting-to-make-netbsd-i386-compile.patch

=== just dumped newish netbsd patch in tree - - you can ignore it for now ==

0032-video4linux2-and-DVB-support-currently-depend-on-lib.patch 0033-Try-to-abort-earlier-if-xinerama-xfixes-support-not-.patch 0034-Beginning-of-ffmpeg-versioned-compilation.patch 0035-Ffmpeg-versioning-in-bdwrite.C.patch 0036-More-libavcodec-versionkng-in-ffmpeg.patch 0037-Even-more-ffmpeg-versioning-in-ffmpeg.C.patch 0038-Nearly-all-versioning-in-ffmpeg.C.patch 0039-Finished-versioning-in-ffmpeg.C.patch 0040-Really-fix-ffmpeg.C-for-ffmpeg-4.4.patch 0041-ffmpeg.git-conditional-fixes.patch 0042-Limit-git-clone-depth-to-1-faster-download-less-spac.patch 0043-More-defines-in-ffmpeg.h.patch

=== ffmpeg 5 subseries, see also 0045

0044-openexr-dynamic-libs-in-termux_dyn.patch 0045-Even-more-ffmpeg-version-ifdefs-in-ffmpeg.C.patch 0046-Hopefully-fix-or-conditional-in-cinelerra-Makefile.patch 0047-Hopefully-fix-dynamic-openexr-via-pkg-config-in-conf.patch 0048-Abort-earlier-if-non-3rd-party-and-openexr-requested.patch 0049-update-termux_dyn-now-detection-of-openexr-moved-to-.patch 0050-Add-usr-include-a52dec-so-libzmpeg3-can-be-compiled-.patch 0051-HIGHLY-EXPERIMENTAL-constify-png_data.patch

=== 0051 can wait... it does not hurt, but does not help either...

0052-Defend-against-missed-lv2-headers.patch 0053-Fix-for-dynamic-libogg.patch 0054-Dynamic-termux-with-ogg.patch 0055-Attempt-at-detecting-OpenEXR-3.patch 0056-Attempt-at-fixing-lv2-detection.patch 0057-Attempt-at-finding-python-on-Freebsd-for-guicast-Mak.patch 0058-pkg-config-based-libs-detection-in-shapewipe-and-tit.patch

== more bsd/termux lib detection,

0059-libbthread-src-copied-to-main-tree-for-Android.patch 0060-Fix-libbthread-on-non-android.patch 0061-Try-to-guard-against-compilation-on-non-termux.patch 0062-dynamic-lv2.patch 0063-termyx_dyn.sh-update.patch

=== finally termux should compile dynamically!

0064-Trim-opencv-build-add-std-c-11.patch

== my tmpfs not very big....

huh?

thanks! On Wednesday, May 4, 2022, Phyllis Smith <[email protected]> wrote:

Checked into GIT from the current Megapile, the following:

0004-mem-and-resource-leaks-in-indexfile-indextate-cppche.patch 0005-cppcheck-in-cursor.C.patch 0006-cppcheck-in-resourcepixmap.C.patch 0007-Add-objrem-target-for-thirdparty-Makefile-removes-on.patch 0008-Freebsd-13-conditional-include-in-exportedl.C.patch 0011-unsigned-long-cast-for-freebsd-in-bctrace.C.patch 0012-dirent64-and-readdir64-aliases-for-freebsd-in-guicas.patch 0013-freebsd-includes-in-plugins-titler.patch 0015-freebsd-in-indexfile.C.patch 0042-Limit-git-clone-depth-to-1-faster-download-less-spac.patch

On Friday, May 6, 2022, Phyllis Smith <[email protected]> wrote:

Checked into GIT from the current Megapile_12, the following:

0020-update-blds-bsd.bld-for-freebsd-13.patch 0021-Attempt-at-adding-freebsd-detect-in-plugins-Makefile.patch 0022-Attempt-at-freebsd-netbsd-detection-in-cinelerra-Mak.patch + + 0046-Hopefully-fix-or-conditional-in-cinelerra-Makefile.patch 0026-Dynamic-termux-configure.patch + + but renamed file to blds/termux_dyn.bld for consistency & will adjust the patches to it myself 0064-Trim-opencv-build-add-std-c-11.patch

\0/ thanks, will test on tablet just now

Andrew

\0/

...

What does the above symbol mean?

...

thanks, will test on tablet just now

hm, I think dynamic blds need to disable libdpx for now...

There are more patches to termux that I have NOT put in from the Megapile, so .... not sure how well the build will work ?

Andrew, I do not understand what WANT_CIN_3RDPARTY is and consequently do not know whether adding patch: 0024-Add-android-only-hacked-Dynamic-makefile-libs.patch (along with 0018) for additional LIBS will affect other people who ae doing their own builds. What I am concerned about is in bold letters below for the addition of the LIBS. Is there an explanation that I can understand?

From cinelerra/Makefile: ifeq ($(WANT_CIN_3RDPARTY),no) LIBS += -lavcodec -lavdevice -lavfilter -lavformat -lavutil LIBS += -lswresample -lswscale PLATFORM = $(shell uname -s) ifneq (, $(filter $(PLATFORM), FreeBSD NetBSD)) LIBS += -lintl -liconv LIBS += -lexecinfo

*+LIBS += `pkg-config --libs OpenEXR`+LIBS += `pkg-config --libs flac`+LIBS += `pkg-config --libs libjpeg`+LIBS += `pkg-config --libs uuid`+LIBS += `pkg-config --libs theora`+LIBS += `pkg-config --libs vorbis`+LIBS += `pkg-config --libs tiff`+LIBS += `pkg-config --libs sndfile`+LIBS += -lgif -l:libtheora.a -l:libvorbis.a -ltiff+LIBS += -landroid-shmem -liconv jpt.a* LIBS += $(shared_libs) LIBS += $(system_libs)

\0/ \0/ (me and Buddy waving wow) Checked into GIT the following (thank you Andrew for these many fixes; please check cinelerra/Makefile to make sure no mistakes) 0010-filegif-addition-of-nearly-all-quntize.c-from-giflib.patch 0018-pkg-config-based-dynamic-openexr-link-fix-for-freebs.patch 0023-experimental-switch-python-in-guicast-to-*python3*-unc.patch -- NOTE below If I remember correctly, Centos was still at *python2 *awhile back and maybe some really old distros. BUT if I do not remember correctly, there may be another good reason for staying at python2? So I hope this is not problematic + OpenCV will still be using python2 by default. 0024-Add-android-only-hacked-Dynamic-makefile-libs.patch 0028-Cinelerra-Makefile-dynamic-section-fix-for-non-andro.patch 0030-Dynamic-termux-libzmpeg.patch 0053-Fix-for-dynamic-libogg.patch On Fri, May 6, 2022 at 12:21 PM Andrew Randrianasulu < [email protected]> wrote:

On Friday, May 6, 2022, Phyllis Smith <[email protected]> wrote:

...

Andrew

...

\0/

...

What does the above symbol mean?

hands up/waving as in 'wow, very cool' (found on irc long time ago)

...

...

...

thanks, will test on tablet just now

hm, I think dynamic blds need to disable libdpx for now...

There are more patches to termux that I have NOT put in from the Megapile, so .... not sure how well the build will work ?

well, it fails if I try to use newly-included dynamic termux build script - sort of expacted, just want to point out easy thing in case you will try it on your tablet before pushing other changes....

Checked into GIT even more from the Megapile: 0027-Hopefully-fixed-dynamic-liba52.patch 0029-Hopefully-fix-dyn-liba52-on-termux.patch 0032-video4linux2-and-DVB-support-currently-depend-on-lib.patch 0033-Try-to-abort-earlier-if-xinerama-xfixes-support-not-.patch 0044-openexr-dynamic-libs-in-termux_dyn.patch 0047-Hopefully-fix-dynamic-openexr-via-pkg-config-in-conf.patch 0049-update-termux_dyn-now-detection-of-openexr-moved-to-.patch 0050-Add-usr-include-a52dec-so-libzmpeg3-can-be-compiled-.patch 0057-Attempt-at-finding-python-on-Freebsd-for-guicast-Mak.patch 0058-pkg-config-based-libs-detection-in-shapewipe-and-tit.patch

...

0010-filegif-addition-of-nearly-all-quntize.c-from-giflib.patch

well, may be we will need rework some of this - I am not sure if you can just slap two files legally. May be I need to put those functions back in one specific file and just add this file (with version check) to build..

*Yeah, I was not too sure about this but it did not affect linux so thought it was needed for BSD.*

...

0018-pkg-config-based-dynamic-openexr-link-fix-for-freebs.patch 0023-experimental-switch-python-in-guicast-to-*python3*-unc.patch -- NOTE below If I remember correctly, Centos was still at *python2 *awhile back and maybe some really old distros. BUT if I do not remember correctly, there may be another good reason for staying at python2?

see also 0057-Attempt-at-finding-python-on-Freebsd-for-guicast-Mak.patch

*Oh, I missed this and should have checked it in at the same time but wanted to check in a group before I got even more confused.*

On Wednesday, May 11, 2022, Phyllis Smith <[email protected]> wrote:

Checked into GIT, LV2 building changes + 4 additional to Calf LV2 plugins that I discovered while testing. They did not cause CinGG to crash EXCEPT when using the Calf LV2 plugins with an AppImage. When using from a self-build, they did not show up in Audio Effects but reported the error message on the terminal window of "unsupported lv2 plugin, missing audio input or output".

0052-Defend-against-missed-lv2-headers.patch 0054-Dynamic-termux-with-ogg.patch 0062-dynamic-lv2.patch 0063-termyx_dyn.sh-update.patch + lv2_blacklist.txt has additionally removed Calf plugins of Fluidsynth, Monosynth, Organ, and Wavetable

This took me awhile to check the above and test so I hope it is correct.

yeah, working one one problem often bring up few more! Thanks, I can't check on termux until most patches are in (a bit too short on 'disk' space for static build) but I'll check on x86 laptop

Working on a Status text file of the other megapile patches.

On Wednesday, May 11, 2022, Phyllis Smith <[email protected]> wrote:

Status of other 18 patches from the Megapile:

*Bluray Truehd:* 0002-Improve-truehd-decoder-encoder-from-ffmpeg.git.patch -- *not quite working; setting aside* *Netbsd Compile;* 0031-Patches-for-attempting-to-make-netbsd-i386-compile.patch -- *25+ files/have to study* *Ffmpeg Versioning:* 0034-Beginning-of-ffmpeg-versioned-compilation.patch 0035-Ffmpeg-versioning-in-bdwrite.C.patch 0036-More-libavcodec-versionkng-in-ffmpeg.patch 0037-Even-more-ffmpeg-versioning-in-ffmpeg.C.patch 0038-Nearly-all-versioning-in-ffmpeg.C.patch 0039-Finished-versioning-in-ffmpeg.C.patch 0040-Really-fix-ffmpeg.C-for-ffmpeg-4.4.patch 0041-ffmpeg.git-conditional-fixes.patch 0043-More-defines-in-ffmpeg.h.patch 0045-Even-more-ffmpeg-version-ifdefs-in-ffmpeg.C.patch *EXR Related:* 0048-Abort-earlier-if-non-3rd-party-and-openexr-requested.patch 0055-Attempt-at-detecting-OpenEXR-3.patch -- *these 2 probably be OK to add* *Png Data:* 0051-HIGHLY-EXPERIMENTAL-constify-png_data.patch -- i*mpossible for me to test/understand* *Libbthread Move:* 0059-libbthread-src-copied-to-main-tree-for-Android.patch 0060-Fix-libbthread-on-non-android.patch 0061-Try-to-guard-against-compilation-on-non-termux.patch -- *big change/have to study*

Andrew, have I missed any? + what is ffmpeg versioning all about and will it require maintenance for every future version and work on all types of linux/ppc/termux/bsd/android/32-bit/other distros forever?

I think those are all I sent (have few more on termux to test). ffmpeg versioning about making fileffmpeg and few other files compile/work with both ffmpeg 4.4 and 5.0/5.1 (hopefully). This is roughly how Einar makes cin-cve compile against wide range of ffmpeg libs... it worked on arm 32 and should be fairly generic, but new major ffmpeg version bump most likely will need additional research....

Checked into GIT, the ffmpeg versioning mods - this is a very useful feature for using one version of ffmpeg but being able to test the upcoming version. Also, mods for allowing for OpenEXR version 3. Patches now included are: 0025-Make-cin-compile-with-ffmpeg-5-unconditionally.patch 0034-Beginning-of-ffmpeg-versioned-compilation.patch 0035-Ffmpeg-versioning-in-bdwrite.C.patch 0036-More-libavcodec-versionkng-in-ffmpeg.patch 0037-Even-more-ffmpeg-versioning-in-ffmpeg.C.patch 0038-Nearly-all-versioning-in-ffmpeg.C.patch 0039-Finished-versioning-in-ffmpeg.C.patch 0040-Really-fix-ffmpeg.C-for-ffmpeg-4.4.patch 0041-ffmpeg.git-conditional-fixes.patch 0043-More-defines-in-ffmpeg.h.patch 0045-Even-more-ffmpeg-version-ifdefs-in-ffmpeg.C.patch 0048-Abort-earlier-if-non-3rd-party-and-openexr-requested.patch 0055-Attempt-at-detecting-OpenEXR-3.patch

Checked into GIT netbsd patch and bld from: 0031-Patches-for-attempting-to-make-netbsd-i386-compile.patch Initially I panicked when I did a grep for 'diff' but today I figured it out as all of the diffs where located in the patch so I was relieved. Later if Andrew makes changes, probably should chmod for netbsd.bld to add executable like the other bld scripts. So now all that is left from Megapile_12 is: 0002-Improve-truehd-decoder-encoder-from-ffmpeg.git.patch (wait for further changes) 0051-HIGHLY-EXPERIMENTAL-constify-png_data.patch (wait for knowledge) 0059-libbthread-src-copied-to-main-tree-for-Android.patch (probably wait for next month for these 3 so I don't make mistake) 0060-Fix-libbthread-on-non-android.patch 0061-Try-to-guard-against-compilation-on-non-termux.patch so, current git should be mostly complete for x86 testing - remaining

patches for termux and my draft of netbsd patches..

Hopefully Andrea will have time to give git a spin...

...

On Sunday, May 15, 2022, Andrea paz <[email protected]> wrote:

The past week and next week I will not have much time; I apologize.

this is ok, have (good) life! I tried to compile CinGG and everything is OK.

I tried to compile with OpenCV taken from CinGG's git (--with-opencv=sta,tar=http://cinelerra-gg.org/download/ opencv/opencv-20200306.tgz) and, although the build gives no errors, the plugins do not show up in CinGG. (I've no Python 2, only Python 3). I tried a new build with the OpenCVs from their git (--with-opencv=sta, https://github.com/opencv/opencv.git) and this time the plugins are there and work in CinGG.

at least something! but git is moving target.. do you have jasper3 or jasper2 lib on your system? Have you tried my modification for opencv_build disabling jasper support there?

NOTE: trueHD does not appear in "BD render" to me unless I uncomment the line in bdcreate.C; why was it commented out?

because we still not sure how to make blu-ray specific truehd mux? it different from normal truehd in ts.... also, thd encoding in ffmpeg currently limited to mono or stereo - no 5.1/7.1

On Sunday, May 15, 2022, Andrea paz <[email protected]> wrote:

...

do you have jasper3 or jasper2 lib on your system? I have jasper2 on my system. Have you tried my modification for opencv_build disabling jasper support there? Sorry, I didn't understand how to do it!

just add this line (-DWITH_JASPER=OFF) without '+' to your opencv_build file in cinegg src root...?

On Sunday, May 15, 2022, Andrea paz <[email protected]> wrote:

...

just add this line (-DWITH_JASPER=OFF) without '+' to your opencv_build file in cinegg src root...?

I tried adding the line with jasper in opencv_build and then compiling CinGG. The result is the same: the OpenCV plugins are not present.

( thanks for trying....

Ho decommentato la riga con truehd in bdcreate.C e provato a fare un rendring con "BD Render...". Render::render_single: Session finished. ** rendered 677 frames in 16.477 secs, 41.088 fps ++ dirname /tmp/bd_20220515-185631/bd.sh + sdir=/tmp/bd_20220515-185631 ++ cd /tmp/bd_20220515-185631 ++ pwd + dir=/tmp/bd_20220515-185631 + PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/lib/ jvm/default/bin:/opt/rocm-5.1.2/bin:/opt/rocm-5.1.2/hip/bin: /usr/bin/site_perl:/usr/bin/vendor_perl:/usr/bin/core_ perl:/home/paz/cinelerra5/cinelerra-5.1/bin + mkdir -p /tmp/bd_20220515-185631/udfs ++ du -cb /tmp/bd_20220515-185631/bd.m2ts ++ tail -1 ++ sed -e 's/[ ].*//' + sz=44414976 + blks=25783 + rm -f /tmp/bd_20220515-185631/bd.udfs + '[' -f /tmp/bd_20220515-185631/bd.meta ']' + mkudffs -b 2048 /tmp/bd_20220515-185631/bd.udfs 25783 filename=/tmp/bd_20220515-185631/bd.udfs label=LinuxUDF uuid=628130f3e97336f3 blocksize=2048 blocks=25783 udfrev=2.01 start=0, blocks=16, type=ERASE start=16, blocks=4, type=VRS start=20, blocks=76, type=ERASE start=96, blocks=16, type=MVDS start=112, blocks=16, type=ERASE start=128, blocks=4, type=LVID start=132, blocks=124, type=ERASE start=256, blocks=1, type=ANCHOR start=257, blocks=25264, type=PSPACE start=25521, blocks=5, type=ERASE start=25526, blocks=1, type=ANCHOR start=25527, blocks=73, type=ERASE start=25600, blocks=16, type=RVDS start=25616, blocks=166, type=ERASE start=25782, blocks=1, type=ANCHOR + mount -t udf -o loop /tmp/bd_20220515-185631/bd.udfs /tmp/bd_20220515-185631/udfs + bdwrite /tmp/bd_20220515-185631/udfs /tmp/bd_20220515-185631/bd.m2ts unknown bluray audio format 0 ch

I have workaround for this, just not sure if it makes any sense without proper ac3/thd interleaving ... currently I think you can try to render just video in ts and ac3 + thd audio separately, then use thdmerge on audio and tsmuxer on video ts + result of thdmerge...

CinGG crashes at the end of rendering; but the bd.m2ts file is produced anyway. This is not read by VLC but works in mpv. As you anticipated only 2 audio tracks were produced: The 2 source files were m2ts (plus a h264 .mov video) with 22 audio channels each.

$ ffprobe bd.m2ts ffprobe version n5.0 Copyright (c) 2007-2022 the FFmpeg developers built with gcc 11.2.0 (GCC) configuration: --prefix=/usr --disable-debug --disable-static --disable-stripping --enable-amf --enable-avisynth --enable-cuda-llvm --enable-lto --enable-fontconfig --enable-gmp --enable-gnutls --enable-gpl --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libdav1d --enable-libdrm --enable-libfreetype --enable-libfribidi --enable-libgsm --enable-libiec61883 --enable-libjack --enable-libmfx --enable-libmodplug --enable-libmp3lame --enable-libopencore_amrnb --enable-libopencore_amrwb --enable-libopenjpeg --enable-libopus --enable-libpulse --enable-librav1e --enable-librsvg --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libssh --enable-libsvtav1 --enable-libtheora --enable-libv4l2 --enable-libvidstab --enable-libvmaf --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxcb --enable-libxml2 --enable-libxvid --enable-libzimg --enable-nvdec --enable-nvenc --enable-shared --enable-version3 libavutil 57. 17.100 / 57. 17.100 libavcodec 59. 18.100 / 59. 18.100 libavformat 59. 16.100 / 59. 16.100 libavdevice 59. 4.100 / 59. 4.100 libavfilter 8. 24.100 / 8. 24.100 libswscale 6. 4.100 / 6. 4.100 libswresample 4. 3.100 / 4. 3.100 libpostproc 56. 3.100 / 56. 3.100 [mpegts @ 0x55cd19a640c0] start time for stream 2 is not set in estimate_timings_from_pts [mpegts @ 0x55cd19a640c0] stream 2 : no TS found at start of file, duration not set [mpegts @ 0x55cd19a640c0] Could not find codec parameters for stream 2 (Audio: ac3 ([131][0][0][0] / 0x0083), 0 channels, fltp): unspecified sample rate Consider increasing the value for the 'analyzeduration' (0) and 'probesize' (5000000) options Input #0, mpegts, from 'bd.m2ts': Duration: 00:00:28.24, start: 0.083422, bitrate: 12584 kb/s Program 1 Stream #0:0[0x1011]: Video: h264 (High) ([27][0][0][0] / 0x001B), yuv420p(tv, bt709, progressive), 1920x1080 [SAR 1:1 DAR 16:9], 23.98 fps, 23.98 tbr, 90k tbn Stream #0:1[0x1100](ita): Audio: truehd ([131][0][0][0] / 0x0083), 48000 Hz, stereo, s32 (24 bit) Stream #0:2[0x1100]: Audio: ac3 ([131][0][0][0] / 0x0083), 0 channels, fltp

I also add cin5.log and config.log in case you ever need it.