using old switch --enable-x265_hidepth build x265 10/12bit ([yes]) Note, i think I'll rebase this on top of Vulkan/etc patches but at some point Phyllis asked if there was way to make x265 build only 8bit variant for faster testing. I think now it is!
Andrew, I have been using the method as documented in the manual, but this patch method is more in keeping with the way things are supposed to be done so will be testing this tomorrow. I am trying to get the easiest things checked off my list first and then upgrade one partition to Debian 13 to check some of your other patches. Thank you! On Sun, Dec 7, 2025 at 10:21 PM Andrew Randrianasulu < [email protected]> wrote:
using old switch
--enable-x265_hidepth build x265 10/12bit ([yes])
Note, i think I'll rebase this on top of Vulkan/etc patches but at some point Phyllis asked if there was way to make x265 build only 8bit variant for faster testing. I think now it is!
The default of x265 multibit works all of the time. With the 2 patches applied, I can not get x265 multibit to NOT work. Tried: --disable-x265_hidepth and --enable -x265_hidepth=no . In looking at the log I see what I expect except always get "using: with-x265_depth". What am I doing wrong?
using: with-xv using: with-vaapi using: with-vdpau using: without-onevpl using: without-vulkan using: without-libplacebo using: without-libzimg using: without-cuda
using: with-nv
using: with-wintv using: with-x10tv *using: with-x265_depth*
So I tried without the "hi" of "hidepth" with no luck: configure: WARNING: unrecognized options: --enable-x265_depth On Sun, Dec 7, 2025 at 10:21 PM Andrew Randrianasulu < [email protected]> wrote:
using old switch
--enable-x265_hidepth build x265 10/12bit ([yes])
Note, i think I'll rebase this on top of Vulkan/etc patches but at some point Phyllis asked if there was way to make x265 build only 8bit variant for faster testing. I think now it is!
пн, 15 дек. 2025 г., 03:15 Phyllis Smith <[email protected]>:
The default of x265 multibit works all of the time. With the 2 patches applied, I can not get x265 multibit to NOT work. Tried: --disable-x265_hidepth and --enable -x265_hidepth=no .
try --without-x265_hidepth ? In looking at the log I see what I expect except always get "using:
with-x265_depth". What am I doing wrong?
using: with-xv using: with-vaapi using: with-vdpau using: without-onevpl using: without-vulkan using: without-libplacebo using: without-libzimg using: without-cuda
using: with-nv
using: with-wintv using: with-x10tv *using: with-x265_depth*
So I tried without the "hi" of "hidepth" with no luck: configure: WARNING: unrecognized options: --enable-x265_depth
On Sun, Dec 7, 2025 at 10:21 PM Andrew Randrianasulu < [email protected]> wrote:
using old switch
--enable-x265_hidepth build x265 10/12bit ([yes])
Note, i think I'll rebase this on top of Vulkan/etc patches but at some point Phyllis asked if there was way to make x265 build only 8bit variant for faster testing. I think now it is!
пн, 15 дек. 2025 г., 03:15 Phyllis Smith <[email protected]>:
The default of x265 multibit works all of the time. With the 2 patches applied, I can not get x265 multibit to NOT work. Tried: --disable-x265_hidepth and --enable -x265_hidepth=no . In looking at the log I see what I expect except always get "using: with-x265_depth". What am I doing wrong?
Strange, probably rebase go wrong? With my git tree I have: ./configure --with-single-user --disable-x265_hidepth using: without-x265_hidepth and it sets WANT_X265_HIDEPTH := no in global_config using: with-xv
using: with-vaapi using: with-vdpau using: without-onevpl using: without-vulkan using: without-libplacebo using: without-libzimg using: without-cuda
using: with-nv
using: with-wintv using: with-x10tv *using: with-x265_depth*
So I tried without the "hi" of "hidepth" with no luck: configure: WARNING: unrecognized options: --enable-x265_depth
On Sun, Dec 7, 2025 at 10:21 PM Andrew Randrianasulu < [email protected]> wrote:
using old switch
--enable-x265_hidepth build x265 10/12bit ([yes])
Note, i think I'll rebase this on top of Vulkan/etc patches but at some point Phyllis asked if there was way to make x265 build only 8bit variant for faster testing. I think now it is!
On Mon, Dec 15, 2025 at 4:17 AM Andrew Randrianasulu <[email protected]> wrote:
пн, 15 дек. 2025 г., 03:15 Phyllis Smith <[email protected]>:
The default of x265 multibit works all of the time. With the 2 patches applied, I can not get x265 multibit to NOT work. Tried: --disable-x265_hidepth and --enable -x265_hidepth=no . In looking at the log I see what I expect except always get "using: with-x265_depth". What am I doing wrong?
Strange, probably rebase go wrong?
With my git tree I have:
./configure --with-single-user --disable-x265_hidepth
using: without-x265_hidepth
and it sets
WANT_X265_HIDEPTH := no
in global_config
I rebased patches manually (there was conflict in configure.ac It sets variables ok, hopefully. for me. I think it sets yes twice, will try to rework/retest it a bit now.
using: with-xv using: with-vaapi using: with-vdpau using: without-onevpl using: without-vulkan using: without-libplacebo using: without-libzimg using: without-cuda
using: with-nv using: with-wintv using: with-x10tv using: with-x265_depth
So I tried without the "hi" of "hidepth" with no luck: configure: WARNING: unrecognized options: --enable-x265_depth
On Sun, Dec 7, 2025 at 10:21 PM Andrew Randrianasulu <[email protected]> wrote:
using old switch
--enable-x265_hidepth build x265 10/12bit ([yes])
Note, i think I'll rebase this on top of Vulkan/etc patches but at some point Phyllis asked if there was way to make x265 build only 8bit variant for faster testing. I think now it is!
On Mon, Dec 15, 2025 at 4:31 AM Andrew Randrianasulu <[email protected]> wrote:
On Mon, Dec 15, 2025 at 4:17 AM Andrew Randrianasulu <[email protected]> wrote:
пн, 15 дек. 2025 г., 03:15 Phyllis Smith <[email protected]>:
The default of x265 multibit works all of the time. With the 2 patches applied, I can not get x265 multibit to NOT work. Tried: --disable-x265_hidepth and --enable -x265_hidepth=no . In looking at the log I see what I expect except always get "using: with-x265_depth". What am I doing wrong?
Strange, probably rebase go wrong?
With my git tree I have:
./configure --with-single-user --disable-x265_hidepth
using: without-x265_hidepth
and it sets
WANT_X265_HIDEPTH := no
in global_config
I rebased patches manually (there was conflict in configure.ac
It sets variables ok, hopefully. for me.
I think it sets yes twice, will try to rework/retest it a bit now.
removed chunk I added initially, it really not needed. Now variable should be set only once
using: with-xv using: with-vaapi using: with-vdpau using: without-onevpl using: without-vulkan using: without-libplacebo using: without-libzimg using: without-cuda
using: with-nv using: with-wintv using: with-x10tv using: with-x265_depth
So I tried without the "hi" of "hidepth" with no luck: configure: WARNING: unrecognized options: --enable-x265_depth
On Sun, Dec 7, 2025 at 10:21 PM Andrew Randrianasulu <[email protected]> wrote:
using old switch
--enable-x265_hidepth build x265 10/12bit ([yes])
Note, i think I'll rebase this on top of Vulkan/etc patches but at some point Phyllis asked if there was way to make x265 build only 8bit variant for faster testing. I think now it is!
With the last 3 patches, I can disable h265_high_depth: to compile is faster and rendering at more than 8-bit depth is not possible. Using the 3 patches but not using the “--disable-x265_hidepth” flag in ./configure, compilation works fine, the time is shorter, and I can render at 10-12 bits. That's OK for me.
Andrew, thanks for rebasing but the error was mine because I had applied the patches to Makefile manually and missed a closing parenthesis. So it works just fine on the older Fedora. But I keep getting segmentation compiler error on fairly new Fedora 42 when compile hidepth, which I think was mentioned somewhere in the mailing list. Will look at tomorrow. On Mon, Dec 15, 2025 at 2:43 AM Andrea paz <[email protected]> wrote:
With the last 3 patches, I can disable h265_high_depth: to compile is faster and rendering at more than 8-bit depth is not possible. Using the 3 patches but not using the “--disable-x265_hidepth” flag in ./configure, compilation works fine, the time is shorter, and I can render at 10-12 bits. That's OK for me.
вт, 16 дек. 2025 г., 03:19 Phyllis Smith <[email protected]>:
Andrew, thanks for rebasing but the error was mine because I had applied the patches to Makefile manually and missed a closing parenthesis. So it works just fine on the older Fedora. But I keep getting segmentation compiler error on fairly new Fedora 42 when compile hidepth, which I think was mentioned somewhere in the mailing list. Will look at tomorrow.
Changelog for x265 4.1 (nearly 1 year old but still latest) mention some compilation issue on clang. https://x265.readthedocs.io/en/master/releasenotes.html You can try to upgrade our copy to 4.1 and see how ( if ) it improves situation on fedora 42.
On Mon, Dec 15, 2025 at 2:43 AM Andrea paz <[email protected]> wrote:
With the last 3 patches, I can disable h265_high_depth: to compile is faster and rendering at more than 8-bit depth is not possible. Using the 3 patches but not using the “--disable-x265_hidepth” flag in ./configure, compilation works fine, the time is shorter, and I can render at 10-12 bits. That's OK for me.
Checked into GIT, the multilib x265_hidepth options change as provided by Andrew. It did compile on Fedora 42 finally -- the laptop may have overheated causing me 3 times in a row to end up with compiler, internal, or segmentation errors. On Mon, Dec 15, 2025 at 6:01 PM Andrew Randrianasulu < [email protected]> wrote:
вт, 16 дек. 2025 г., 03:19 Phyllis Smith <[email protected]>:
Andrew, thanks for rebasing but the error was mine because I had applied the patches to Makefile manually and missed a closing parenthesis. So it works just fine on the older Fedora. But I keep getting segmentation compiler error on fairly new Fedora 42 when compile hidepth, which I think was mentioned somewhere in the mailing list. Will look at tomorrow.
Changelog for x265 4.1 (nearly 1 year old but still latest) mention some compilation issue on clang.
https://x265.readthedocs.io/en/master/releasenotes.html
You can try to upgrade our copy to 4.1 and see how ( if ) it improves situation on fedora 42.
On Mon, Dec 15, 2025 at 2:43 AM Andrea paz <[email protected]> wrote:
With the last 3 patches, I can disable h265_high_depth: to compile is faster and rendering at more than 8-bit depth is not possible. Using the 3 patches but not using the “--disable-x265_hidepth” flag in ./configure, compilation works fine, the time is shorter, and I can render at 10-12 bits. That's OK for me.
ср, 17 дек. 2025 г., 00:20 Phyllis Smith <[email protected]>:
Checked into GIT, the multilib x265_hidepth options change as provided by Andrew. It did compile on Fedora 42 finally -- the laptop may have overheated causing me 3 times in a row to end up with compiler, internal, or segmentation errors.
I *STRONGLY* recommend to take this laptop to doctor ... err, repairman.
On Mon, Dec 15, 2025 at 6:01 PM Andrew Randrianasulu < [email protected]> wrote:
вт, 16 дек. 2025 г., 03:19 Phyllis Smith <[email protected]>:
Andrew, thanks for rebasing but the error was mine because I had applied the patches to Makefile manually and missed a closing parenthesis. So it works just fine on the older Fedora. But I keep getting segmentation compiler error on fairly new Fedora 42 when compile hidepth, which I think was mentioned somewhere in the mailing list. Will look at tomorrow.
Changelog for x265 4.1 (nearly 1 year old but still latest) mention some compilation issue on clang.
https://x265.readthedocs.io/en/master/releasenotes.html
You can try to upgrade our copy to 4.1 and see how ( if ) it improves situation on fedora 42.
On Mon, Dec 15, 2025 at 2:43 AM Andrea paz <[email protected]> wrote:
With the last 3 patches, I can disable h265_high_depth: to compile is faster and rendering at more than 8-bit depth is not possible. Using the 3 patches but not using the “--disable-x265_hidepth” flag in ./configure, compilation works fine, the time is shorter, and I can render at 10-12 bits. That's OK for me.
participants (3)
-
Andrea paz -
Andrew Randrianasulu -
Phyllis Smith