vp9 is not Supporting profile 3

[pokuri...@gmail.com]

I have configure the codec with high bit depth(libvpx 1.6.1)
while executing the vp9 encoder i am not able to execute the video with profile 3 and i am getting the error Failed to initialize encoder: Invalid parameter Profile > 1 not supported in this build configuration
    i am using Linux platform
Command arguments are : ./configure --enable-vp9-highbitdepth and the make
for executing the encoder commands are : vpxenc Input.yuv -o test.webm -w 352 -h 288 --profile=3

[James Zern]

Hi,

On Mon, Jan 30, 2017 at 2:30 AM, <pokuri...@gmail.com> wrote:
> I have configure the codec with high bit depth(libvpx 1.6.1)
> while executing the vp9 encoder i am not able to execute the video with
> profile 3 and i am getting the error Failed to initialize encoder: Invalid
> parameter Profile > 1 not supported in this build configuration

This works for me using a build from the release tag v1.6.1. It's
possible you have a stale vpx_config.h, you might want to clear your
tree and reconfigure.

> i am using Linux platform
> Command arguments are : ./configure --enable-vp9-highbitdepth and the make
> for executing the encoder commands are : vpxenc Input.yuv -o test.webm -w
> 352 -h 288 --profile=3
>

Note you'll most like need --bit-depth=10 or --bit-depth=12 here.

[Debargha Mukherjee]

For raw yuv source, use the ​--input-bit-depth=<n> option to specify the source bit-depth 

along with one of ​--i420, ​--i422, --i444 or --i440 to specify the color sampling, ​in 

addition​ ​to​ ​​--width=<w>​,​ ​​--height=<h>​​ ​and​ ​​--fps=<f>​​ ​options.

--
You received this message because you are subscribed to the Google Groups "WebM Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webm-discuss+unsubscribe@webmproject.org.
To post to this group, send email to webm-d...@webmproject.org.
Visit this group at https://groups.google.com/a/webmproject.org/group/webm-discuss/.
For more options, visit https://groups.google.com/a/webmproject.org/d/optout.

[pokuri...@gmail.com]

Hi,

I have tried with  command vpxenc input.yuv -o testOut.webm -w 352 -h 288 --best --max-q=0 --min-q=0 --input-bit-depth=12 --i422
but the compiller throwing error: Error: Unrecognized option --input-bit-depth=12

Usage: vpxenc <options> -o dst_filename src_filename

Options:
  -D,       --debug                        Debug mode (makes output deterministic)
  -o <arg>, --output=<arg>                 Output filename
   .
   .
   .
some more option i am getting 
--input-bit-depth is not available in this

On Monday, January 30, 2017 at 7:28:50 PM UTC, Debargha Mukherjee wrote:

For raw yuv source, use the ​--input-bit-depth=<n> option to specify the source bit-depth 

along with one of ​--i420, ​--i422, --i444 or --i440 to specify the color sampling, ​in 

addition​ ​to​ ​​--width=<w>​,​ ​​--height=<h>​​ ​and​ ​​--fps=<f>​​ ​options.

On Mon, Jan 30, 2017 at 11:21 AM, 'James Zern' via WebM Discussion <webm-d...@webmproject.org> wrote:

Hi,

On Mon, Jan 30, 2017 at 2:30 AM,  <pokuri...@gmail.com> wrote:
> I have configure the codec with high bit depth(libvpx 1.6.1)
> while executing the vp9 encoder i am not able to execute the video with
> profile 3 and i am getting the error Failed to initialize encoder: Invalid
> parameter Profile > 1 not supported in this build configuration

This works for me using a build from the release tag v1.6.1. It's
possible you have a stale vpx_config.h, you might want to clear your
tree and reconfigure.

>     i am using Linux platform
> Command arguments are : ./configure --enable-vp9-highbitdepth and the make
> for executing the encoder commands are : vpxenc Input.yuv -o test.webm -w
> 352 -h 288 --profile=3
>

Note you'll most like need --bit-depth=10 or --bit-depth=12 here.

--
You received this message because you are subscribed to the Google Groups "WebM Discussion" group.

To unsubscribe from this group and stop receiving emails from it, send an email to webm-discuss...@webmproject.org.

[James Zern]

On Mon, Jan 30, 2017 at 10:30 PM, <pokuri...@gmail.com> wrote:
> Hi,
>
> I have tried with command vpxenc input.yuv -o testOut.webm -w 352 -h 288
> --best --max-q=0 --min-q=0 --input-bit-depth=12 --i422
> but the compiller throwing error: Error: Unrecognized option
> --input-bit-depth=12
>
> Usage: vpxenc <options> -o dst_filename src_filename
>
> Options:
> -D, --debug Debug mode (makes output
> deterministic)
> -o <arg>, --output=<arg> Output filename
> .
> .
> .
> some more option i am getting
> --input-bit-depth is not available in this
>

That seems to confirm your build is not a high-bitdepth enabled build.
That option is protected by CONFIG_VP9_HIGHBITDEPTH (enabled with
--enable-vp9-highbitdepth). You might want to try doing another build
in a clean checkout.

[pokuri...@gmail.com]

I have enabled --enable-vp9-highbitdepth while configuring and cross checked in configure make doc which will be generated after successful configuration

command used for that is: ./configure --enable-vp9-highbitdepth

[James Zern]

On Tue, Jan 31, 2017 at 2:50 AM, <pokuri...@gmail.com> wrote:
>
> I have enabled --enable-vp9-highbitdepth while configuring and cross checked
> in configure make doc which will be generated after successful configuration
> command used for that is: ./configure --enable-vp9-highbitdepth
>

Something doesn't seem to line up with the binary you're running for
your test. You could try to put a #error within any of the #if
CONFIG_VP9_HIGHBITDEPTH blocks within vpxenc.c to double check your
build.

>
>
>
> On Tuesday, January 31, 2017 at 12:12:56 PM UTC+5:30, James Zern wrote:
>>
>> On Mon, Jan 30, 2017 at 10:30 PM, <pokuri...@gmail.com> wrote:
>> > Hi,
>> >
>> > I have tried with command vpxenc input.yuv -o testOut.webm -w 352 -h
>> > 288
>> > --best --max-q=0 --min-q=0 --input-bit-depth=12 --i422
>> > but the compiller throwing error: Error: Unrecognized option
>> > --input-bit-depth=12
>> >
>> > Usage: vpxenc <options> -o dst_filename src_filename
>> >
>> > Options:
>> > -D, --debug Debug mode (makes output
>> > deterministic)
>> > -o <arg>, --output=<arg> Output filename
>> > .
>> > .
>> > .
>> > some more option i am getting
>> > --input-bit-depth is not available in this
>> >
>>
>> That seems to confirm your build is not a high-bitdepth enabled build.
>> That option is protected by CONFIG_VP9_HIGHBITDEPTH (enabled with
>> --enable-vp9-highbitdepth). You might want to try doing another build
>> in a clean checkout.
>

[pokuri...@gmail.com]

do VP9 suports 12bit yuv 422 chroma format..??

[James Zern]

Hi,

On Wed, Feb 1, 2017 at 2:24 AM, <pokuri...@gmail.com> wrote:
> do VP9 suports 12bit yuv 422 chroma format..??
>

Yes, in profile 3. From [1]:
// 00: Profile 0. 8-bit 4:2:0 only.
// 10: Profile 1. 8-bit 4:4:4, 4:2:2, and 4:4:0.
// 01: Profile 2. 10-bit and 12-bit color only, with 4:2:0 sampling.
// 110: Profile 3. 10-bit and 12-bit color only, with 4:2:2/4:4:4/4:4:0
// sampling.

[1] https://chromium.googlesource.com/webm/libvpx/+/master/vp9/common/vp9_enums.h#30

[pokuri...@gmail.com]

Hi Thank you for your analysis..
In the previous mail you asked me to cross check with #error in vp_encoder.c, I have done that exercise, while encoding error message should get popped up but i haven't got the error message.

After configuring with  ./configure --enable-vp9-highbitdepth
The following files got generated docs-x86_64-linux-gcc.mk, tools-x86_64-linux-gcc.mk .. etc. with all these files config.log also got genarated.

I am able to see the Flag CONFIG_VP9_HIGHBITDEPTH = yes in all the following make files docs-x86_64-linux-gcc.mk,tools-x86_64-linux-gcc.mk,libs-x86_64-linux-gcc.mk , but i haven't seen this in config.log file
will this create an issue..??

[James Zern]

Hi,

On Wed, Feb 1, 2017 at 8:30 PM, <pokuri...@gmail.com> wrote:
> Hi Thank you for your analysis..
> In the previous mail you asked me to cross check with #error in
> vp_encoder.c, I have done that exercise, while encoding error message should
> get popped up but i haven't got the error message.
>
> After configuring with ./configure --enable-vp9-highbitdepth
> The following files got generated docs-x86_64-linux-gcc.mk,
> tools-x86_64-linux-gcc.mk .. etc. with all these files config.log also got
> genarated.
>
> I am able to see the Flag CONFIG_VP9_HIGHBITDEPTH = yes in all the following
> make files
> docs-x86_64-linux-gcc.mk,tools-x86_64-linux-gcc.mk,libs-x86_64-linux-gcc.mk
> , but i haven't seen this in config.log file
> will this create an issue..??
>

config.log is unlikely to mention anything regarding this flag,
vpx_config.h would end up having the CONFIG_VP9_HIGHBITDEPTH
reference. This should be set to 1 given the configure arguments. If
it is and you're not hitting the #error you likely have another copy
in your include path.