Why does the 'detelecine' filter exist? It is weak and clumsy.
The 'fieldmatch' filter followed by 'decimate' do a better job and are robust.
By continuing to carry obsolete filters, FFmpeg emasculates itself. That makes FFmpeg appear to be
chaotic and ties up names that could be, should be reu
On Wed, May 28, 2025 at 7:00 PM Mark Filipak
wrote:
> Why does the 'detelecine' filter exist? It is weak and clumsy.
>
> The 'fieldmatch' filter followed by 'decimate' do a better job and are
> robust.
>
> By continuing to carry obsolete filters, FFmpeg emasculates itself. That
> makes FFmpeg app
On 28/05/2025 15.04, Paul B Mahol wrote:
On Wed, May 28, 2025 at 7:00 PM Mark Filipak
wrote:
Why does the 'detelecine' filter exist? It is weak and clumsy.
The 'fieldmatch' filter followed by 'decimate' do a better job and are
robust.
By continuing to carry obsolete filters, FFmpeg emasculat
On Wed, May 28, 2025 at 7:07 PM Mark Filipak
wrote:
> On 28/05/2025 15.04, Paul B Mahol wrote:
> > On Wed, May 28, 2025 at 7:00 PM Mark Filipak >
> > wrote:
> >
> >> Why does the 'detelecine' filter exist? It is weak and clumsy.
> >>
> >> The 'fieldmatch' filter followed by 'decimate' do a bette
I recently used the detelecine command in this thread:
https://ffmpeg.org/pipermail/ffmpeg-user/2025-May/059249.html
I got strange results with the `fieldmatch+decimate` combo, where decimate
wasn't removing the correct frame.
I also got strange results with the `pullup` filter.
This may just be
On 5/28/2025 12:00 PM, Mark Filipak wrote:
By continuing to carry obsolete filters, FFmpeg emasculates itself.
How?
That makes FFmpeg appear to be chaotic and ties up names that could
be, should be reused to better effect.
There are probably extremely few instances where a name would rationa
On 28/05/2025 15.39, Alex Xu wrote:
I recently used the detelecine command in this thread:
https://ffmpeg.org/pipermail/ffmpeg-user/2025-May/059249.html
I got strange results with the `fieldmatch+decimate` combo, where decimate
wasn't removing the correct frame.
I also got strange results with
On Wed, May 28, 2025 at 9:19 PM Mark Filipak
wrote:
> On 28/05/2025 15.39, Alex Xu wrote:
> > I recently used the detelecine command in this thread:
> > https://ffmpeg.org/pipermail/ffmpeg-user/2025-May/059249.html
> >
> > I got strange results with the `fieldmatch+decimate` combo, where
> decima
On Wed, May 28, 2025 at 9:25 PM Mark Filipak
wrote:
> On 28/05/2025 16.33, Carl Zwanzig wrote:
> > On 5/28/2025 12:00 PM, Mark Filipak wrote:
> >
> >> By continuing to carry obsolete filters, FFmpeg emasculates itself.
> >
> > How?
>
> Developers have to dream up new names because what they are d
On 29/05/2025 12:57 pm, Mark Filipak wrote:
>I couldn't get the frame numbers burned into the pictures
Did you specify the full pathname? For example, in Windows 10, when
adding a two-filename text overlay to identify the two sides in a
split-screen comparison of unstabilised vs stabilised ver
I got those "unnecessary" args from
https://trac.ffmpeg.org/wiki/Encode/YouTube They aren't important.
I think you might be mistaking CRF for CFR (constant frame rate). CRF is a
measure of compression, it shouldn't drop and duplicate frames. frames.
https://trac.ffmpeg.org/wiki/Encode/H.264#a1.Cho
On 28/05/2025 21.27, Alex Xu wrote:
I got those "unnecessary" args from
https://trac.ffmpeg.org/wiki/Encode/YouTube They aren't important.
I think you might be mistaking CRF for CFR (constant frame rate).
Nope.
CRF is a
measure of compression, it shouldn't drop and duplicate frames. frames.
I did this:
ffmpeg^
-i g:\detelecine_input.mkv^
-vf fieldmatch,decimate^
-c:a aac^
-c:v libx264^
g:\fieldmatch_decimate_only_no_r.mkv
I couldn't get the frame numbers burned into the pictures, so I counted them as
I single stepped.
There were no dropped frames. There were no repeated fram
On 29/05/2025 00.26, Carl Zwanzig wrote:
On 5/28/2025 2:37 PM, Mark Filipak wrote:
There are probably extremely few instances where a name would rationally
be reused;
Name even a single such instance, Carl. Name reuse to a better filter
could be covered by the
documentation.
Your mission is
By the way, frame 173 (at running time 7.215) is the first frame following a shot change. There's
nothing wrong with 'fieldmatch,decimate'. You need to look elsewhere -- your text overlay, maybe? --
to determine the problem.
I did this:
ffmpeg^
-i g:\detelecine_input.mkv^
-vf fieldmatch,deci
And one quick note, my external video player (IINA and VLC) seems to not
like it if I don't include "-pix_fmt yuv420p" Otherwise I get strange
screen tearing issues.
On Wed, May 28, 2025 at 9:30 PM Alex Xu wrote:
> I tried running your exact command and got the same result. Dropped frame
> 173 a
Am 28.05.25 um 21:00 schrieb Mark Filipak:
Why does the 'detelecine' filter exist? It is weak and clumsy.
The 'fieldmatch' filter followed by 'decimate' do a better job and are
robust.
By continuing to carry obsolete filters, FFmpeg emasculates itself. That
makes FFmpeg appear to be chaot
On 5/28/2025 2:37 PM, Mark Filipak wrote:
There are probably extremely few instances where a name would
rationally
be reused;
Name even a single such instance, Carl. Name reuse to a better filter
could be covered by the
documentation.
Your mission is failed.
My challenge was to Carl.
Nop
I tried running your exact command and got the same result. Dropped frame
173 and a later duplicated frame.
I've attached 15 frames here of the results.
https://drive.google.com/drive/folders/139rjLqsxIuAO_3yzVNe4dD7-gI9YPRgn?usp=drive_link
I've attached 5 frames each of the original input, usin
On 28/05/2025 16.33, Carl Zwanzig wrote:
On 5/28/2025 12:00 PM, Mark Filipak wrote:
By continuing to carry obsolete filters, FFmpeg emasculates itself.
How?
Developers have to dream up new names because what they are doing, what they are writing, has a
name, but that name has already been
On 28/05/2025 17.28, Paul B Mahol wrote:
On Wed, May 28, 2025 at 9:25 PM Mark Filipak
wrote:
On 28/05/2025 16.33, Carl Zwanzig wrote:
On 5/28/2025 12:00 PM, Mark Filipak wrote:
By continuing to carry obsolete filters, FFmpeg emasculates itself.
How?
Developers have to dream up new names
On Wed, May 28, 2025 at 9:35 PM Mark Filipak
wrote:
> On 28/05/2025 17.27, Paul B Mahol wrote:
> > On Wed, May 28, 2025 at 9:19 PM Mark Filipak >
> > wrote:
> >
> >> On 28/05/2025 15.39, Alex Xu wrote:
> >>> I recently used the detelecine command in this thread:
> >>> https://ffmpeg.org/pipermai
Thanks for your feedback, I tried your command but added a frame counter:
ffmpeg -i detelecine_input.mkv -vf
"fieldmatch=order=tff:combmatch=none,decimate,drawtext=fontfile=Arial.ttf:
text='%{frame_num}': start_number=0: x=(w-tw)/2: y=h-(2*lh):
fontcolor=black: fontsize=20: box=1: boxcolor=white:
Got the same problematic result with
ffmpeg -i detelecine_input.mkv -vf
"fieldmatch,decimate,drawtext=fontfile=Arial.ttf: text='%{frame_num}':
start_number=0: x=(w-tw)/2: y=h-(2*lh): fontcolor=black: fontsize=20:
box=1: boxcolor=white: boxborderw=5" -c:v libx264 -crf 18 -c:a aac -b:a
192k -pix_fmt
On 28/05/2025 18.15, Paul B Mahol wrote:
On Wed, May 28, 2025 at 9:35 PM Mark Filipak
wrote:
On 28/05/2025 17.27, Paul B Mahol wrote:
On Wed, May 28, 2025 at 9:19 PM Mark Filipak
On 28/05/2025 15.39, Alex Xu wrote:
I recently used the detelecine command in this thread:
https://ffmpeg.org/pi
On 28/05/2025 18.36, Alex Xu wrote:
Got the same problematic result with
ffmpeg -i detelecine_input.mkv -vf
"fieldmatch,decimate,drawtext=fontfile=Arial.ttf: text='%{frame_num}':
start_number=0: x=(w-tw)/2: y=h-(2*lh): fontcolor=black: fontsize=20:
box=1: boxcolor=white: boxborderw=5" -c:v libx2
чт, 29 мая 2025 г., 01:38 Mark Filipak :
> On 28/05/2025 18.15, Paul B Mahol wrote:
> > On Wed, May 28, 2025 at 9:35 PM Mark Filipak >
> > wrote:
> >
> >> On 28/05/2025 17.27, Paul B Mahol wrote:
> >>> On Wed, May 28, 2025 at 9:19 PM Mark Filipak <
> markfilipak.i...@gmail.com
> >>>
> >>> wrote:
On 28/05/2025 19.33, Andrew Randrianasulu wrote:
чт, 29 мая 2025 г., 01:38 Mark Filipak :
...
Generally I prefer to have two versions of similar things, so when one
broke at least I have plan B or some easy way to compare them.
So, I conclude you dislike deprecation. But I assume you want FFm
Just to be clear, in the other thread, I got detelecine to work (with the
wrong params?) but couldn't get fieldmatch+decimate to work. But this may
have been due to my files being particularly exotic.
And there's also another inverse detelecine filter called "pullup".
Also, how do you do: configu
On 28/05/2025 17.27, Paul B Mahol wrote:
On Wed, May 28, 2025 at 9:19 PM Mark Filipak
wrote:
On 28/05/2025 15.39, Alex Xu wrote:
I recently used the detelecine command in this thread:
https://ffmpeg.org/pipermail/ffmpeg-user/2025-May/059249.html
I got strange results with the `fieldmatch+dec
On 28/05/2025 18.07, Alex Xu wrote:
Thanks for your feedback, I tried your command but added a frame counter:
ffmpeg -i detelecine_input.mkv -vf
"fieldmatch=order=tff:combmatch=none,decimate,drawtext=fontfile=Arial.ttf:
text='%{frame_num}': start_number=0: x=(w-tw)/2: y=h-(2*lh):
fontcolor=black
I put it there since it was in Huspnet's original command. I don't think it
matters, same problem occurs if I omit it or use -r 24000/1001 instead.
On Wed, May 28, 2025 at 3:41 PM Mark Filipak
wrote:
> On 28/05/2025 18.36, Alex Xu wrote:
> > Got the same problematic result with
> >
> > ffmpeg -i
Heh-heh. I don't know what happened to my previous post. My apologies.
On 28/05/2025 18.36, Alex Xu wrote:
Got the same problematic result with
ffmpeg -i detelecine_input.mkv -vf
"fieldmatch,decimate,drawtext=fontfile=Arial.ttf: text='%{frame_num}':
start_number=0: x=(w-tw)/2: y=h-(2*lh): fontc
On 28/05/2025 18.46, Alex Xu wrote:
I put it there since it was in Huspnet's original command.
Ahem!
I don't think it
matters,
Of course it matters. You're taking 30/1.001fps, detelecining to 24/1.001fps, then telling FFmpeg to
make 24fps from it. Of course, every 1000th frame is repeated.
When I said, "doesn't matter" I meant that the issue of duplicated frames
174+175 and skipped 173. Yes, I understand that incorrectly specifying the
fps can cause other issues though.
Anyways, I tried:
ffmpeg -i detelecine_input.mkv -vf
"fieldmatch,decimate,drawtext=fontfile=Arial.ttf: text='%{fr
Whoops, looks like I got cut off, the above command still has same issue of
duplicated frames 174+175 and skipped 173
On Wed, May 28, 2025 at 4:18 PM Alex Xu wrote:
> When I said, "doesn't matter" I meant that the issue of duplicated frames
> 174+175 and skipped 173. Yes, I understand that incor
On 28/05/2025 19.18, Alex Xu wrote:
Anyways, I tried:
ffmpeg -i detelecine_input.mkv -vf
"fieldmatch,decimate,drawtext=fontfile=Arial.ttf: text='%{frame_num}':
start_number=0: x=(w-tw)/2: y=h-(2*lh): fontcolor=black: fontsize=20:
box=1: boxcolor=white: boxborderw=5" -c:v libx264 -crf 18 -c:a aac
37 matches
Mail list logo
