2025-08-22 10:25
michaelni changed the topic of #ffmpeg-devel to: Welcome to the FFmpeg development channel | Questions about using FFmpeg or developing with libav* libs should be asked in #ffmpeg | This channel is publicly logged | FFmpeg 8.0 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
00:04
Kimapr has quit [Ping timeout: 244 seconds]
00:08
Kimapr has joined #ffmpeg-devel
00:40
damian101_ has joined #ffmpeg-devel
00:42
damian101 has quit [Ping timeout: 240 seconds]
01:13
ubitux has quit [Ping timeout: 250 seconds]
01:15
ubitux has joined #ffmpeg-devel
02:28
mkver has quit [Ping timeout: 244 seconds]
03:09
MillerBOSS has quit [Remote host closed the connection]
03:13
MillerBOSS has joined #ffmpeg-devel
03:34
jamrial has quit []
03:36
Martchus_ has joined #ffmpeg-devel
03:37
Martchus has quit [Ping timeout: 244 seconds]
03:43
Guest18 has joined #ffmpeg-devel
03:48
Guest18 has quit [Quit: Client closed]
04:07
mkver has joined #ffmpeg-devel
04:07
damian101_ has quit [Ping timeout: 264 seconds]
04:10
damian101_ has joined #ffmpeg-devel
04:31
witchymary has quit [Ping timeout: 248 seconds]
04:43
witchymary has joined #ffmpeg-devel
04:57
witchymary has quit [Ping timeout: 245 seconds]
05:10
witchymary has joined #ffmpeg-devel
05:34
Kei_N_ has joined #ffmpeg-devel
05:34
Kei_N has quit [Read error: Connection reset by peer]
05:58
GewoonLeon has joined #ffmpeg-devel
06:45
System_Error has quit [Remote host closed the connection]
06:53
System_Error has joined #ffmpeg-devel
07:13
j45_ has joined #ffmpeg-devel
07:13
j45 has quit [Ping timeout: 260 seconds]
07:13
j45_ is now known as j45
07:14
j45 has joined #ffmpeg-devel
07:14
j45 has quit [Changing host]
07:15
System_Error has quit [Remote host closed the connection]
07:21
System_Error has joined #ffmpeg-devel
07:29
damian101_ has quit [Ping timeout: 240 seconds]
07:34
NullSound has joined #ffmpeg-devel
07:35
NullSound has quit [Client Quit]
07:57
mkver has quit [Remote host closed the connection]
07:57
mkver has joined #ffmpeg-devel
08:38
rvalue- has joined #ffmpeg-devel
08:38
rvalue has quit [Ping timeout: 240 seconds]
08:43
rvalue- is now known as rvalue
10:16
j45_ has joined #ffmpeg-devel
10:17
j45 has quit [Ping timeout: 260 seconds]
10:17
j45_ is now known as j45
10:17
j45 has joined #ffmpeg-devel
10:17
j45 has quit [Changing host]
10:32
GewoonLeon has quit [Remote host closed the connection]
11:14
jamrial has joined #ffmpeg-devel
11:22
System_Error has quit [Ping timeout: 272 seconds]
11:32
System_Error has joined #ffmpeg-devel
11:43
Everything has joined #ffmpeg-devel
11:56
MetaNova has quit [Remote host closed the connection]
11:58
MetaNova has joined #ffmpeg-devel
12:03
Everything has quit [Quit: leaving]
12:06
<
ramiro >
Traneptora: thanks, I'll have a look
12:43
<
Lynne >
haasn: ping on a temperature setting in libplacebo
14:29
mkver has quit [Ping timeout: 245 seconds]
14:44
System_Error has quit [Remote host closed the connection]
14:51
System_Error has joined #ffmpeg-devel
15:17
Guest91 has joined #ffmpeg-devel
15:18
Guest91 has quit [Client Quit]
15:49
Everything has joined #ffmpeg-devel
16:03
minimal has joined #ffmpeg-devel
16:16
markh has quit [Remote host closed the connection]
16:23
markh has joined #ffmpeg-devel
16:27
<
haasn >
Lynne: can you remind me what API you would expect for this?
16:41
<
Lynne >
haasn: just a generic setting like colorspace or contrast
16:49
Everything has quit [Ping timeout: 264 seconds]
16:51
Everything has joined #ffmpeg-devel
16:52
damian101 has joined #ffmpeg-devel
16:54
sepro4 has joined #ffmpeg-devel
16:55
desmond-netint has joined #ffmpeg-devel
16:55
Teukka has quit [Read error: Connection reset by peer]
16:56
sepro has quit [Ping timeout: 256 seconds]
16:56
sepro4 is now known as sepro
16:58
Teukka has joined #ffmpeg-devel
16:58
Teukka has joined #ffmpeg-devel
16:58
Teukka has quit [Changing host]
17:06
damian101 has quit [Ping timeout: 245 seconds]
17:08
<
haasn >
I mean we already have a generic setting
17:08
<
haasn >
so do you literally just want the ability to set it in kelvin?
17:16
mkver has joined #ffmpeg-devel
17:17
ccawley2011 has joined #ffmpeg-devel
17:19
<
Lynne >
yeah, pretty much
17:19
<
Lynne >
the existing setting also cannot be set via the generic option string API
17:22
CounterPillow is now known as CountPillow
17:28
<
haasn >
can I recruit you to write the docs change?
17:29
<
haasn >
in theory it can
17:29
<
haasn >
:extra_opts=temperature=-1
17:29
<
haasn >
it's just a wonky scale
17:30
<
haasn >
in retrospect I regret normalizing it to -1..1
17:31
<
Lynne >
could you lower the minimum kelvin limit of 2500 or 3000 down to at least 1500k?
17:31
<
haasn >
can you find an approximation curve for getting the CIE x/y white point from the CCT for such low temperatures?
17:31
<
haasn >
the one we have is only accurate down to 2500K
17:35
<
Lynne >
the one in libplacebo looks to be accurate within 1000k to 10000k
17:36
<
Lynne >
though the switchover happens at 7k, whereas in what I found for the exact formula, it happens at 4k
17:38
<
Lynne >
and the author does say that its reasonably accurate within 1000 to 10000k
17:41
rvalue has quit [Read error: Connection reset by peer]
17:42
rvalue has joined #ffmpeg-devel
17:42
damian101 has joined #ffmpeg-devel
17:46
ccawley2011 has quit [Ping timeout: 252 seconds]
17:48
<
haasn >
is that the same curve as the one we use? cba dissecting those polynomials
17:49
rvalue has quit [Remote host closed the connection]
17:49
<
haasn >
or did you mean that's a different estimation you found
17:49
rvalue has joined #ffmpeg-devel
17:51
<
Lynne >
this is the code that ACES cites and uses
17:51
<
Lynne >
the kang2002 estimation says its accurate within 1667k to 25000k
17:52
<
Lynne >
the 2013 version goes all the way down to 1000k, but it pretty much abandons estimations and seems to just compute it altogether
17:54
<
Lynne >
I'm happy with 1667k min, its well within candelight
17:59
<
haasn >
says it's accurate within 1000 to 15k
18:00
<
haasn >
that's more than enough as far as I care
18:00
<
haasn >
I'll update our code tomorrow, it's getting late
18:00
<
Lynne >
cool, thanks
18:06
damian101 has quit [Ping timeout: 264 seconds]
18:20
damian101 has joined #ffmpeg-devel
18:48
Everything has quit [Ping timeout: 245 seconds]
18:50
rvalue has quit [Quit: bmV2ZXJnb25uYWdpdmV5b3V1cG5ldmVyZ29ubmFsZXR5b3Vkb3du]
18:50
Everything has joined #ffmpeg-devel
19:05
rvalue has joined #ffmpeg-devel
19:11
ccawley2011 has joined #ffmpeg-devel
19:27
ccawley2011_ has joined #ffmpeg-devel
19:31
ccawley2011 has quit [Ping timeout: 265 seconds]
19:52
damian101 has quit [Ping timeout: 240 seconds]
19:54
<
haasn >
Lynne: this formula doesn't return D65 for 6500
19:55
<
haasn >
it returns something completely different, CIE xy = (0.3134941075102412, 0.32366253911989207)
19:56
<
haasn >
D65 should be (0.31271, 0.32902)
19:56
<
haasn >
seems like D65 is not actually
_on_ the planckian locus
19:56
<
haasn >
that sure would explain why we don't get it back
19:56
<
haasn >
maybe our old formula was "tweaked" to return the actual D illuminant series
19:58
<
jamrial >
BtbN: i get a "Credentials are incorrect or have expired." message on an otherwise blank page when i try to open any c.f.o link
20:03
<
haasn >
Lynne: ah, seems like the confusion here is daylight locus vs blackbody (planckian) locus
20:03
ccawley2011_ has quit [Ping timeout: 244 seconds]
20:03
<
haasn >
they are parallel
20:04
<
haasn >
Lynne: based on our documentation, we want the daylight locus, not the planckian locus
20:04
<
haasn >
"Computes the CIE xy chromaticity coordinates of a CIE D-series illuminant with the given correlated color temperature."
20:10
<
mkver >
The server seems to have issues.
20:11
<
kierank >
damn, I wanted to tweet about your removing MMX
20:13
Sean_McG has joined #ffmpeg-devel
20:14
<
mkver >
It's not that important and I actually only started assembly last week, so it is very likely that my code is suboptimal and embarrassing.
20:14
<
mkver >
It would also attract attention to the fact that we are breaking ABI by not issuing emms in mmx functions.
20:15
<
kierank >
your work is excellent
20:15
<
kierank >
I think it's important cleanup to get rid of emms
20:15
<
kierank >
As it does cause problems on some systems
20:16
<
jamrial >
kierank: the fact we call it, or that we don't call it always where we should?
20:16
<
kierank >
the former is less of an issue I guess
20:17
<
BtbN >
jamrial: it seems the server has run out of RAM. I gave it a bit more cache to MariaDB the other day. Guess I misjudged.
20:18
fjlogger has quit [Remote host closed the connection]
20:18
<
mkver >
For me the problem is that we don't abide by the ABI, leading to strange issues like the one 57a29f2e7dd2374 fixed; and that emms_c gets called at places where it is unnecessary for many codecs.
20:19
fjlogger has joined #ffmpeg-devel
20:20
fjlogger has quit [Remote host closed the connection]
20:20
fjlogger has joined #ffmpeg-devel
20:27
fjlogger has quit [Ping timeout: 264 seconds]
20:28
fjlogger has joined #ffmpeg-devel
20:30
ccawley2011_ has joined #ffmpeg-devel
20:31
Everything has quit [Quit: leaving]
20:51
ccawley2011__ has joined #ffmpeg-devel
20:54
ccawley2011_ has quit [Ping timeout: 250 seconds]
21:02
<
haasn >
Lynne: I think I will just introduce a second function
21:02
<
haasn >
not to break backcompat
21:25
ccawley2011__ has quit [Ping timeout: 252 seconds]
21:43
LainIwakura has joined #ffmpeg-devel
21:45
desmond-netint has quit [Remote host closed the connection]
21:51
ccawley2011__ has joined #ffmpeg-devel
22:01
desmond-netint has joined #ffmpeg-devel
22:04
Kimapr_ has joined #ffmpeg-devel
22:04
Kimapr has quit [Remote host closed the connection]
22:08
<
kasper93 >
could anyone approve #20619, if it's ok? Those are small patches, don't want to keep them in limbo
22:10
LainIwakura has quit [Quit: Client closed]
22:22
sepro has quit [Ping timeout: 245 seconds]
22:29
Raz- has quit [Ping timeout: 256 seconds]
22:31
tufei has joined #ffmpeg-devel
22:32
LainIwakura has joined #ffmpeg-devel
22:36
Kimapr_ has quit [Remote host closed the connection]
22:36
Kimapr_ has joined #ffmpeg-devel
22:42
Raz- has joined #ffmpeg-devel
22:52
CountPillow has quit [Quit: Bye.]
22:52
CounterPillow has joined #ffmpeg-devel
22:52
ccawley2011__ has quit [Ping timeout: 248 seconds]
22:56
CounterPillow has quit [Client Quit]
23:00
CounterPillow has joined #ffmpeg-devel
23:03
ccawley2011__ has joined #ffmpeg-devel
23:09
ccawley2011__ has quit [Read error: Connection reset by peer]
23:44
leo60228 has joined #ffmpeg-devel
23:44
av500 has joined #ffmpeg-devel
23:45
kurufu_ has joined #ffmpeg-devel
23:46
chainik155 has joined #ffmpeg-devel
23:46
marcj has joined #ffmpeg-devel
23:47
adema has joined #ffmpeg-devel
23:47
Kwiboo- has joined #ffmpeg-devel
23:47
kurufu has quit [Read error: Connection reset by peer]
23:47
jkkm has quit [Read error: Connection reset by peer]
23:47
chainik15 has quit [Quit: Ping timeout (120 seconds)]
23:47
jkkm has joined #ffmpeg-devel
23:47
_av500_ has quit [Ping timeout: 250 seconds]
23:47
vriska has quit [Ping timeout: 250 seconds]
23:47
kurufu_ is now known as kurufu
23:47
chainik155 is now known as chainik15
23:49
Kwiboo has quit [Ping timeout: 250 seconds]
23:49
DodoGTA has quit [Ping timeout: 245 seconds]
23:52
rossy has quit [Ping timeout: 265 seconds]
23:52
rossy has joined #ffmpeg-devel
23:55
minimal has quit [Quit: Leaving]