Current Server Time: 20:51 (Central Europe)

#apertus IRC Channel Logs

2022/01/06

Timezone: UTC


06:11
Bertl_oO
off to bed now ... have a good one everyone!
06:11
Bertl_oO
changed nick to: Bertl_zZ
08:27
se6astian
good day
12:48
Bertl_zZ
changed nick to: Bertl
12:48
Bertl
morning folks!
16:40
vup
Bertl: can it ever happen that you need to combine a B frame with a A frame one higher?
16:40
vup
for the sample footage se6astian uploaded some time ago for anuejn this seems to work better tm
16:47
Bertl
yes, if for some reason the exposure gets synced to the B frame, then you need to do that
17:16
anuejn
ouch that is bad :(
17:19
Bertl
it should be easy to prevent this from userspace before starting a recording though
17:19
Bertl
i.e. stop live acquisition, wait for odd frame, resume live acquisition
17:20
Bertl
(probably three or four bash lines ;)
17:26
se6astian
sounds like a good addition to the startup script
17:31
Bertl
maybe
17:31
Bertl
off for now ... bbl
17:31
Bertl
changed nick to: Bertl_oO
18:31
anuejn
hm... I guess we dont really mean the same thing
18:31
anuejn
what we are experiencing cant be detected automatically in a trivial way?
18:32
anuejn
it can be distinguished by looking at the results though but thats tedious?
20:21
se6astian
have you observed that actually happening so far?
20:27
anuejn
i guess?
20:27
anuejn
your sample footage looks better if we decode it "wrong"
20:31
Bertl_oO
anuejn: we are likely talking about the same issue
20:31
Bertl_oO
the problem here is the sequencer
20:33
Bertl_oO
the sequencer doesn't care wheter we have an A or B frame, it simply syncs to a frame change
20:33
Bertl_oO
if that happens to be a B->A transition, then everything is fine and you will get A and B frame from the same exposure
20:37
se6astian
interesting!
20:38
se6astian
so how could this be tested/verified?
20:39
se6astian
the 4 line bash script you suggested would run on the beta?
21:01
Bertl_oO
yes, and it would read out the frame counter register
21:02
Bertl_oO
wait for the right moment and then enable the acquisition
22:11
se6astian
that is fil_reg 15 ?
22:16
se6astian
and accessing the frame counter register can be done like in https://github.com/apertus-open-source-cinema/axiom-firmware/blob/main/software/scripts/axiom_raw_mark.sh I assume?
22:51
Bertl_oO
capture is turned on/off via the button overrides in capture reg 15, yes
22:51
Bertl_oO
the frame counter is in the hdmi generator register block
22:53
Bertl_oO
axiom_raw_mark does not access the frame counter, it just configures static overlay entries