Current Server Time: 01:57 (Central Europe)

#apertus IRC Channel Logs

2018/07/02

Timezone: UTC


02:15
rton
left the channel
02:43
jucar
joined the channel
03:08
jucar
left the channel
06:00
Lordvan_
changed nick to: LordVan
08:36
se6astian|away
changed nick to: se6astian
08:58
se6astian
changed nick to: se6astian|away
09:14
sebix
joined the channel
09:14
sebix
left the channel
09:14
sebix
joined the channel
09:48
se6astian|away
changed nick to: se6astian
10:18
BogdanXOR
joined the channel
10:33
rton
joined the channel
10:47
RexOrCine|away
changed nick to: RexOrCine
11:23
BogdanXOR
left the channel
11:37
BogdanXOR
joined the channel
12:06
TofuLynx
joined the channel
12:07
TofuLynx
Good Morning!
12:08
TofuLynx
se6astian: Thanks for the photos! However the photos are glitchy, even in darktable. I converted using raw2dng. Did I miss any step?
12:17
TofuLynx
Nevermind, Fixed :)
12:17
TofuLynx
I had to use --swap-line
12:48
BogdanXOR
darktable vs rawtherapee - what are your thoughts on this? I have just some (I'd say limited) rawtherapee experience, so I'd like to take the opportunity to get more insights into these (or any other open source & free alternatives) as to decide on a path to follow
12:48
BogdanXOR
so what do you prefer / recommend, which is the better bet?
12:49
BogdanXOR
you = everyone here, feel free to share input
12:58
se6astian
both are good :)
12:58
se6astian
TofuLynx: great to hear
13:31
TofuLynx
BogdanXOR: never used rawtherapee, but darktable is good
14:10
comradekingu
Used both, darktable for me
14:11
comradekingu
Depends how many controls you want
14:17
BogdanXOR
could you elaborate on that (amount of controls)?
14:19
comradekingu
from what i remember rawtherapee didnt offer as much adjustment
14:22
BogdanXOR
I see. I should give darktable a spin then
14:22
BogdanXOR
any insights on the speed of rendering the image changes, between the two?
14:31
se6astian
I think rawtherapee is offering just as much control/development options by now
15:09
se6astian
added a header margin-top on wiki
15:09
se6astian
hopefully this will help structure pages a bit better
15:10
se6astian
please check and report back if you think its too much/too little
15:21
TofuLynx
https://wiki.apertus.org/index.php/AXIOM_Beta/Camera_Structure
15:21
TofuLynx
is this layout expected?
15:22
TofuLynx
the images seem a bit off
15:24
Bertl_zZ
changed nick to: Bertl
15:24
Bertl
morning folks!
15:24
TofuLynx
Hello Bertl :)
15:35
LordVan
left the channel
15:39
TofuLynx
Unrelated to apertus, anyone here is expert in network systems, namely Mobile Networks?
15:49
ansari
joined the channel
15:49
ansari
left the channel
15:55
BogdanXOR
left the channel
15:56
BogdanXOR
joined the channel
15:56
sebix
left the channel
15:58
BAndiT1983|away
changed nick to: BAndiT1983
16:32
ymc98_2
joined the channel
16:35
ymc98_1
left the channel
16:45
TofuLynx
BAndiT1983: Are you there?
16:47
BAndiT1983
hi TofuLynx
16:47
TofuLynx
Hey :)
16:47
TofuLynx
Remember the thing I said with the nested for loops?
16:48
BAndiT1983
yes
16:48
TofuLynx
The acceleration I made so far that is on github improved GEDI from 67ms to 9ms. Now, I used nested for loops to remove that if statements, and it is now just 5ms :)
16:49
BAndiT1983
very good
16:49
BAndiT1983
which optimizatin flag do you use?
16:49
BAndiT1983
*optimization
16:49
TofuLynx
none for now, just openMP
16:50
TofuLynx
when I use O3 or O2 it doesnt affect the speed at all, which is strange
16:50
BAndiT1983
in release mode?
16:50
TofuLynx
yeah
16:50
BAndiT1983
one of the next steps is to analyze whole processing pipeline times
16:51
TofuLynx
Okay
16:51
BAndiT1983
file loading -> conversion -> debayering etc.
16:51
TofuLynx
File loading and saving is memory bound, right?
16:51
BAndiT1983
to find more culprits, it's a preparation for dng folder loading
16:51
TofuLynx
so it cant really be accelerated, correct?
16:51
BAndiT1983
more disk bound, in my opinion
16:52
BAndiT1983
yep, meant by file loading the steps which are taken after it'S in the memory
16:52
TofuLynx
I was referring to disk memory, sorry xD
16:52
BAndiT1983
maybe memory mapping would be useful for it
16:52
TofuLynx
hmm such as extractor?
16:53
BAndiT1983
yes
16:53
TofuLynx
Understood
16:53
BAndiT1983
there are several loops here and there, but i don't want to use SIMD at the moment, for byte swapping and processing
16:54
BAndiT1983
which is possible option for the future, as processors have special instructions for such cases like ours
16:57
BAndiT1983
quick reboot
16:57
TofuLynx
But is it really neccessary such speed by SIMD at loading?
16:58
BAndiT1983
loading?
16:58
BAndiT1983
changed nick to: BAndiT1983|away
16:58
TofuLynx
Isn't the memory mapping stuff related to the loading pipeline?
16:59
BAndiT1983|away
changed nick to: BAndiT1983
16:59
BogdanXOR
left the channel
16:59
BAndiT1983
memory mapping is used for bigger files
17:00
BAndiT1983
simd was mentioned as possible improvement for extractor
17:00
TofuLynx
hmm okay
17:00
TofuLynx
I will use the improvements made in GEDI for bilinear and for shoodak
17:01
TofuLynx
then I will research for further improvements in the GEDI code and SHOODAK code
17:01
TofuLynx
and then I will document the new algorithms which will mark the end of what I proposed in my proposal. Then I can start working on that task
17:02
BAndiT1983
excellent
17:02
TofuLynx
Later I would like to do the video playing
17:02
BAndiT1983
alright, i will join this task maybe soon, have first to sort out my work etc.
17:04
TofuLynx
Great :)
17:06
BogdanXOR
joined the channel
17:06
TofuLynx
I will be right back
17:07
BAndiT1983
ok
17:07
BAndiT1983
changed nick to: BAndiT1983|away
17:09
BAndiT1983|away
changed nick to: BAndiT1983
17:16
BogdanXOR
left the channel
17:17
BAndiT1983
changed nick to: BAndiT1983|away
17:18
BAndiT1983|away
changed nick to: BAndiT1983
17:48
se6astian
meeting in 12 minutes
17:53
TofuLynx
Back
17:57
ArunM
helloooo
17:58
Bertl
hello ArunM!
17:59
TofuLynx
Hello ArunM :)
18:00
ArunM
Bertl are you available after the meeting?
18:00
Bertl
sure
18:00
se6astian
good evening everyone
18:00
se6astian
again please PM me now to be put in the sharing/reporting order
18:01
se6astian
nmdis199-: contacted me beforehand to be excused from todays meeting
18:01
se6astian
and first up is ArunM, please go ahead
18:02
ArunM
Hello everyone
18:02
ArunM
Currently programming the Generator part the design
18:02
ArunM
Here is the architecture
18:02
ArunM
https://docs.google.com/document/d/1Pv46Nvl1mE6RruxbesQJu-GaRoef1IV2VUDQu5oFSJo/edit?usp=sharing
18:02
ArunM
The architecture is yet to be discussed with BertL, already started programming because I wanted to check the possible patterns that can be generated
18:02
ArunM
Now having a better idea of the design and no problem if it have to be changed completely
18:03
ArunM
During simplification of old code i got stuck in changing one design for another and then changing next module.
18:03
ArunM
Finally relaxed with it and hopefully the documentation of old code will be submitted before next week evaluation.
18:03
ArunM
along with the generator module
18:03
ArunM
Here is the list of old patterns for the camera https://docs.google.com/document/d/1Ph_kRRllu43dGQx5Iw6vgpNmFYtquQEu-2h4mNFIS3c/edit?usp=sharing
18:04
ArunM
New patterns will be updated and discussed by tomorrow only.
18:04
ArunM
done reporting Captain :)
18:05
TofuLynx
How did you make the diagrams?
18:05
ArunM
draw.io
18:05
TofuLynx
oh nice :)
18:06
ArunM
also i'll be writting a detailed description b4 discussing
18:06
ArunM
the architecture
18:06
BAndiT1983
offline alternative to draw.io is yed
18:06
Bertl
sounds good!
18:06
se6astian
many thanks!
18:06
se6astian
ArunM: all done?
18:07
ArunM
yes
18:08
se6astian
right, TofuLynx you are next in line
18:08
TofuLynx
Okkey
18:08
TofuLynx
So my task for the last week was to accelerate Bilinear, GEDI and SHOODAK demosaicers
18:09
TofuLynx
I used openMP and my first approach
18:09
TofuLynx
which was giving each color channel to each thread
18:09
TofuLynx
made a significant improvement. 67ms to 27ms in case of GEDI debayer, on my PC
18:10
TofuLynx
however, BAndiT1983 quickly pointed that I could distribute the threads inside each loop
18:10
TofuLynx
I took this approach and it improved even more the speed. 67ms to 9ms in GEDI
18:11
TofuLynx
Yesterday and today I noticed that there was a condition in every loop that I thought it was slowing things down
18:11
TofuLynx
so I used nested for loops, which openMP also accelerates nicely
18:11
TofuLynx
At the end, it improved from 67ms to 5ms, which is great.
18:12
TofuLynx
In fps, it is equivalent to 15fps to 200fps
18:12
se6astian
very nice!
18:12
TofuLynx
Now, I will do a last look at the code
18:13
TofuLynx
and see where further improvements can be made
18:13
TofuLynx
Also, those benchmarks numbers can be "misleading"
18:13
TofuLynx
because, as openMP distributes the workload for the CPU threads
18:14
TofuLynx
my CPU has a fairly high number of threads, which is 12
18:14
TofuLynx
a more common computer only has four threads
18:14
TofuLynx
Nonetheless, it's a significant improvement
18:14
TofuLynx
That's all, se6astian :)
18:14
se6astian
great, thanks
18:15
se6astian
can you already estimate a 4096x3072 image processing time?
18:15
TofuLynx
the benchmarks were done with a 4096x3072 image
18:15
se6astian
ah, excellent
18:16
TofuLynx
so it's basically 200fps for a more than 4K video
18:16
TofuLynx
on the CPU, if we left the workload to the GPU, it would be even faster
18:16
TofuLynx
But that is not supported, for now
18:16
alexML
that sounds a lot faster than current hdmi4k / raw2dng stuff :D
18:16
TofuLynx
how fast is that?
18:17
alexML
you need to leave the computer overnight for a tiny clip :P
18:17
TofuLynx
wow xD
18:17
BAndiT1983
don't forget, that it'S just debayering, so numbers will shrink
18:18
TofuLynx
yeah, we have to have a fast memory pipeline
18:20
se6astian
anyone else who wants to report
18:20
se6astian
it would be your turn now
18:20
Bertl
ymc98_2: ping?
18:21
Bertl
rahul_: ping?
18:21
se6astian
I will use the time to report a bit myself: the next team talk video and article are almost ready for release
18:21
TofuLynx
supragya_: is this for logging?
18:21
BogdanXOR
joined the channel
18:21
se6astian
release should be in the next couple of days
18:21
TofuLynx
what are the topics of the team talk video?
18:22
se6astian
the move to the new office
18:22
se6astian
and you :)
18:22
se6astian
the gsoc students and projects
18:22
TofuLynx
Great! :)
18:22
se6astian
plus some other minor updates/news
18:23
se6astian
twitter posts about your recent progress:
18:23
se6astian
https://twitter.com/ApertusOSCinema/status/1011555886609944576
18:23
se6astian
910 impressions
18:23
se6astian
https://twitter.com/ApertusOSCinema/status/1013044232360210432
18:23
TofuLynx
what are impressions?
18:23
se6astian
4,729 impressions
18:24
se6astian
it means that it appeared on someones screen
18:24
TofuLynx
Oh cool!
18:24
BAndiT1983
and a comment for magiclantern forum :D
18:25
TofuLynx
the thing with canon 80d? xD
18:26
RexOrCine
Present
18:26
se6astian
ok that should conclude this weeks meeting then
18:26
se6astian
many thanks to all participants
18:26
se6astian
and rex in particular! :D
18:28
se6astian
TofuLynx: what do you think of the task of creating some screencaptures of OC in general and what you added so far?
18:34
TofuLynx
I can handle that :)
18:34
TofuLynx
For now, there's just one difference
18:34
TofuLynx
the ProcessingTest UI
18:34
TofuLynx
It was revamped by BAndiT1983
18:39
BogdanXOR
left the channel
18:59
se6astian
changed nick to: se6astian|away
19:00
RexOrCine
changed nick to: RexOrCine|away
19:16
TofuLynx
left the channel
19:20
BogdanXOR
joined the channel
19:55
se6astian|away
changed nick to: se6astian
20:09
BogdanXOR
left the channel
20:10
BogdanXOR
joined the channel
20:40
lexano
left the channel
20:41
se6astian
changed nick to: se6astian|away
20:43
BogdanXOR
left the channel
20:48
se6astian|away
changed nick to: se6astian
20:54
BogdanXOR
joined the channel
20:54
lexano
joined the channel
21:07
BogdanXOR
left the channel
21:41
BAndiT1983
changed nick to: BAndiT1983|away
21:41
BAndiT1983|away
changed nick to: BAndiT1983
21:56
se6astian
off to bed
21:56
se6astian
good night
21:56
se6astian
changed nick to: se6astian|away
22:12
BAndiT1983
changed nick to: BAndiT1983|away
23:02
BogdanXOR
joined the channel
00:15
BogdanXOR
left the channel
00:25
BogdanXOR
joined the channel
00:29
BogdanXOR
left the channel