Current Server Time: 00:28 (Central Europe)

#apertus IRC Channel Logs

2020/12/28

Timezone: UTC


23:01
BAndiT1983
changed nick to: BAndiT1983|away
00:11
mumptai
left the channel
02:16
eppisai
joined the channel
05:52
Bertl_oO
off to bed now ... have a good one everyone!
05:52
Bertl_oO
changed nick to: Bertl_zZ
07:02
eppisai_
joined the channel
07:04
eppisai
left the channel
07:04
eppisai_
changed nick to: eppisai
07:14
Spirit532
left the channel
07:15
Spirit532
joined the channel
07:26
eppisai
left the channel
07:27
eppisai
joined the channel
07:40
se6ast1an
good day
07:46
AndroUser
joined the channel
07:46
AndroUser
left the channel
07:47
AndroUser
joined the channel
07:48
AndroUser
changed nick to: abcd
07:52
eppisai
left the channel
07:52
eppisai
joined the channel
08:00
eppisai
left the channel
08:00
eppisai
joined the channel
08:17
eppisai
left the channel
08:21
se6ast1an
vup was any progress made with https://github.com/apertus-open-source-cinema/axiom-firmware/issues/179 ?
08:33
mumptai
joined the channel
08:38
BAndiT1983|away
changed nick to: BAndiT1983
10:54
abcd
left the channel
10:57
abcd
joined the channel
11:38
mumptai
left the channel
11:39
mumptai
joined the channel
12:06
xthenode
joined the channel
12:07
xthenode
happy monday!
12:08
xthenode
i was able to build a GL app on the mac so i wonder why SDL2 crashes...
12:13
xthenode
the goal is to create a skinable client for web, mobile, desktop and command line
12:14
xthenode
if one doesn't exist...
12:16
xthenode
left the channel
12:31
BAndiT1983
xthenode: what is the purpose of the client? maybe i've missed a part of logs where it's described
13:49
Bertl_zZ
changed nick to: Bertl
13:49
Bertl
morning folks!
13:59
mumptai
left the channel
14:01
Shohei
joined the channel
14:01
se6ast1an
good day Bertl
14:05
Bertl
thanks! and a good one to you as well!
14:11
Shohei
Hello!I'm participating here for the first time. Nice to meet you : )
14:13
Bertl
Hello Shohei! nice to see you on IRC!
14:14
Bertl
looking forward to hear more about the FOSS/OH movement in Japan!
14:15
Shohei
Thank you for replying for the mails, Bertl, se6bast1an : )
14:15
Shohei
se6ast1an : )
14:16
Shohei
By all means!It may be slow to write, but I'll do my best.
14:17
Bertl
no worries, IRC is a very patient medium
14:17
Bertl
many discussions here span over days, because folks live and work in different timezones
14:17
Shohei
I see!
14:18
Bertl
which also means, if you have a question, just drop it here, and idle around (or check the public IRC logs)
14:18
Bertl
somebody will answer sooner or later but it might not be immediately
14:19
Shohei
I see! Thank you!
14:21
Shohei
Is everyone chatting on #apertus only, or are there more specialized channels?
14:22
BAndiT1983
changed nick to: BAndiT1983|away
14:22
Bertl
while there are some other 'related' channels, the main action is here
14:22
Bertl
for example the meeting today will happen on this channel
14:23
Bertl
there are also private messages which work just fine if only two people want to coordinate or so
14:23
Shohei
I got it, thank you. I will check the public IRC logs to catch up with you : )
14:24
Bertl
you're welcome!
14:25
Shohei
It's a big help : )
14:51
bluez
joined the channel
14:55
BAndiT1983|away
changed nick to: BAndiT1983
14:58
se6ast1an
hi Shohei! great to see you here
14:59
se6ast1an
you can read the public logs Bertl mentioned here: http://irc.apertus.org/
15:15
Shohei
Great to see you here, too! se6ast1an : ) Thank you very much for the link!
15:22
BAndiT1983
changed nick to: BAndiT1983|away
15:28
mumptai
joined the channel
16:03
Bertl
no meeting today?
16:03
se6ast1an
why not? :)
16:03
se6ast1an
MEETING TIME!
16:03
se6ast1an
who is here?
16:03
Bertl
is here ...
16:03
Shohei
is here
16:04
metal_dent[m]
present
16:04
bluez
* bluez is here
16:05
se6ast1an
great
16:05
se6ast1an
metal_dent[m]: do you have something to share and start the reports?
16:05
metal_dent[m]
Yes!
16:06
se6ast1an
please do
16:06
metal_dent[m]
Last week I worked on multiple tasks. Firstly, I helped Aman with the checkbox code, which is merged now.
16:07
metal_dent[m]
Then I stumbled upon my old PR ':D where I tried to reduce the redundant code and now waiting for BAndiT1983|away 's approval
16:08
se6ast1an
amans PR: https://github.com/apertus-open-source-cinema/AXIOM-Remote/pull/24
16:08
se6ast1an
video: https://user-images.githubusercontent.com/54789531/103020612-bf1d0000-456e-11eb-9566-c41b178acc67.mp4
16:09
se6ast1an
metal_dent[m]: did you push a revised PR for review already?
16:09
metal_dent[m]
We have also created a new task (T1215) to add a desaturation filter on the remote display, I have added the shader code which works fine on the camera preview image (Bertl's image)
16:09
se6ast1an
great
16:10
metal_dent[m]
<se6ast1an "metal_dent: did you push a revis"> Yes, you can see the new commit there (sorry for the other mess as there was a lot of conflicts ':D )
16:10
se6ast1an
link please :)
16:11
metal_dent[m]
https://github.com/apertus-open-source-cinema/AXIOM-Remote/pull/17
16:11
metal_dent[m]
T1215: https://lab.apertus.org/T1215
16:11
se6ast1an
thanks
16:12
se6ast1an
butt the last commit to that PRis from May 23rd
16:12
metal_dent[m]
That's it from my side!
16:12
Bertl
nice
16:13
metal_dent[m]
This is the last commit:
16:13
metal_dent[m]
https://github.com/apertus-open-source-cinema/AXIOM-Remote/pull/17/commits/d0b109667d0c379507207d04c4936bff080cad06
16:15
se6ast1an
very good, thanks!
16:15
se6ast1an
bluez: please go ahead next
16:15
bluez
sure.. thanks!
16:16
xthenode
joined the channel
16:16
bluez
so last week Bertl helped me get up to speed with vhdl and the tcl side of vivado
16:17
bluez
this will be helpful with the final testing setup
16:18
xthenode
re: xthenode: what is the purpose of the client? a generalized client to be used a control which can be mapped to a particular platform...
16:18
bluez
which we will hopefully start this week
16:18
bluez
had to lot to learn on the vhdl side ':D
16:18
xthenode
in this case axiom...
16:19
Bertl
xthenode: we are currently having our weekly meeting
16:19
bluez
that's it from me!
16:19
Bertl
xthenode: please wait till it is your turn to report :)
16:20
Bertl
bluez: thanks!
16:20
se6ast1an
great, bluez, many thanks!
16:20
se6ast1an
Shohei: would you also like to introduce yourself quickly?
16:21
se6ast1an
maybe some people followed the email conversation with you already but live is always better :)
16:21
Shohei
Thank you, se6ast1an :)
16:23
Shohei
Hello, everyone. I'm mechanical engineer in Japan, and also running small startup focusing on open hardware.I'm very interested in this project, and personally joined : )
16:24
xthenode
left the channel
16:24
se6ast1an
great, thanks Shohei!
16:24
Bertl
great to have you here!
16:24
se6ast1an
I was already thinking of things we could work on together in the enclosure/CAD area
16:25
se6ast1an
lets discuss this soon when you have a better overview of everything
16:25
se6ast1an
vup you also have some news for us!
16:25
se6ast1an
please go ahead
16:26
vup
yeah, well not too much got done over the last week
16:26
vup
anuejn and I worked a bit more on the wavelet compression
16:26
Shohei
I will be happy to join you!
16:27
vup
we have managed to write a python compressor and a python decompressor that actually work in the sense, that the python decompressor can read the output of the python compressor and produce a somewhat sensible result
16:27
vup
there are still a few problems even there (the input and output image do not match exactly for some reason, even when doing lossless compression)
16:28
vup
and furthermore, the python decompressor currently produces a completely broken image, when feeding it the input from the gateware compressor
16:29
vup
currently we think this is due to wrong rounding and too few bits for the intermediate values in the gateware compressor
16:29
vup
but we have to experiment there a bit more to get it working
16:29
vup
nonetheless, we have figured out a buffering / processing scheme for the gateware that does not use too many bram's / should be quite efficient with them, so thats a big progress
16:30
se6ast1an
very cool!
16:30
se6ast1an
any news on the usb3 front as well?
16:31
vup
well no real news, but we looked at the ft602 again and noticed a very nice feature where it can transmit setting various uvc settings back to the fpga via i2c, which would make for a great sidechannel
16:32
vup
also I looked at the firmware of the ft601 again and we will hopefully dump the firmware of the ft602 in the next couple of days and maybe we can figure out how to add such a sidechannel for the ft601 aswell
16:32
vup
(the are probably the same chip with just different firmware anyways)
16:32
se6ast1an
ah, great!
16:33
vup
oh also one thing about the wavelet compression, currently it does not need any RAM bandwidth
16:33
vup
so the compression can just be done on the output path (maybe to the usb3 module) and things like the HDMI preview still "just work"
16:34
se6ast1an
how do you buffer the image currently?
16:34
se6ast1an
or how big buffers are required by the different compression algorithmn steps?
16:34
vup
well for the wavelet transformation one only ever needs 6 rows of the image at once (atleast for a single step)
16:35
vup
so those just get put into bram's in the fpga
16:35
vup
so currently you needs about 12 rows of the image
16:35
Bertl
for the FTDI it would probably be most interesting to get bandwidth allocation working (on the USB side) not sure how complicated that would be though
16:36
vup
se6ast1an: so about 72KB if I am not mistaken
16:37
vup
Bertl: yeah, the usb2 part probably supports isochronous transfers, as the "officially" programmable ftdi mcu's support that
16:37
vup
for usb3 we will have to see
16:37
vup
but yeah that would also be very nice, I agree
16:37
se6ast1an
sounds good, rather performance friendly
16:38
Bertl
alternatively we can consider switching to the cypress EZ-USB chips
16:39
se6ast1an
there is a fx3 dev kit: https://www.cypress.com/documentation/development-kitsboards/cyusb3kit-003-ez-usb-fx3-superspeed-explorer-kit that could easily be put on a plugin module carrier if that makes sense?
16:39
vup
se6ast1an: yep, I was quite surprised it can actually be done this way, we probably will need a bit more buffers in the future, but at maximum a factor of 1.5 more I think (we currently only do grayscale, we probably should split the colors for the bayer images)
16:40
vup
yeah, the fx3 is certainly a interesting option aswell
16:40
vup
or even just a ecp5 (but thats of course a lot more work)
16:40
Bertl
se6ast1an: well, a rather large plugin module :)
16:40
se6ast1an
sure, but size is of no concern for development no?
16:41
se6ast1an
or would you create a custom plugin module with the chip right away?
16:41
vup
a 0.8mm pitch bga should be doable in 4 layers probably, and with the new 6 layer oshpark runs it should certainly be possible
16:42
Bertl
I think it would make more sense to design a plugin featuring the FX3
16:42
se6ast1an
great, fine for me :)
16:42
vup
well if there are no more questions, I am done for today
16:43
se6ast1an
many thanks vup!
16:43
se6ast1an
quick updates from me:
16:43
se6ast1an
not that much happened as I was away over christmas and returned yesterday
16:43
se6ast1an
a package with hardware for bluez and metal_dent[m] was posted just before christmas though
16:43
se6ast1an
should arrive in a week or 2
16:44
se6ast1an
in production news the oshpark panels have been sent to the fab
16:44
se6ast1an
and nextpcb has started to discuss the stackup with us
16:44
se6ast1an
so fingers crossed
16:45
se6ast1an
the proposed prepreg thickness would be a perfect impedance match but some things need to be confirmed/corrected first
16:45
se6ast1an
I started looking into dissipative ESD foam suppliers again today
16:45
se6ast1an
contacted 2 companies
16:46
se6ast1an
and manfred sent me pictures of a new CNC mill he wants to build
16:46
se6ast1an
which would be suitable for milling such foam most likely
16:46
se6ast1an
the tormach can of course also do it
16:46
se6ast1an
but overkill for such a soft material
16:47
se6ast1an
as I am on holidays I should have a good amount of time to work on things over the next week
16:47
se6ast1an
so if you have tasks for collaboration let me know
16:47
se6ast1an
done from my side
16:48
se6ast1an
Bertl: please be the closer!
16:50
Bertl
well, not too much happened here this week
16:51
Bertl
mostly busy with x-mas and similar
16:52
Bertl
I did update the TE0714 sata plugin
16:52
Bertl
because there was some interest to build that and I made sure that the design is current and the bom can easily be dreived
16:53
Bertl
*derived
16:53
se6ast1an
so no changes to the design itself?
16:53
Bertl
I also updated Vivado to the latest version and created a test bitstream for se6ast1an to play with
16:53
se6ast1an
ah yes, forgot to report about thta
16:54
Bertl
that's basically it
16:54
se6ast1an
what I want to test is the feature of the CMV12000 where it can be exposed while data is being read out
16:54
se6ast1an
this should allow maximizing exposure time
16:55
se6ast1an
but it comes with some drawbacks and artefacts affecting pixels read out values
16:55
se6ast1an
which I want to document properly
16:55
se6ast1an
the custom gateware bertl built should enable doing that
16:56
se6ast1an
but needs further testing/analysis so I returned the beta hardware to bertl so he can work on it
16:57
se6ast1an
Bertl: the sata plugin module design did not contain any design changes I assume?
16:57
se6ast1an
*update
16:57
Bertl
no
16:58
se6ast1an
right
16:59
se6ast1an
ok then, anyone else who wants to report/share/comment on anything specific?
17:01
se6ast1an
MEETING CONCLUDED!
17:01
se6ast1an
many thanks everyone participating!
17:01
BAndiT1983|away
changed nick to: BAndiT1983
17:10
Bertl
off for now ... bbl
17:11
Bertl
changed nick to: Bertl_oO
17:20
Shohei
left the channel
18:16
bluez
left the channel
20:16
intrac
left the channel
20:17
intrac
joined the channel
20:44
lexano
left the channel
20:46
lexano
joined the channel
20:49
lexano
left the channel
20:50
lexano
joined the channel