Current Server Time: 15:19 (Central Europe)

#apertus IRC Channel Logs

2017/03/16

Timezone: UTC


00:00
Bertl
check out the C tools which access registers
00:00
BAndiT1983
also it's called everytime and maps over and over, which is a little disturbing
00:00
BAndiT1983
which tools do you mean?
00:01
Bertl
like the cmv_snap3
00:01
BAndiT1983
se6astian showed me also the code where registers are read for DNG export
00:01
Bertl
wb Alvis!
00:03
Alvis
left the channel
00:03
Alvis
joined the channel
00:05
BAndiT1983
this methods should be centralized i think, as there are many helpers in there
00:07
BAndiT1983
it would reduce the code and would simplify maintenance in the future, but there are many more things which can be straighten
00:08
Alvis
left the channel
00:08
BAndiT1983
would it be sufficient to do memory mapping at dameon start?
00:09
Bertl
probably, but note that you might want to limit the mappings to ranges which are actually used
00:10
BAndiT1983
how does the repeating call of mmap affects performance?
00:10
Bertl
I don't think it has any effect on performance at all
00:10
Bertl
it will create more TLB entries though
00:11
BAndiT1983
according to the cmv script there are 4 ranges which are used for spi registers
00:12
Bertl
really?
00:12
BAndiT1983
i would like to reduce calls as far as possible, so the things are managed in a central place, so lower overhead and higher performance
00:12
BAndiT1983
https://github.com/apertus-open-source-cinema/beta-software/blob/master/beta-scripts/cmv.func
00:13
Bertl
only the cmv_reg is actually SPI mapped to registers
00:14
BAndiT1983
are others not used at the moment?
00:14
Bertl
they are used, but there is no SPI involved
00:15
BAndiT1983
alright, but as they are mapped it doesn't matter on the linux end
00:15
Bertl
correct, check out ug585 for a zynq memory map
00:18
Bertl
(page 112)
00:19
BAndiT1983
already there ;) , got there and in the next second you message appeared
00:19
BAndiT1983
*your
00:19
Bertl
the two PL ranges are where we can put registers
00:20
Bertl
(i.e. virtual registers inside the FPGA)
00:20
BAndiT1983
what is mapped besides CMV registers?
00:21
Bertl
https://wiki.apertus.org/index.php/CMV12000_Register_Blocks
00:22
BAndiT1983
you've shown me that already ;) but i was wondering if there are also other things mapped besides the sensor
00:22
Bertl
in the first two tables, the sensor is only one line :)
00:23
BAndiT1983
ah, so the caption of the page is misleading a bit
00:23
Bertl
yes, very much
00:23
arpu
left the channel
00:24
Bertl
https://wiki.apertus.org/index.php/CMV12000_to_HDMI_registers
00:24
Bertl
this one too, no idea why folks keep calling it CMV12k
00:24
BAndiT1983
hm, i think i can really encapsulate some things and there would be some kind of API
00:24
BAndiT1983
wiki has to be reviewed again, but now against technical specs
00:25
Bertl
also note that this map is from the AXIOM Alpha and has certainly changed in the Beta
00:25
Bertl
some blocks were added, like for the pong stuff
00:26
Bertl
others were slightly rearranged (bits and their meaning changed)
00:26
BAndiT1983
as i don't have the hardware it doesn't really matter, but it's a good material for my reflection on how to proceed
00:29
Alvis
joined the channel
00:30
RexO
Definitely keep on top of Wiki entries as and when.
00:33
BAndiT1983
i'm missing a little bit breadcrumbs in apertus wiki
00:33
BAndiT1983
or to be precise, categories
00:34
BAndiT1983
like in normal wiki at the bottom, so one can look up all the things in the category in one go
00:34
BAndiT1983
https://en.wikipedia.org/wiki/Bayer_filter
00:34
BAndiT1983
https://en.wikipedia.org/wiki/Category:Color_filter_array
00:35
Bertl
did you try adding a [[Category:xxxx]] link?
00:35
BAndiT1983
i would never find registers without the link from Bertl
00:35
Bertl
that's how wikipedia does it IIRC
00:36
BAndiT1983
apertus wiki doesn't seem to be categorized
00:37
RexOrCine
There was a version two main page list that came up at one point, but it never got implemented.
00:37
BAndiT1983
this is appended in the normal one -> [[Category:Digital photography]]
00:37
BAndiT1983
[[Category:Image sensors]]
00:37
BAndiT1983
[[Category:Color filter array]]
00:37
arpu
joined the channel
00:38
RexOrCine
https://wiki.apertus.org/index.php?title=NewMain_Page
00:38
RexOrCine
Came up here - https://lab.apertus.org/T122
00:38
BAndiT1983
oha, much better
00:39
Bertl
so you don't like the colorful blocks on the current main page?
00:39
RexOrCine
That would mean abandoning the images on the main page, which I'd probably disapprove of, but it would be good to have an option available for people to reconfigure things
00:40
BAndiT1983
you don't have to abandon images, but we should consider to add categories
00:41
RexOrCine
I agree.... I even thought that new main page might be good at the bottom of the main page, but there were warning signs attached, and I didn't know what to make of that ie. "Don't edit this or it'll break"
00:41
BAndiT1983
seems to work, i get Categories: Beta Hardware | Registers at the bottom
00:42
BAndiT1983
we need just meaningful names for that stuff, then it's much easier to move in the same area to get more info
00:42
Bertl
so what's the problem then? :)
00:43
BAndiT1983
i don't know where the parts belong to, you said it's partially from alpha
00:44
Bertl
in general, you can assume that everything which doesn't explicitely say 'Beta' somewhere is ffrom the Alpha
00:44
Bertl
*from*
00:44
BAndiT1983
https://wiki.apertus.org/index.php/CMV12000_Register_Blocks
00:44
BAndiT1983
there should be 2 categories at the bottom as test
00:44
BAndiT1983
don't know why the category pages are editable, they usually shoudn't be
00:45
Bertl
you can lock pages down, probably that's what wikipedia does
00:48
BAndiT1983
the page has to be created once, just entered summary and and content as "Registers"
00:48
BAndiT1983
now they should stay blue
00:51
BAndiT1983
so, bed time for me, see you later
00:51
BAndiT1983
changed nick to: BAndiT1983|away
00:52
Bertl
nn
00:58
illwieckz
left the channel
00:59
illwieckz
joined the channel
00:59
illwieckz
left the channel
00:59
illwieckz
joined the channel
01:19
jucar
joined the channel
01:19
intracube
changed nick to: intracube_afk
01:25
illwieckz
left the channel
01:25
illwieckz
joined the channel
01:25
illwieckz
left the channel
01:25
illwieckz
joined the channel
03:17
Rex0r
left the channel
03:48
Alvis
left the channel
04:05
Bertl
off to bed now ... have a good one everyone!
04:06
Bertl
changed nick to: Bertl_zZ
04:11
Alvis
joined the channel
05:01
jucar
left the channel
05:26
Kshitij
joined the channel
05:27
Kshitij
left the channel
05:27
usmankhan
joined the channel
05:31
usmankhan
Hello, Bertl, I am not sure how we can use a single comparator to build ADC. Maybe, we can use a successive approximation ADC which uses a comparator but in that case the control logic will be provided by the FPGA
05:34
usmankhan
The disadvantage, however, (compared to a traditional flash ADC) is that it can handle a typically maximum frequency of 100kHz
05:36
usmankhan
http://www.electronics.dit.ie/staff/tscarff/dac_adc/adc_da2.gif
05:41
usmankhan
left the channel
05:57
Spirit532
left the channel
06:00
ItsMeLenny
joined the channel
07:00
usmankhan
joined the channel
07:00
Alvis
left the channel
07:02
niculescu_vlad
joined the channel
07:03
niculescu_vlad
left the channel
07:09
ItsMeLenny
left the channel
07:10
usmankhan
left the channel
07:33
niculescu_vlad
joined the channel
07:55
Spirit532
joined the channel
08:06
se6astian|away
changed nick to: se6astian
08:11
IldarValiev
joined the channel
08:18
niculescu_vlad
left the channel
08:43
dimaursu16
left the channel
08:44
dimaursu16
joined the channel
09:23
Spirit532
left the channel
09:47
Bertl_zZ
changed nick to: Bertl
09:47
Bertl
morning folks!
09:48
Bertl
Hey IldarValiev!
09:49
Bertl
usmankhan: the idea is to drop the ADC and have one or two comparators which compare the output voltage to the reference voltage generated by the DAC
09:52
IldarValiev
left the channel
10:16
RexO
left the channel
10:34
Rex0r
joined the channel
11:51
Bertl
off for now ... bbl
11:51
sagnikbasu95
joined the channel
11:51
Bertl
changed nick to: Bertl_oO
11:52
BAndiT1983|away
changed nick to: BAndiT1983
12:10
comradekingu
left the channel
12:10
comradekingu
joined the channel
13:20
RexOrCine
left the channel
13:42
Spirit532
joined the channel
13:59
BAndiT1983
changed nick to: BAndiT1983|away
14:42
se6astian
changed nick to: se6astian|away
14:43
se6astian|away
changed nick to: se6astian
15:08
Rex0r
left the channel
15:13
usmankhan
joined the channel
15:14
usmankhan
Ok, I get the idea. However, we won't be able to implement the modified adaptive PID algorithm since we don't know the precise error values.
15:15
usmankhan
Also, I remember from one of the earlier chats where you said that we should be able to check if the error exceeds a certain threshold and we should be able to log/report that. That won't be possible with this approach too
15:16
usmankhan
Anyway, I will modify the proposal as per your comments and then discuss further
15:17
usmankhan
In the meanwhile, please let me know if you have any other comments
15:19
usmankhan
left the channel
15:22
jucar
joined the channel
15:27
Bertl_oO
usmankhan: as I said, maybe investigate both options and evaluate the pros and cons ... note that we need to connect the ADC somehow and also note that we need at least 8 ADCs for the planned power board
15:34
Alvis
joined the channel
15:37
usmankhan
joined the channel
15:37
usmankhan
Alright, makes sense. I will do that and get back to you
15:50
jucar
left the channel
15:55
intracube_afk
changed nick to: intracube
16:04
usmankhan
left the channel
16:12
se6astian
changed nick to: se6astian|away
16:54
se6astian|away
changed nick to: se6astian
16:54
usmankhan
joined the channel
17:10
sagnikbasu95
left the channel
17:49
elllkk_
joined the channel
17:53
elllkk_
left the channel
17:57
Elbehery
joined the channel
18:16
sagnikbasu
joined the channel
18:19
Bertl_oO
off for now ... bbl
18:24
Alvis
left the channel
18:33
Alvis
joined the channel
18:39
Alvis_
joined the channel
18:42
Alvis
left the channel
19:17
stancho
joined the channel
19:45
sagnikbasu
left the channel
20:02
stancho
left the channel
20:03
Alvis_
left the channel
20:05
Bertl_oO
changed nick to: Bertl
20:05
Bertl
back now ...
20:08
se6astian
changed nick to: se6astian|away
20:13
anuditverma
joined the channel
20:15
niculescu_vlad
joined the channel
20:20
dimaursu16
left the channel
20:34
dimaursu16
joined the channel
20:34
dimaursu16
left the channel
20:34
dimaursu16
joined the channel
20:36
Alvis
joined the channel
20:39
niculescu_vlad
left the channel
20:39
niculescu_vlad
joined the channel
20:39
anuditverma
left the channel
20:50
niculescu_vlad
left the channel
21:01
BAndiT1983|away
changed nick to: BAndiT1983
21:02
Alvis
left the channel
21:05
Alvis
joined the channel
21:38
Alvis_
joined the channel
21:39
Alvis
left the channel
22:36
Alvis
joined the channel
22:39
Alvis_
left the channel
22:40
Alvis_
joined the channel
22:42
Alvis
left the channel
23:05
Alvis_
left the channel
23:06
intracube
left the channel
23:13
intracube
joined the channel
23:17
Alvis_
joined the channel
23:25
intracube
changed nick to: intracube_afk