00:06 | Legendin | left the channel | |
00:12 | dmjnova | joined the channel | |
00:38 | Bertl_zZ | changed nick to: Bertl
| |
00:38 | Bertl | back now ...
| |
00:57 | intracube | Bertl: se6astian|away: a small comment about the wording of the perks on the main indiegogo page;
| |
00:58 | intracube | it could be interpreted by people not familiar with indiegogo/crowdsourcing that they're obligated to buy the camera when they choose one of the €350 perks
| |
00:58 | intracube | rather than it giving them an entitlement to get the camera for a discount
| |
00:59 | intracube | I was a bit unsure about that when I first read it
| |
00:59 | comradekingu | How would you change it?
| |
01:00 | intracube | and there have been comments from at least one other person (on magiclantern forum) who also seems to have misunderstood
| |
01:00 | intracube | comradekingu: it's the "and est. €2,300 due at ship of axiom beta" that could be reworded
| |
01:01 | intracube | 'payment due' in english usually means some form of a bill
| |
01:03 | intracube | http://www.magiclantern.fm/forum/index.php?topic=11787.msg130295#msg130295
| |
01:07 | intracube | maybe the number of people actually confused is very low, idk
| |
01:12 | comradekingu | depends where you imply the comma
| |
01:12 | comradekingu | but it isnt actually when it ships, is it?
| |
01:17 | comradekingu | "When the axiom beta is due to ship" is better i think
| |
01:26 | Bertl | intracube: the problem is, the perks become immutable once they are chosen by at least one contributor
| |
01:26 | Bertl | i.e. for obvious reasons, they cannot be changed
| |
01:28 | intracube | comradekingu: not really, it still doesn't make clear that:
| |
01:28 | intracube | - the final purchase of the camera is completely optional
| |
01:28 | Bertl | but I agree, wording could have been better ... then on the other hand, we are not native English speakers, and we try to ask native speakers for help on every occasion, it probably just wasn't available when it was written
| |
01:28 | intracube | - the buyer doesn't have to buy the camera as soon as they become available
| |
01:28 | intracube | but as Bertl says, it can't be changed now
| |
01:29 | Bertl | also note that indigogo perks and descriptions are quite limite
| |
01:29 | Bertl | +d
| |
01:29 | Bertl | (in number of characters that is)
| |
01:30 | intracube | Bertl: yup, no worries :)
| |
01:30 | Bertl | I'm not worried :)
| |
01:31 | intracube | figure of speech :P
| |
01:31 | Bertl | I know *G*
| |
01:31 | Bertl | btw, can anybody help me understand what the idea of 'will sandenbergh' is? (comments section)
| |
01:33 | intracube | Bertl: what bit are you confused by?
| |
01:34 | Bertl | I do understand the feature list and vote on part
| |
01:34 | Bertl | but does he mean on indiegogo? in combination with a contribution/perk?
| |
01:35 | wescotte | joined the channel | |
01:35 | intracube | I don't know. possibly he means after the current campaign ends.
| |
01:36 | Bertl | hmm ...
| |
01:36 | intracube | create a list on apertus.org and allow people to vote for most wanted features
| |
01:37 | intracube | which might be the basis for anther funding campaign
| |
01:38 | intracube | the idea sounds unnecessary right now
| |
01:38 | Bertl | we already do regular feature voting (although it was a little slow/lame the last month)
| |
01:38 | Bertl | (lame because we didn't change the questions :)
| |
01:38 | intracube | right now I guess the focus for you is on the fundamental 'must have' features
| |
01:38 | intracube | 'would like' features come much later
| |
01:39 | Bertl | yes, I'm constantly urging folks to put features they would like to see at some point on the wiki
| |
01:39 | intracube | extra lens mounts, different sensors would be too much variation at this stage
| |
01:39 | intracube | I'm surprised there are three sensor options already
| |
01:39 | Bertl | because right now I have to be super critical about every feature because it _will_ cost extra time and money
| |
01:39 | Bertl | and we do not want to finish the Beta in five years :)
| |
01:41 | intracube | but it would be good to know what features most people want for the longer term
| |
01:42 | Bertl | yes, precisely
| |
01:42 | intracube | an OLPF would be right near the top of my list
| |
01:43 | Bertl | what's the problem with that?
| |
01:43 | intracube | obviously testing can be done without, but for usable footage it's important
| |
01:43 | Bertl | I mean, what is the problem with simply adding one?
| |
01:43 | intracube | aliasing/false colour/moire
| |
01:43 | intracube | no problem
| |
01:43 | Bertl | so why add it as 'core part' to the Beta?
| |
01:43 | intracube | except finding a supplier and getting a filter that closely matches the sensor resolution
| |
01:44 | Bertl | so that is something we can certainly outsource
| |
01:44 | intracube | and have it physically fit in front of the sensor
| |
01:44 | Bertl | (to the community that is)
| |
01:44 | intracube | Bertl: agreed
| |
01:44 | intracube | it's still a core component, imo
| |
01:45 | Bertl | it is a quite expensive part nevertheless, so nothing everybody will want to have integrated or so
| |
01:45 | intracube | yep
| |
01:46 | Bertl | and we won't be able to test a variety of solutions, otherwise we would spend all our money on OLPF samples :)
| |
01:46 | Bertl | so, while it might be an important part, it is better addressed by the community
| |
01:46 | intracube | maybe you could get free samples if OLPF manufacturers think they'll get more business?
| |
01:47 | Bertl | what we can think about, and if it is something which is important to you, is to devise test methods to make it easy to compare different solutions
| |
01:49 | Bertl | something like the Beta can only scale if we actually manage to leverage to potential of the community here
| |
01:49 | Bertl | i.e. devise some tests which provide enough data to classify a solution
| |
01:50 | Bertl | have everybody interested acquire a specific solution (coordinated efford) and test it with the devised test system for classification
| |
01:50 | Bertl | feed back the data and identify excellent choices as well as get rid of problematic solutions
| |
01:51 | intracube | yep
| |
01:58 | intracube | time for bed
| |
01:58 | Bertl | sweet dreams!
| |
01:59 | intracube | night
| |
01:59 | intracube | changed nick to: intracube_afk
| |
02:49 | jucar1 | joined the channel | |
02:56 | jucar | left the channel | |
04:01 | wescotte | left the channel | |
04:24 | Bertl | 90k \o/
| |
04:44 | se6astian|away | changed nick to: se6astian
| |
04:45 | Bertl | morning se6astian!
| |
04:45 | se6astian | good morning
| |
04:45 | se6astian | still working away on the AXIOM Beta? :)
| |
04:45 | Bertl | yup
| |
04:46 | se6astian | great!
| |
04:49 | Gegsite | joined the channel | |
05:57 | Gegsite | left the channel | |
07:17 | g3gg0 | joined the channel | |
07:53 | Bertl | off to bed now ... have a good one everyone!
| |
07:53 | Bertl | changed nick to: Bertl_zZ
| |
07:57 | g3gg0 | left the channel | |
08:50 | aquarat2 | joined the channel | |
08:54 | aquarat | left the channel | |
09:12 | Legendin | joined the channel | |
09:58 | g3gg0 | joined the channel | |
10:32 | Legendin | left the channel | |
10:44 | Legendin | joined the channel | |
11:03 | Gegsite | joined the channel | |
12:47 | Bertl_zZ | changed nick to: Bertl
| |
12:47 | Bertl | morning folks!
| |
12:57 | Legendin | Heya Bertl!
| |
12:58 | Legendin | Am I right in assuming you live in the states, or?
| |
12:59 | Gegsite | left the channel | |
13:02 | Bertl | no, you're wrong, I'm in Austria (Europe)
| |
13:42 | Bertl | off for now ... bbl
| |
13:42 | Bertl | changed nick to: Bertl_oO
| |
15:02 | ctag | left the channel | |
15:27 | Legendin | left the channel | |
15:48 | Bertl_oO | changed nick to: Bertl
| |
15:48 | Bertl | back now ...
| |
16:40 | daFred | joined the channel | |
16:44 | daFred | it's almost time to cool some bottles of champagne...
| |
16:46 | Bertl | almost, yes :)
| |
17:10 | se6astian | changed nick to: se6astian|away
| |
17:10 | Gegsite | joined the channel | |
17:27 | g3gg0 | still a few k€ to gain
| |
17:28 | g3gg0 | 5 days and nearly 8k€
| |
17:28 | g3gg0 | the last boost calms down, hope it wont stagnate and stop at 99k€
| |
17:28 | g3gg0 | :)
| |
17:29 | Bertl | it is the weekend
| |
17:29 | Gegsite | I would donate, but still confused
| |
17:30 | Bertl | so let's hear what confuses you :)
| |
17:30 | Bertl | maybe we can clear that up once for all
| |
17:30 | Gegsite | not mutch so will dont recall to me that 8k€ :)
| |
17:31 | Bertl | oookay?
| |
17:31 | Gegsite | mmmkay
| |
17:32 | Gegsite | I'm reading now the axiom beta
| |
17:32 | Bertl | means, no clue what you are trying to communicate?
| |
17:35 | Bertl | anyway, if you have questions regarding the AXIOM Beta, your potential contribution or anything related to the development, please do not hesitate to ask
| |
17:45 | comradekingu | heia g3gg0 from german op boards?
| |
17:46 | Gegsite | I will be probably Mega supporter in a few days :P
| |
17:50 | comradekingu | 93% it says
| |
17:55 | daFred | are there engineering grade cmosis CMV12000 available? Thinking of a long therm "burn in" test as soon as possible. So we have maybe 6 month test when the first batch is ready. We could put the latest image on the webpage...
| |
17:58 | Bertl | we got one for the Alpha, but I think they are not readily available anymore
| |
17:58 | g3gg0 | comradekingu: german, right. magic lantern dev.
| |
18:00 | comradekingu | nice
| |
18:00 | comradekingu | I think i tried that on my dads point and shoot canon
| |
18:00 | comradekingu | no that was chdk
| |
18:02 | Bertl | g3gg0: btw, has there been some support from Canon for ML?
| |
18:02 | g3gg0 | bertl: not at all
| |
18:02 | g3gg0 | not even a word about us
| |
18:04 | g3gg0 | comradekingu: yeah, CHDG is related to ML
| |
18:04 | g3gg0 | the chips share the same architecture
| |
18:04 | g3gg0 | registers, devices etc
| |
18:05 | g3gg0 | at least the DIGIC-chip internals are related
| |
18:05 | g3gg0 | on-board devices like codecs, sensors, ADCs and so on are different
| |
18:05 | g3gg0 | http://magiclantern.wikia.com/wiki/Register_Map
| |
18:05 | g3gg0 | this is what we know about the digic's on-die peripherals
| |
18:06 | g3gg0 | ADCs are external and we are not sure how to program them. we learnt a lot from experiments too, but a datasheet would help quite a lot more :)
| |
18:07 | Bertl | g3gg0: I suspected so, but they should have realized by now that ML is not the enemy, on the contrary, I'm pretty sure many folks buy those cameras because of ML
| |
18:08 | g3gg0 | Bertl: its complicated for a company like canon to give a statement about us... they dont control us. so they cannot be sure that we will not act against their plans
| |
18:08 | Q_ | g3gg0: Are those ADCs from some known manufacturer?
| |
18:08 | g3gg0 | see the 1DC 1DX discussion..
| |
18:08 | g3gg0 | Q_: yes, ADI
| |
18:08 | Bertl | so how is ML designed ATM, can you give me a quick intro?
| |
18:09 | Q_ | g3gg0: ADI should have datasheets for them?
| |
18:09 | g3gg0 | Q_: http://magiclantern.wikia.com/wiki/ADTG one of them in a 600D
| |
18:09 | g3gg0 | Q_: no, its a customer-specific part. not even technical support has access to information what this chip does
| |
18:09 | Q_ | Oh.
| |
18:09 | g3gg0 | Bertl: well, its quite complex. let me explain....
| |
18:10 | g3gg0 | Bertl: stock canon firmware has a flag in its configuration area that tells the bootloader to check the memory card for a binary AUTOEXEC.BIN that is being loaded to a predefined memory address and executed
| |
18:11 | Bertl | okay (LOL @ name)
| |
18:11 | g3gg0 | Bertl: now using a forged firmware update file we rewrite this flag. (firmware update does not contain a new firmware, but just custom flashloaders that write the flag)
| |
18:11 | Bertl | so the firmware/updates are not signed or anything, yes?
| |
18:12 | g3gg0 | then we have two different methods of hooking into boot: a) copying firmware parts into RAM, patching it, hooking a task and continue execution
| |
18:12 | g3gg0 | or b) patching cache content so the flash "looks" different and so patch any arbitrary memory location
| |
18:12 | g3gg0 | flashloaders are signed
| |
18:13 | g3gg0 | but we know how to.....
| |
18:13 | Bertl | okay :)
| |
18:13 | g3gg0 | thats the story. that way we can hook into normal firmware execution
| |
18:13 | Bertl | so basically you get control early and patch up the existing firmware with custom parts, yes?
| |
18:13 | g3gg0 | of course we have to hijack some ram areas for ourself etc
| |
18:13 | g3gg0 | yeah
| |
18:14 | g3gg0 | not much needed by the way
| |
18:14 | g3gg0 | normal magic lantern only needs startup patches
| |
18:14 | g3gg0 | during runtime we can simply add features by calling firmware routines
| |
18:15 | g3gg0 | sometimes we have to patch RAM pointers, but normally no firmware content has to be modified
| |
18:15 | g3gg0 | if we have to "patch" firmware content for some features, we mostly do that with cache hacks
| |
18:15 | g3gg0 | the ARM's CP15 interface allows arbitrary access to cache content
| |
18:16 | g3gg0 | lock cache pages, place our patches there, noone notices
| |
18:17 | Bertl | understood, and development tools for the cpus (what do they use) are available somewhere?
| |
18:17 | Bertl | ah yes, okay, that explains
| |
18:17 | Bertl | is there a danger/potential for canon to lock you out?
| |
18:17 | g3gg0 | sure
| |
18:17 | g3gg0 | they never did :)
| |
18:17 | Bertl | or is the design "so good" that they have no way to block you, like on the xbox :)
| |
18:17 | g3gg0 | you mean A20 ?
| |
18:17 | g3gg0 | nah they have the chance to block us by signing .FIR = firmware update a lot better
| |
18:17 | Q_ | Afaik, canon known about it and doesn't care.
| |
18:18 | g3gg0 | Q_: yep thats the case
| |
18:18 | g3gg0 | they dont love us, they dont hate us
| |
18:18 | Bertl | yeah, that's good, but it might suddenly become a problem
| |
18:18 | g3gg0 | we dont cause trouble, people like ML, so they stay nice and dont actively suppress ML
| |
18:19 | Bertl | of course, there is no automatic firmware update (yet) so probably not critical for existing cameras
| |
18:19 | g3gg0 | Bertl: who knows what happens in the future, yes
| |
18:19 | g3gg0 | t.b.h.. from the firmware design... canon cameras never were made for video recording
| |
18:20 | g3gg0 | reworking all the sensor IO driver stuff will give us perhaps 100 fps without trouble.
| |
18:20 | g3gg0 | but they do it... uh oh... in a so "unconvinient" way that you can be lucky to get 60 fps
| |
18:21 | g3gg0 | toooo much reinitializations. every frame is set up like a single frame shot. all settings written again etc..
| |
18:21 | comradekingu | g3gg0: i would spend more time focusing on being profitable if i were canon. Magic lantern is the only ting making me consider canon cameras
| |
18:22 | g3gg0 | comradekingu: not sure... dont think they ever noticed our work in their money pockets
| |
18:23 | comradekingu | they must notice their own business model not working out
| |
18:23 | g3gg0 | the coolness factor of a custom firmware hack like ML is absolutely high. but this affects maybe 2% of their customers
| |
18:24 | Bertl | okay, so how does development work in ML? are features added/unlocked because they have been discovered or is it more development on user demand?
| |
18:24 | comradekingu | this is the paradox though, i value magic lantern, but i dont want to support canon
| |
18:24 | Bertl | a good point for the AXIOM I guess :)
| |
18:25 | g3gg0 | Bertl: both. its our hobby. if we find things out that look cool, we try to make it a feature that users can use. if users ask for addition or changes, we try to add them.
| |
18:25 | comradekingu | yes, thats why axiom makes sense. I figured this when working on the opensource site
| |
18:26 | g3gg0 | e.g. raw video. people then asked for raw(-data based) histograms and alex added it
| |
18:26 | Bertl | cool! so what are the most requested features?
| |
18:26 | g3gg0 | 4k raw video
| |
18:26 | g3gg0 | definitely
| |
18:27 | Bertl | and that works at a reasonable frame rate (i.e. 30 FPS or so)?
| |
18:27 | g3gg0 | people also love: focus peaking, raw histograms, zooming in some areas to check focus
| |
18:27 | g3gg0 | 24fps or a multiple
| |
18:28 | Bertl | where does the data go?
| |
18:28 | ctag | joined the channel | |
18:28 | g3gg0 | tahts waht i remember of video guys asking for
| |
18:28 | g3gg0 | in ML?
| |
18:28 | Bertl | 4k@24FPS is a lot of data if uncompressed
| |
18:28 | g3gg0 | yeah
| |
18:28 | g3gg0 | in ML we only have the chance to write it onto CF/SD cards
| |
18:29 | g3gg0 | a good CF in canon cameras can get up to 110 MB/s
| |
18:29 | Bertl | and you have a single card, yes?
| |
18:29 | g3gg0 | we have a few hundred megabytes of RAM, so we can buffer a few seconds
| |
18:29 | g3gg0 | so there are two ways of recording:
| |
18:30 | g3gg0 | some 1080 resolution recording as RAW without limit. so you can record minutes
| |
18:30 | Q_ | There are at least models that have more than 1 slot.
| |
18:30 | g3gg0 | others do only 5 seconds recordings with max resolution
| |
18:30 | Bertl | okay, let me do some math and correct me when I go wrong:
| |
18:30 | g3gg0 | depending on your storyline this is possible.
| |
18:30 | Q_ | But I think the camera goes off if you open the door?
| |
18:31 | Q_ | Or did they fix that?
| |
18:31 | g3gg0 | 5D3 has SD + CF
| |
18:31 | Bertl | 4k = 4096x3072 or something like that, at probably 10-12bit, yes?
| |
18:31 | g3gg0 | you cannot open the door while the camera is recording
| |
18:31 | g3gg0 | eos has 14 bit
| |
18:31 | g3gg0 | but 12 is okay
| |
18:31 | Bertl | so roughly 16 gigabit per frame uncompressed raw
| |
18:32 | comradekingu | can you hook up the CF pins to a P-ata interface?
| |
18:32 | Bertl | no, sorry, too fast
| |
18:32 | g3gg0 | 18Mib/frame
| |
18:32 | g3gg0 | MiB
| |
18:32 | Bertl | 160 megabit/frame
| |
18:32 | g3gg0 | lets calc with 20, yeah
| |
18:33 | Bertl | or toughly 20Mbyte/frame
| |
18:33 | Bertl | *roughly
| |
18:33 | Bertl | right?
| |
18:33 | g3gg0 | you would need a few CFs in parallel
| |
18:33 | g3gg0 | yep
| |
18:33 | Bertl | okay, so I presume you do some lossless compression on the fly?
| |
18:33 | g3gg0 | nope. wait!
| |
18:34 | g3gg0 | we *can not* do 4k video yet
| |
18:34 | g3gg0 | people demand it :)
| |
18:34 | Bertl | ah, okay, that explains a lot :)
| |
18:34 | g3gg0 | we can only deliver 1080 yet
| |
18:34 | Bertl | because I think the memory doesn't help you either
| |
18:34 | g3gg0 | or higher resolutions but only a few seconds of video due to the large buffers :)
| |
18:34 | Bertl | for obvious reasons, as you can hardly get a second stored in a few hundred megabytes
| |
18:35 | g3gg0 | 2k is doable iirc
| |
18:35 | Bertl | we did some calculations for the beta, which has 1 Gigabyte of memory
| |
18:36 | Q_ | Do any of them have HDMI?
| |
18:36 | Bertl | and you can get away with 128M for linux, if you need to
| |
18:36 | g3gg0 | due to the raw workflow upscaling to e.g. 4k isnt a big problem and you still get better footage than 4k compressed video
| |
18:36 | Bertl | which leaves almost 900Mbyte for buffering
| |
18:36 | g3gg0 | (from what i remember video guys telling us)
| |
18:37 | g3gg0 | but you could add four CF slots :)
| |
18:37 | g3gg0 | or SSD-SATA ports
| |
18:37 | g3gg0 | just one hypothetical example:
| |
18:37 | Bertl | yes, 6G sata ports on the Z-7015, that's the idea (once we get there)
| |
18:38 | Bertl | 93k \o/
| |
18:38 | regnirps | Do like Mac Pro. Put flash on a system bus.
| |
18:39 | g3gg0 | four CF-slots with 100MiB/s write rate each, hardware controlled I/O from FPGA domain. this should get you 24fps with 4k stable.
| |
18:39 | g3gg0 | just the storage size, lets say 512 GiB is limiting you
| |
18:39 | Bertl | barely if at all
| |
18:40 | daFred | short input: 6655 EUR to go .... this is better than watching TV ...
| |
18:40 | Bertl | 4k,12bit,24FPS (to take your example) gives 452MByte/s
| |
18:40 | Bertl | daFred: definitely
| |
18:40 | regnirps | But I think non-volative RAM (the feroelectric stuff. Samsung is going to be producing - among others) will displace flash in a year or so, so plan for it?
| |
18:41 | g3gg0 | uh i get 430 MiB/s
| |
18:41 | g3gg0 | 432
| |
18:41 | Bertl | yes, you are right
| |
18:41 | g3gg0 | but still doable with nowadays CF
| |
18:41 | Q_ | MiB vs MB
| |
18:41 | Bertl | I had the 30FPS in mind
| |
18:42 | g3gg0 | ah ok :)
| |
18:42 | Bertl | but that's actually 540MByte/s
| |
18:42 | g3gg0 | if you play with these thoughts, i can recommend using the .mlv format for video frames
| |
18:43 | g3gg0 | it allows you span the data across several media if needed
| |
18:43 | Q_ | So if you want to record an hour, you're at 1.6 TB.
| |
18:43 | g3gg0 | e.g. on 5D3 we can use CF plus SD for writing video frames
| |
18:43 | Bertl | sounds good, does it handle random access, indexing and checksumming somehow?
| |
18:43 | g3gg0 | some post processing tools already know this format and the workflow is evolving
| |
18:44 | g3gg0 | Bertl: random access: yes. it is block oriented. every data block (metadata or video/audio) contains a µs timestamp
| |
18:45 | g3gg0 | Bertl: blocks can appear in any order, in any file and PP tools have to index them.
| |
18:45 | Bertl | okay
| |
18:46 | g3gg0 | Bertl: i had in mind that files are split at 2G / 4G, a "recording" can contain of multiple files and due to buffering the data from camera isnt necesarily in sequence
| |
18:46 | g3gg0 | checksums are not implemented. good point.
| |
18:47 | g3gg0 | easy to add if required
| |
18:47 | Bertl | that is always good to hear
| |
18:48 | g3gg0 | we have a lot of metadata blocks that are predefined
| |
18:48 | Bertl | regarding the raw data itself, what data order/sequence does/can it support?
| |
18:48 | g3gg0 | so if you experiment with raw video, contact me, we can get a minimal solution very quickly
| |
18:49 | g3gg0 | it is (right now) hardcoded to 1-16 bpp where the data is stored as follows:
| |
18:49 | g3gg0 | feed the raw stream into a shift register, 16 bit wide and store the 16 bit words as LE
| |
18:49 | g3gg0 | example
| |
18:50 | g3gg0 | http://magiclantern.fm/forum/index.php?topic=11787.msg128638#msg128638
| |
18:50 | g3gg0 | canon always uses RGGB iirc
| |
18:50 | g3gg0 | so the first line will be RGRGRGRG and the second GBGBGBGB
| |
18:50 | Bertl | okay, but it doesn't (at the moment) support reordering or similar?
| |
18:51 | Bertl | e.g. swapping columns or rows
| |
18:51 | g3gg0 | i am currently reworking the RAWI metadata block to support defining various other modes
| |
18:51 | g3gg0 | ah i see
| |
18:51 | g3gg0 | if you need smth like that, let me know.
| |
18:52 | g3gg0 | so you would need a field that denotes the orientation of the sensor data
| |
18:52 | Bertl | it might come very handy with the Cmosis sensors, if you don't want/need to reconstruct the image in the camera
| |
18:52 | g3gg0 | sure, thats easily implemented
| |
18:52 | g3gg0 | it only has to taken into account when converting MLV->DNG
| |
18:52 | Bertl | because they send data as 0, 128, 256 ... 1, 129, 257, ...
| |
18:53 | g3gg0 | ?
| |
18:53 | Bertl | that's columns
| |
18:53 | g3gg0 | they interlace colums?
| |
18:53 | Bertl | kind of, there are up to 64 LVDS channels on existing Cmosis sensors
| |
18:54 | Bertl | and they basically split a line (or several) up into evenly spaced chunks
| |
18:54 | g3gg0 | the best then would be providing a pixelmap block
| |
18:54 | g3gg0 | https://docs.google.com/spreadsheet/ccc?key=0AgQ2MOkAZTFHdHJraTVTOEpmNEIwTVlKd0dHVi1ULUE&usp=drive_web#gid=1
| |
18:55 | g3gg0 | on the first page you see an example file layout
| |
18:55 | g3gg0 | files have to start with a header
| |
18:55 | g3gg0 | contains basic infos about its content
| |
18:55 | g3gg0 | then info blocks that describe the payload format, the lens settings etc
| |
18:56 | g3gg0 | and later you will see video / audio blocks and info blocks in any order
| |
18:56 | Bertl | okay, how are the blocks chained?
| |
18:56 | g3gg0 | there are only two blocks - beside video data of course - that are important. file header "MLVI" and raw data description "RAWI"
| |
18:57 | g3gg0 | every block has the same basic header format
| |
18:57 | g3gg0 | see tab "Structures"
| |
18:57 | g3gg0 | first four bytes: block type, like "MLVI", "LENS", "RAWI" etc
| |
18:57 | g3gg0 | then four bytes block size
| |
18:58 | g3gg0 | then 8 bytes timestamp in µsec resolution
| |
18:58 | Bertl | okay, I guess it would be great to get a checksum field into most of those structures
| |
18:58 | g3gg0 | thats the basic structure shared with all blocks (except file header)
| |
18:58 | g3gg0 | yeah
| |
18:59 | g3gg0 | 99% of our data streams were not checksum-able, so it didnt come into mind to do so ;)
| |
18:59 | Bertl | the actual checksumming doesn't need to be fixed in advance, but a 64bit field would help a lot
| |
19:00 | g3gg0 | i'd add the fields at the end of each block
| |
19:00 | Bertl | we can easily do some crc polynomial on the data, which then can be combined if needed
| |
19:00 | g3gg0 | would keep compatibility with all tools
| |
19:01 | Bertl | end is fine as well I guess
| |
19:01 | g3gg0 | i'd add a CRCI block that explains which CRC to use
| |
19:01 | g3gg0 | or a field in MLVI
| |
19:01 | Bertl | that's an excellent idea
| |
19:01 | Bertl | the CRCI block
| |
19:01 | g3gg0 | then we can opt-in the CRC and also stay open for new algorithms if necessary at all
| |
19:02 | Bertl | it would allow to use case specific checksums, from simple to cryptographically strong
| |
19:02 | g3gg0 | you think about signing?
| |
19:02 | Bertl | well, if authenticity is important, it might make sense
| |
19:03 | Bertl | for example, we had a very interesting request from a reporter
| |
19:03 | Bertl | who asked if we could do encryption, so that not even he would be able to decrypt it, only the new company "at home" (mostly for war zones)
| |
19:03 | Bertl | *news
| |
19:04 | Bertl | when they want to prevent tampering or similar
| |
19:04 | g3gg0 | haha cool
| |
19:04 | g3gg0 | wait
| |
19:05 | g3gg0 | http://magiclantern.fm/forum/index.php?topic=10279.0
| |
19:05 | g3gg0 | RSA
| |
19:06 | g3gg0 | *RSA enryption support
| |
19:07 | Legendin | joined the channel | |
19:07 | Bertl | there you go :)
| |
19:08 | g3gg0 | so you can point him to us ;)
| |
19:09 | g3gg0 | there is close to no request for this, but its there and can be polished if someone needs that
| |
19:10 | Bertl | yeah, those are special cases after all, but it is a fair reason
| |
19:11 | g3gg0 | afk, son demands his dad :)
| |
19:11 | Bertl | okay, I have to take a nap, I'm rather exhausted ... thanks a lot for the very informative and pleasant chat, I'll be back later tonight ...
| |
19:12 | Bertl | send my best to your son :)
| |
19:12 | Bertl | changed nick to: Bertl_zZ
| |
19:12 | g3gg0 | sure, sleep well. we will see us :)
| |
19:12 | g3gg0 | cya
| |
19:14 | troy_s | Bertl_zZ: I love that encryption idea.
| |
19:14 | troy_s | That's damn brilliant to have a reporter sign their footage and encrypt it.
| |
19:29 | comradekingu | g3gg0: maybe it could be combined with https://www.martus.org/
| |
19:45 | daFred | left the channel | |
19:51 | Gegsite | left the channel | |
19:59 | acycke | joined the channel | |
20:04 | acycke | left the channel | |
20:45 | se6astian|away | changed nick to: se6astian
| |
20:56 | mars_ | 95k, so close
| |
20:58 | se6astian | I think we can make it!
| |
20:58 | mars_ | yes, we can!
| |
20:59 | se6astian | are we austrians or austricans?!
| |
20:59 | se6astian | or austricants :)
| |
20:59 | mars_ | cans!
| |
21:08 | wescotte | joined the channel | |
21:11 | wescotte | Wow! Great last two days!
| |
21:18 | wescotte | left the channel | |
21:30 | g3gg0 | re
| |
21:31 | g3gg0 | anyone having a plot of the money flow from the beginning?
| |
21:32 | se6astian | http://crowdlogs.com/project/axiom-beta-the-first-open-digital-cinema-camera
| |
21:36 | mars_ | cool
| |
21:44 | Bertl_zZ | changed nick to: Bertl
| |
21:46 | Bertl | back now ...
| |
21:47 | g3gg0 | wow!
| |
21:47 | g3gg0 | thx se6astian
| |
21:48 | Bertl | where crowdlogs is of course not realtime, and we seem to have entered the second hot phase, probably thanks to your (ML) excellent PR
| |
21:53 | se6astian | yes you can clearly see in the graph when ML started encouraging people to support AXIOM
| |
21:53 | se6astian | and when this was placed on the ML frontpage
| |
21:54 | se6astian | I think (even though we are not 100% there yet) that without you (ML) we would not have made it
| |
21:54 | dmjnova | wow, we broke a lot of their stats
| |
21:54 | se6astian | ok time for bed, already up since 5:30AM :)
| |
21:55 | se6astian | night night!
| |
21:55 | se6astian | changed nick to: se6astian|away
| |
21:56 | Bertl | sleep well!
| |
21:59 | g3gg0 | yeah now when seeing the plot. WOW
| |
21:59 | g3gg0 | impressive support through our userbase
| |
22:00 | g3gg0 | sleep well, se6astian
| |
22:00 | Bertl | yes, so I wouldn't worry about your Beta cameras :)
| |
22:01 | g3gg0 | yay :D
| |
22:01 | Bertl | have you decided (as developers) what configurations you'd prefer?
| |
22:02 | g3gg0 | not at all yet :)
| |
22:03 | g3gg0 | where a1ex is looking into getting the most out of the sensor, i tend to look into raw recording
| |
22:03 | Bertl | because I think it would make sense to focus on a specific variant and try to get early betas to you ASAP
| |
22:04 | Bertl | (which may later be exchanged against final versions)
| |
22:05 | g3gg0 | so i'd stick to a1exs opinion. he looked very deep into sensor details and how good those are
| |
22:05 | daFred | joined the channel | |
22:06 | daFred | left the channel | |
22:06 | Bertl | so a preference for the CMV12k probably on MicroZed as backend then, yes?
| |
22:07 | g3gg0 | gigabit ethernet is working on beta?
| |
22:07 | Bertl | yes, but not at full performance without FPGA support (which isn't there yet)
| |
22:07 | Bertl | you get about 560-580k
| |
22:08 | g3gg0 | ah okay
| |
22:09 | Bertl | but gigabit is way to slow, even at full performance for recording
| |
22:09 | Bertl | 10G would kind of work for low quality :)
| |
22:10 | g3gg0 | nah just hoped to see some more bandwidth output than SD card :D
| |
22:10 | g3gg0 | dont have an atomos
| |
22:11 | Bertl | we are working on a solution for that as well, no worries
| |
22:12 | Bertl | if you're tricky, you can use both SD cards, internal and one on the external SD interface at up to 50M/s
| |
22:12 | Bertl | (with fast cards)
| |
22:12 | g3gg0 | that would be a solution, yeah
| |
22:13 | g3gg0 | i did a lot of magic with ARM cores, but never touched FPGAs
| |
22:14 | g3gg0 | so i am curious how both worlds could be combined for reasonable performance
| |
22:15 | g3gg0 | but thats solvable :)
| |
22:19 | g3gg0 | ok BTT. sensor type: i have no clue which one to start with. i guess the one which promises best quality
| |
22:21 | wescotte | joined the channel | |
22:21 | wescotte | left the channel | |
22:21 | Bertl | that is a good answert, but I have no clue which one that would be :)
| |
22:22 | Bertl | the KAC looks quite interesting on paper but we haven't had a chance to test it yet
| |
22:22 | Bertl | the Cmosis has been tested, so you probably know the quality better than we do
| |
22:23 | Bertl | anyway, nothing which needs to be decided tonight, but maybe have a chat about that in the next week or so
| |
22:24 | g3gg0 | i'd definitely stick to a1ex word
| |
22:24 | g3gg0 | you make four dark shots and he tells you how good the sensor is :)
| |
22:25 | wescotte | joined the channel | |
22:28 | Bertl | he had the KAC on his chart, but I do not remember if it was just projected from the datasheet or the result of actual testing
| |
22:29 | wescotte | left the channel | |
22:30 | Bertl | in any case, danieel should be able to take some samples with the KAC, as he has built a camera with that sensor, IIRC
| |
22:30 | g3gg0 | i'll speak to him soon
| |
22:30 | g3gg0 | a1ex has itnernet trouble atm
| |
22:30 | Bertl | ouch
| |
22:30 | g3gg0 | should be fixed during next week
| |
22:33 | Bertl | great!
| |
23:03 | Q_ | So I wanted to look at the raw samples from the CMV12000 that someone posted on the ML forum and points to footage.apertus.org, but that requires a login?
| |
23:13 | Bertl | that's mostly because it is read/write access at the moment, do you know what particular file it is/was, I can probably make it accessible without password
| |
23:14 | Bertl | or is it a general link to 'footage', because in this case, we should wait for se6astian
| |
23:21 | Q_ | The link was to: http://footage.apertus.org/AXIOM%20Alpha%20footage/ML/
| |
23:22 | Q_ | I wanted to look at those dark frames.
| |
23:28 | Bertl | okay, give me a second
| |
23:34 | troy_s | Bertl: Congrats.
| |
23:37 | troy_s | Bertl: That chart shape isn't ML. That shape is every single funding campaign ever.
| |
23:39 | Bertl | really?
| |
23:39 | Legendin | left the channel | |
23:39 | troy_s | Bertl: Really
| |
23:39 | troy_s | Almost creepily so
| |
23:40 | troy_s | Bertl: Every single campaign successful or failure, that hits a threshold of contributions looks precisely like that curvr
| |
23:40 | troy_s | Go figure.
| |
23:40 | troy_s | Bertl: Looking for another sample
| |
23:40 | Bertl | so it is coincidence that the knee points match the ML announcements?
| |
23:42 | troy_s | More or less.
| |
23:42 | troy_s | Every crowdfun goes through a surge, a plateau, then a surge
| |
23:42 | troy_s | There was an exhaustive discussion when Ubuntu edge happened
| |
23:42 | troy_s | It is a very familiar s curve
| |
23:46 | troy_s | I think there is an odd correlation between entry curve and exit too
| |
23:49 | Bertl | okay, in any case, great to have the ML folks on-board
| |
23:51 | Bertl | Q_: files will show up here: http://vserver.13thfloor.at/Stuff/AXIOM/ALPHA/ML/
| |
23:52 | Bertl | I will notify se6astian tomorrow to find a proper solution, the server uses nginx and I have no clue where this beast is configured
| |
23:53 | Bertl | it will take some time till all files are there
| |
23:54 | troy_s | Bertl: It is great to have ML on the same side. Some top minds there.
|