| 23:36 | slikdigit | hmmm. It could be hard coded bins could be tags instead, with reasonable per tag views
|
| 23:36 | slikdigit | and search
|
| 00:41 | fsteinel_ | joined the channel |
| 00:44 | fsteinel | left the channel |
| 01:34 | fsteinel_ | changed nick to: fsteinel
|
| 02:18 | troy_s | slikdigit: Yes something like that is plausible. Hence why it is critical to think through the logging side of applications
|
| 03:01 | intracube | left the channel |
| 03:09 | Bertl | off to bed now ... have a good one everyone!
|
| 03:09 | Bertl | changed nick to: Bertl_zZ
|
| 06:52 | niemand | joined the channel |
| 07:03 | niemand | left the channel |
| 07:13 | niemand | joined the channel |
| 07:35 | niemand | left the channel |
| 08:38 | Bertl_zZ | changed nick to: Bertl
|
| 08:38 | Bertl | morning folks!
|
| 09:08 | baldand_ | joined the channel |
| 09:10 | baldand | left the channel |
| 10:45 | niemand | joined the channel |
| 11:30 | niemand | left the channel |
| 11:50 | niemand | joined the channel |
| 12:03 | davidak | joined the channel |
| 12:22 | niemand | left the channel |
| 12:56 | danieel | troy_s: what is the best format for alpha channel? (separate stream - would a grayscale tiff do?)
|
| 12:57 | niemand | joined the channel |
| 13:12 | intracube | joined the channel |
| 13:34 | niemand | left the channel |
| 13:36 | intracube | re. editing tools, kdenlive has a basic clip bin - select in/out points on a longer clip then drag the preview over to the bin to create a virtual clip
|
| 13:37 | intracube | and the project format is xml so there's the possibility of converting to some standard EDL format
|
| 13:37 | intracube | cc troy_s, slikdigit ^
|
| 13:39 | intracube | which... I guess you guys already probably know :)
|
| 13:44 | troy_s | danieel: An associated alpha TIFF would do, yes.
|
| 13:44 | troy_s | Anything associated.
|
| 13:44 | troy_s | (As in no PNG.)
|
| 13:45 | Bertl | troy_s = PNG hater :)
|
| 13:45 | troy_s | intracube: The issue is it is designed as an always online NLE = broken by design.
|
| 13:45 | danieel | cant do PNG, would be slow as hell
|
| 13:45 | troy_s | Bertl: Yep. Express a colored green gas emitting light with a PNG. See you in the afterlife... :)
|
| 13:46 | danieel | so the alpha tiff is just a plain tiff, or some extra metadata is necessary to treat it as mask channel?
|
| 13:46 | troy_s | (As in only associated alpha can express light as it exists.)
|
| 13:47 | troy_s | danieel: Are you using it for lines?
|
| 13:47 | troy_s | danieel: Or trying to darken a region?
|
| 13:47 | danieel | also - is a lut common on 8 bit alphas? or we should aim for 16bits - its only for edges
|
| 13:47 | danieel | using it for keying
|
| 13:48 | danieel | but there is some water, so it is semitransparent
|
| 13:49 | Bertl | troy_s: you're aware that PNG supports associated alpha and 16bit depths?
|
| 13:49 | troy_s | danieel: Sorry... This is for post then?
|
| 13:50 | danieel | yes, post
|
| 13:50 | troy_s | Bertl: Nope. Junk format. No one uses it. Don't care that they revised it.
|
| 13:50 | intracube | troy_s: oh ok, I don't know what that means in practice here
|
| 13:50 | troy_s | danieel: Then you only need two formats. EXR and DPX. Keep your whole post pipe in EXR.
|
| 13:51 | danieel | exr is always 16bit float or can it do other?
|
| 13:51 | troy_s | danieel: Once you are in EXR, at the head, stay there until the very end.
|
| 13:51 | troy_s | danieel: 16 or 32 float.
|
| 13:52 | danieel | it is R3D for the video... that gets untouched, alpha is matched to it - so not sure how to call that input phase of post... but exr would do probably as well
|
| 13:52 | troy_s | danieel: Always linear. There are plenty of metadata options if needed, including custom.
|
| 13:52 | Bertl | the fact that many use it isn't any indication of the quality, no? just take GIF, which is probably the worst format ever (except for the microsoft bitmap format of course)
|
| 13:52 | danieel | at least it looks pro and who can not handle it, is not enough pro for our data :)
|
| 13:53 | troy_s | Bertl: No. No indication. Many say RGB is a colorspace. ;)
|
| 13:53 | lab-bot | sebastian created T268: Exposure Compensation. http://lab.apertus.org/T268
|
| 13:53 | troy_s | danieel: So you want to pull a key off water?
|
| 13:53 | danieel | Bertl: I like png for reference frames, but for actual video (even archive) i would not use it
|
| 13:53 | troy_s | Bertl: So yes... Fitting example. :)
|
| 13:53 | troy_s | danieel: Reference frames = DPX
|
| 13:54 | danieel | reference stills - like lossless
|
| 13:54 | troy_s | Basically when you have to communicate a transfer curve or final frame, that is precisely what DPX was designed for.
|
| 13:54 | troy_s | And if you are doing any sort of manipulation, that is what EXR was designed for. It extends well beyond that, of course.
|
| 13:55 | danieel | i do not think it is yet normalized to 0...1 as nontransparent...fullytransparent, so it curve does not matter yet
|
| 13:55 | danieel | *its
|
| 13:56 | troy_s | For me it is a no brained.
|
| 13:56 | troy_s | Brainer
|
| 13:56 | troy_s | Just go to EXRs at the head for any and all post work, then DPX for archival frame purposes.
|
| 13:57 | troy_s | (You could of course also go to DPX, then EXR as needed and leave untouched frames as DPX)
|
| 13:57 | danieel | can exr/dpx use compression? the keys/alpha might have same colored regions
|
| 13:57 | danieel | well compressible
|
| 13:57 | troy_s | It really is a helluva logical path that will keep your data pristine.
|
| 13:57 | troy_s | Lossless yes.
|
| 13:58 | danieel | yes, lossless
|
| 13:58 | troy_s | EXRs can use three types IIRC.
|
| 13:59 | troy_s | DPXs, being typically log, have log-light compression.
|
| 13:59 | troy_s | As well as RLE and possibly others
|
| 13:59 | danieel | i would wonder if i can pick any compression algo specified... and everybody would be able to read it
|
| 14:00 | troy_s | That can certainly be an issue. You should design to your post pipe and make sure everyone understands it.
|
| 14:00 | danieel | i am not designing the post pipe
|
| 14:00 | troy_s | If all you have is a bunch of ProRes twiddlers, then you have to design to that.
|
| 14:00 | danieel | i need to deliver the original r3d with a matching alpha/key stream
|
| 14:01 | troy_s | No, but I am saying you are somewhat constrained by the others.
|
| 14:01 | danieel | yes, that might be the case as one of the options for format was QT :) hard to explain that to some people, that it is not a codec, just a container
|
| 14:02 | troy_s | Oh yes. The joys of the cascading "Why!?!"
|
| 14:02 | danieel | so it seems we are back at the tiff sequence, simple to read by anybody
|
| 14:02 | danieel | who do exr would merge the r3d an the tiffs to their packed exr
|
| 14:03 | niemand | joined the channel |
| 14:03 | troy_s | danieel you will like this on "Why" http://youtu.be/wMFPe-DwULM
|
| 14:04 | troy_s | So is exr
|
| 14:04 | troy_s | The real issue is more complex
|
| 14:04 | troy_s | If you want a proper comp, you need linear radiometric light composites.
|
| 14:05 | troy_s | What are they comping in?
|
| 14:06 | troy_s | AE?
|
| 14:06 | danieel | we do not care really, we create and supply the alpha data
|
| 14:16 | troy_s | Well... Then go for the EXR
|
| 14:17 | troy_s | Also note that the alpha is inherently bound to the RGB in an associated format. Hard to go back and forth without care.
|
| 14:18 | troy_s | danieel: Very challenging for example, to go from associated values of RGBA of 1.2,0.3,0.0,0.0 back and forth. :)
|
| 14:19 | danieel | there is no going back, the data is real world capture
|
| 14:30 | troy_s | danieel: So did they ship you the footage and want the plates back?
|
| 14:32 | danieel | sort of yes
|
| 14:33 | Bertl | off to grab some groceries ... bbl
|
| 14:33 | Bertl | changed nick to: Bertl_oO
|
| 14:33 | troy_s | danieel: Complex shot or something?
|
| 14:34 | danieel | if semitransparencies rate for complex then definitely
|
| 14:35 | troy_s | LOL
|
| 14:44 | niemand | left the channel |
| 16:22 | niemand | joined the channel |
| 16:36 | se6astian|away | changed nick to: se6astian
|
| 16:42 | se6astian | good evening
|
| 16:43 | niemand | left the channel |
| 16:54 | niemand | joined the channel |
| 16:57 | davidak1 | joined the channel |
| 16:59 | davidak | left the channel |
| 17:14 | davidak1 | hello se6astian
|
| 17:14 | davidak1 | i really like your video :3
|
| 17:15 | davidak1 | it would be nice to see some random videos of different parts of the project.
|
| 17:16 | davidak1 | like cnc milling, firmware programming and testing, funny bugs etc
|
| 17:18 | Bertl_oO | developers eating a bowl of rice? :)
|
| 17:19 | se6astian | great you liked it :D
|
| 17:19 | se6astian | we will try to do more of this daily developer life in the future :)
|
| 17:19 | niemand | left the channel |
| 17:23 | Bertl_oO | davidak1: I'm currently eating my lunch :)
|
| 17:23 | se6astian | can I stop by and film you eating?
|
| 17:23 | se6astian | though cooking it would have been more interesting to film? :)
|
| 17:23 | Bertl_oO | sure, no problem
|
| 17:24 | Bertl_oO | but you have to be quick, it's almost gone
|
| 17:26 | se6astian | damn, last bite slow motion :D
|
| 17:38 | Bertl_oO | changed nick to: Bertl
|
| 17:49 | niemand | joined the channel |
| 18:04 | intracube | +1 for an unedited video of the first Beta case being milled :)
|
| 18:30 | niemand | left the channel |
| 18:37 | Bertl | off for a nap ... bbl
|
| 18:37 | Bertl | changed nick to: Bertl_zZ
|
| 19:02 | niemand | joined the channel |
| 19:35 | aombk | joined the channel |
| 19:36 | aombk | greetings from liberated greece :P
|
| 19:40 | niemand | left the channel |
| 19:47 | se6astian | hello!
|
| 19:47 | se6astian | I heard you will hire all fired state officers back
|
| 19:48 | niemand | joined the channel |
| 19:53 | aombk | all unconstitutionally fired state officers yes
|
| 20:03 | aombk | among other things. you disagree?
|
| 20:06 | alexML_ | left the channel |
| 20:07 | alexML | joined the channel |
| 20:18 | designbybeck_ | left the channel |
| 20:26 | designbybeck_ | joined the channel |
| 20:48 | troy_s | I give up.
|
| 21:15 | lab-bot | BAndiT1983 committed rOPENCINE5d4c48680f5e: - Prototype: WIP, Commiting for backup (authored by BAndiT1983).
|
| 21:21 | troy_s | slikdigit: Good god. That was pathetic.
|
| 21:26 | aombk | what are you talking about
|
| 21:37 | Bertl_zZ | changed nick to: Bertl
|
| 21:37 | Bertl | back now ...
|
| 21:41 | niemand | left the channel |
| 21:43 | se6astian | time for bed
|
| 21:49 | se6astian | changed nick to: se6astian|away
|
| 21:53 | Bertl | good night!
|
| 21:56 | slikdigit | troy_s: what was pathetic?
|
| 22:02 | troy_s | slikdigit: the discussion over in PiTiVi. God.
|