API tools faq
paste
TheProfosist

YATTA 2011.02.27

TheProfosist
Mar 3rd, 2011
164
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 26.86 KB | None | 0 0
raw download clone embed print report
  1. 13:55:16] <~TheProfosist> no sorry your going to need that also but its not that
  2. [13:55:39] <@IppE> We'll get to that
  3. [13:55:58] <~TheProfosist> http://athome.pastebin.com/djEEDzye
  4. [13:55:59] <@IppE> So I just copypaste the analysis pass part and replace the tfm part on my script with it
  5. [13:55:59] ot4ku [[email protected]] has joined #@Home
  6. [13:56:08] <~TheProfosist> u need a script like that
  7. [13:56:22] <@IppE> As I expected
  8. [13:56:34] <~TheProfosist> load that into megui
  9. [13:56:54] <~TheProfosist> and click the queue analysis pass button
  10. [13:57:09] <~TheProfosist> basically what it does is play through the file
  11. [13:57:14] ot4ku [[email protected]] has quit IRC:
  12. [13:57:27] ot4ku [[email protected]] has joined #@Home
  13. [13:57:33] <@IppE> and does what the script tells it to, simple enough
  14. [13:57:49] <~TheProfosist> yea
  15. [13:58:05] <@IppE> I wonder what this does to the ED though
  16. [13:58:23] <@IppE> Its all jittery and the text has something that looks like interlacing
  17. [13:58:26] ot4ku [[email protected]] has quit IRC:
  18. [13:58:30] <@IppE> But the bg appears to be progressive
  19. [13:58:37] ot4ku [[email protected]] has joined #@Home
  20. [13:59:47] <~TheProfosist> once that done tell me cause ill send you anther script that will help in clueing us in on wether we can just decimate or if well need VFR
  21. [14:00:19] <@IppE> Why would it be VFR
  22. [14:00:21] ot4ku [[email protected]] has quit IRC:
  23. [14:00:32] ot4ku [[email protected]] has joined #@Home
  24. [14:00:47] ot4ku [[email protected]] has quit IRC:
  25. [14:00:58] ot4ku [[email protected]] has joined #@Home
  26. [14:01:20] <~TheProfosist> if it has some 29.970 segnment and 23.976 or if it can all be decimated to 23.976
  27. [14:01:57] <@Hale> <~TheProfosist> hey Hale u would know how to do a analysis pass outside MeGUI right <-- analysing what?
  28. [14:02:09] <@IppE> OP is telecined, hale confirmed this, episode itself seems to be progressive with pulldown flags (so its 23.976fps with frame doubling to get it at 29.970) and the ED is something
  29. [14:02:20] <@IppE> Or am I being silly again?
  30. [14:02:43] <@Hale> episode looked 30p progressive to me
  31. [14:02:47] <@Hale> I may be wrong tho
  32. [14:03:04] ot4ku [[email protected]] has quit IRC:
  33. [14:03:12] <~TheProfosist> so were going to have to use VFR then
  34. [14:03:15] ot4ku [[email protected]] has joined #@Home
  35. [14:03:25] <@IppE> well the leght didn't change with decimate but fuck
  36. [14:03:33] <@IppE> lets check for VFR
  37. [14:03:45] <@Hale> well, lemme see if I can find something with constant motion first to check for duplicates
  38. [14:04:01] ot4ku [[email protected]] has quit IRC:
  39. [14:04:11] ot4ku2 [[email protected]] has joined #@Home
  40. [14:04:44] ot4ku2 [[email protected]] has quit IRC:
  41. [14:04:54] <~TheProfosist> http://athome.pastebin.com/5ZNv7NVE
  42. [14:05:07] <@Hale> frame 20450 is a pan at 30p, but I'll try to find some actual animation too
  43. [14:05:41] <~TheProfosist> use that script you dont even have to let it run because as soon as i starts itll throw out what we need the timecodes file
  44. [14:05:48] <~TheProfosist> k thanks hale
  45. [14:05:57] <@Hale> don't use this crap <_>
  46. [14:06:49] <~TheProfosist> ok fine but i still use it to check as i am not exactly versed yet in YATTA and ive forgotten how to check for frame rate changes
  47. [14:06:53] <@Hale> 26868 - clouds moving in the background
  48. [14:07:34] <@IppE> Well you guys try to figure out what the fuck to do with it, I'll watch osu! mp world cup now :3
  49. [14:08:04] <~TheProfosist> would you like to explain to him YATTA? if its needed for this?
  50. [14:08:06] <@Hale> it's supposed to be your job >_>
  51. [14:08:13] <@IppE> I'm bad at this
  52. [14:08:22] <@Hale> then get better ;)
  53. [14:08:48] <@IppE> Gah
  54. [14:08:50] <@Hale> 34038 <-- ok.. there's some 24p stuff
  55. [14:09:18] <@Hale> in other words - you'll have to dig through the episode and examine sections yourself
  56. [14:09:32] <@IppE> Wonderful
  57. [14:09:33] <@Hale> pulldown flags lied
  58. [14:09:52] <@IppE> So how exactly do I tell the difference between a 24fps and 30fps on frame by frame basis
  59. [14:11:02] <~TheProfosist> ill brb Hale. Hale is far mor knowledgeable at this than i anyway. and Hale go ove YATTA again im going to forsure read the logs and use it on my next interlaced source.
  60. [14:11:57] ot4ku [[email protected]] has joined #@Home
  61. [14:20:44] ot4ku [[email protected]] has quit IRC: Quit: Reconnecting
  62. [14:20:55] ot4ku [[email protected]] has joined #@Home
  63. [14:21:39] <@Hale> oh, sorry - got distracted here for a second
  64. [14:22:39] <@Hale> <@IppE> So how exactly do I tell the difference between a 24fps and 30fps on frame by frame basis <-- well basically.. you check if you see a single duplicate in a row of five frames (better to check for 2 in ten tho since patterns can change in-between)
  65. [14:22:40] ot4ku [[email protected]] has quit IRC:
  66. [14:23:28] <@Hale> if the pulldown flags weren't broken, you'd see them as c-only sections, but well.. you're fucked in this case
  67. [14:23:52] ot4ku [[email protected]] has joined #@Home
  68. [14:24:01] <@Hale> assuming you'd use yatta, that is
  69. [14:24:06] <@IppE> can't I just srestore the whole thing to 23.976 ;_;
  70. [14:24:37] <@Hale> sure, but you can't expect help from me in that case >_>
  71. [14:24:50] <@IppE> What if I do a test
  72. [14:25:18] <@Hale> testing what? -.O
  73. [14:25:31] <@IppE> just run the fmt tdecimate thing for the whole video, if theres nothing stupid going on and everything syncs etc.
  74. [14:25:34] <@IppE> Isn't that ok?
  75. [14:25:45] <@junh1024> i just use TFM and then do a step by step in vdub.
  76. [14:26:35] <@IppE> Cause I don't see any duplicates with tfm and tdecimate in the script
  77. [14:27:05] <@Hale> I just pointed out 2 different sections .. oO
  78. [14:27:15] <@IppE> well I do but idk asdindfgnf
  79. [14:27:23] <@IppE> I thought it was the animation being slow
  80. [14:28:06] ot4ku [[email protected]] has quit IRC:
  81. [14:28:57] <@Hale> hold on, lemme check which episode I extracted here
  82. [14:29:46] <@Hale> that'd be ep3 from the second iso
  83. [14:29:50] ot4ku [[email protected]] has joined #@Home
  84. [14:30:15] <@Hale> extract, run through ymc, check sections I mentioned, come back with results
  85. [14:30:37] <@IppE> ymc'tte nani?
  86. [14:31:00] ot4ku [[email protected]] has quit IRC:
  87. [14:31:05] <@Hale> oh, right.. you probably don't even have that ^^'
  88. [14:31:55] ot4ku [[email protected]] has joined #@Home
  89. [14:32:20] <@Hale> http://ivtc.org/ <-- snatch recent yatta build, extract, fire up ymc.exe, define your avisynth plugin path, load d2v into the window and tell me when you got to that point
  90. [14:33:04] ot4ku [[email protected]] has quit IRC:
  91. [14:33:22] <@IppE> I'll be doing this to ep1 btw
  92. [14:34:07] <@Hale> well...
  93. [14:34:27] <@Hale> I'd need to find something similar for ep1 then <_<
  94. [14:34:33] <@IppE> I can find a few spots with continuous animation, just tell me what to do with them
  95. [14:34:36] <@Hale> mendou~
  96. [14:35:17] <~TheProfosist> dont you have to separate fields so that you can find the different framerates or something like that?
  97. [14:35:19] <@Hale> again, check for 1 dup in 5 / 2 dups in 10 frames
  98. [14:35:35] <@IppE> ok
  99. [14:35:36] ot4ku [[email protected]] has joined #@Home
  100. [14:35:44] <@IppE> So what do I do after loading the d2v in
  101. [14:35:46] ot4ku [[email protected]] has quit IRC:
  102. [14:35:46] ot4ku [[email protected]] has joined #@Home
  103. [14:36:00] <@Hale> uhm no, TheProfosist?
  104. [14:36:25] <@IppE> Why would you seperate filds on progressive content?
  105. [14:36:31] <@IppE> fields
  106. [14:37:57] <@Hale> you tick crop on chop off the border noise (configure), then tick tfm+telecinemetrics and configure to use either bottom field first, or top field first, depending on which one gives you c/n matches only, afterwards you tick scxvid and finally tdecimate
  107. [14:38:25] <@Hale> define your path to save, fire it up and wait for things to complete
  108. [14:38:41] ot4ku [[email protected]] has quit IRC:
  109. [14:41:25] <@IppE> I don't see a scxvid here
  110. [14:42:50] <@Hale> http://athomeencoding.com/team/index.php?dir=TheProfosist/
  111. [14:43:03] <~TheProfosist> sorry i was thinking of finding the pulldown which shouldnt be needed
  112. [14:43:23] <@Hale> http://athomeencoding.com/team/TheProfosist/
  113. [14:43:31] <@IppE> And I dump this in avisynth plugins folder or?
  114. [14:43:39] <@Hale> yatta dir iirc
  115. [14:43:53] <@IppE> k
  116. [14:44:07] ot4ku [[email protected]] has joined #@Home
  117. [14:44:34] <@IppE> Nope, still can't see it
  118. [14:44:55] <@Hale> did you restart ymc?
  119. [14:44:59] <@IppE> yes
  120. [14:45:12] <@IppE> put it in avisynth folder and it works now
  121. [14:45:18] <@Hale> alright
  122. [14:46:16] ot4ku [[email protected]] has quit IRC:
  123. [14:46:26] ot4ku [[email protected]] has joined #@Home
  124. [14:46:27] <@IppE> http://gyazo.com/781add78f363d5bd6b4b084116e88771.png
  125. [14:46:39] <@IppE> This k?
  126. [14:47:04] <@Hale> yeah. which field mode did you end up with?
  127. [14:47:11] <@IppE> BFF
  128. [14:47:13] <@Hale> k
  129. [14:47:31] <@Hale> the start and wait a while =\
  130. [14:47:32] <@IppE> Though the way I usually check field order told me its tff
  131. [14:47:33] ot4ku [[email protected]] has quit IRC:
  132. [14:47:35] ot4ku [[email protected]] has joined #@Home
  133. [14:47:42] <@IppE> But I guess this is different from it
  134. [14:48:11] <@Hale> well it doesn't even matter for the final product. just collecting information on the source here
  135. [14:48:25] <@IppE> mm
  136. [14:48:30] <@IppE> Just wondering
  137. [14:51:25] <@IppE> While we're waiting, about the x264 settings
  138. [14:51:47] <@IppE> Is it really necessary to force level 5.1
  139. [14:53:06] <~TheProfosist> yes
  140. [14:53:07] <@Hale> right, but it screws with users trying to play back stuff via hardware
  141. [14:53:43] <@IppE> Like me :3c
  142. [14:53:51] <@IppE> Though thats just for 1080p stuff
  143. [14:54:24] <@IppE> But with DVD stuff, if you leave out the forced level it'll probably end up being something else
  144. [14:54:33] <@IppE> Even with the 16 reframes
  145. [14:55:08] <@Hale> 4.0 iirc
  146. [14:55:11] <@IppE> Not that it matters much as you can skip level check with dxva decoding but imo forcing things for no reason is silly
  147. [14:55:31] <@Hale> you can skip level check with dxva decoding <-- oh, I didn't know about that one
  148. [14:55:38] <@IppE> yeah
  149. [14:55:55] <@Hale> altough it doesn't really matter either since resolutions most likely are too funky for some of the chipsets anyway
  150. [14:56:14] <@IppE> thats just nvidia being retarded
  151. [14:56:26] <@Hale> right, same situation tho
  152. [14:56:56] <~TheProfosist> just leave it in. ive thought of doing hardware/dxva compliant release but that wouldnt let us use the setting for best results. ive also considered doing 2 releases a quality one and a lower quality 720 one thats DXVA compliant but idk if i want to cater to those people
  153. [14:56:59] <@IppE> http://gyazo.com/5c0def5aca8a91404aafae7836f0632d.png
  154. [14:57:25] <@IppE> No no no, going for dxva compliancy would mean forced 4.1 level in most cases
  155. [14:57:48] <@IppE> I meant leaving it out completely, so it'll be compliant if it happens to end up like that
  156. [14:57:52] <~TheProfosist> yea basically a BD profile maybe a little better
  157. [14:57:52] <@IppE> Like with DVD stuff
  158. [14:57:54] <@Hale> uhm why? levels are donwward-compatible
  159. [14:58:11] <@Hale> any chip that can decode 4.1 also includes 4.0 and below
  160. [14:58:13] <~TheProfosist> where is that screenshot from by the way
  161. [14:58:31] <@IppE> mpc-hc video filter settings window
  162. [14:59:15] <~TheProfosist> ahh ok i use CoreAVC
  163. [14:59:28] <~TheProfosist> makes sence how i wouldnt have seen that window
  164. [14:59:36] <@IppE> Indeed
  165. [15:00:45] <~TheProfosist> u going to switch once they addded DXVA support?
  166. [15:00:56] <~TheProfosist> if that ever happenes it been promised for over a year
  167. [15:01:09] <@IppE> I'm comfy with the mpc-hc's internal filter
  168. [15:01:31] <@IppE> And if it fails I just switch over to ffmpeg-mt
  169. [15:01:48] <@IppE> If that fails to play it smoothly, I either re encode the video or just delete it :3c
  170. [15:02:07] <~TheProfosist> i use all external filters cause they produce better results madVR rocks as a renderer
  171. [15:02:22] <@IppE> I'm stuck with EVR Custom
  172. [15:02:27] <@IppE> But I'm ok with that
  173. [15:02:46] <~TheProfosist> u have a GPU right? why not use madVr
  174. [15:03:00] <@IppE> All computers these days have a GPU
  175. [15:03:03] <@IppE> Silly
  176. [15:05:02] <~TheProfosist> well a non onboard GPU even the sandybridge ones work i think though i havnt tested it
  177. [15:05:23] <@IppE> This calls for a DS quote
  178. [15:05:50] <@IppE> ""Intel" and "fast GPUs" flow together like a river of bricks."
  179. [15:06:13] <@IppE> But anyway, I read a while ago that madVR doesn't support DXVA
  180. [15:06:22] <~TheProfosist> nope
  181. [15:06:38] <~TheProfosist> cause DXVA goes directly through the GPU
  182. [15:07:19] <@IppE> Thus I won't be using madVR for now
  183. [15:08:07] <~TheProfosist> if coreAVC ever starts supportsing DXVA it might work out without an update to madVR dont quote me on this though
  184. [15:08:32] <@IppE> I'm happy with my dxva and evr custom
  185. [15:08:39] <@IppE> I see no real reason to change
  186. [15:10:00] <~TheProfosist> better results?
  187. [15:10:11] <@IppE> marginal to me
  188. [15:10:22] <~TheProfosist> true marginal
  189. [15:10:36] <@IppE> everything is marginal
  190. [15:10:59] <@IppE> Now I shall proceed to encode this AC3 track to HE-AACv2 and see if it sounds horrible
  191. [15:11:20] <@Hale> >implying it may not sound horrible to others
  192. [15:11:46] <@IppE> Well if it sounds obviously horrible
  193. [15:12:02] <@IppE> Horrible enough for me to notice it being horrible
  194. [15:12:05] <@IppE> etc.
  195. [15:12:08] <~TheProfosist> how the YATTA junk going?
  196. [15:12:39] <@IppE> http://gyazo.com/6303fff0e1294ee99db69a0805bea280.png
  197. [15:12:40] <~TheProfosist> *stuff not junk
  198. [15:12:42] <@IppE> aka slowly
  199. [15:12:54] <~TheProfosist> hmm ok
  200. [15:12:58] <@IppE> Mainly because I'm encoding a video atm
  201. [15:13:52] <@Hale> nah, it's single-treaded anyway
  202. [15:14:14] <@IppE> And theres a lot of threads to go
  203. [15:17:34] <~TheProfosist> i wish huffy made betyter use of threads im only useing 30% of my cpu
  204. [15:17:49] <~TheProfosist> im getting .43 FPS
  205. [15:18:09] <@IppE> I got 0.08fps on that nnedi3 upscale
  206. [15:18:13] <@IppE> deal with it
  207. [15:18:19] <@Hale> how about you trim() and encode multiple instances?
  208. [15:18:46] <~TheProfosist> would AVI synth MT help? i could go 64bit but not all the filters i use are supported
  209. [15:19:24] Hale feels like he's beeing ignored again
  210. [15:19:41] IppE pats Hale
  211. [15:19:44] <@IppE> There there
  212. [15:20:26] <~TheProfosist> i can do that?
  213. [15:20:37] <@IppE> I doubt he'll bite
  214. [15:20:39] <~TheProfosist> will i have to recombine the avi afterwards?
  215. [15:23:08] <~TheProfosist> i sorry hale ive never used the trim() command i know that may be amazing
  216. [15:24:13] <@IppE> Just a heads up, the 5.1 seems to be a proper one
  217. [15:25:14] <~TheProfosist> thats great to hear
  218. [15:25:37] <@IppE> Or at least theres clear differences between channels
  219. [15:25:58] <@IppE> And they seem to make sense too
  220. [15:26:14] <@junh1024> TheProfosist: would AVI synth MT help? < last time I tried, MT didn't speedup on most filters
  221. [15:26:16] <@Hale> now cross-check as to if the live downmix to stereo differs from the actual stereo mix
  222. [15:26:51] <@junh1024> there's more than one way to downmix, just like there;s more than one way to skin a cat
  223. [15:27:08] <@junh1024> ale: how about you trim() and encode multiple instances? < i think thats what THE_RYUU's parallell encoding script does.
  224. [15:27:50] <@Hale> sure; but if it differs vastly, there's no reason not to include both of em
  225. [15:30:00] <@IppE> Well I can't really do a comprehensive test based on what it sounds like, seeing as I only have a 2.0 setup
  226. [15:30:04] <@IppE> http://gyazo.com/cc5b604bbe3d087668ef63e39b346452.png
  227. [15:31:06] <@IppE> Theres a part with humming and footsteps, quite obvious where right
  228. [15:31:14] <@Hale> uhm.. I think you're missing the point here. I told you to check if the 2.0 mix sounds similar or vastly different to the 5.1mix downmixed on demand
  229. [15:31:29] <@IppE> They sound pretty much the same to me
  230. [15:31:57] <@IppE> the 5.1 has few parts where the direction of the sound is bit more obvious and realistic
  231. [15:33:52] <@IppE> But again thats just me and I'm not an audiophile or anything
  232. [15:36:10] <~TheProfosist> junh1024: do you have a link?
  233. [15:42:09] <~TheProfosist> hale do you know what junh was talking about?
  234. [15:42:36] <@Hale> yea
  235. [15:42:59] <@Hale> just some automated scripting to take care of this task
  236. [15:44:11] <~TheProfosist> mind linking me to it so that i can tak a look my google searches resulted in nothing
  237. [15:44:28] <@Hale> no idea, try doom9
  238. [15:46:02] <@IppE> http://forum.doom9.org/search.php?searchid=5586598&pp=25
  239. [15:46:07] <@IppE> Start searching
  240. [15:47:18] <~TheProfosist> would it be easier for you ton just tell me how to use trim() hale?
  241. [15:48:57] <@IppE> Trim(first frame to include, last frame to include)
  242. [15:49:11] <@IppE> For example, 500 frame clip, split it half
  243. [15:49:33] <@Hale> ^-- take in mind that frame 0 is the first one, not frame 1
  244. [15:49:40] <@IppE> yeah was going to say that
  245. [15:50:06] <@IppE> So its Trim(0, 249) and Trim(250,500) right
  246. [15:50:20] <@Hale> avisource("mysource1.avi")+avisource("mysource2.avi") <- for splicing
  247. [15:52:27] <~TheProfosist> so for 4 cores split into 4 sections or 6 like x264 does?
  248. [15:52:36] <~TheProfosist> ahh ok easir than i thought
  249. [15:53:01] <@junh1024> this one
  250. [15:53:02] <@junh1024> http://pastebin.com/7TV0CCAz
  251. [15:56:14] <~TheProfosist> hmm idk eve know what to do with that i may just stick with what Hale and IppE were saying
  252. [15:56:31] <@IppE> its a batch script I believe
  253. [15:56:40] <@IppE> So one would stick it in a .bat file
  254. [15:59:06] Shoutmon [[email protected]] is now known as Shoutmon[AFK]
  255. [16:00:30] <~TheProfosist> hmm and my guess trim goes right after you load the video right?
  256. [16:01:55] <@Hale> you're better off wrting a new script, loading your previous script via import("myscript.avs") and throw in the trim there
  257. [16:01:56] tp7 thought to say 'it depends' but it would be to obvious
  258. [16:02:23] <@Hale> ugly, but gets the job done
  259. [16:04:29] <@IppE> So the yatta thing got done
  260. [16:04:33] <@IppE> what now
  261. [16:04:34] <~TheProfosist> ok so new scrip would look like
  262. [16:04:34] <~TheProfosist> import("myscript.avs")
  263. [16:04:34] <~TheProfosist> trim(0,5800)
  264. [16:05:15] <~TheProfosist> brb
  265. [16:07:14] <@Hale> <@IppE> what now <-- well.. load it into yatta, save and backup, right click pattern guidance > make sure all three of them are ticked (since you know it's a vfr source, otherwise you'd untick c only) > click pattern guidance and see what happens
  266. [16:10:05] <@IppE> lots of "Matching impossible" and a few Matching failures
  267. [16:10:39] <@Hale> right, but ignore that window for now (don't close it tho).
  268. [16:10:54] <@Hale> take a look at the pattern section in the main window
  269. [16:11:27] <@Hale> see how it changed the matching to follow a specific pattern while striking out one out of five?
  270. [16:11:49] <@IppE> yeah
  271. [16:12:15] <@Hale> basically what it did was loop through each possible pattern combination for a given section and apply the one which came with the lowest metrics
  272. [16:12:29] <@Hale> now here's the deal with the error window
  273. [16:12:53] <@IppE> I found a interlaced frame in the episode itself btw
  274. [16:13:07] <@IppE> Is this YATTA's doings or something to be concerned of
  275. [16:13:18] <@Hale> each of the entries in there were secions which, for whatever reason, couldn't be recognized correctly and the algorithm fell back to tfm alone
  276. [16:13:42] <@Hale> well.. it's proof for the source beeing partly telecined
  277. [16:14:45] <@IppE> Why do the mastering people do this
  278. [16:14:54] <@Hale> saa
  279. [16:15:39] <@IppE> So now, wtf am I supposed to do with this
  280. [16:15:58] <@Hale> basically, what it tells you now, is that the opening is telecined, parts of the episodes are telecined (while others are 30p progressive) and the ending is.. well.. no idea
  281. [16:16:37] <@IppE> So I check which parts need which treatment?
  282. [16:16:41] <@Hale> for startes you'd go throug the entries in the error window and check why the algorithm couldn't figure out what's going on in a given section
  283. [16:16:51] <@Hale> right, but that comes later
  284. [16:17:05] <@Hale> or rather, you do that while you work yourself through the source
  285. [16:17:55] <@IppE> So, the first thing in the log is "Matching impossible (too short section): 35815
  286. [16:18:01] <@IppE> How did I do what
  287. [16:18:19] <@Hale> basically pg hickups for a reason. either the scenechange algorithm threw in a new section although the scene sitll runs (ctrl+q here to kick out the following section)
  288. [16:18:40] <@Hale> or the sc algo didn't insert a section although one was needed
  289. [16:19:00] <@Hale> or one scene's pattern overlaps with another one during a fade
  290. [16:19:28] <@Hale> in which case you can either be lucky and the match, or you're fucked and have to figure out where to split both patterns
  291. [16:19:33] <@Hale> *they
  292. [16:20:11] <@Hale> double-click that entry and see what's going on in this case
  293. [16:20:57] <@IppE> So I got into the first thing
  294. [16:21:05] <@Hale> right, what do you see?
  295. [16:21:11] <@IppE> A progressive frame
  296. [16:21:15] <@Hale> ^^''
  297. [16:21:17] <@IppE> previous and next ones are too
  298. [16:21:27] <@IppE> but beyond that is interlaced
  299. [16:21:31] <@IppE> at least one frame
  300. [16:21:35] <@Hale> ok, open up the tools window and tick the sections tab
  301. [16:22:08] <@Hale> now navigate forth and back in this section to figure out where it starts/ends
  302. [16:22:28] <@Hale> do the section marks match with what you see?
  303. [16:23:08] tehKitten [[email protected]] has quit IRC: Quit: G'Bye~ Byaa~
  304. [16:23:30] <@IppE> It starts with 2 progressive frames, then an interlaced one
  305. [16:23:34] <@IppE> Then more progressive
  306. [16:23:40] <@Hale> that's not what I mean
  307. [16:23:48] <@IppE> I don't know what you mean
  308. [16:23:49] <@IppE> tbh
  309. [16:24:17] <@Hale> is the section mark which defines the stat of the section on spot whith the actual start of the scene?
  310. [16:24:29] <@IppE> no
  311. [16:24:44] <@Hale> so the start point it missing?
  312. [16:25:08] <@IppE> theres a ton of sections on the same scene
  313. [16:25:25] <@Hale> ah, one of these
  314. [16:26:03] <@IppE> So basically I just have to make sure theres only one section for each scene or?
  315. [16:26:11] <@Hale> ok make sure that the section starts where the scene starts and then navigate forward kicking out all further sections in that scene by pressing ctrl+q
  316. [16:26:19] <@Hale> yes and no
  317. [16:26:34] <@Hale> in theory each scene should follow one and only one pattern
  318. [16:26:50] <@Hale> however, in the real world, sometimes the pattern changes for no apparen reason
  319. [16:26:56] <@Hale> apparent
  320. [16:27:34] <@IppE> And angle change counts as scene change or
  321. [16:28:05] <@Hale> assuming your section isn't fucked up, and you make sure there's only one section entry in it, ctrl+alt+g should fix said section by re-applying pg to this very section
  322. [16:28:27] <@Hale> depends.. if it's part of a different animation it may come with a new pattern
  323. [16:28:36] <@Hale> but usually it's still the same one
  324. [16:29:36] <@Hale> when you have a perfect source (for example one of the recent ts broadcasts), you'd run pg and have only one pattern throught the whole episode
  325. [16:29:53] <@Hale> older shows authored on dvd aren't that easy tho =(
  326. [16:30:05] <@IppE> http://gyazo.com/88afd6cbd32718a61d9b0dfd77b483df.png
  327. [16:30:15] <@IppE> So this would be scene changing trough blend?
  328. [16:30:21] <@Hale> exactly
  329. [16:31:08] <@Hale> add a section here and run pg anew ctrl+alt+g > should fix it and possibly change the last frame in the section to p
  330. [16:31:34] <@Hale> which makes sense, since it can't refence to the next frame
  331. [16:31:57] <@IppE> ok
  332. [16:32:04] <@Hale> get the idea?
  333. [16:32:21] <@IppE> So this http://gyazo.com/4db558d7b97440ded38a91c3cc62a244.png and this http://gyazo.com/50fff3b9f0146c746f2a72d1ff91b161.png count as same scene as its only a angle change?
  334. [16:32:56] <@Hale> uhm no, that'd also be two scenes
  335. [16:33:04] <@IppE> awesome
  336. [16:33:22] <@IppE> So I just go trough the list of errors and figure out wtf is really going on?
  337. [16:33:28] <@Hale> pretty much
  338. [16:33:38] <@IppE> I think I got it now
  339. [16:34:50] <@Hale> assuming the section is telecined it should apply a pulldown pattern and mark one out of five for decimate (the one which is striked out); assuming it "thinks" the section is progressive, it'll apply a c-only section
  340. [16:35:10] <@Hale> however: you're still to set no-decimates yourself
  341. [16:35:38] <@Hale> there's right-click > somwhere > c to vfr, tho
  342. [16:38:00] <~TheProfosist> sweet i got the trim thing working
  343. [16:38:33] <@Hale> you can also set the first few frame in a section manually by hitting s, and apply the correct pattern by calling crtl+t afterwards, but that's a different approach which I tend to avoid
  344. [16:38:37] <@IppE> Correct me if I'm wrong, but when I press play it plays untill it finds a part that it doesn't wtf to do with?
  345. [16:38:56] <@IppE> doesn't know*
  346. [16:39:20] <@Hale> not quite right.. it plays until the metrics go lunatic
  347. [16:39:43] <@Hale> which can either mean broken pattern or z0mg dct blocking
  348. [16:39:59] <@Hale> *dtc
  349. [16:41:37] <@Hale> now back to playing this vn
  350. [16:50:07] <@IppE> Is it possible to go trough the whole thing manually, setting sections at scenechanges, and have it figure out the patterns from that?
  351. [16:50:54] <@Hale> sure, but I doubt that'd be faster
  352. [16:51:34] <@IppE> Well it would be a lot simpler and I'd prefer that
  353. [16:51:55] <@IppE> As I said, I got time
  354. [16:52:20] <@IppE> Especially now that I got tuesday and wednesday a day off froms school so I don't have to idle there
  355. [16:52:28] <@Hale> s/would be faster/makes any sense whatsoever/
  356. [17:00:45] <@IppE> Anyways, the bg
  357. [17:00:58] <@IppE> I tried just applying nnedi3 to it and the text looked fine
  358. [17:01:09] <@Hale> what text?
  359. [17:01:11] <@IppE> the bg suffered a bit but not drastically
  360. [17:01:16] <@IppE> the credits
  361. [17:01:22] <@IppE> on eding
  362. [17:01:40] <@IppE> was supposed to say ed but apparently it turned to bg along the way
  363. [17:01:46] <@IppE> I should probably go to bed
  364. [17:01:57] <@Hale> either hunt down the old r1 discs, or apply 60i_to_24p
  365. [17:02:26] <@Hale> assuming it it 24, >_>
  366. [17:02:34] <@Hale> otherwise qtgmc, I guess?
  367. [17:02:46] <@IppE> Well the bg looks progressive
  368. [17:02:56] <@IppE> but then again so did the episode itself a while ago
  369. [17:03:07] <@Hale> progressive can be both, 24p or 30p.. derp
  370. [17:03:14] <@IppE> yes yes
Advertisement
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!