1 Make a DCP with subs using subtitle edit.
3 Look at http://liblqr.wikidot.com/en:manual
7 Small instance $0.085 ph
8 Sintel Trailer 1080p @ 200000 Mbps
9 1247 frames @ 24fps ie 51.96s
12 High-CPU medium $0.186 ph
13 Sintel Trailer 1080p @ 200000 Mbps
14 1247 frames @ 24fps ie 51.96s
17 High-CPU extra-large $0.744 ph
18 Sintel Trailer 1080p @ 200000 Mbps
19 1247 frames @ 24fps ie 51.96s
23 Transfer out $0.120 per GB
28 Write still j2ks straight to a MXF.
29 md5_data to use openssl
30 Write all j2ks straight to a MXF? Possible?
33 Format name in ~/.dvdomatic screws up with spaces; use ID or something
34 Thumbnails are poorly named
36 Restartable jobs somehow
38 Nice error when trying to thumbnail with no content.
39 Destroy _buffer_src_context / _buffer_sink_context
40 Don't start later jobs when one breaks.
41 Compute time remaining based on more recent information.
56 Screen 1: "1.37" masking preset, projector only has DCI 133 preset.
58 With 1480x1080 alignment in DCI 133: bottom you see purple, yellow; top purple; left and right no lines
59 With 1480x1080 alignment in DCI Flat: outside masks, but you see bottom purple, yellow; left/right all; top purple
62 Screen 2: no real masking preset, projector has DCI 133 and DCI 137
76 Screen 3: projector has DCI 1.38
83 films-0.6: Dolby Countdown looks as though it's 3D. THX Terminator 2 fucked
84 (these on default settings)
85 fq/gradfun --- no obvious effect
86 hqdn3d --- pretty good denoising
87 ow --- no obvious effect
88 tn --- interesting; much noise reduction, bad artefacts on movement, colour tint even in black
91 Benchmark SWS options: lanczos ?
92 hqdn3d=0:0:6 ? (turn off chroma/luma blurring)
94 Lanczos; no visible effect on Ghostbusters.
97 THX_Monster with master Intel Core 2 Duo E4600 (2.4GHz), slave Intel Core i3 M350 (2.27GHz)
98 1920 x 1080 original -> DCI Flat
101 [Gbit: gigabit ethernet rather than 100Mbit]
102 [im-mod: after modification to memcpy RGB data then to RGB -> XYZ in the encode thread
103 [hack1]: after modification to pass YUV and to swscale in the encode thread (includes im-mod)
104 [hack2]: modified hack1
105 Time Seconds FPS Speedup relative to 1 local
106 1 local: 20m57 1257 0.19 x 1
107 2 local: 11m24 684 0.35 x 1.84
108 2 local [im-mod]: 13m13
109 2 local + 1 slave: 6m34 394 0.61 x 3.19
110 2 local + 2 slave: 5m13 313 0.77 x 4.02
111 2 local + 4 slave: 5m05 303 0.79 x 4.15
112 2 local + 4 slave [Gbit]: 2m50 170 1.41 x 7.39
113 2 local + 4 slave [Gbit,im-mod]:2m33
114 2 local + 4 slave [Gbit,hack1]: 3m20
115 2 local + 4 slave [Gbit,hack2]: 2m22
116 1 local + 8 slave [Gbit]: 2m28 148 1.62 x 8.49
117 2 local + 8 slave [Gbit]: 2m41 161 1.49 x 7.81
118 2 local + 8 slave [Gbit,im-mod]:2m35
123 Encode + Image create 1m27
124 Encode + Image create (memcpy, not convert) 53s.
126 THX_Monster with master Intel Core i3 M350 (2.27GHz), slave Intel Core 2 Duo E4600 (2.4GHz)
127 1920 x 1080 original -> DCI Flat
132 4 local [im-mod]: 2m53
133 4 local + 2 slave [Gbit]: 2m22
134 4 local + 4 slave [Gbit]: 2m21
135 4 local + 4 slave [Gbit,in-mod]:2m21