I rented it. (I know, it's bad... :oops:) :lol:
But I can tell it was field shifted (if I can trust the small function Incredible provide once to "analyse" the stream - Interlacing2Reader). |
If it's field shifted, TFM() should be used instead as it'll be a lot faster than using TDeint with tryweave=true. TFM can be found in the TIVTC package. Chances are that you could even use TFM on progressive material without any damage.
|
Good to know. I will use it if the next box set is the same. Thanks !
|
Quote:
BTW I see you needed 3 hrs per episode to encode at half resolution without the slowest part of the script, so maybe the speed I got it's ok: I need 4 hrs but at full resolution and with the "scriptclip" line (I found full resolution increase by almost 100% the encoding time, is it normal?)... I thought there was a way to encode ALMOST in real time (about 1 hrs for 40 mins episode), but maybe it's impossible with all the filters my script is using. Do you think I can have some boost with the Mt.dll filter at the top of the script? |
Quote:
Quote:
Quote:
|
Quote:
You could try wrapping LRemoveDust and LimitedSharpenFaster inside MT like this: Code:
dgdecode_mpeg2source("E:\Lavoro KVCD\lost.d2v", cpu=4, idct=7) |
Thanks you both guys, I'll try your suggestions right now :D
EDIT: Mt didn't speed up the encoding, neither did adding the crop line to the function. But adding the converttoYUY2().converttoYV12().LRemoveDust_YV12(1 7,1) line (I read it in the thread about LS) did result in a boost. So I'm having some other tries before comparing screenshots |
Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.