View Full Version : A question about rip HDTV
wolfsoft
19th January 2005, 09:03
How can I get the correct frames from HDTV which is 1920*1088I?
Which method above is right?
SeparateFields()
SelectEven()
Decimate(5)
or
Telecide(order=1)
Decimate(5)
It's meaning that either we can only get the even field as a frame, or get two fields to form a frame? Whether the fields in HDTV are different with each other? We should keep its aspect as 1920*1088, or resize it as 960*540?
I have touch HDTV in only several days, please teach me about this. Thanks:)
Guest
19th January 2005, 09:14
HD is no different from all the other video. If the material is telecined film, use IVTC. If it is video, then deinterlace.
You need to do some research here. It's all been answered millions of times.
wolfsoft
20th January 2005, 00:27
I have found this thread in the whole forum. At least there are three ways to deal with 1080i.
1: SeparateFields, then get even field as single frame.
2: SeparateFields, then bob to get a frame.
3: Deal with as DVD.
That's why I ask here because I want to know which way is right for 1080i.
For neuron2's words it looks like that it should deal with as DVD, isn't it?
Richard Berg
20th January 2005, 02:56
#3 isn't fundamentally different from #1 or #2, since as neuron2 said, HD is just high-res video. You also need to consider #4, "IVTC"...but again as he said, it depends on your source.
wolfsoft
20th January 2005, 06:04
Maybe I should give a chart to explain my mean.
As HDTV it's frame formed by field as this:
frame A B C D E
fieldt at bt ct dt et
fieldb ab bb cb db eb
t=top,b=bottom
#1 I get the top field, then like this:
frame at bt ct dt et
Then resize the top field to 960x540.
#2 I get the top field and bob it, get the even frames, then it would be:
frame A B C D E
fieldt at bt ct dt et
fieldb at bt ct dt et
The frames is bobed by top field
#3 Of couse, in my means, "deal with as DVD" is IVTC.
Which method is correct?
Mug Funky
20th January 2005, 06:10
which one is correct depends on how much time you have. full IVTC is pretty slow on such massive sources (i'm on a 5 year old machine), so separatefields().selecteven().decimate() would be much faster and give a similar result, though you'll be stuck with the knowledge that you've thrown out half the resolution.
i'd be inclined to IVTC and decimate it.
i don't get what you mean by bobbing off the top field... bobbing is shorthand for full-rate deinterlacing, which uses both fields. surely separatefields.bob will just give you a crappy output for no gain?
wolfsoft
20th January 2005, 06:19
Originally posted by Mug Funky
which one is correct depends on how much time you have. full IVTC is pretty slow on such massive sources (i'm on a 5 year old machine), so separatefields().selecteven().decimate() would be much faster and give a similar result, though you'll be stuck with the knowledge that you've thrown out half the resolution.
i'd be inclined to IVTC and decimate it.
i don't get what you mean by bobbing off the top field... bobbing is shorthand for full-rate deinterlacing, which uses both fields. surely separatefields.bob will just give you a crappy output for no gain?
"Bob" just was the way I saw in forum. I don't mind how much time I should cost but want to know the best way to deal with HDTV. Thank you for giving my directions.
Everyone:thanks:
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.