1080i broadcast film --> 1280x720 @ 23.976fps

[txporter]

I am considering resizing my 1080i captures to 1280x720 to save space.

Two questions:

1) is there a good way to estimate/measure what I will be losing by resizing? My main TV is a Samsung HL-S5087W. I assume that whatever I feed it will be upscaled to 1080p. So I guess I will now effectively being going 1080i-->720p-->1080p.

2) If I decide to go ahead with resize to 1280x720, when should I resize in my avisynth script? This is what I was planning:

MPEG2Source("test.d2v",cpu=3)
TFM().TDecimate()
Undot()
Bicubicresize(1280, 720)

[poisondeathray]

No good way because there are too many variables involved. e.g. quality of player or tv upscaling algorithm, encoding settings, bitrate used, source quality etc. etc...

You have to go through the process from start to finish and play it to be sure

If it was a TV cap, chances are it doesn't have that much detail to warrant 1080p

[txporter]

No good way because there are too many variables involved. e.g. quality of player or tv upscaling algorithm, encoding settings, bitrate used, source quality etc. etc...

You have to go through the process from start to finish and play it to be sure

If it was a TV cap, chances are it doesn't have that much detail to warrant 1080p

So basically, if it looks good to me....that's really all that matters. Ok, got it.

What about when to resize? Re-encode is faster if I resize before TIVTC, but I wasn't sure if that was recommended or not.

[poisondeathray]

you can't do a straight resize before tivtc , fields will get mucked up . It's similar as trying to do interlaced resize. It has to be an interlace aware method