-
The need to render interlaced?
I understand how interlaced footage works and all that, but there’s just one thing I don’t understand.
Let’s pretend we set up a 30fps comp with nothing but stuff made in After Effects (shapes made by masks, text, etc) and we’re going to end up using it in Avid with some footage we shot or something like that. Obviously none of those AE elements were ever interlaced, so why the need to render it with fields? I always hear people saying that if you plan to import it into Avid or your final output is going to be back to tape then you should render it with fields, but I don’t undererstand why.
If you render it with fields in After Effects, doesn’t it just take each of the 30fps and split each frame into 2 fields (so that when those 2 fields are put together they make a perfect replica of a whole frame, as opposed to footage shot with in 30i where two fields put together are actually at two different points in time)?
But if you render it without fields in AE, when you import it into Avid, wouldn’t it do that exact same process of splitting each frame into 2 fields when laying it out to tape? So why would it be better to do it in AE when it’s just going to do it in Avid anyway (or when you render it out to a tape)?
I guess the whole point of this is, why the need to render with fields in AE in a situation like this?
-
21 Replies
-
When you render with fields, each resulting frame contains two fields. Each field represents a different moment in time, each one 1/60th of a second apart. (Half of 30fps frame rate). So the position of a moving object in field 1 is different to it’s position in field 2. The effective result is 60 positions in time in one second.
If you rendered non-interlaced frames and then split a resulting frame into two fields, both fields would represent the same moment in time, or 30 positions in one second.
Field rendered motion appears smoother.
-
But if you make all of your graphics in after effects, and you make it at 30fps, then where would it get those 60 different points in time? Wouldn’t it only have 30 points in time (since it’s 30fps) and just split each of those in half…therefore having field 1 and field 2 of each frame be the same point in time?
-
it will split the difference
-
Rendering to fields in AE doubles the frame rate of your comp; rendering alternating frames to fields in your outputted movie.
As for why your Avid guy doesn
-
I mean I always see people on here saying to render with fields if it’s going to go to video, which is what I don’t understand, because it will automatically get converted to fields during output anyway. That’s why I don’t see why it matters if you render it out of AE with fields or not.
Maybe I’m not explaining correctly. I understand that when you capture video at 30i, it actually captures 60 different points in time…each field is from a different “picture”. But when you’re creating stuff in AE (and not using any video) that you’re eventually going to put into avid, why render with fields? What is the benefit since rendering to fields will just create two seperate fields taken from the exact same frame instead of from a different “picture”? Because if you don’t field render, and you output to tape, won’t outputting to tape do that exact same process of taking each of the 30 AE frames and separating them into 2 fields?
-
I understand technically alright, but artistically I don’t understand why you would want to. Just like video, it’s too smooth. 30p to me looks natural whereas render fields looks too clean like an old video switcher.
Just my useless opinion.
Chris Smith
https://www.sugarfilmproduction.com -
“….because it will automatically get converted to fields during output anyway”
Actually, your NLE does all the heavy-lifting in this respect; converting you 30p to 60i behind the scenes for you. Visually you end up with a movie that appears to change it’s state 30 times a sec, but your actually looking at a 60i stream (unless your previewing on some high end gear)
“…when you capture video at 30i, it actually captures 60 different points in time.”
Well, no. You can capture footage at a native 30p with a resulting video file composed of solid progressive frames.
“…rendering to fields will just create two seperate fields taken from the exact same frame instead of from a different “picture”
AE is _litterally_ doubling your framerate when you select “render to fields” in your output dialog. Try rendering 2 different DV files from the same comp, one as 30p and the other as 60i, and drop them into your favorite NLE. Set your playback output to your NTSC monitor and check out the results.
“… won’t outputting to tape do that exact same process of taking each of the 30 AE frames and separating them into 2 fields?”
Depends on how your NLE handles it, some do a better job than others.
60i and 30p look quite different, and many clients want the “live-at-5” smooth motion 60i look… some may want the more “cinematic” 30 or 24p look, I’d be prepared to deal w) both.
-
I still don’t think it’s being understood what I’m asking.
Okay, In after effects I have just a circle that I created in AE that I’m going to have move around the screen. It’s a 30fps project. I’m going to export this, put it into avid, and export it onto a regular old 30i (or 60i, however you want to say it) tape. So I’m starting with 30 full frame pictures. Nothing more, nothing less.
Doing either of these will give the exact same result…
1. I render out of AE WITH fields. It takes my first frame and doubles it. It takes my second frame and doubles it. Etc. I now have 60 frames, with every two frames being the same. I take that into avid and output it onto a VHS tape or DV tape or something comparible. The tape now has 60 fields per second. But every 2 fields are the exact same (except one is upper and one is lower)…field 1 is derived from frame 1 and field 2 is derived from frame 2…but since frame 1 and 2 are the same, it’s like they were derived from the exact same frame.
2. I render out of AE WITHOUT fields. I import it into avid to export out onto VHS, DV, etc. That means that avid then has to make fields for me to put to tape. But the thing is, it does the exact same thing as AE. It will make field 1 from frame 1 and field 2 from frame 1. Again, now we have both fields being derived from the exact same picture, which is exactly what after effects does when they double the frame rate.
So as you see in both cases you end up with 60 total fields and in both cases the fields will end up exactly the same. In both cases the first “picture” is used to get both field 1 and field 2, “picture two” (frame 2) is used to get both field 3 and field 4, frame 3 is used to get both field 5 and 6, etc.
Again, remember that this isn’t camera footage I’m talking about…there was never footage that started with 60 different points in time. It was a 30 frame AE composition. So this is the case in which I ask why does it matter whether AE renders out the frames or Avid does it, because they both will end up outputting the exact same 60 fields…
-
[Sandman] “I mean I always see people on here saying to render with fields if it’s going to go to video, which is what I don’t understand, because it will automatically get converted to fields during output anyway. That’s why I don’t see why it matters if you render it out of AE with fields or not.”
Well, it depends on what you mean by “converted to fields during output”. If you animate something in AE then render to frames, it will be broadcast/displayed as fields on TV, but each field will be identical, giving the less-smooth progressive look.
However, if you animate something in AE then render to fields, it will also be broadcast/displayed as fields on TV, but each field will be different, giving the smoother interlaced look.
And not all of us render to fields when going to video. It’s an aesthetic choice. The technical choice is upper or lower first: one is right and the other is wrong, depending on your gear.
[Sandman] “… What is the benefit since rendering to fields will just create two seperate fields taken from the exact same frame instead of from a different “picture”?
That’s not true. If you actually animate something in AE, on rendering AE figures out (since it is interpolating, after all) what each field will look like and renders them differently. If you doubt this, drag a 29.97 frame-rendered and a 29.97 field-rendered animation into a 59.94 fps comp and see the difference.
[Sandman] “Because if you don’t field render, and you output to tape, won’t outputting to tape do that exact same process of taking each of the 30 AE frames and separating them into 2 fields?”
Not if you want the two fields to be different images. The outputting NLE won’t create different fields (60 fields interlaced) where there were identical fields (30 frames progressive) beforehand.
Note that I’m talking about animations created in AE, where AE creates the motion.
When it comes to footage, say that has just been color-corrected, here are some notes:
– if the footage was 30p, you might only be able to create the different fields with a plugin such as Twixtor or FieldsKit. Only then can you render as fields, where each field is different.
– if the footage is 30p, both fields would be identical whether you rendered fields or frames … if you didn’t use one of the above plugins.
– if the footage is 60i, and you choose not to separate fields on import, you can render as frames and the footage will not change. It will still be 60i. Just don’t do any animation in the comp, because it would be 30p over 60i footage. Odd..
– if the footage is 60i, and you separate fields on import, if you render as frames, you’ll get slightly soft 30p. If you render as fields, you’ll get fields, but I’m not sure if the footage will be slightly soft.My apologies if I’ve repeated anything or left anything out.
-
I just read your last post …
[Sandman] “Doing either of these will give the exact same result…
1. I render out of AE WITH fields. It takes my first frame and doubles it. It takes my second frame and doubles it. Etc. I now have 60 frames, with every two frames being the same. I take that into avid and output it onto a VHS tape or DV tape or something comparible. The tape now has 60 fields per second. But every 2 fields are the exact same (except one is upper and one is lower)…field 1 is derived from frame 1 and field 2 is derived from frame 2…but since frame 1 and 2 are the same, it’s like they were derived from the exact same frame.
No, sorry. It renders line 1 of 30p imported footage, then line 2 of 30p imported footage and so on … but since those lines are identical points in time, the footage still appears 30p. But anything actually animated in AE will be rendered with line 1 different from line 2 (and so on), since AE has the capability of interpolating its animation to create images between frames, that is, it creates new images for the fields.
[Sandman] “2. I render out of AE WITHOUT fields. I import it into avid to export out onto VHS, DV, etc. That means that avid then has to make fields for me to put to tape. But the thing is, it does the exact same thing as AE. It will make field 1 from frame 1 and field 2 from frame 1. Again, now we have both fields being derived from the exact same picture, which is exactly what after effects does when they double the frame rate.”
Avid does not make fields, it splits a frame into even and odd slices, then plays one before the other depending on its preferred field order. Both fields will be derived from one frame, but if that frame is made up of an image where line 1 is slightly different (in time) from line 2 (as the animated circle would be if field-rendered), then you have 60i video. If the animated circle were frame-rendered, then both fields will be identical in time.
When you animate anything in AE and render to fields, it actually interpolates motion between the frames of a comp. If something moves x=10 pixels per frame, it renders (assuming upper field first for example) line 1/3/5 etc. at x=0 pixels, then line 2/4/6 etc. at x=5 pixels, then line 1/3/5 at 10 pixels, and so on, interpolating the motion as 5 pixels per field.
Then your NLE takes this comb-like frame and splits it up in time.
(Note that when I wrote “identical” before, I meant identical in time.)
Does that help?
Reply to this Discussion! Login or Sign Up