Interlacing de-interlaced animation?
 

Hello all,
I am in the process of developing a childrens music program for a client. I decided to go with 3D animation/ live action. The biggest problem I am facing now, is creating the animation in an interlaced format. It can take hours or days, just to render the animation in a normal frame mode. To render it in interlaced format at least doubles or triples that time.

So I say to myself, "Self- what if I create it in de-interlaced frame format, and then export it to my GL1? When I capture it back from the cam, would it be properly interlaced?" What do you think?
Keith

You know, Keith, this is a good question... I think you should try it one time and tell us how it looks!

Ok... I tried it, and it looks fair on the monitor, except for the jitters and flickers. I had to try it because it is taking almost 10 to 18 minutes a frame to render in interlaced format. I still need to find a better way.
Keith

Only PAL DV has it's fields reversed (lower field first). Any decent
program should allow you to render that. I know Lightwave 3D
does. Did you know you can just mix interlaced and non-interlaced
footage? To the camera it is all the same. Non-interlaced footage
looks fine on my interlaced tv.

Try it.

p.s.
DV stores all of its information in one frame. Whether
it is interlaced or not. So, it does not matter which it
is (for the camera).

Rob,
I'm using Infini-D, and it allows me to render odd or even fields. It also has the option for fast second field and reduced flicker. The problem is, that it is taking anywhere from 10 to 18 minutes per frame to render it Interlaced. I'd like to finish this BEFORE I retire.

Someone else suggested that I render out 60 fps, and take it into After Effects to interlace it. I don't have it. That is why I was asking about exporting to the cam, and re-importing it. It was ok, but not good enough.
Keith

This is being targeted for Broadcast. Thus, the reason for interlaced, broadcst safe, and all the other stuff. It was much easier when I was editing for the web :)
Keith

They cannot tell whether your material will be interlaced or not?
Why, because when they either export from your DV tape it will
be read interlaced or when they use another output it will be
interlaced too! You should have no problems!

Hmmm... Are you sure about that?
Keith

I am pretty sure.... consider it this way. If I output non-interlaced
(ie, progressive) material to tape and I then go and attach my
XL1s to my TV, what will happen? Since my TV only understands
interlaced the camera will make sure it is interlaced. Let us define
interlaced for a moment here:

interlaced is the definition that half of the image belongs to one
field and the other half to another field. Each field is displayed
one after another at 50 or 60 fields per second. Now DV stores
this image just as we would a progressive image. Thus the first
line in the image is from one field and the second line from the
other field! THIS IS IMPORTANT. DV basically treats the image
the same, whether it is from interlaced on non-interlaced source.
The only difference is in a little flag in the stream that indicates
which of the two it is!!

So what should happen if you play a progressive/non-interlaced
image with your camera? Well.... the camera will just output
non-interlaced. So it will send the first field of your frame (even
though in theory it has no different fields) in the first 1/30th or
1/25th of a second and the second field in the other 1/30th or
1/25th of a second. This should not produce any "interlacing"
artifacts which we know from viewing interlaced material on
the computer screen because there IS NO time difference in
the original recording (it was non-interlaced, remember?). Your
TV will display it apart in time, but you should not be able to see
this.

Try it out. I have been shooting progressive/non-interlaced
(or frame mode as they call it on the XL1s) for 6 months now and
it looks perfect on interlaced TVs.... Again, the ONLY difference
between interlaced and non-interlaced is the TIME DIFFERENCE
between the fields when acquired. Your camera (or whatever
device is playing, converting or exporting your DV stream) is
responsible for producing a valid TV COMPATIBLE SIGNAL! Not
your original recording. DV streams only store the color of the
pixel and where it is located (next to audio and some other
information). It does not store timing signals and other broadcast
stuff. That is the responsibility of the reproducing device!

But the best way is always to try and see it for yourself. Render
a minute animation in progressive/non-interlaced mode, output
that to your camera and play it on a regular TV.... should look
perfect!

I hope I explained it better to you now.