[Chris Brett]

Hi Kevin

— much appreciated — yes — I’m sure you are right without even checking — seems pretty obvious now you say it — 1.46 is PAL anamorphic right ?

— all the best ——– Chris Brett

========================================

[Chris Brett]

Thanks Guys — much appreciated — chris —

[Chris Brett]

— yes I’m very much with Ted on that one — never throw any picture information away until you really have to ….

— all the best ——-chris

===================================

[Chris Brett]

… obviously I dont get the upload piccy thing — Ill have another try ……..

—————–c—

[Chris Brett]

I reckon the best way, if possible, is to use live action footage as a displacement map on your comp — add little floaters / leaves whatever on top and animate to fit in and sell the shot …. this can look very convincing.

Use of this technique can be seen as per attached

chris brett // uk

[Chris Brett]

Stuart

—- I know this is stating the obvious but if its easy to test this way I would be inclined to regrade another test section with the same technical settings but much less colour adjustment, say just lower the red slightly or some other minor tweak and see if this problem goes away — also work in the highest bpc available.

If the problem goes away it may be that too much picture info has been taken out to cope with the background.

If the problem persists check the technical specs.

—— chris brett // uk

===========================

[Chris Brett]

Hi Erik

Another option might be as below

1 ) make 2 copies of the backround layer so you have 3 in total

2 ) Comp the lettering over the background on the bottom layer …. this will give you lettering in front of the trees which might sound like I’ve missed the point but there is method here …

2 ) Key the trees ( etc ) in the middle layer over the bottom layer using the top layer as a track matte – adjusting the luminance levels on the track matte using ‘curves’ to control the opacity and spread of the key.

The advantage of this method is that you can use it where its hard to get a good key and also in comps where you want to create subtle effects.

Sometimes with trees and its useful to have a bit of random transparency detail as it helps to give the idea that subject is traveling behind and that helps sell the shot.

If you want to try this its sometimes worth having a few copies of the keyed trees ( or whatever ) and set the levels differently to catch various parts of the image – for example you might want one key a bit ‘soft’ and transparent for blurry details out of focus leaves etc and another layer on top with a more general ‘tighter’ key possibly followed by another layer with a much harder ‘crushed’ black and white key to make sure main main branches trunk etc are good and solid without ‘holes’ in them.

This technique can be very useful if your subject is moving fast ( ish ) and the trees are a bit fuzzy and hard to key — sometimes you can get away with loads using this method…..

I’m sure Daves technique will work just fine but this one is worth knowing for those awkward shots.

chris brett // uk

===========================================

[Chris Brett]

Hi Andrew

— heres a few things to consider…

( Ive rotobrushed the face of the subject (so I have just his face) and then Ive rotobrushed out the face of the climber )

No need to roto out the face of the climber unless you are going to patch some background over it to remove it before applying new face — which in some cases is useful as there is more scope

( I am having some problems with motion tracking. Because the subjects in both videos are moving, )

it sounds like you are having problems because the movement of the 2 shots is different … to a certain extent they have to match and the nearer the better.
However if they roughly match but not quite good enough it might work better if you take a still of the new face just before ho blinks – patch that on forst then just patch on the few frames where he blinks as another layer.
In fact it will probably be better to do this anyway as the poster is effectively a ‘still’.
In fact I might be tempted to get a still from the background ( poster ) and a still of the new face into PS and get them to work there first as a singlr frame comp then it should be easy…….

( Ive used motion tracking to create a motion path for the face of my subject and targeted the path to a null object. Ive done the same with the body of the climber in the poster – also targeting the motion to a null object. Is this the right way to do it? )

Newish to Ae and not done this one myself but pretty sure the way to do it is …. track the poster / apply data to null / then pick-whip new face to null

Anyway theres loads of good tutorials around that can tell you much better than I .. checkout Video-copilot / Video 2 Brain / and Andrew Devis on the Creative Cow site

………… it should be fairly easy when you know how

Chris Brett /// uk

==========================

[Chris Brett]

Hi Gerard

…… not entirely sure if I get it but I reckon you need to get your cycle to run on ‘threes’.

If the anim has 8 drawings per second and you want it to look ok in a 25 fps setup it makes sense ( to me ) to multiply each position in the cycle by 3 ie 3 copies of first drawing then 3 copies of second etc etc….

…. then that way ( although its not an exact match ) the number of video frames to complete the cycle is roughly the same as the frame rate of the comp assuming your comp is set to 25 fps.

chris brett // uk

=====================

[Chris Brett]

Hi Mark

…. at the risk of stating the obvious….

If the encoded file looks large enough to make you think it may contain all the data in the first place then it might just be an import glitch in which case am wondering what happens if you use ‘interpret footage’ to set the frame rate to 25fps …. ?

If the file size is much smaller than you might expect then its probably an encoding issue.

— not a clear answer but if you havent checked it so far its a start………..

Chris Brett // uk