[Filip Vandueren]

Just a quick reminder that you can press the period on the numeric keypad to do an audio-only Ram-preview.

[Filip Vandueren]

Expressions can’t help you in limiting the effect of a plugin to a certain region.
you’ll need to pre-comp.

But to speed things up: Perhaps you can use some kind of 3d-proxy to animate (for example a normal text-layer with the same dimensions)
It won’t have any depth but for rotating and positioning, it should suffice.

Then use simple pick-whip expressions, or even copy-pasting of keyframes, to link it’s rotation, scale, positions etc. to the Invig-object, and hide the text.

Pickwhipping works perfect across comps, just open up the two timelines side by side.

I sometimes do the same with a null-object with sliders that controls a plug-in in a sub-comp.

[Filip Vandueren]

Expressions can’t help you in limiting the effect of a plugin to a certain region.
you’ll need to pre-comp.

But to speed things up: Perhaps you can use some kind of 3d-proxy to animate (for example a normal text-layer with the same dimensions)
It won’t have any depth but for rotating and positioning, it should suffice.

Then use simple pick-whip expressions, or even copy-pasting of keyframes, to link it’s rotation, scale, positions etc. to the Invig-object, and hide the text.

Pickwhipping works perfect across comps, just open up the two timelines side by side.

I sometimes do the same with a null-object with sliders that controls a plug-in in a sub-comp.

[Filip Vandueren]

Decided to have some fun:

here’s the gist of my idea for stereo, with even some doppler added 😎

the audio is a motor at constant speed,
the distance from “+” to L and from “+” to R is calculated and used to lower the volumes and is also added to the time-remapping.
It’s factored by a “speed of sound”, so you get accurate delays and hence automatically a doppler if it moves.

for 5.1, you would do the exact same thing, place some more ‘mics’.
Perhaps use a “layer”-expression-control on a null-layer, to select which mike you want to hear.
Let all the audio-expressions reference that layer-selector, and you get a more elegant solution.

project-files are here:

https://www.vandueren.be/forumstuff/doppler/

BTW, this works in 3D.

[Filip Vandueren]

Decided to have some fun:

here’s the gist of my idea for stereo, with even some doppler added 😎

the audio is a motor at constant speed,
the distance from “+” to L and from “+” to R is calculated and used to lower the volumes and is also added to the time-remapping.
It’s factored by a “speed of sound”, so you get accurate delays and hence automatically a doppler if it moves.

for 5.1, you would do the exact same thing, place some more ‘mics’.
Perhaps use a “layer”-expression-control on a null-layer, to select which mike you want to hear.
Let all the audio-expressions reference that layer-selector, and you get a more elegant solution.

project-files are here:

https://www.vandueren.be/forumstuff/doppler/

BTW, this works in 3D.

[Filip Vandueren]

Well, if you disregard stuff like the doppler effect and ambience, and just want a mixdown of the relative levels in relation to some virtual microphones, it wouldn’t be too hard to do right now.

Just duplicate all the layers that have sound and pre-comp them.
duplicate that pre-comp once for every speaker and then write a simple expression that calculates the distance from the layer to the position of the mic, subtract that value from the current dB-level.
Add in a damping-co

[Filip Vandueren]

Well, if you disregard stuff like the doppler effect and ambience, and just want a mixdown of the relative levels in relation to some virtual microphones, it wouldn’t be too hard to do right now.

Just duplicate all the layers that have sound and pre-comp them.
duplicate that pre-comp once for every speaker and then write a simple expression that calculates the distance from the layer to the position of the mic, subtract that value from the current dB-level.
Add in a damping-co

[Filip Vandueren]

Well, if you disregard stuff like the doppler effect and ambience, and just want a mixdown of the relative levels in relation to some virtual microphones, it wouldn’t be too hard to do right now.

Just duplicate all the layers that have sound and pre-comp them.
duplicate that pre-comp once for every speaker and then write a simple expression that calculates the distance from the layer to the position of the mic, subtract that value from the current dB-level.
Add in a damping-co

[Filip Vandueren]

Well, if you disregard stuff like the doppler effect and ambience, and just want a mixdown of the relative levels in relation to some virtual microphones, it wouldn’t be too hard to do right now.

Just duplicate all the layers that have sound and pre-comp them.
duplicate that pre-comp once for every speaker and then write a simple expression that calculates the distance from the layer to the position of the mic, subtract that value from the current dB-level.
Add in a damping-co

[Filip Vandueren]

[Midnight Oil] “I was thinking about importing the images as a sequence, but I don’t know how to make it randomly jump around with dissolves between them. If I could do that, I could just start the sequence at different times on the grid.”

If that scenario is random enough for you, where you have 1 sequence that just is looped,

then why not just first make one or two comps with the 15 frame-dissolves, (takes about 5 seconds with the sequence layers command)
render them out, re-import and interpret the movie to loop 99 times.

Now you can drop those movies in your layers and with the “pan-behind”-tool you can choose your start-point.
If you don’t want to alter the startpoints manually, enable time-remapping and add a random static value to each layer’s current time.

—

An all-expression way could also use time-remapping:

set up a pre-comp with al your images sequenced.
make a new comp with two copies of that precomp.

let’s say you want to fade in a layer over 1 second, hold for one second, then fade to the next, hold for one second etc.

create some opacity keyframes for the top-layer to reflect that animation:0-100 in 1 sec, hold 1 sec, then 100-0 down, hold 1 sec, then put a final 0% keyframe.
add a loopOut(type = "cycle", numKeyframes = 0)-expression so the keyframes get repeated.

give the time-remapping this expression:


holdTime=4; //hold each frame 4 seconds;
holdOffset=0; // offset 0 seconds
maxTime=source.duration-source.frameDuration;
seed=Math.floor((time+(holdTime-holdOffset))/holdTime);
seedRandom(seed, timeless = true);
random(maxTime)


do the same for the bottom layer, but here, change the holdOffset-value to 1

copy that pair of layers for each square and scale/position them.

Keep in mind that being a true random sequence, you will get duplicate frames from time to time.