[Ben Rollason]

How about replacing the five videos with five nulls.

Then make a second set of five nulls and position them along a line.

Now drop your video in there and add two expression-layer sliders. Call them myLayer1 and myLayer2 or something. Set these for each video so that the first one references a null on the circle and the second one references a null in the line.

Now set an expression slider “crossblend”. This should be a value between 0 and 100. You can clamp it by adding the following expression. t = effect(“crossblend”)(“Slider”); t = (t>100?100:(t<0?0:t));

and put the following code in the position attribute...


myBlend = effect(“crossblend”)(“Slider”)/100;
l1 = effect(“myLayer1”)(“Layer”);
l2 = effect(“myLayer2”)(“Layer”);

P = ( l1.transform.position*myBlend + l2.transform.position*(1 -myBlend) );


Ben.

this is the average between the two points.

http://www.benrollason.com

[Ben Rollason]

If I understand you correctly, you’re trying to get the video layers to concertina out. They start stacked and they fold out as though they were one long strip?

One method is to stack all of your layers on top of one another, then change their anchorpoints to alternating sides. The first layer would have its anchor point on the far left for instance, the second on the far right, the third on the far left again and so on. The position will need to be adjusted after this so that the layers sit on top of one another again.

The anchor point represents the hinge where the layers attach to one another.

Next, starting at the top layer, make it the child of the layer beneath and so on right down to the bottom.

Put the following code into the yRotation property of each layer except the bottom one.


thisLayer.parent.transform.yRotation *-1;


Now when you rotate the lowest layer, all the layers above will rotate too, but in alternating directions and as the layers are parented to one another, the stack will appear to fold out.

Ben.

http://www.benrollason.com

[Ben Rollason]

That was just the ticket.

Thanks for your help.

Ben.

http://www.benrollason.com

[Ben Rollason]

Create an expression slider, call it “wiggle amount”, then paste this into the rotation property…

transform.rotation + noise(time*30)*effect(“wiggle amount”)(“Slider”);

or this might also do the trick…

transform.rotation + random(effect(“wiggle amount”)(“Slider”)) – effect(“wiggle amount”)(“Slider”)/2;

or even..

transform.rotation + Math.sin(time*80)*effect(“wiggle amount”)(“Slider”);

it depends on the effect you want. For the dart effect, make sure you move the layer’s anchor point right down to the end that hits the ‘board’.

Hope that helps.

http://www.benrollason.com

[Ben Rollason]

I don’t think you can pull the mask point values straight out of a mask path. I’ve never heard of it anyway.

What you can do though is copy the mask path, paste it into the position attribute of a null layer (or any layer with visibility turned off). It doesn’t matter where the keyframes fall in time.

Then add an expression slider to the layer you want to move along the path. I called my slider “percentage”, then paste the following code into the position attribute of the same layer.


tPath = thisComp.layer("null").transform.position; //change this to the null layer you just created!
tLastFrame = tPath.key(tPath.numKeys).time;//get the time of the last keyframe
tFirstFrame = tPath.key(1).time; //get the time of the first keyframe

tPercentage = effect("percentage")("Slider")/100; //read in the value from the percentage slider
//the next line converts that percentage into a point in time between the first and last keyframe.
tPercentageAsATime = (tPercentage*(tLastFrame-tFirstFrame)) + tFirstFrame;

myPos = tPath.valueAtTime(tPercentageAsATime); //read the position at that point in time and put it into this layer's current position. Hey presto!


Now you can have multiple layers all referencing the same path, but moving along it at different times, courtesy of their percentage slider.

Hope that helps.

-Ben.

http://www.benrollason.com

[Ben Rollason]

Thanks Adam. That sounds like a great idea. I’ll give it a go.

http://www.benrollason.com

[Ben Rollason]

nVidia 7800 GT here also. That could be it.

Ben.

Fink
https://www.fink-base.com

[Ben Rollason]

This is correct. Widescreen images are anamorphic (they use a non-square pixel). In fact, even 4:3 TV images are slightly anamorphic too. 720×480 is an NTSC DV size.

It all seems a bit weird, but it is normal. When you play it on a TV it will look fine. If you are using Quicktime, you can use the movie properties dialog box to adjust the display size to account for the non-square pixels on a square pixelled (computer) screen.

In Quicktime (I think you need Pro) go to Window -> Show movie properties. When the dialog box comes up, click on the video track, click the visual settings button and then adjust the display size to 720 x 405. Then save your movie. It doesn’t do anything to it, other than change the way that it is displayed to be 16:9.

Hope that helps,

Ben.

Fink
https://www.fink-base.com

[Ben Rollason]

All of this is subject to the old adage that you can’t put detail back. It’s like trying to unblur an image. You can remove some of the noise, using noise filters, but it will usually simply serve to reveal the lack of underlying resolution in what you shot, making a kind of blurry, blocky mess.

The problem with underexposing and then ‘stretching’ the gain in post is that it pulls out the noise as well as the image. It’s effectively doing the same as the camera does to compensate low light. It’s all about the signal to noise ratio in what you shot and underexposing isn’t going to help that.

The best bet really is to shoot a less grainy image – which means you need more light. I know sometimes that’s simply impossible, but often a little ingenuity can help matters a great deal. Gather together whatever household lighting you can – Maybe change the bulb in the room from a 40 watt to a 100 or 150 watt. If this isn’t an option, you could try making a homemade reflector out of white card or tinfoil to reflect available light back onto your subject. You need to try and light your subject to the point where the camera doesn’t use any gain to expose it correctly. It’s the gain that causes the grain. Gain amplifies the image, but it amplifies the noise too and only more light can improve your signal to noise ratio.

Also, use the white balance feature of the camera – you’ll find it’s much easier to turn your image back to an orangey tungsten look in post (use an orange layer with the color blending mode and adjust the opacity of that layer) than it is to pull any colour out of an image with an orange tungsten cast.

As a post-production tweak, I’d recommend lifting the gamma, using the levels or curves effect. This alters the mid grey point, effectively stretching the shadows a little. It will make your image seem a bit brighter. A gentle colour grade often helps too.

The other alternative is to go with your strengths and use the graininess of what you shot as a feature.