Kevin Camp
Forum Replies Created
-
something like this may work for you:
endTime = 5 // end time in seconds;
if (time < endTime) { loopOut(); }else{ value; } you could link the endTime variable to a slider on a null, then link your other end time variables to the same slider, thus linking all your end time values together. so your first line might look something like this: endTime = thisComp.layer("Null 1").effect("Slider Control")("Slider"); the stop point in the loop will be abrupt, and it will 'jump' to the last keyframed value. this can be acounted for if needed, or avoided if anticipated in the loop keyframe timing. Kevin Camp Designer – KCPQ, KMYQ & KRCW -
i’m not too familiar with image ready…. but, what type of file did you export from image ready? if it was an amiated gif, i’m not sure how ae interprets gifs (i’ll see if i can test it), so that may be an issue.
can you export from image ready as an image sequence? if you can get a psd, png or tif sequence with alpha, ae can easily bring those in and use them as compiled footage.
other options may be to export a black and white matte from image ready to use as a track matte for the aimation in ae, or try to key the areas that need to be transparent.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
in addition to not needing to run under powerpc emulation you will also be able to utilize all your ram (rosetta had to restrict the ram available to apps under emulation) and better use your available processors (cs3 now has a multi-processor render feature, prior versions only had a few mp aware effects). note that prefered ram for multi-processing is 2gb per processing core, so you may want to purchase more ram when upgrading.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
that is a nice expression, thanks mike (and filip).
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i think the way you did it was the right way… and you did use an expression, wiggle()…
the reason i would do it that way, is you’ll have better control over how much ‘natural’ wiggle you want to re-introduce to the shot.
if you wanted to experiment some, you could try the smoother feature (window>smoother) on the keyframes that were generated by the tracker. note, that the smoother is destructive, it will alter the keyframes that were generated. but it would mostlikely reduce the amount of correction that the tracker did.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
this cow article from a few months ago reviewed a product from convergent design that could take the hdmi out from a hdv camera to capture signifiacantly better video than the recorded hdv signal.
it sounded like a good way to get good quality video from a less expensive camera. and since you’d be shooting in a studio, capturing directly to the computer shouldn’t be a huge issue… although depending on the compression you may need a fast disk array for hd.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
you would need to be more specific about how you want things to animate if you wanted help creating an expression or finding a script that would work…
you might also look at echospace, a plugin that is very good at duplicating layers and fairly easily creating complex arrays of those layers. it might be the easiest way to go.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
[bogiesan] “for those of us who appreciate what you have done and have absolutely no clue what any of this means”
i’ll try to explain it….
x = linear(index, 0, thisComp.numLayers, 0, thisComp.width);
multiplier = thisComp.layer(“Null 1”).effect(“Slider Control”)(“Slider”);
z = Math.abs(Math.sin(index * Math.PI / thisComp.numLayers)) * multiplier;
[x,value[1],z]the first line, x = linear(index, 0, thisComp.numLayers, 0, thisComp.width); is to establish the horizontal positioning of the layers. the linear() function allows a value (in this case the layer’s index) that falls between one set of values to be plotted based on another set of values. so it takes the layer’s index, looks at where that falls between 0 and the number of layers (thisComp.numLayers), then places the layer based on where it would be between 0 and the width of the screen (thisComp.width).
the second line, multiplier = thisComp.layer(“Null 1”).effect(“Slider Control”)(“Slider”); is to set a mulitplier that will be used to adjust the z position and link that to a slider value on a null. linking to a remote slider makes it easy to adjust the amount of z offset of all the layers on the fly. this value will be used in the next line.
the third line, z = Math.abs(Math.sin(index * Math.PI / thisComp.numLayers)) * multiplier; is to generate the z position of the layer. when you want curves, you’ll usually be looking at sine functions (Math.sin()). the position again needed to be based on the index of the layer but as it related to the number of layers (thisComp.numLayers), so that needed to be within the sine function. pi (Math.PI) was needed to adjust the frequency of the wave (to make the sine wave a single wave, rather than a bunch of waves). to use just the portion of the sine wave that was round (i didn’t want thesides to gradually recede) i needed to separate just the positive values of the sin curve (the top of the wave) by placing the entier sine function within an absolute value function (Math.abs()). finally, since the absolute value of the sine fuction returns a value between 0-1, the mulitplier is needed to increase that value to generate noticeable z postion offset.
the last part, [x,value[1],z] just brings the values for x and z into a 3 demensional position vector [x,y,z] where the y is just the current value of the layer (value[1].
for the second expression:
spacingAdjust = thisComp.layer(“Null 2”).effect(“Slider Control”)(“Slider”);
y = linear(index, 0, thisComp.numLayers, 0 – spacingAdjust – (height / 2), thisComp.height + spacingAdjust- (height / 2));
value + [0, y]the first line is just to set a valiable for fine tuning the x positioning of the layers and link them to a slider on a null to make it easy to adjust all layers from one adjustment, very similar to what was done in the first expression.
in the next line, y = linear(index, 0, thisComp.numLayers, 0 – spacingAdjust – (height / 2), thisComp.height + spacingAdjust- (height / 2)); the linear() function was used again to distribute the layers based on their index. similarly to the first expression it takes the layer index, looks at it as a value between 0 and the total number of layers, and then plots the position based on the comp’s height. i needed to make an adjustment for the layers height, that’s where the (height / 2) comes in. i hadn’t really needed this in the first expression, since the layer were pretty small, expanding the distribution range wasn’t that neccesary. but, when made the first comp, i had made it full size and i needed to make an adjustment to the range by a full layer height (half off the top and half off the bottom). in addition the adjustment (spacingAdjustment) needed to be factored in.
the final line, value + [0, y] just brings the new y value in and adds it to the current position of the layer.
i hope this helps people out… expression are really pretty cool, i just started getting into them about 6-8 months ago and really wish i had started several years ago… i had previously only used expressions for randomness (random() and wiggle()), and thought you’d have to be a programmer to do anything else. but since joining the cow last year, i’ve managed to take my 1982 6th grade applesoft basic programming skills and cow knowlege (with a few other sites, like dan ebbert’s motionscript.com and jjgifford.com) and really started having some fun with expressions.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
two ways… if you only need the image to appear once, the easy way is to trim the layer by sliding the left or right layer handles in the timeline to set the in and out points. you can find more info in the help file, search trim layer.
if it needs to appear more than once, you could use the method above and duplicate the layer, repositioning the layer where it needs to appear again. or you could use opacity settings to show and hide the layer. the easiest way to work this way is to set the keyframes to hold keyframes (these are keyframes that won’t allow interpolation of values between keys)…
to do this, create an opacity keyframe (whatever value you want it to start as), then go to animation>toggle hold keyframe. now when you add new keyframes they will be hold keyframes too.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i’m on a mac, but my pref is in users/[username]/library/peferences/adobe/after effects/[version]/
the name of the file is ‘Adobe After Effects 8.0 Prefs’
i hope that helps…
also, opengl preview rendering and nucleo preview rendering may not be compatible… that is you can use one or the other but not both. i know with cs3, multiprocessing cannot be used for previews if opengl is set to always on…
Kevin Camp
Designer – KCPQ, KMYQ & KRCW