Mylenium
Forum Replies Created
-
Well, though this sounds harsh, I can only advise you to get more RAM. You will never be able to get beyond a certain degree of complexety and realism if you don’t give your programs breathing room. There’s no way around that. Picture this: If you map an 8000×8000 pixel image onto a “Sky sphere” with 1km diameter, each pixel only covers 3140m/8000 = 0,35m in the horizontal direction, which isn’t that much the closer you get to the sphere (and 1km diameter is a small sky sphere). I myself barely ever use textures below 4000 pixels wide for small, plana mapped objects and in many situations that is merely sufficient when you are getting close to the object. If you use lower res textures, then blur them to hide the original pixel structure, but this may not be feasible. Assuming you only need a sky reflection, it would work. In that case you may wish to add your actual sky later in compositing to circumvent certain problems when doing it all inside C4D.
Mylenium
[Pour Myl
-
Mmh, why not simply use more character value/ character offset animators? Define the alphabet ranges and the number ranges and then choose a suitable mode for the animator like “intersect”. This way you can get both. You’d simply have to animate the Start/ End or Offset to define on which characters there is an overlap between values.
Mylenium
[Pour Myl
-
No, you cannot use specific hardware related CoDecs with custom pixel sizes. Such CoDecs are always optimized specifically for the hardware to allow realtime playback throughout and only support standard resolutions (NTSC/ PAL DV in your case). To render your clip at its current size, you need to use a generic Co’Dec such as DivX or no compression at all.
Mylenium
[Pour Myl
-
Mmh, if you have shots of the shaleves you can simply mape them onto 3D layers in a sub-comp to form a box, then drag that composition in your main comp, enable “collapse transformations” and animate it with 3D properties intact. Otherwise I think that is more of a job for a 3D program – doesn’t need to be extrmely detailed, just some weapon dummys rushing in with lots of motion blur. The row closest to the camera most likely could be shot for real using toy/ replica guns.
Mylenium
[Pour Myl
-
As big as you can. ;o) 8000 pixels wide is not too much.
Mylenium
[Pour Myl
-
[Laszlo Kovacs] “Oops, should’ve posted this in the new expression forum?”
No problem. Your post will be relocated eventually. For the time, there is nothing wrong with posting here. Anyway, I’m not really sure what you are hoping to achieve. If, as you say, the pages are already designed and not random at all, there can be no such thing as using random animation. The only simplification I see is that you could automate the positional transitions by cross-checking the positions for each setup. You’d still have to define the timing somehow, be it just by using a slider to animate everything. You could achieve some further automation by checking for the in and outpoints. It might look something like:
pos_a=thisComp.layer(“Website a Element X”).position.valueAtTime(outPoint);
pos_b=thisComp.layer(“Website b Element Y”).position.valueAtTime(inPoint);move=linear(time,inPoint,OutPoint,pos_a,pos_b);
[move]
This does not figure in any layer space transforms to pin the elements in place on the website plates, so you will have to add toWorld() and toSurfacePoint() I haven’t wrapped my head around this entirely, so maybe someone else has some simpler ideas.
Mylenium
[Pour Myl
-
[Laszlo Kovacs] “Oops, should’ve posted this in the new expression forum?”
No problem. Your post will be relocated eventually. For the time, there is nothing wrong with posting here. Anyway, I’m not really sure what you are hoping to achieve. If, as you say, the pages are already designed and not random at all, there can be no such thing as using random animation. The only simplification I see is that you could automate the positional transitions by cross-checking the positions for each setup. You’d still have to define the timing somehow, be it just by using a slider to animate everything. You could achieve some further automation by checking for the in and outpoints. It might look something like:
pos_a=thisComp.layer(“Website a Element X”).position.valueAtTime(outPoint);
pos_b=thisComp.layer(“Website b Element Y”).position.valueAtTime(inPoint);move=linear(time,inPoint,OutPoint,pos_a,pos_b);
[move]
This does not figure in any layer space transforms to pin the elements in place on the website plates, so you will have to add toWorld() and toSurfacePoint() I haven’t wrapped my head around this entirely, so maybe someone else has some simpler ideas.
Mylenium
[Pour Myl
-
When I tried this morning, it worked, but I used fixed values. But you may be right. It’s possible that the cursor does not react to this and I overlooked it.
Mylenium
[Pour Myl
-
Put your greyscale layer on top of another layer. On the lower layer from the blendmodes/ matte modes simply choose Luma Matte.
Mylenium
[Pour Myl
-
Combining effects with collapsed transforms can always cause any number of problems. You may need to add yet another nesting level by precomposing the pages with the effect first, then in another comp rotating them and finally with collapsed transforms have them rendered in your other comp.
Mylenium
[Pour Myl