Jump to content


  • Content Count

  • Joined

  • Last visited

  1. seems like it would be more straight forward to just rotate each joint on the Y axis individually rather than use a joint goal. if you want it to spiral out you will also have to offset the rotation animation of each joint by a few frames and on do. in the case of splitting the ring you'll want to detach those edges from the ring. my suggestion is to delete one of the face loops as it goes up the puck (this will leave a small gap, maybe you can hide this in the back of the animation. Then you will most likely have to look into weight painting to get the last joint to not effect the puck past the gap split edge. sorry I can't be more specific.
  2. Thanks! I feel like I'm close to getting the proper output now. not quite what I need. from what I can tell I did what you said. cross product of spline 1 tangent with position of rail spline 2 then universal node set to normal to normalize the result of that. I hooked up the output from that into the rotation and it seems pretty close but its not identical to the way that align to spline tag works with rail input. I have an xpresso setup here if anyone can dig in and figure out what the solution would be. www.paraportable.net/files/AlightToSpline_Xpresso_RAIL_V3.rar
  3. doesn't seem to quite work. I'm trying to make a rail style orientation control. so the setup is like this here: http://www.base80.co..._spline_xpresso but instead of using tangent, which does orient pretty well but it can produce a flipping problem in certain cases I'd like to use a rail curve. so the setup would use another curve and another spline node to output a second position. then I need to find the vector or how the position from curve one points to the corresponding position on curve two. here is my file setup www.paraportable.net/files/AlightToSpline_Xpresso_RAIL.rar basically the output of the new matrix the one defined by the rail would go into the matrix2HPB and then that would pipe to the object rotation.
  4. how to setup in Xpresso so that I can input 2 point positions and figure out what the angle of rotation is between two points. I was looking at matrix functions to do this - http://en.wikipedia.org/wiki/Rotate_matrix not sure if it can be done in one node or if i need a few? I haven't seen any xpresso nodes yet which will output a rotation given 2 point positions of (x,y,z) vectors?
  5. im pretty sure nevo uses XSI for most projects... so it wouldnt be "sketch and toon" per say. It looks like mostly its just a colored surface shader set to 100 illumination and a -red colored light set with a high contrast, the contrast could be added in post though... there maybe an XSI sort of toon pass i remember seeing somewhere they were talking about some custom shaders using a sampler info node which basically a kind of edge detection where you can give things a kind of outline. all this could have been done as black and white then just add the red in post.
  6. I thought thats what forums were for?
  7. yea if its EASIER it must be BETTER right? In fact I'm going to give up home cooked meals... its just too hard... I'm only going to eat fast food from now on. actually scratch that, fast food is too hard... I'm only going to drink protein shakes and blend all my food and suck it through a straw. I'm also giving up working out, cause lifting those weights is hard. And painting, I'm ONLY using paint by numbers... its just way easier I don't even have to think about it... i can make a kitten a horse a landscape... I mean what else would I possibly want to paint other than whats there for me in those kits??? nah but the reverse is true to also though... if there is a tool out there that does exactly what you need it too, you don't need to reinvent the wheel. there seems to be a big battle where people are saying my 3D app is BETTER than your 3D app... which is about as intelligent as saying my dad is better than your dad. Sure there are very real differences between the apps... some of them work on on OS others not, some of them have certain plug-ins or render engines. And well those differences don't necessarily make one thing "BETTER" it just makes it more suited to your personal needs. It's probably true that more motion graphics people place emphasis on FASTER and EASIER being better... and well so it makes the most sense to select a 3D application that fits that criteria.
  8. well i just chime in with my 2 cents for what they are worth. I've been using Maya for motion graphics for quite some time. It was really really hard at first but like anything you grow to understand it more. and ive found over the years ive built a toolkit for myself for doing things easier. Bottom line is I would say you can do just about anything with the various 3D apps out there, but some of them lean towards a particular way or method of working. I like Maya because I can really get in under the hood and set things up custom for myself, thats not to say you can t do the same in cinema, but it just from the software architecture standpoint it leans toward pre built functions and plug ins. Its like if you have the option to use a plug in you probably will because its easier... but if you dig deeper down on how things are working sometimes you'll feel that being able to create a customizable effect is more interesting. Just about any app has a script box where you can customize your 3d scene in various ways. There has been a lot of hype about c4d being able to export camera's etc.. to seamlessly integrate into AE, but I've been doing this with Maya for long before you could do it with c4d. but seems like the business and marketing plan of c4d is to really hit the mo-graph market they go through considerable length to make these sort of features for you "out of the box". That said I personally I use maya, because I want absolute bare bones control over every element, but for many people c4d is the way to go.. just depends on what your trying to do and what type of person you are. If its strictly to add a little bit of 3d flair to an otherwise 2d ae motion graphics piece.. id say its maybe the easiest for someone who wants to use 3D but not really learn 3D.
  9. will it? I usually use it because its easier to control the color from an expression. I dunno, I just tried a test, where i stacked 20 fills on top of a text layer... zoomed in to test the edges and didnt notice any "crunchyness". I guess I have always been suspicious of using fills and loosing the quality of the edges.. but I dont have the proof to back it up.
  10. ahhhh... nice to have that ahaa moment. I see by pre-comping both forground stuff and background stuff... you can achieve a clean look using straight multiply. sheesh good example. thanks for the schoolin
  11. yea good point... still confused though. looks like i get crunchy alpha even by the method you suggest. unless im doing something wrong. im going to prep some files to explain this better please note that you will have to save these images and open them in pshop or like... the browser doesnt show the alpha properly. this image shows the standard (how it SHOULD look) www.paraportable.net/images/good_straight.tif this image shows a render using the method you descibed above (or at least i think it does) www.paraportable.net/images/precomp_straight.tif still crunchy ive posted the ae project as well if you want to tell me mabye im not following your method to the T. www.paraportable.net/images/renderBackgroundAlpha.zip
  12. ah ok great response i forgot all about set matte... i spend more time in 3d so i tend to think in those terms more. as a side note i found a way to streamline this even more.... since i really just want NO ALPHA in the background layer, instead of precomping, applying matte effect, setting the precomp as the alpha source.. what ill do is use the set matte effect on the background layer that i want with no alpha, and in the effect ill pick THAT BACKGROUND layer, select alpha then ill invert the matte. this is exactly what i was looking to do. just makes it one step easier than precomping stuff etc....
  13. ah FANNY if you would only follow your own advice, i could spend the other half of my day animating instead of teaching you how to animate ok well sorry for the premature post. I figured the answer. Ill post the solution just for anyone else who was wondering. and well there is something tricky going on as well which explains the hang-up i was having. also there is some chrunchyness associated with the STRAIGHT alpha channel which is unfortunate but necessary i guess. so what i did was make a black solid the size of the comp background and used that layer as a luma matte for everything that i wanted to remain black in the alpha. well when you do this ae it makes your layer totally disappear? On a hunch i decided to render this out anyways and see if it looked right. lo and behold when you render it with a straight alpha channel your background comes back! and your alpha channel is there hidden as a straight alpha channel so when you import it back in you can choose to ignore or interp as straight. pretty much what i wanted. seems like it would be handy if there was an effect that would allow you to apply to the layers you wanted to render hidden in the alpha channel. this way works, but makes the comp extra complicated as i have to make luma mattes for everything i want visible in the rgb but hidden in the alpha. oh well cant get everything eh? by the way the only other side effect of this method i noticed was by rendering to a straight alpha channel the borders of the alpha appear a bit chrunchy in the RGB channels which makes not so good for client presentations... oh well.
  14. anyone know how to render solids/layers in ae so that they appear in the RGB channels but do not "contribute" or add white to the alpha channel. In other programs this is sometimes called "matte opacity" where you can set certain layers to render visibly in the RGB channels, but render BLACK in the alpha. I guess the best way to describe this is that i have a bunch of text layers and i also have an image serving as the background. I would like to render to a qt and i want this image plane to be visible in the background, but still have the alpha ONLY from the text layers. in the past what i would normally do is render the text layers on top of the image with no alpha, then i would render again hiding the background image and this time with alpha. that way i would end up with 2 files, one that appears correctly and one that has the alpha. well this seems a bit redundant as it causes me to render twice and then i have 2 files to keep track of. Just trying to streamline the workflow a bit so that i can render once, and have everything appear correctly visually if i ignore the alpha but then if i need i have that hidden alpha channel embedded in the qt so i can easily overlay the text over a different background. anyone know how to do this?
  15. Don't forget the newly formed Rebus Farms... former DK and Superfad peeps http://www.rebusfarms.com/
  • Create New...