- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - EIAS FAQ Last updated: 11/19/98 22:27:30 This document was automatically generated by web server software developed by LAFFEY Computer Imaging. Updates to this document can be retrieved from LAFFEY Computer Imaging: http://www.laffeycomputer.com Special thanks to Donovan Rittenbach for his efforts toward this FAQ. Visit his website at: http://www.sirius.com/~zen7 You are looking at very recent data. This downloadable version is updated with every 5 submissions. The online version is updated with every submission and is fully searchable at: http://www.laffeycomputer.com Due to the fact that web users may contribute items at will LAFFEY Computer Imaging cannot accept liability for any damages resulting from the content of this FAQ. USE AT YOUR OWN RISK Please be aware that this FAQ is edited periodically, but some undesireable or obscene posts may get through at times. All that said... Enjoy the FAQ! Please contribute your own questions and answers and help the EIAS FAQ thrive! Contributions may be added via the web at: http://www.laffeycomputer.com - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Animation ------------------------------------------------------------------------------- Q:How do I use velocity curves? A:During one of my most recent projects in EI, I was required to animate a fourteen second animation of two motorcycles chasing each other through a virtual city. I hope my experiences will help you to avoid some of the problems I encountered while using EI's velocity curves. I initially created the path of the first bike using an effector and ran through a test of the animation and the thing I noticed was that the bike sped up and slows down before and after various keyframes. This is because when I looked at my velocity window there were up and down curves in my animation. This meant that for the motorcycle to get from one keyframe to the next it had to go faster or slower to get to the next keyframe. TIP----------------------------- |You can customize what values you see in the velocity |window. If you are only working with a position value, turn |off all the other ones getting in your way using the |'Custom' popup. It will also allow you to make your curves |taller using the 'Size' popup. -------------------------------- In the Velocity window, I changed the positions of the keyframes on the keyframe posts (my word) to get all the numbers at the top to be the same. The effect produced was that at the keyframes currently set, the speed of the bike was the same. The next thing was to move the keyframes left and right (by dragging the red keyframe post) so that the spikes and dips between the keyframes were minimized. WARNING------------------------- |If you drag a keyframe post in the Velocity window it |will create a NEW keyframe. If you want to keep things |simple with very few keyframes, drag your keyframes |from within the project window. -------------------------------- By getting rid of the velocity spikes, the bikes were now going the same speed all the way through. The same had to be done with the second bike and camera. TIP----------------------------- |A camera linked to an effector is much easier than trying to |animate the position and reference point of a camera. It |also allow you to use off center rotation points like real |camera rigs. Try it, you'll like it. -------------------------------- Sometimes you will find that you get very obstinate keyframes that just don't want to flatten out. One thing that I've found that works is to insert keyframes at every frame surrounding a wierd spot and look at your velocity curves. Then after you are done flattening out the offending parts, you can delete the keyframes that aren't necessary to correct the problem. The process of 'flattening' out the keyframes was the most intensive part of the process but it made all the difference in the world. Hope this helps out. Submitted by: Michael Jackson 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How can I smooth camera paths? A:One thing that can help smooth motion paths is if you set up all your key frames , then (in the camera velocity window) change them from bezier to natural cubic, then back to bezier again. Why? well in editing your keyframes you can get some extra loops & bumps that you may not want & cannot easily see, making them Natural cubic knocks out those irregularities, then switching it back to bezier gives you back your control handles again. Submitted by: Scott Wells Animax Design 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Any tips on camera paths? A:1. You may want to use as few keyframes as possible, and use the bezier handles to fine-tune the motion. 2. Check the velocity window of the camera. If you want a constant speed as you fly through the canyon, level out the velocity curves. 3. Work on one thing at a time. First get the camera fly through motion right and save your project. Then work on banking, twisting, etc (I find this the most comfortable way to work...others may have different suggestions that work better for you). Submitted by: D. Hsu 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Any ideas on spicing up flying logos? A:The best advice I can give is - constant movement/motion. Even after the "pay-off" it is nice to see something moving, otherwise, people think that the show is over. 3D is great fun for giving perspective, so use the most of it, however, don't neglect the 2d design. If you are using "3D" and "Design" as your words of interest, how about really using that 3D space. That is not to say go crazy. Slow dramatic movements around the set will always look elegant. I will assume that this would be the beginning of your presentation, so maybe you could darken the surroundings, while brightening the logo to simulate the beginning of the show/movie. Last but definitely not least, make a nice transition into the next scene without fading out or dissolving. It is always fun to go through an element into a "new dimension". Submitted by: MikeT 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How can I make cool cities without 2000 hour renders? A:>If I was going to do a city, I am curious about modelling the buildings? Maybe not put real rooms in, but maybe put luminance maps or something? Would one just put a single floor map on and tile it so it looked like many floors? I just want to avoid obvious tiling as much as possible. What if I wanted to do a fly through of a city, ending at some spotlights and a sign, or city hall or something? What is the best way to do something like that? Model the buildings close buy with exceptional detail, and those far way with little? I am curious, because it seems like it would be easy to get a high polygon count really fast....do you know of any techniques for minimizing polygon counts in this instance?< That's exactly what the city fly I did was - twisting through the streets of a fantasy future city, zooming up the side of an Art Deco city hall, and ending on the roof with a panoramic view of the city at sunset. I modeled the city hall and a plaza in front of it with a fair amount of detail; the buildings along the path with just enough detail that they didn't all look the same; and the background buildings were just stretched cubes. Start with very simple shapes and set up your flight path. Then watch a test render a few times and you'll see where stuff needs more detail. Remember that in an animation you can get by with far less detail than a still needs. I did the same with textures. The buildings along the path had fairly detailed maps and city hall was very detailed since it appeared in extreme close-up. I was able to use many of the building textures a few times; that cut down the workload quite a bit. Each building was a single non-tiling texture with color, bump and glow maps. For the bg buildings I made a few variations of glow-mapped window lights - no color or bump at all - and applied them to the cubes at different scales, rotations and start points. Worked fine - with the bg popping in and out of view and constantly moving even I couldn't spot the repetitions. The best way to keep polygon counts low is to use textures instead of modeling whenever possible. Especially in an animation, its amazing how much detail you can simulate with good textures. Submitted by: Screaming Metal mac3d@earthlink.net http://home.earthlink.net/~mac3d/ 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Bugs ------------------------------------------------------------------------------- Q:Is there a texturemap bug in 2.8R? A:Yes, Texturemaps cannot be of the size 640x480. You need to either resize them in Photoshop or AfterEffects, or clip a few pixels off either dimension. This bug may apply to textures with a single dimension of 640 pixels as well. (I had it crash like this once.) Submitted by: Joe Laffey LAFFEY Computer Imaging joe@laffeycomputer.com http://www.laffeycomputer.com 9/2/98 18:50:39 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Are there any import bugs in 2.8R? A:The prefs in 2.8R are out of whack. Even if you change the inherit settings for importing models the change will not "stick." You can either choose "Default Preferences" from the Edit menu (I think) before import files... Or you can import them and then go to the group link window of a single model and hold down the option key as you click inherit rotation, scale, and position. Note that for most operation in EI holding down the option key while you change a setting or click a checkbox will apply the setting to all the models in the project. This is a fast way to set the motion blur of all your objects to motion vector. Submitted by: Joe Laffey LAFFEY Computer Imaging joe@laffeycomputer.com http://www.laffeycomputer.com 9/2/98 18:57:57 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:What is an error #-1309? A:I recently had this error (or something very similar) after bringing a project from home to work. What seemed to be the error was a bad master material. If you have any master materials, do some tests by assigning then to simple shapes, and then render each shape separately until you get the error or it goes away. That will be your bad master material. Worked for me, anyway... but I could be way off, too. Worth a try. Submitted by: Bear Weiter Wombat Interactive bear@wombative.com http://www.wombative.com/bear/ 9/4/98 16:17:17 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Camera & Rendering ------------------------------------------------------------------------------- Q:How Can I use Hi-Res Models for rendering and Low-Res for animating? A:As a work around you can keep 2 sets & swap them before you render. I've done this several times & it works well. If you hold the option when you click "render", ei will write out the camera temp to the camera directory (those of you who have slave cameras probably already use this instead of rendertrauma). Camera will launch, which you should immediately pause, go to the finder & get info on the camera temp file & lock the file, & then quit camera. Once you've done that, swap your 'hi-res' fact files into the folder where your low res are & move the low res ones to another folder. Its important that the names of the models & the objects contained in each fact file are identical other than poly count (name, number of objects, etc...). Double click camera & it will automatically start rendering the Camera file in its directory, now with the hi-res models. I've done this with a great deal of success when using the "~~" model cycling feature. Since Camera loads them one at a time, but EI wants to load them all in ram, I found having extremely low res versions made working in EI much more pleasant, while still getting the being able to use the hires versions when rendering. Submitted by: Scott Wells Animax Design swells@tiac.net swells@aol.com animax@aol.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How do I texturemap cycled models? A:You have to understand that the dxf cycling option is only suitable for certain types of animation. Fully jointed, seamless characters animated in Hash & exported to EI would be impossible to texture in a meaningfull way. And since EI is not aware of the nature of the cycling geometry, your textures will 'slide' as your model moves. If, however, you just wanted to do a characters head this way there is a bit of a workaround on the texture issue. I promised to post this a few weeks ago & will do so now: What you will need is a program that can do morphs/distortions (e.g. Elastic Reality, Morph, Movie FLo). I'd highly recommend Movie Flo, having used all three, it is by far the superior product in terms of power & quality. First - create the texture map for your model, using the first model in the animation for a base. Next, in Photoshop or some paint program, you need to make a grid that you will map on your animated models that is the same res as the texture map. The grid should be such that it has a main grid in one line color and a sub grid of higher resolution (subdivides each square of the main grid into 4 smaller squares) in a lighter color. Number each main grid with a unique number (starting at say 001, & increasing left to right/ top to bottom) and letter each sub grid with "a-d". The point of all this is to be able to track where the geometry moves within your grid once its mapped to your animation. The grid needs to be big enough (pixel dimension) so you can read the numbers & letters when you render the animation. Those of you who have taken Zax class will remember the grid he used to demonstrate how to create textures that map correctly to your model. This takes that principle a step further. Save the grid out as an Image file & import it into EI, mapping it to your "~~Parent" using what ever mapping method you desire (spherical seems to work best for me). Render out the animation once from the front & once from the side (if you object's geometry is asymmetrical or animates in an asymmetrical fashion - say a lopsided smile, you may need to render other views as well). You'll want to resave the Image file animations as quicktime so they will be easily accessible for reference. In your morphing/warping package, import the grid image. In the background you want to have the quicktime movies of your model open in movie player & the info windows open so you can reference frame numbers. Or, to reduce clutter, you could run them though premier & ad timecode to them. >From this point, the basic Idea is (referencing the quicktime movies) to create the shapes of your major masses on the model referencing where those shapes pass though your grid. For a face, you'd create outlines for the jaw, mouth, eyes, eyebrows, nose, etc.. by noting how the geometry intersects the lines of the grid mapped to the face & duplicating it with control lines in the morphing program. Once you've created all the outlines to your satisfaction, you want to walk through the qi movies & find "key frames"of extreme motion. at these points, take not of the frame, move to that frame in your morphing program & modify your outlines, (again referencing the way the geometry has moved & where it now intersects your grid). Continue to do this for the entire sequence. Some experimentation may be required so the texture distorts in a "natural" way. Movie Flo has an unbelievable amount of control over both the region of influence of each control line and the quality of the final image (no pixelating), so I would recommend it highly over other programs. Once you've finished, swap your grid for your texture map & render out the sequence. Convert it to image format & replace the grid with your new movie as the parents texture. Render your animation out of EI again for a check, then make any adjustments necessary in your morph program. A couple of things to note: If your working with a face, starting with a closed mouth & going to an open mouth wont give the morph program enough information to make a believable open mouth (you'll get stretched lips!)- the same could be said for the eyes. Plan your animations accordingly & be prepared to paint more texture maps that you might do true morphs between. Also, programs like MovieFlo allow you to have different images on layers, but warp them locally & globally, so you could break your texture into layers to avoid problems with mouths & eye areas. Hope this makes sense. It seems complicated, but its not that bad & works very well. If anyone needs clarification, feel free to ask. Submitted by: Scott Wells Animax Design swells@tiac.net swells@aol.com animax@aol.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How can I take a project to be rendered on a slave machine without using RenderTrauma? A:Check batch render box in EI, when you hit "go" and assign a name and destination a Camera.temp file will be written into the Jobs folder. Take these camera.temp files along with appropriate project, model and texture files to the computer you installed the slave on and drag and drop the camera.temp to Camera app. and rendering starts. There might be an easier way to do this, but this has been working for me. Submitted by: Jon JP623@aol.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How can I get outlined objects and shadows with an alpha? A:Use the visibility check marks next to the objects that you do not want visible (The background, etc.) Check "Shadow Object Only" in the Group Info dialog for any object that you do not want to see, but you do want to cast a shadow. Check "Generate Shadow Mask" in the Group info dialog for any object (e.g. the background) that you do not want to show up, but you do want to "catch" a shadow. Go to the Group Color dialog for these shadow catching objects and set their group color to black with no specular value. ( You may use another color if you wish. This is the color that the shadows show up as in the final composite.) Be sure you choose Millions+ in the Render dialog and if you use QuickTime make sure you choose Millions+ for the codec settings. If you desire to you may have the fog control fog out to an alpha value. Thus, objects beyond a certain point could vanish from the composite. This can be gradual or sudden depending of the settings in the starting and ending fog values. Submitted by: Joe Laffey LAFFEY Computer Imaging joe@laffeycomputer.com http://www.laffeycomputer.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Any tips on rendering and disk fragmentation? A:If you have Camera on the same disk as your rendering is being saved to... (i.e. the temp files (shadows, reflections, etc.) are on the same disk as the rendering. You will get some SERIOUS disk fragmentation in a very neat pattern. You can see it with Speed Disk or Mac Tools. The system tries to allocate contiguous blocks for files and each frame is written after the tempfiles are written. When the temp files are removed, there is not quite enough space to re allocate the temp files in the same place on the disk. Thus the temp files get written AFTER the frame.. Then the whole thing starts over... Render an animation on the same drive as your Camera and take a look at it in Speed Disk. You'll see what I mean... Freaked me out the first time I saw it. This also fragments the hell out of your animation making it play back very poorly. Submitted by: Joe Laffey LAFFEY Computer Imaging St. Louis, MO joe@laffeycomputer.com http://www.laffeycomputer.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How do I make my rendered image sharper? A:EIAS texture mapping sharpness The default texture map and anti-aliasing settings in ElectricImage tend to create "soft" images that have no temporal aliasing artifacts. This softness is sometimes not desired. There is great flexibility in the renderer and sharp maps can be created without resorting to high anti-aliasing settings (eg 32x32 oversampling). The first step in creating sharper images is to change the anti-aliasing mode from Oversample to Adaptive in the render control window. The anti-alias setting should be at its default of 4x4 and the sampling level should be at its default of 1x1. Upon rerendering, you will notice that image is much sharper. To increase the sharpness further, adjust the Min Adaptive Sampling Threshold in the render control window. It defaults at 0 but try it at 32, 64, 128 and finally, 255. When both the min and max are at 255, the renderer is in its Average anti-aliasing mode. This will yield the sharpest imagery, but may result in temporal aliasing artifacts (ie. crawl or buzz). You may find that a lower setting yields an acceptable level of sharpness with no temporal problems. A second feature that can be used to improve the sharpness of texture maps is the Softness: pulldown in the texture map window. Changing it to None from Sharp may actually yield better results. Submitted by: Kaos 2/1/97 14:11:34 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Why are my camera temp files SO HUGE? A:There are several things that can cause big control files. The first is deformations. If you have used deformations, especially when you've applied deformations to a parent of many children, the deformation record is HUGE. This also applies to deformations applied to plug-ins, where the deformation record is created twice and, I believe, four times for plug-ins that have children or are children of deformed parents. (This is the cause of one of the bugs where sometimes deformed children appear as though they are deformed twice as much as they should be.) Another main cause for large control files is the record stored by each plug-in. When EI writes the control file, this record is repeated for every frame. It's also repeated for every plug-in, so if you have a scene with lots of plug-ins you're getting a lot of duplicate information for every single frame. There are ways of minimizing this amount of information though. I can't speak for the other plug-in developers, but when the Invigorator reads your Illustrator file, it reads in and stores the whole file. This includes anything on Invisible layers and any bitmaps you've placed on the page, and any bitmaps you've used as a background for tracing over. If you leave these things in your Illustrator file the plug-in record for each frame becomes the size of the Illustrator file times the number of Inivgorators you've used. Multiply this by the number of frames and viola! you've used up a couple of Gigs. Just in case you're wondering why the hell the Invigorator is storing all the bitmapped data instead of throwing it out when it reads in the file... it's a loophole. We decided early on to keep all the data in a file for when we added more features in the future and you would thereby be able to open old projects and do more with them at a later date. However, we forgot that "all data" meant the bitmaps too, so these slipped by. This will be corrected in a forthcoming version, but in the meantime, just remember to delete them before you save your file. If the size of your AI file is small you should be fine. It's only when it's over a meg and you create lots and lots of Invigorator models that you will get Camera temp steriod meltdown. Submitted by: Zax Dow Zaxwerks zax@zaxwerks.com http://www.zaxwerks.com 9/8/98 11:3:30 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Deformations ------------------------------------------------------------------------------- Q:Tips for multiple bend deformations A:Firstly: The min/max co-ordinates refer to the region along the axis that the bend will occur eg. if you are deforming the x axis values of -10 & 10 would bend the object over a range of twenty units centered on 0. clicking the boxes means that the deformation will not occur outside this region, not that these numbers are now meaningful (I naively assumed they were irrelevant if the box was not ticked :)). Secondly: You want to move consecutive bends along the axis to be deformed (with the position variable). You can then get a pretty nice result with sharp bends and are able to create some really outrageous curves that you can follow with for instance a unrolling frieze (which I was working on when I worked all this out). Submitted by: Nigel Caughey PlanetGo nigel@planetgo.co.nz http://www.planetgo.co.nz 5/28/97 7:11:41 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Getting Work ------------------------------------------------------------------------------- Q:How can I get a job doing 3D animation? A:Good question, there are several steps and ways. For example, Disney offers a test for all potential artists. You must go to one of these tests and your work will be judged by their experts. In my opinion this is fair. Another step is through your Art Center or wherever it is that one studies art. They should have placement info posted. Warner Bros Animation looks at Portfolios, Reels and art backgrounds as do the rest. Word of advice,,, if you want to work in the film industry then you need to relocate yourself to where the film industry is. There are many artists of all sorts ringing the phones off the hook everyday at all major studios. Not only do potential employers want to see your work, they want to see you. It's very rare that someone outside the area gets hired unless he/she already has an excellent and well known track record in their work history or fresh out of animation/art school. As far as contracting, that all depends on the production house you are working for. 3D Job opportunities are not limited to the film industry, there are also game developers, TV Networks and smaller production houses all over the country. One thing that I feel uncomfortable doing is sending my artwork to agencies unless I'm there with the artwork. This is the #1 way to get your hard work and ideas stolen. I have seen this happen at every level. There are good agents one can acquire, but then you must see that agent on a regular basis. These agents are normally found in offices right off the lot which brings you back to the fact that you must live in or near the studio area. I know it hurts, but I have also seen people get great positions because of mommy and daddy being way up in the company. On a good note 9 out of 10 artists will get hired at one level or another. Sometimes it's enough to get your foot in the door by doing pion work. We all have to start somewhere. I recommend that you also get along with the Software company of whose product you're using and see what they can offer. Attending shows such as Siggraph, NAB, MacWorld, etc. are invaluable to your success. Regardless of the expense, the contacts you can make there will pay for everything quickly. On a last note I also suggest feeling around your own area if the prospect of moving out to the big city is too much for you at this time. I'm posting the Job Hotline # at Warner Bros. I give no guarantee of any jobs being offered, but I did notice the job board offering jobs for 2D and 3D animators at Warner Feature Animation. The # is: 818-954-5400. Follow the instructions from there. I hope this helps a little. Submitted by: Erik Dekhoda ErikTek1@aol.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Any newbie tips on dealing with clients? A:Let me explain the points I was trying to make. Be conscious of the image you are portraying to people as you communicate with them. 'Stuff like that' portrays you as some one who is unsure. Which is fine in this forum asking for advice, but something like 'or related subjects' would have conveyed the same message very professionally. When you are dealing with clients you need to make them feel confidant that you can do the job they need. The second half of that is Don't Bullshit. People will usually know. You only have one reputation and it's better to be honest and tell someone you can't do something or you're not sure than to say 'No Problem' when you have no idea what you're doing. People will respect you're honesty and trust you for bigger and better things. Of course that's just my opinion, I could be wrong. Submitted by: kory Fox Sports 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Any more tips for newbies seeking work? A:If you are showing me your reel, never introduce it with an apology. If it merits an apology, why should I bother to watch it? Be proud of your work. Of course, if there is a technical problem beyond your control you might wish to explain (not apologize). If the tools you have access to are limited, don't use that as a crutch or an excuse. Creativity and craftsmanship transcend tools. Use what ya got to the best of your ability. Submitted by: adam g 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Still more tips for getting work? A:<> Nobody wants to be the first person in. It is a self perpetuating cycle, apparently, that sometimes you aren't busy, because you don't look busy! Ever go walking by a store and looked in, and all you saw was the clerk at the back of it, and nobody else, and even though the store looked interesting you didn't go in, because you didn't want the clerk to come over a "help you" and try and sell you something, because you like to shop in peace? Its the same thing with our pianist. He always puts a couple of bucks in his tip jar at the start of the night, or else he won't make jack. Nobody wants(except for me) to be first. Anyways, that's my two bits. Submitted by: Donovan Rittenbach CineOptic Productions 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Memory Management ------------------------------------------------------------------------------- Q:Tell me about Memory Allocation and EI... A:"Could not allocate transparent map: 4566754" is an error message too well known. The result of this error is almost always a compromise, and only leaves a few options. Solution 1: Buy more memory. Although this is the most costly, it makes rendering easier in all respects and prevents a lot of frustration. Ultimately this is the solution to memory problems altogether. Solution 2: Since spending more money is not a high priority on anyone's list, another option is to minimize the project, or in other words make compromises with quality. This can be time consuming and is surely a cause of headaches. Solution 3: Optimize the memory usage so that you know the memory being used is being used at its potential, before you follow any of the previous solutions. As a guide, the following formulas can be used to figure the basic memory requirements for your specific project. Polygons 1 Polygon = 100 bytes 1 Transparent Polygon = 300 bytes ElectricImage displays the number of polygons in the project if you select Statistics from the File menu. Keep in mind that this statistic does not take into account the sub-polygons Camera creates when rendering the surfaces of complex polygons. So this can only be used as an approximation. The exact number of polygons rendered is displayed in the render information while Camera is rendering. You can arrive at a close estimate of transparent polygons by adding the polygon statistics from each individual transparent group. Then by subtracting the number of transparent polygons from the project's total number of polygons, you will have an approximation of the number of non-transparent polygons. Multiply the number of non-transparent polys to 100, and then multiply the number of transparent polys to 300. Add these two values together and you have a good idea of how much memory the model requires to render. Example: Project Total: 122504 Transparent Total: 10432 3,129,600 Non-Transparent Total: 112072 11,207,200 Total Memory Required: Over 14 Megabytes Shadows ((Buffersize[in pixels]^2) * 4[bytes per pixel])/1024 Don't let this formula daunt you. It is really quite simple. Shadows are measured in pixels, similar to a texture map. The default size for a shadow map is 1280x1280 pixels. This value is set in the Light Info Window by the Buffer Size. Since the shadow map is a perfect square, the Buffer Size edit box only requires a single value. Each pixel of the shadow map requires 4 bytes of memory. This is calculated as follows: 1280 * 1280 = 1,638,400 * 4 = 6,553,600 In order to convert this value into kilobytes divide by 1024: 6,553,600/1024 = 6,400 or 6.4 megabytes The equation in its entirety can be viewed as: ((1280 * 1280) * 4)/1024 Texture Maps Summation = X * Y * 16 bytes Mip = 1.3333 * X * Y * 4 bytes Total map size = largest texture1 + largest texture2 + largest reflection map size. What this means is that the total map size is the amount of memory allocated to every map in the project. So even the smallest maps get the largest memory allocation. This is contrary to the idea that the physical size of an image in megabytes determines the amount of memory needed, although there is a correlation. Largest texture: 1000x1000 pixels Second largest texture: 500x500 pixels Largest reflection map: 750x750 pixels Largest texture memory: 1000 * 1000 * 16 bytes = 16megs Second largest memory 500 * 500 * 16 bytes = 8megs Largest reflection map: 750 * 750 * 16 bytes = 12megs Total Map Size 36megs So every map that is in the project (including the largest and second largest) are allocated 36megs. This also means that all smaller texture maps can be as large as the second largest without impacting memory usage. This is not an additive process, but rather maps are swapped in and out of this 36 meg allocation as they are needed. When transparency related maps are involved, the memory needed for those maps is allocated at the beginning of the render in addition to the normal map alloaction. This has potential to really take up a lot of memory and is something the user should be aware of, and avoid if possible. Memory optimization Polygons: The best way to optimize polygons is to use ElectricImage's import and export features. Using the import features you can combine co-planar faces, merge edges and vertices that reside within a certain distance of each other, and merge duplicate polygons. This often works magic with complex models and there have been cases where the polygon count has been dropped by half without reduction in quality. Be careful with the merge features because they do have the potential to impact the quality of a model. The export features include options to save only polygons. This is useful if you have a model that has line and point artifacts floating around. This will strip everything but the polygons and leave you with a clean model. Shadows: Quite often the buffer size for a shadow can be decreased without an apparent impact on the render quality. This is especially true for background shadow casting lights that are not directly the center of attention. Textures: There are several ways to optimize memory use for textures. One way is to find the 2 largest texture maps and the largest reflection map and scale them down using Photoshop. Trial and error will bring you to the lowest resolution setting for each of the largest maps without loosing quality. (Remember that a map that is too small will blur. So when you reach the blurring point you know that it is too small.) Another way to optimize memory usage for maps is to hierarchically link all groups with the same texture together. This will not actually make more memory available, but will save the time Camera takes to swap textures in and out of memory. The last way to modify memory usage for textures is to use CameraExtender. This application allows the user to set special preferences for Camera. One of these preferences is to 'Load All Textures'. In effect, every map used in the project is loaded into memory from the beginning of the render. This will almost always take up more memory, but renders much faster. 'Load All Textures' can be used to save memory in cases where there is only one texture used, and it is used multiple times in the project. Generally, Camera is able to locate instances and load each texture only once. However, when only one map is used multiple times, it takes the role of both the largest and second largest texture. This results in a render that takes twice the memory needed. If you use 'Load All Textures' you will load this single texture file from the start and it will be forced to instance the texture. System: Often it is the system that keeps the Camera from reaching its maximum efficiency. There are several things a user can do to improve render performance through the system level. Turn off all extensions. (Quicktime is needed for 2.5.2 versions and higher) Turn off extras such as menu clock and background tile. Turn off 'Calculate Folder Size', 'Info header', and files size, date, and type. Camera: The memory allocated to Camera has been a minor controversial issue. Some prefer to leave a lot for the Finder to avoid 'Type 1' errors, while others will leave a minimum amount. The problem is that the system memory fluctuates. Probably the best thing to do is to see 'About this Macintosh'. This will show the current amount of memory being used by all running applications. Be sure to do this just before your render so you get the current statistics. When you allocate memory to Camera, leave room for the Finder to fluctuate a couple of megabytes and the rest can go to rendering. This will help avoid errors. Note: Camera should be the only application active when it is rendering for the best results. Do no use renderama unless you need to batch process renders. Renderama tends to slow Camera down because of the feedback process. If all else fails, solution 1 is almost always the answer. However, it is always a good idea to contact us at ElectricImage. After all the cause of the out of memory error could be a bug such as a memory leak. Good Luck, and remember: "You can never be too beautiful, have too much money, or have too much RAM!" -Etienne Taylor Submitted by: Steven Walker ElectricImage 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How can I have maximum memory for Camera to render? A:You might try holding down the control key (or is it the option key.. I never remember) when clicking the Go button in the Render dialog. This will allow you to specify where to save a cameratemp file that is not invisible. Let the Camera start.. Then choose pause from the camera file menu. Next go to the Finder and lock the cameratemp in the get info box. Then go back to the Camera and choose quit.. Then look at your largest unused block in the About this Mac window and set your Camera partition to this size minus a meg or two for system use... (Might not be a bad idea to restart Mac here.) Then double click camera. It will start the render again and hopefully it will have more memory than before... Be sure to unlock and delete the cameratemp before you do another render.. >To Joe, Frank (or others), one follow-up on this whole issue: Last night I tried the control-"Go" trick and it worked, except for one very important issue: EI didn't quit. The camera attempted to fit into the 12 Mb left over but didn't kill EI like it usually does. Is there any way around this, or do EI and the camera have to be in the same folder for this to>work? When you use the control-click EI does not quit.. Which is sometimes good... But if you are running out of RAM once Camera launches switch to it immediately and choose "Pause" from the File menu. Then go to the Finder locate the camera.temp file and lock it in the Get Info box. Then go back to Camera and quit Camera. Next Quit EI. Then restart Camera and it will render the camera.temp file in its folder... Kind of a pain but it works... Submitted by: Joe Laffey, LAFFEY Computer Imaging St. Louis, MO joe@laffeycomputer.com http://www.laffeycomputer.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Miscellaneous Tips ------------------------------------------------------------------------------- Q:Tell my about units in EI... A:I've read several answers that were correct about this whole granger units thing, but not quite as clear as they could be. A Granger unit is not equal the size of the first object imported, but to one of the _units_ in which that model was created. For instance, if I import a human figure that is modeled in inches, my Granger units equal one inch and my model will be about 72 Granger units high. If I then import a model of a building that was modeled in meters, each meter will be scaled to one inch, and I have a nice desk-top scale model of the building next to my full sized human. If I then import a third model, this one of a human figure modeled using feet as the standard unit, it will come in as a six inch "action-figure". I usually work at full scale in Form*Z, using decimal inches (I used to work in an old machine shop and got used to thinking in 1000ths of an inch. I should just bite the bullet and go metric). If I need to use a third-party model, I first import it into Form*Z and make sure it is properly scaled. This can be a bit of a pain if a model has been created with metric units, or as I have seen often, with no regard to scale what-so-ever using what ever units the modeling program used as defaults. That seems to only be a problem with organic type models, as more mechanical objects usually require some kind of numerical precision. So anyway, what that all boils down to is that for me, one Granger unit always equals one inch. Most Granger related plug in stuff is either strait GUs or GUs/Sec. Submitted by: Frank McIntyre 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Where can I find models online for download? A:Online Models URL: http://www.baraboo.com/3dcafe/models/ Submitted by: Anonymous 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Modelers ------------------------------------------------------------------------------- Q:Any tips on using Amapi with EI? A:Some Amapi file conversion hints: If you export a file and its corrupted, or won't open at all in other apps, just keep re-exporting. This usually fixes it. You can export the same file to the same format several times and other apps may read all, some, or none of the copies correctly. More of an irritation than a real problem but be ready for it. Amapi exports get much less stable as files get bigger. If your Amapi files are more than a couple MB in size you'll probably have to export piece-by-piece. Be sure to give Amapi a lot of RAM if you want to import 3DS files. Amapi needs over 80MB to open a 6MB 3DS file dependably. Be sure to 'Force Backfaces' when bringing Amapi exports into Infini-D. Don't use Amapi's FACT export. Use DXF instead and let ElectricImage do the conversion. Trust me on this. Very small objects will not import or export properly. Amapi's 3DMF export occasionally inverts the X axis. The only solution I've found is to export as DXF instead. >What caught my eye is that it outputs directly to POV. So for everyone >who has posted requests for POV export, here is it is. That's the reason I bought it in the first place. Unfortunately Amapi exports POV and some other formats with the coordinate handedness flipped. This means you have to model a mirror image of the object you want to export*. I should have been warned since the Amapi guys didn't even know there was a Mac version of POV... Amapi (the full version) is a great modeler with some really aggravating quirks and bugs. I've been using it a lot and been annoyed most of the time, but I keep going back for more ;) A new version was supposed to ship sometime in December. (1996--should be out in Jan 97) * unless you export an entire scene and use camera transformations in POV. You can't imagine how much trouble this is until you try it. Submitted by: Screaming Metal mac3d@earthlink.net 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Is there a FormZ mailing list? A:Yes. Send a email to FormZ-List-On@softmotion.com. No subject or message body text is needed. To sign off, send a email to FormZ-List-Off@softmotion.com To mail to the list send to FormZ-List@softmotion.com Submitted by: Rene Hedemyr rene@softmotion.com SoftMotion AB, Sweden 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Other Programs ------------------------------------------------------------------------------- Q:How great is Hash and it's compatibility with EI? A:The models work quite well considering Hash's variable resolution output & EI's excellent (and quick!) dxf import. I've never had a problem on EI's end with Hash models - once they look good in Hash, they always look good in EI. Some experimentation with the smoothing angle is ususally required. >Also, if I'll be rendering with EI, do I need the full Animation Master version? < No, the only thing that the full version gets you is more rendering effects in Hash (motion blur, network rendering, depth of field). If you want to render in EI, this doesn't matter to you, the rest of the two packages are identical. Submitted by: Scott Wells Animax Design swells@tiac.net swells@aol.com animax@aol.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:What's the best way to use Detailer with EI? A:The following notes are designed to make EI 2.7 and Fractal Design Detailer work together. Assuming that the geometry to be painted is inside EI follow these steps. 1) Select the desired group (one group only) and export as 3DMF. 2) In Detailer Import the 3DMF file. Set the 3d view to the most comfortable size memory permitting. 3) Now you need to decide what is the best method for texture map projection that you will use inside Electric Image that best fits the model. Keeping this in mind there are couple of things that you should not do inside Detailer. A) You CANNOT use box mapping. The ways that Electric Image uses cubic maps and the way that Detailer does it are different. 4) In Detailer initiate a new texture(texture,bump,highlight, or glow) map with the desired projection. For the first map that you define align, it with OBJECT AXIS. All subsequent maps align with first defined map. For spherical projections define maps that have equal number of ROWS and COLUMNS. Decide early how big your map should be. You can rez down but not up too much without sacrificing quality. 5) Using your artistic talents and art direction, paint the texture, bump and other maps. In Detailer, set the bump map to be INVERTED and value of 1. Set the specular, diffuse, and reflection values to approximately planned values in ElectricImage. 6) When complete, export the maps as Photoshop 3 documents or PICT files. 7) Quit Detailer and launch Photoshop or equivalent application. Load the texture and bump map files from the previous step. 8) Insert copy/paste the grayscale bump map into the ALPHA channel of the texture map. Export the resulting 32 bit image as EI IMAGE format. 9) Launch the EI project. Select the group that was painted and go to its texture window. If spherical projection was used: Set the spherical projection with following values. Rotation X 0.0 Rotation Y 90.0 or -90 Rotation Z 90.0 or -90 If necessary set transformation to Z -1.0 or 1.0 to nudge the map Turn off all repeats except NEGATIVE Z Set WRAP ANGLE TO 360 Set BAND ANGLE to 180 For cylindrical maps: Align to either TOP or BOTTOM assuming that in Detailer cylindrical map was defined along the OBJECT axis Turn off all repeats except NEGATIVE Z Set WRAP ANGLE TO 360 Set BAND ANGLE TO 180. Assuming the map you painted contains a bump map in the alpha channel: Set Alpha to BUMP Set BUMP VALUE to 1.0 For complex shapes, such as a head of character, it is best to use spherical map projection. Submitted by: (EI tech support I think) 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Is there an After Effects mailing list? A:Yes. Send a email to AE-List-On@softmotion.com. No subject or message body text is needed. To sign off, send a email to AE-List-Off@softmotion.com To mail to the list send to AE-List@softmotion.com Submitted by: Rene Hedemyr rene@softmotion.com SoftMotion AB, Sweden 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Plug-Ins ------------------------------------------------------------------------------- Q:How can I keep that damned Dicing dialog from showing up every 2 seconds? You can save out your diced model as a FACT and re-import it. This way it is already diced. You might have two versions of your model, a low-res for animating, and a hi-res for rendering. See elsewhere in this FAQ for techniques... The same technique can be used for standard shapes as well. Submitted by: Joe Laffey LAFFEY Computer Imaging joe@laffeycomputer.com http://www.laffeycomputer.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Reference Material ------------------------------------------------------------------------------- Q:Is there an EI mailing list? A:Yes. Visit http://www.toolfarm.com/resources/eilist.htm for details... Anonymous Submitted by: Reference Material Q:How Can I use Hi-Res Models for rendering and Low-Res for animating? - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:What are some good books on character animation? A:Here's four of my favorites -- The Illusion of Life(The Illusion of Life---ISBN-0-7868-6070-7) - Frank Thomas & Ollie Johnston Animation - Preston Blair Animation from script to Screen - Shamus Culhane The Animator's Workbook - Tony White Submitted by: George Maestri maestri@ix.netcom.com 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Where can I get a PIXAR animation tape? A:>I have not seen any pre Toy Story Pixar stuff except Luxor Jr. I would however, love to see it. Can you point me towards a source for this tape? Is Luxor Jr. on it as well?< the Tape has: 1. Luxo (not Luxor) Jr. (1986) 2. Reds Dream (1987) 3. Tin Toy (1988) 4. Knickknack (1989) I also recommend you "The Worlds greatest animations" from Expanded entertainment the toll free # is 800-996-toon Submitted by: Rolando 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:Are there any books on special effects or ILM? A:There a book out called. Industrial Light & Magic Into the Digital Realm ISBN 0-345-38152-1 The book covers the history of ILM, leaving off with Star War New Hope. Great pictures of John Knoll work using EIAS. It a must have! Submitted by: David Sunder 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Rumors ------------------------------------------------------------------------------- Q:Is anyone gonna submit a rumor? A:Hopefully so... Submitted by: Anonymous 9/8/98 10:44:17 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Secret Techniques ------------------------------------------------------------------------------- Q:What are those hidden Advanced features of EI? A:ElectricImage Animation System Version 1.5 Advanced Features (NB: Many of these no longer apply to the current version of EI) This document describes some advanced features in version 1.5 of ElectricImage Animation System (EIAS). These features are either experimental in nature or were added to the software after the manual was printed. These features have not been as thoroughly tested as the rest of EIAS and should be used with caution. Niether printed documentation nor telephone support is available for these features. Questions, comments, suggestions and bug reports should be sent to Mark Granger on America Online (user ID "Granger"). The purpose of providing this documentation and support files is to allow some new and experimental features of EIAS to be made available to advanced users without having to wait for the normal beta cycle to be completed. The success of this process will determine whether documentation will be provided for advanced features in future versions of EIAS. This documentation has three sections. The first section describes how to access the solid texture shaders built into Camera, the second section describes the CameraExtender application, and the third section describes some hidden features of the system. Built In Solid Texture Shaders Plugin code support is planned to be added to EIAS in future versions of the software. These plugins will provide a method of adding solid texture shaders to Camera. As an experiment, several solid texture shaders were hard coded into Camera before the plugin code support was developed. These shaders consist of the most common types of solid textures. In version 1.5 of EIAS, a method was devised to make the shaders available to users with a fairly simple (although limited) interface. Each shader is activated by applying a specially named texture map to a group. The names each begin and end with a dollar sign which makes them easy to recognize. The contents of each texture's image is not used for mapping. Instead, a single alpha channel is calculated which is used to mix between the texture's material color and the selected RGB channel. Some of the solid textures also generate a normal perturbation which is used to generate a bumpy surface. Solid textures can be used in the same way as image map textures. The Alpha channel of the map is generated by the solid texture function. The RGB channels of the texture come from the Group Color of the Texture Filter material picker. The Alpha and RGB popup boxes work in the same way they do when an image map texture is applied to a group. The Bump Factor can be used to scale the bump generated by the some of the solid textures. Unless otherwise described, the alpha channel of the Group Color of the Texture Filter material picker is used to scale the alpha channel which is generated by the shader. Each of the solid textures is described below: $Noise$ generates a simple 3D noise pattern. The shader also generates bump normals which can be used to give any model a randomly bumpy surface. $Stucco$ is a modification of the $Noise$ shader. It has the effect of producing a stucco surface. This shader also produces bump normals. $Marble$ is the most commonly used of all solid texture shaders. It creates a simulated marble surfaces with turbulent bands of color. Bump normals are not generated by this shader. $Wood$ is another very common solid texture. Light brown or orange for the texture color and dark brown or black for the surface color yields the best results. Use Cylindrical mapping for this shader to create concentric rings of wood. Bump normals are not generated by this shader. $Fractal$, if used to darken the diffuse channel of a group, creates a good weathered surface appearance. Bump normals are not generated by this shader. $Clouds$ is a modification of $Fractal$. The alpha channel of the Group Color in the Texture Filter material picker is used as a cutoff value for the fractal noise. The Bump Scale is used to scale the fractal noise. Bump normals are not generated by this shader. This shader works best when the texture's RGB channels are white and the group's surface color is blue. The Alpha popup should be set to Decal and the RGB should be set to Surface. $Waves$ produces a concentric series of wave rings. The rings move outwards from the center over time. The alpha channel of the Group Color in the Texture Filter material picker is used to control the speed of the waves. Zero creates stationary waves. 255 creates waves which move out at the speed of one cycle ever 255 frames. The lower the value the faster the waves move outwards at each frame. The shader generates bump normals which will make the mapped surface appear to be wavy. $Mandelbrot$ is a mandelbrot set generator. This shader is currently not very useful because it can only be used to blend between two colors rather than picking a color from a table. Since the bottom left corner of the map is located at [0.0, 0.0], the map must be offset and scaled so that the bottom left corner of the map is in the center of your group's extents. In addition, the texture application mode can be forced by including $Flat$, $Cylindrical$, $Spherical$, $Cubic$ or $Planar$ in the name of any image map. $Cubic$ is used for cubic mapping for textures and environment mapping for reflections. You can create your own cubic environment map by generating the six sides of a cube in Photoshop from photographed images. They should be taken from a camera position facing front, facing backwards, facing left, facing right, tilted up and tilted down from some home position. Each image must have the same resolution and must have a square aspect ratio (the same X and Y resolution). Append the six images into one six frame animation and apply it as a flat reflection map to your group. Rather than trying to take six 45 degree field of view images from the same exact location, you could try using a single wide angle image and then using Photoshop to generate the six approximate sides of a cube. When rendered, the approximation should not be very noticeable. The only thing you should be concerned with is making the edges of the cube match so that seams are not visible in the rendered image. $Planar$ is used for blurred refraction mapping for textures and environment mapping for reflections. To generate a refraction map, you can render an image of a scene from the same camera location with the group to be mapped turned off. The image should be of the same resolution as the final rendered image. Include $Planar$ in the name of the map and then apply it as a flat texture map to your group. By raising the Mip Factor, you can create the appearance of a fogged piece of glass. The map's scale should be 1.0 for each axis and the offset should be 0.0 for each axis. You can create the effect of refraction by changing the offset slightly. The offset value is scaled in pixels. ElectricImage CameraExtender Application CameraExtender is a small application which can be used to change some special features of Camera. The application modifies the ElectricImage Preferences file located in the Preferences folder in the System folder. When this program is launched, a modal dialog box appears with several options. When the OK button is clicked, the special features which have been changed are updated in the preferences file. Selecting Cancel will cause any changes to be ignored and the preferences file will not be modified. Default will set all the features to their default values. Revert will revert all of the features to their original values before CameraExtender was launched. Reserved Memory sets the amount of application memory set aside by Camera for use by System routines. When Camera is launched, it allocates all of the available application memory for its own use. The reserved memory will not be allocated and is available for use by system routines such as those used to display the preview window and status boxes. The reserve memory is never smaller than 64KB and free memory for Camera should never fall below 32KB. 128KB (131072 bytes) is the default reserved memory. You should never need to change this value unless you see Camera's free memory falling below 32KB (which has never happened to our knowledge). Motion Blur Shutter Angle adjusts the length of trails behind motion blurred points and lines. By default, EIAS uses 1.0 for this value which means that the length of each trail is the distance between the point or line's position at the last frame and its position at the current frame (0 in this box selects the default shutter angle of 1.0). To simulate the shutter of a more realistic camera, a value of less that one could be used. 0.5 would be a good approximation of many film cameras. Values greater than 1.0 or negative values can be used to create special effects. Preload All Texture & Reflection Maps, when checked, causes Camera from loading all mapping images into memory for each frame and prevents it from swapping them from the disk. This could force Camera to require more memory to render an image but may make the rendering go much faster if there are many different textures or reflections in the scene. If you have a large amount of memory allocated to Camera (over 32MB) this option is highly recommended because it could save you many hours of rendering time. If less memory is available, the decision will have to be made on a job by job basis. You should use this option whenever you think Camera is spending a lot of time swapping textures and can easily fit them all into memory. Store Polygon Normals In Memory is another feature which can eat up more memory but can give a small rendering speed increase. This option has the most effect when there are a lot of intersecting polygons in the scene. When this option is unchecked, the surface normal of each polygon is recalculated every time it is used. When checked, the surface normals are calculated once and stored in memory for each polygon. You could try using this option when you have 64MB or more allocated to Camera. Supersample Filter Transparencies will super sample all groups which have filter transparencies or transparency maps assigned to them. Supersampling must also be enabled in the Render Control dialog box for transparencies to be supersampled. Supersampling will cause textures to be appear to be four times as sharp but can take four times as long to render. This option should only be used when you wish to make a transparency map sharper. It should be used with caution because all transparent groups must be supersampled if this option is checked. ElectricImage Animation System Hidden Features ElectricImage In ElectricImage, if a group has Precision Shadows turned on and Cast Shadows turned off, the Gap and Transition values will be forced to zero. This will make shadows cast on the group by other groups begin very close to the surfaces casting the shadows. If the group is a ground plane, for example, a thin wall sitting on the ground will cast a shadow beginning precisely at its base. This trick can also be used for any group which does not need to cast a shadow on another group. Transporter In Transporter, hold down the Option key and click the 'Assume Correct Normals' check box to change it to 'Force Correct Normals'. This will cause all imported models to have their surface normals to point in the same direction. This does not prefer clockwise or counterclockwise orientation for polygon vertices, it simply matches polygon edges so that all polygons which touch will have the same orientation. This option is recommend because it does not take much time when importing models (you will probably not notice it) and it can improve the shading of many imported models which were created in AutoCAD or other programs which do not have clockwise or counterclockwise polygon vertex ordering. This option will probably be turned on all the time in the future version of Transporter and the Assume Correct Normals option will be removed. This option does not appear to have any negative side effects Projector Projector has some hidden features which can only be accessed by pasting in two new menu resources with ResEdit. Use ResEdit to open "Projector Menus" and Copy the menu resources. Open a copy of Projector and Paste the menus into it. Choose to replace the menus with the same IDs as the ones you are pasting. Save the changes to Projector and Quit ResEdit. When you run the modified Projector application, two of the pop-up menus will be different. The Alpha Channel Display menu in the Preferences window now has two new options: Invert will display the all four channels (Alpha, Red, Green and Blue) in the display buffer with the alpha channel inverted. This may be useful on some display boards which require inverted alpha channels for overlaying graphics over video. Display will draw only the alpha channel of the image as a gray scale picture on the screen. This is very useful for quickly viewing the alpha channel of the image. You can use the Overlay mode to view only the RGB channels of an image. The Alpha Mode menu in the Composite Image open file box now has six new options: Screen will mix the background image with the foreground image. Each channel is calculated as follows: result = (255 - background) * (255 - foreground) / 255. Multiply will mix the background image with the foreground image. Each channel is calculated as follows: result = background * foreground / 255. Screen Composite calculates the red, green and blue channels in the same way Screen does but the result alpha channel is chosen as the largest of the background and foreground alpha channels. Multiply Composite calculates the red, green and blue channels in the same way Multiply does but the result alpha channel is chosen as the largest of the background and foreground alpha channels. Screen Layer calculates the red, green and blue channels in the same way Screen does but the result alpha channel is chosen as the sum of the background and foreground alpha channels clipped at 255. Multiply Layer calculates the red, green and blue channels in the same way Multiply does but the result alpha channel is chosen as the sum of the background and foreground alpha channels clipped at 255. All EIAS Applications An FKey resource has always been present in all ElectricImage applications which can be used to grab an image of the entire screen which Shift-Command-9 is pressed. The FKey will create an EIAS Image file if the Caps Lock key is up and will produce a PICT file if the caps lock key is down. If the Control key is down, the cursor will not be hidden while the screen snapshot is taken. You may find this FKey useful if you are producing documentation or an animation of the screen contents. If multiple images are taken, each will be numbered. You can use the Edit command in Projector to append the individual frames into an animated sequence. You can use ResEdit to copy and paste the FKey into other applications or your System file so that it will be available at all times. You can also change the resource ID number so that it does not conflict with some other FKey you already have installed. There are public domain INITs which you can download which will allow you to activate this FKey while a menu is pulled down. Submitted by: Electric Image, Inc. 117 East Colorado Boulevard Suite 300 Pasadena, CA 91105 Voice: (818) 577-1627 Fax: (818) 577-2426 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:What are some of the hidden features in EI? A:This is an attempt to unveil all the hidden features in EIAS v2.7. Please fill in the missing ones. Ctrl - Go = Select which Camera to render with Alt - Windows Menu = Reset your windows into the default positions Alt - File Menu = Misc new features in the File menu Alt - Select = Misc new features in the Select menu Alt - Keyframe Menu = Misc new features in the Keyframe Menu Alt - Checkboxes, PopupMenus = Switch all checkboxes, menus for all the objects Alt - ObjectLock, Visibility = Switch all lockers, checkmarks under the selected objects Alt - Assume correct normals = Force correct normals Cmd - MoveIcon (pointing diagonal) = Adjust angel for the move tool constrains Cmd - PlugInIcon = show some information about the plugins Cmd - KeyframeCell = select entire row ~~parent = model cycling PageUp, PageDown (Shift) - Scroll Up, Down (fast) HomeKey = Zoom to all objects F2, F3, F4 = Cut, Copy, Paste Procedural Textures = $Clouds$, $Cubic$, $Mandelbrot$, $Noise$, $Stucco$, $Wood$ $Wave$ Each file saved from EIAS contain information about who saved it. Use ResEdit to look at it. Submitted by: Johan Aberg ingvar.ericson@programpaketet.se 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How can I use average anti-aliasing? A:Most of you probably know this trick, but it is good to keep in mind when you are rendering still images: The default anti-aliasing type in EI is "Oversample":. This anti-aliasing type is great for producing animated video for NTSC but causes still images to appear fuzzy. You can make the image appear sharper by using the "Adaptive" antialiasing mode, but to make the images really sharp, you should try "Average" antialiasing. You cannot select "Average" anti-aliasing from the pop-up menu in the Render Control window. Instead, select "Adaptive" and set both the Min and Max Adaptive Sampling Thresholds to 255. When you render your image, you will see that the anti-aliasing mode is now "Average" in Camera's information status window. The resulting image will appear a lot sharper than either Oversample or standard Adaptive anti-aliasing. This really helps when using EI for pre-press work or to build user-interface elements for use in multimedia. Submitted by: Mark Granger VP R&D Electric Image, Inc. 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Special Effects ------------------------------------------------------------------------------- Q:Where can I get flame and spark images or movies? A:The ElectriKinetics CD has nice animated looped flame movies. Editor's Note: It does have excellent flame animations. Very nice for doing torches, etc. not to mention some excellent tutorials and models, including a fully articulated skeleton. There is also a CD called Pyromania available. Submitted by: Anonymous 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How do I make colored light rays for stained glass? A:Use a projection map in version 2.8. Render an image from the position of the light. Match the cone angle of the camera to that of the light. Then use this image as a projection map for the light. Enable volumetrics and you're set. Shadows may need to off for the projector light to get the beams through the window. The is a tutorial on this on the 2.8 CD. Submitted by: Anonymous 11/19/98 22:27:30 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Texture Mapping ------------------------------------------------------------------------------- Q:Can I create environment reflection maps myself? How? A:This is a trick I picked up on the AOL forum almost a year ago. It has saved me many hours of rendering time, and allowed me to make short deadline projects look better than they would have otherwise. (These would be the maps that actually reflect their surroundings, not "Weird Hi-con".) There is a nifty thing you can do when using environmental reflection maps if you don't need the reflection to change over time. If you set your Data folder to a custom location then render an image that contains objects with environmental reflection maps, Camera will put files called environment0.temp, environment1.temp etc there as it renders. You can then pause Camera and rename these files $cubic$anyname. Later, put them in the reflection map slot on the texture map window. Now camera will use your $cubic$ environmental map each time without having to re-render six images for the reflection in every frame. This speeds things up tremendously. The cubic reflection map is just six square image files in a row, so you can easily build your own cubic environmental reflections. Just make a six frame animation file with Projector and name it $cubic$something. I always forget to write down the order of the images for each face, but it only takes a few minutes of testing to figure it out. I've built up a small library of environmental reflections that can be used in many situations; things like Living Room Reflection, Desert Reflection, Orbiting Earth Reflection, Industrial Reflection, High Altitude Reflection etc. I seem to remember that this was not intentionally designed into the program. It just turned out that because of the way EI works, you can trick it into using custom files for the environmental reflection map. Thus there is no documentation of this "feature". It sure would be nice if we had this kind of cubic mapping for all textures, not just reflections. I'm no programmer, but it seems like most of the coding for such a thing must be done already since the reflection thing works so well. Submitted by: Frank McIntyre 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Q:How can I make a moving chrome sheen? A:>>Just for the record, environment maps cannot be animated in terms of rotating the map over time. You can, however, use an Image animation file of any length as an animation map<< It is a lot easier (and saves disc space) to link all your objects, including camera and lights, to an effector (null obj.) and rotate the effector once you have set up the animation. This will give you that nice "moving sheen" even when the flying logo or whatever you are animating is standing still (in relation to your camera). Submitted by: Mario Fuchs Softmotion Stockholm, Sweden 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------------------------------------------------------------------------------- Video & Film Output ------------------------------------------------------------------------------- Q:What is field interlacing? How do I do it? A:This is an issue that has confused so many people (including myself at one time) it should just be uploaded weekly. What is field interlacing? Televisions draw their screen images in two fields (or passes). The image is drawn every other line and then goes back and draws the other lines, (even/odd, even/odd, and so on and so on....) faster than the human eye can see. This dates back to black and white televisions. When some of the first tvs were created there were not two separate fields and the screen was drawn from top to bottom. This caused a problem because the top of the screen image would start to fade before the bottom of the image. Hence someone (Mr. TV) came up with the idea of using two passes to draw the image. (enter present day) Does my computer screen draw in two passes? No. The computer monitor is of better quality and can draw pictures from top to bottom with no problem. (HD tvs can do this also.) This becomes important later. If video is 29.97 FPS, why does everyone suggest rendering at 59.94 FPS? AH HA. When you render your animation non-interlaced you are creating frames at 29.97 FPS non-interlaced. Hence, there is only enough information for one field. (Actually the television will try and fill in the gaps to the best of its ability.) The reason for rendering at 59.94 FPS is to give After Effects (or other converters) enough information to build both FULL fields. (Render out of AE at 29.97 with interlacing on. Media 100 users should choose upper fields first.) As a side note. When you render from EI as interlaced, it should take about as long as it would take to render at 59.94. Although I receive better results rendering at 59.94 and then field rendering in After Effects. If you are creating animations for the web, kiosks, or CD-ROMS render at 29.97 FPS and forget about interlacing. With today's tvs does interlacing really matter? The jury seems to be out on that one. In MY experience rendering with enough information for both fields produces a much cleaner/sharper image. I will not render out to tape at 29.97 non-interlaced. The choice is yours. I don't have the time to render at 59.94. Uh.....okay. There is another choice to save time. Some people render their animations at 24 FPS and then use After Effects with a 3:2 pulldown. (24 FPS being films frame rate.) This can create a unique look but doesn't always look that great with things like flying logos. Cut the techno stuff, how do I get the best looking animations to tape. 1. Render out of Electric Image at 59.94 FPS. 2. Import your animation into After Effects with a comp set to 29.97 FPS. 3. Render out of After Effects at 29.97 FPS with Field Interlacing chosen. (You may need to experiment to find out if you need Upper Fields first or Lower Fields first. It can vary from output board to output board.) 4. Lay off to tape. If you (or anyone) has any questions/comments fire away. Submitted by: Jason Hill Creative Director Digital Imagery Productions jhill@ns.net 1/1/97 00:00:00 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - END OF FAQ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - End of the EIAS FAQ Last updated: 11/19/98 22:27:30 This document was automatically generated by web server software developed by LAFFEY Computer Imaging. Updates to this document can be retrieved from LAFFEY Computer Imaging: http://www.laffeycomputer.com Special thanks to Donovan Rittenbach for his efforts toward this FAQ. Visit his website at: http://www.sirius.com/~zen7