Search the Community

It is a tool we wrote internally. It is a separate app that we wrote that is a basic converter to translate to the controller hardware we use, and we added some features to help with the real world issues of acceleration and deceleration limits of the actuators. I hope to at some point develop this stuff into an A:M plugin(s) that allows me to see in realtime on a timeline an overlay of an acceleration curve with the current position over time spline. This would allow me to animate to the limits preset into the plugin and make sure that nothing is moving in a way it can't in reality without having to save out and bring back in. When we did this during our first tests we just created an Excel project that did all of this conversion.

scenes from un-aired pilot for a TV series starring William Schallert and trouble-making cartoon character named "Philbert." This was primarily a project of Friz Freleng. Live action directed by Richard Donner, later of "The Omen," "Superman" and "Lethal Weapon" fame.

This became a good news, bad news, good news story. I successfully reproduced your example in v19e. GOOD Used the technique in a car steering project but every time the project was saved, closed and re-opened, the constraint aiming the steering bone roll handle at the null went wacky and needed repairing meaning it couldn't work using Netrender. BAD Used your technique to properly position a null under the car's front axle, then constrained a second null to translate to and orient like the front axle null, then constrained the roll handle of the steering bone to point at the second null. This always survived the project close and re-open. GOOD Thanks again for your efforts Robert.

I have a question that I have just begun to notice. I have applied ver 19j and now each time I save a project I have a new set of folders for each save. Each save has several sub folder. The main folder is titled the name of the project and then a series of numbers that I could not decode. Under the folder there are usually two folders. The first is a capital letter such as E with a dash, so E-. Its contents are then followed with another folder with the name of the project and its folder has the project file. The second main folder under the top is labeled the name of the pc such as Steve-PC. Under its folder are several folders. one has a listing the the textures. One has a folder with all the decals, another has all the rotoscopes, another has maps, etc. Each save seems to vary with the content. As I load my project into AM it may take 3-5 minutes to load the file. I am wondering if I need to reinstall a earlier version. So my question to eveyone has anyone else noticed these folders as well? I thought it might be a problem with this machine but it is doing to my other copy of AM as well. Steve

I wonder if those views are stored in the Windows registry. Perhaps not there... perhaps somewhere else where we can manipulate it. And if so, perhaps a fun C++ project to put together.

Yessir, this was done using a linear workflow (working in 1.0 gamma until the final image...which is 2.2 gamma). I'll make a quick video tomorrow and post it along with the Project.

Yes, another fur/hair question. For each of these contests of ours I endeavor to learn/try something new. This time it's hair. I am at the point of trying to create an animal skin (what caveman would be caught out in public without his high fashion cave fur?). I'm trying to get the pattern of a decal to color the hair. I've followed the steps, but all I'm getting is a solid brown, not the spotting I'm looking for. my fur decal the result the test project ProjectTest.prj

Will post project after more tweaking is there is any interest and I fix the lighting

With apologies for file size. This is first assembly edit of project. Any critical feedback very welcome indeed. I've spotted somethings that will need to be altered but would be very keen to hear about any you may have noticed ? regards simon BEWARE !!! The sound is very grating Sequence 1.mov

Hi Paul, Very interesting project! I am also interested in printing bigger parts but have not found an affordable solution for me up to now. may i ask which printer do you use? Actually i am waiting for some service Like sculpteo and such offering bigger sizes Mike 1 to 1,5 meters..

Thought I would give an update on this project. I has been a magical mystery learning tour for sure. I've learnt a bunch on modeling in a:m and how that translates to the 3d printing world. At some point I will out a tutorial together, but for now I've got to plow forward with the seagull. The biggest challenge with this is it's shear size. Standing over 11 inches tall this guy is big. Add to that I need to model it hollow so I can fit the servos inside and this rapidly turns into a mechanical engineering project. At first I thought I could just model the body in halves and assemble from there. But that meant modeling in screw mounts and then breaking the body up into four pieces, making 8 prints total for the body alone. As a test I did one half and assembled to try and figure out servo locations. The result is this It quickly became clear that breaking up a model in am, then trying to reassemble I. The physical world wasn't an easy process. The solution is what I am embarking on now. Model the body as one piece, but model in trap doors on the side under the wings and the back. While I still have to work in screw holes, I'm not going to worry about threading since I will use brass gnurls that I can heat set into the pla plastic. For a little stress release I printed off a whole model of the seagull in tiny size. Here he is in all is plastic glory...

Yeah, my desk is a bit of a clutter. Basically my office is an homage to disney. The shelves and walls are covered in disney parks and movie props (spent way too much time and money in auctions!). As far as my seagull project, I've spent the week printing one half of the body. Each half had to be broken down into 3 pieces in order to fit on the printer. Each chunk is taking a full day to print, and gobbling up filament like it is going out of style. I've gone thru a whole spool already and just order two more to make sure I can get the body printed. I figure I will put the one half together, then dry fit servo mounts and build a basic arduino controller to make sure everything is copacetic before moving on. Also added the wings in to the model, and figured what the hell, I'll animate those as well. That makes movement on the head, beak, legs, wings, and tail to rig and program. Might get all that done before the snow flies!

It's been quite a while since I've been on the forums. I've been heads down on my latest 3D printing project. This one is actually the one that got me into 3D printing to begin with. If any of you have been to Disney World, Epcot in Orlando you might remember the Finding Nemo ride, and the animatronic seagulls outside the entrance. I have been wanting to build one of these for the longest times, and this project was the reason for buying the 3D printer I did, since it was the best I could afford that would print the largest model. So a few months ago I started modeling one in A:M and this is where I'm at so far. I've been able to print the head and beaks, which is broken up into four separate parts. The legs are in three parts. The body is too large to even print on my printer and will have to be broken up into at least four separate parts. On the animatronic front, it will have animated beak, head, tail, and legs. It will be a challenge for sure. The head alone I spent over two weeks on and printed at least five versions before I was satisfied to move on. The legs I got better on, and have only needed to do two copies. The body needs to have mounts perfectly aligned for all of the servos to add the movement to, and I'm sure I'm going to go through a spool of filament before I get it right. Once all that is done, I move onto the electronics where I need to design and build a controller and program that. It will be based on an arduino, we'll see how many a fry in the process!

It's a very simple test project. It is baffling because the same project worked before I changed it to render with AO. IT has to be the same version of Netrender since everything is coming from the AMRoot directory on the server

Is it a complex project or just a very simple one (start with the second of cause ). It really is just a guess but are you running the same version of A:M/Netrenderer on both systems? Are you using GPU features in the project? I think I could even run different versions of A:M, but I am just not sure at the moment. Best regards *Fuchur*

If a Project loads and renders OK in A:M, what makes a render node decide it is "Invalid"? The nodes that are on my server computer still render it fine, but the remote node can not. On the Render Server display the nodes is just perpetually stuck at "Loading frame" with no error reported.

I got the p2p services "started" The problem with Homegroups seemed to be that each computer had one established but they were not the same Homegroup. I killed one of them and joined the other and now I can indeed see shared folders remotely. However, my remote Netrender node still is "unable to load Project." I did this and now there is a user named "Remote01" on my main computer but... that doesn't seem to change anything about my remote computer.

Trying to finish off the Learning curve, using V18g in OSX 10.85. Made some modifications to one of the models, saved it as an iteration. Looked ok visually so tried to test it it as a render using the action window previously worked in. When I tried to do a quick render, the process seemed to stall, with the message appearing, "calculating the visible patches". That went on for some minutes and eventually aborted that render. Tried again, the render crashed the program. Opened up again, same project, same settings, same problem with calculating the visible patches. Went out to do some food shopping and leave it to get on with it. Just checked progress and, the render was going through but, instead of taking 2 seconds or less per frame to render, it was taking 20 minutes. Closed out AM, rebooted and just tried again, its back to taking 2 seconds or less per frame, and not ( touch wood) crashing out. Is this a known problem ? I've experienced it before with the extended calculation time but thought it must be an error on my part. This happened just after modifying the model but, only realtively minor mods saved after each change. regards simon

I think I see what is going on. You need to share the folder that holds the project and the the project will render too. All my projects are in a folder called "AM_Data" Right click the folder select "Properties" click on the sharing tab(Your use name for windows should be in the box as owner). Your name should be all you need if that is what you log into every computer on your network as the same log in name. But if you log in with different names on other computers you need to add that name too or "everyone"(There is risks to using everyone anyone on the network can over write or delete with out logging in) Now hit share This will give you a path to the folder(this is mine) \\jason-gaming-pc\AM_Data Now when you open a project on render server user that path. \\jason-gaming-pc\AM_Data\so\data\Act III\Seq03-Departures\Sc03 When you set the save path you also need to do the same thing. \\jason-gaming-pc\AM_Data\so_render\Act III\Seq03-Departures\Sc03

Where are the project files placed? Netrender will not share the project files (or textures, etc.) over the network or something. It needs to be accessable by every computer in your network by itself best by using the same network drive letter or you have to replace the path-string "c:/project/myproject/" by "g:/project/myproject/" or something like that. See 6.) for where to do that. Or can you already access it like that on your windows file explorer using the same path and it can not access that? Best regards *Fuchur*

So far, all of my remote nodes are still "unable to open project"

Chris, I'm a bit late to the feedback party but I enjoyed your short film. You kept me interested (with minor exception) to the very end.** Others have commented on the style/look and the one thing in that arena that I initially thought could use some finessing is the contrast. The idea being that if we were to place a dot of light on the object/area where we want the audience to look all other factors should support that. Easier said than done.. I know! In reviewing the short to look for specific examples to illustrate what I'm talking about I was a little surprised to see that your setup was much more on point that my memory had recalled. In particular, I thought you incorporated some nice use of silhouetted shapes and that helped in reading the action tremendously. I do think that the girls face could be lit more perhaps simply by adjusting the ambiance intensity. There are some instances where the girl looks rather sinister and while this might be intentional I'd say that the shadows/lighting should have a source and one in the film is very obvious... the robot/creature's face screen. The contrast then for purpose of mood might then largely be controlled by the background and environment which if darker would make the characters -pop- out of the scene*** Robert suggested Jeremy Brin's book and I wholeheartedly agree. If there is only one thing you get from his book however I believe it should be the methodology of turning off (or even deleting) all lights except one and then perfect that light to the greatest extent possible. When we think we have that light set ideally then turn it off (maybe renaming it in a way that helps us recall it's primary purpose. TIP: In current A:M releases we can Right Click and Duplicate Lights. This is a useful way to finesse lighting starting from a known location/setting. For instance, we might duplicate a primary light and move it slightly to the left or right and then adjust it's color or intensity. This is how such stylistic lighting as red hues on one side of characters face and blue on the other can be created. The primary light then lights the main part of the face so we can accentuate the facial expression and help it read. Moving on to the story a little. Love the shots/replays of the pratfalls echoing the main action as seen on the robot's screen. Nice. I really like the twist ending but agree that it isn't entirely clear what is happening. This is too bad because it's a nice surprise and I think unexpected by most in the audience (hence the reason some aren't quite following). How to accentuate? One idea might be to consider the girls eyes... especially after spending so much time peering into the -eye- of the robot creature. After the switching off of the girl (which could use a little more highlighting... perhaps zooming in to a closer close up?) a view of the girls eyes as they 'turn off'... changing from normal eyes to the black eyes in the subsequent scene. This could then be echoed in the other characters in the final shots... their black unliving eyes could also be emphasized as the camera moves into/through the scene. Lot's of opportunity to really sell the ending there! ** I seem to recall a few transitions that had me slightly losing interest. The one that stands out in my mind is where the interior changes to exterior just prior to the characters going outside. The establishing shot is important but I wonder how that might be played up a little so as not to slow down the pace too much... slowing is good... but it seems to slow a bit too much. One idea might be to start the film with the establishing shot of the exterior of the building so that when we see them exit later there is no question of where they are. They start in the building... exit and consider the stars and the heavens... and then... into the finale. This is a nice opportunity here as most of the film is inside the building. It is this climactic movement when they move outside of building that sets up the punchline/ending. *** I'm still not sure exactly what I might do with regard to background, lighting and such if this was my project but my initial thought is to render without a background so that the background can be manipulated independently (via compositing) to more easily control the mood and focus of the scene. For purists that want everything rendered -in camera- this can still be done by inserting a 'color card' at the back of the scene which could be manipulated; even with decaled images where needed. One reason that compositing works well with backgrounds is that blur can be added independently to control the DOF (Depth of Field) and pull the focus forward to where it is needed. A final thought regarding the comments on 'flatness'. It's always good to consider at least three levels of depth in any scene: - Background - Foreground - Character space It can also help to move the camera slightly off center and then point it at the characters. Then converging lines in the background, foreground and throughout the scene can have a chance to direct the viewer's eye to where it needs to be from shot to shot and scene to scene. Having the camera straight on (or perpendicular to the BG, FG and Character planes) is often best used for comedy whereas more dynamic camera angles enliven the environment by allowing the audience to feel as if they could move through the scene. A thought here is that your story starts out as light hearted and even comical... with the robot creature echoing the shots on his face screen... so those moments lend themselves to flatter camera setups. It might be interesting to see how you might transition from the flatter camera to more dynamic as the story unwinds to it's ending. All this... just random observations. Not necessarily suggesting anything should be implemented. Bottom line: Nice job. I was entertained. Keep up the good work and I'll look forward to your future animations.

Hi Robert, a good starting point might be that: It is a shame, that the pdfs are not available anymore here, but there is still nice information in there. And here the PDFs are still available: Important: All computers need to be in the same network of cause! It is very helpful, if you have the render-data at a place where every computer can access it. Best is a NAS or another type of network drive. If you do not have something like that, you can still copy the data (A:M files, textures, etc.) over to each computer on the same position like for instance "C:/am_project/my_project.prj". Important: Do not copy it on your desktop, since the path will vary from computer to computer if the computer user name is different(!). In general it is pretty easy: 1.) Start RenderServer_64.exe on one of your computers. Make sure it has at least one Core assigned, which is not occupied by a Messenger/Slave. Theoretically it is not necessary, but I think it is the best option. 2.) Now you should read out the IP-address and Port of this computer / the Render Server. You should be able to see that by going to "Tools > Options > Connection". 3.) You can use the other cores of the computer (renderserver) to render too. 4.) Double-Click on RenderMessenger_64.exe in the A:M folder and use the wizard. There your NetRender Server should be automatically set to the IP address you got from step 2. If not, put it in. Same with the port. 5.) On "User defined data" you do not have to put something in, but it is a good idea to write something in that helps you to identify the computer with. That makes it easy to find out, if something is happening with that computer (this is not very hard to guess if you have only 1 or 2 computers in use, but I did it with 12 computers before and than this is really helpful). 6.) If the path of the project files (etc.) is different for the computers (so you could not use a NAS / network drive with the same letter (P:/, etc.) on the computers or you could not copy the files to the same position, the next step can help you. I would try to avoid it, but you can try it out here. 7.) Now start the next computer and install A:M on it. You do not need an own license for that computer or something. After install, just click on "RenderMessenger_64.exe" and follow the Wizard here too. Be sure to put in the right IP and Port of the RenderServer now. 8.) Check the connection of the Rendermessenger. You can see that, in the task bar (next to the clock on Windows). If the icon shows a connection (both "lines are next to eachother without a gab inbetween). If that worked, go back to the Renderserver and have a look. In the lower area of the RenderServer-Window you should now see the other computer-Netrenderslaves and the one of your own computer. 9.) If that is the case, the connection is working and you should be pretty ready to go :). After that just open Render Messenger on both computers as often as you've got cores (or virtual cores if you use HT/SMT). Only on the Netserver computer leave one to the server. So if you have a 4 core computer with HyperThreading (HT, Intel) or Simultaneous Multithreading(SMT, AMD) you can use 8 instances on the other computer and 7 on the the server (if the server is a 4 core too). 10.) Now on the Render Server go to "File > New". This will open a new window (Pool1 or something like that). In that at the top right-click and choose "New Job". A wizard will pop up. Give it a name (this is only for you if you render a couple of different jobs in a row for instance you can see what you do there... for instance "Peter's day out - Walk animation Scene 01 - Take 01" or something like that is a good idea, especially since I leave the jobs in there in general even if they have finished. Browse for the project-file (or chor) on the next button. Be sure, that the path is available on all computers in the same way (or use the option of 6.). Select which camera and chor in the project you want to use below that. In general you can leave OpenMP off and GPU is only needed if you use for instance FakeAO/SSAO or something like that. But first I would just try it with an very easy project. 11.) Next put in the render settings you want or use a preset. At the path at "Filename" make sure that the path is available for any computer. You can render to your own c:/ or a net-drive. For the net drive be sure you can access that from any computer. Anything else should be clear. It is mostly the normal render settings you already should know. 12.) After that the new job should be available in the Pool1-Wndows. Now save the job on the server somewhere. (or a net drive... this does not need to be available to other computers) 13.) Now drag and drop the Instances in the lower area of the Window to the lower part of the Pool1 window. After that, the Netmessengers on the computers should start Renderslaves and render the things. You should be able to watch that on the computer displays and you should be able to see that on the REnder Server too. It refreshes every few settings. If something is not working, you should get an information about errors and so on. If you get some, check what it might be on the computer or by having a look at the render server. Notice: In the Pool1 Window the right mouse button is key. There are a couple of different options there which are really helpful (and pretty straight forward). I hope I could help a little. If you have any kind of question or problems let us know. Best regards *Fuchur* PS: And this might be interesting for a lot of cores too: https://www.patchwork3d.de/start-am-netrender-185-en