verb

I answered my own question. 1. Collect all the job files into a folder using the "collect" function within animator 2. Create a new sparse disk image using "disk utility" (mac) make it a decent size it'll need to be big enough to take the final rendered files (sparse disk image mean's the actual image file will only be as large as the stuff you put on it but can contain as much data as you want - you need to set the size of the disk when you're setting it up in disk utility) 3. Put the collected job on the disk image 4. open the job from the disk image and setup your network render using the current machine you're working on, save the output file to the new disk image. 5. copy the disk image to the new master machine 6. copy the folder "eias app folder/renderama jobs/your new output job" to the same directory in the new master machine 7. open renderama and run the job Note. if you're using the master to also render, you'll need to set this up as a slave - it won't work if you use the "local" slave option in renderama, this is because any shaders won't be found in the local machine because it'll have a different disk location to the original machine, where as actual slaves seem to have all files copied to perform the render.

So, I'm hoping someone may be able to help. I'm working off site for the holidays - obviously with my USB key. I've got a large job ready to render. I've got my render machines back at my studio 1000 km's away (that's 1600 miles for some of you) Can I set up a network job on one computer then move it to another computer to be the master? I don't want to leave my laptop awake for the next 3 days, I can remote access my studio machines so would rather just leave them running while I lounge by the pool!

yep, I completely deleted the other version. The problem seems to be with the project file. I merged the old project into a new one and the jobs are now showing up so I guess it's sorted. But it would be good to know why it's done this...

Is there any reason why exported network jobs aren't showing up in the job queue in Renderama? I've tried both 10.9 and 10.8 machines running version 9.1 & 9.0 of eias but I can't get the jobs to show in the queue. I've come back to a job after a few months as the client wants more videos (yay) but I can't network render these out. From memory all I do is select the "enable network rendering" check box, save the location and the job shows up in the queue, am I missing something here? thanks for any help!!

1. The first thing you have to do is set up an account with Amazon, or login with your existing account. 2. Follow this guys guide on setting up the virtual machine (http://www.judpratt....ec2-renderfarm/) I think you only need to watch the first one or two videos, make sure you learn how to create an AMI image, you use this once you've finished setting up the first slave. He's setting up a different render farm so you can ignore everything past the first or second video. If you want to keep it free be sure to read (http://aws.amazon.com/free/) and make sure you understand what is free and what is not. There are plenty of CPU options if you want to pay, kind of handy if you're in a rush and need to finish a job. One you've set up your first slave you can create as many nodes as you like and you can set up new nodes with different CPU's, you're not locked into the cpu you set up on the first instance. 3. Set up your security group to only allow for the remote admin port, and the slave port, you do this under "security groups" in your Amazon services dashboard, Jud briefly goes into this on his video. 4. Log into your windows machine, Jud explains how to connect so if you watch his videos it should be pretty easy. I'm on os x so I use CoRD (http://cord.sourceforge.net) great program it works a treat. 5. Once you're logged into windows use internet explorer to download a better browser ;) 6. download eias 9 7. Create a short cut for slave so that the application opens on start-up. (I don't use windows and can't remember how to do it - google it) 8. go into the start menu and go to the control panel, choose security/firewall and turn it off. I couldn't get slave to get through it so gave up on it. Besides you control access to the machine using amazon security groups. 9. test access to the machine, type the DNS name of your machine into renderama with the port number (the same url you use to connect to the virtual machine, Jud shows you how to find this) Make sure your new slave renders your frame :) 10. Once you've got your machine rendering, make sure there's not camera control file in the EIAS folder (I assume everyone knows how to find this file) and log out of the remote desktop app. 11. Find your new machine instance in your EC2 control dashboard, right click or control click the instance (it can be running) and choose "Create Image (EBS AMI)" give it a name like EIAS render node or whatever. This will capture an image of your set up slave machine. Give it 10 minutes. 12. Go to "images/AMI" in your amazon EC2 dashboard and find your new machine image, Right click/control click the image and choose "Launch". This will bring up a dialog box, the "number of instances" is the number of slaves, so if you want 10, make it 10. Instance type: "T1 Micro" is the only free one. It has a geek bench score of about 2500, but only 650 mb of ram, so if you need more ram you'll have to pay. continue on until you get to "configure firewall", and choose the setting you created with the renderama ports, you might only have one setting/option here. Continue on to review and press "Launch". This will launch 10 new instances (or whatever number you created) 13. Go to "instances" in your amazon EC2 dashboard, there you'll find your 10 slaves, give them a few minutes to boot up, you'll need to right click each one and choose "connect" to get the DNS name to type into renderama, these machines will be identical to the first slave you made so should work the same, you might need to give it 5 minutes for each machine to boot up and open the slave app. 14. If renderama doesn't connect or can't connect, log into each machine and make sure the app has opened and your slave short cut is working. I've only been using this for 24 hours so don't have much trouble shooting experience, but at this stage If I'm in a pinch to finish a job, I think this is a great option. The only other thing I'd say is if you've got a lot of data (models, texture maps etc) for renderama to upload to each slave, I recommend setting up the first slave with the job already uploaded to the slave to save bandwidth and the time to upload. Lastly, you get 730 hours per month, so if you set up 10 machines, thats 73 hours with 10 machines running, So when you've finish with the machines be sure to shut them down by right clicking on each of the instances and choosing "stop" otherwise you'll be up for a hefty bill. Have fun.

after further research you get 730 hours per month, so if you have 16 machines that's 45.625 hours of free rendering time per month, worth considering if you've got a job you're trying to get out the door.

Okay, I got this working it was a problem with the virtual machine's firewall settings, turning off the firewall within the virtual machine fixed the problem (the amazon firewall is still running) Anyone interested in this, I followed this dudes post to get the basic virtual machine set up and running (http://www.judpratt.com/tutorials/ec2-renderfarm/) then simply downloaded and installed eias within the virtual machine, set up the slave, and entered the amazon dns name and port into renderama, it now works a treat. I think Amazon have a free service for a year, so far I've set up 4 machines, each machine has a geek bench score of about 2500, my mac pro is about 14000 so 4 machines is kind of like adding an extra machine to my work flow, I'm not sure yet what the limit is for a free user, but I'm going to find out :)

Has anyone else tried to set up rendering cores on an Amazon EC2 virtual machine? They have a year worth of free computing, I thought I'd give it a go. I've been able to set up the machine and install eias (on windows server 2008) but rama can't seem to access the machine, I've opended all TCP and UDP ports with no luck.

awesome thanks S. Houtzager! And Tom, thanks for answering all my never ending novice questions :)

Hi Tom, I'm changing the timing in the render dialog - is this the wrong place to do it? I did just noticed a "render: all frames" or "range of frames" is this how I should be rendering? sorry if I've wasted your time!

I set up a hackintosh a few weeks ago with a 3770k processor for $680 to use as a render box. In terms of rendering it's only a tiny bit slower than my 2010 8 core Mac Pro so I couldn't be happier. It's been sitting there for 3 weeks smashing away frames with no complaints what so ever - running mountain lion 10.8.2. Anyone interested, I used this guys build and guide as a basis (including the nice bitfenix prodigy case)

Not really sure if anyone else has come against this problem, I'm rendering a large output into several small chunks i.e. 0 - 2 secs 2 - 4 secs etc. The problem I'm having is the video I'm using as a texture map always starts on frame 1, no matter when I start the rendering in the animator timeline. So, if I start rendering at 6 seconds, the video texture always starts at 0 seconds. So when I sync up all the different chunks in After Effects the texture jumps around instead of a nice smooth playback. It would be awesome if the video texture started at the same time/frame as the render. So, if I render a frame at 6 sec camera also starts the video file at 6 seconds. Anyway maybe a feature for a future release.

Awesome thanks so much for your help Tom!

I want to create depth maps with anti-aliasing, I found this old post in another forum where one of the users explains how to create depth maps using luminous white objects and a black fog. So I made all my objects luminous white (I assume I do this luminance/glow tab) and made a black fog in the "world info" dialog using linear mode. I basically just get a black image, I'm obviously messing this up! Any help would be appreciated :)

I've also been doing some tests as I've got a large project to render with limited resources. In my testing, it's quite a bit quicker to allocate 1 camera to each CPU thread - in my case an 8 core mac pro so 16 threads. (I also tested 1 camera per core but 1 per thread is significantly faster) Here are my test results - 16 Cameras given 1 CPU thread per camera Allocated 1 camera to 1 single thread, Frame 78 took 2 hours 36 minutes to render 720p at 3/4 size. This config gives me 16 frames in 156 minutes. - 1 camera given all 16 threads - multi core Allocated 1 camera the full 16 threads, frame 78 took 18min 42 seconds to render 720p 3/4. This gives me 16 frames in 283 minutes So if the multi core camera rendered 16 frames it would have taken 283 minutes. Whereas allocating 1 camera to each CPU thread x 16 threads give me 16 frames in 156 minutes - almost 50% less time. Of cause this is based on one specific frame - 78, each frame will have a different render time. This was all done on a mac 10.8.2 in 32bit mode. Ram allocation didn't play much of a part - the scene only uses about 250mb. It might be different on a PC. Anyway feel free to pick holes in my crappy test situation, but until I'm shown this is wrong, I'll be rendering 1 camera per thread!

Okay, finally got round to testing this. You are correct! The still frames from rama import perfectly before they're stitched. I'll keep rendering straight from animator until it's fixed...

Yes, so changing the energy saver prefs so the computers don't sleep works fine, but it would be better if the app kept the machine awake - having 3 or 4 slaves sitting around on 24/7 is a drain on power.

Anyone else having issues with slaves under mountain lion? If a slave is running OS X 10.6 the computer doesn't fall asleep while the camera is rendering and the whole operation works as expected. If a slave is running OS X 10.8 (10.8.1 and 10.8.2) the slave falls asleep and the render errors in renderama, the slave camera continue to renders when it's woken, however it is no longer talking to it's master thus renderama no longer uses the slave. The master doesn't fall asleep during a render.

Hi Tom, yep all IP address match on restart. I deleted the preferences on the host machine for renderama and that seems to have fixed it, all machines now rendering :)

thanks Tom, I'll try deleting prefs. The machines are connected with ethernet connected using manual addresses - I don't see this being a problem because all other services work and have working for years.

Yep, all the machine's have a static ip address. the weird thing is, each slave will work if it's first in the list so I know it's kind of working.

I've been using renderama for days with no problems on 3 or 4 machines. I even got it working from work to home using port forwarding. Today however renderama is only using one slave - the first one in the list, and I can't fix it. If I delete the first slave, and stop and start the render it moves to the next one in the list but won't just use them all at once. I've attached a screen grab of renderama only using one slave and showing the other lazy idle slaves just sitting there. Is there a preference somewhere or a way to reset the application? I've restarted all the machines still no luck, completely deleted eias and reinstalled it on the host machine with no luck, any ideas?

no problem I'll shoot one over in the next day or so...

I have found the source of this, When rendering the z-depth file with renderama, the file is only 28mb and after effects won't import it. When rendering straight out of animator the file is over 700mb and it's works a treat. When I open both files in animator, they look identical, perhaps renderama isn't telling the camera to render all the 32bit 3d depth information and associated channels? Not really too much of a problem given the z-depth files are quick to render if you turn off retracing and thus the animation doesn't need to be network rendered - it's pretty quick on my local machine.

I'm having problems getting After Effects to read my image.img files with image.eiz files. I import the .img file and the associated .eiz file is in the same folder with the same name (different extension) but I get an error - "After Effects error: Bad Auxiliary info ( 46:: 32 ) I can get it to work if both the .img and .eiz files are single frame files, but as soon as I render each as an animation I get this problem. Any Ideas what I'm doing wrong?