I’m researching rendering engines for a animation project, along the lines of Elephants Dream (but longer), and I’d like to any hear from any users the pros and cons they may have encountered while using Indigo.:yes::no:
Any constructive responses would be appreciated. Thx. 
I haven’t used Indigo for very long, but I really can’t see how you could hope to use it on an animated film project. Its strength is doing extremely realistic still renders. To achieve that, a render might take 40 hours or more! Can you afford that length of time per frame?
If you wanted to take shortcuts and speed the render times up, I think you would basically lose any benefit Indigo would bring to the film - ie. extreme photorrealism. In that case, you really should go for a less realistic but more efficient renderer. Most renderers are created with animation in mind. Indigo wasn’t.
@lisae: Yes, the rendering times are long. However, one reason I’m doing this poll and P/C thread is that if Indigo does become the “renderer of choice” my colleagues and I will build a suitable render farm. We hope to “raise the bar” a bit in realism in an animated film. Feedback from Indigo users will help aid in our decision.
Thx for your info.
Well, I appreciate the fact that you have a goal like that, but before my co-workers and I ever take a 3D project on, we ask ourselves (or our clients) the question - “why should this be animation and not live action”?
If you want realism, it’s cheaper, faster and easier to film it as live action. You would only do an ultra-realistic 3D film if it is completely impossible to achieve what you want with live action. Can you do this with your film? If so, forget animating it. Don’t waste a big part of your life on something that is ultimately futile.
This is the reason you don’t see people routinely “raising the bar” in realism in animation. It’s a waste of time and money. Use animation on projects where the strengths and benefits of animation shine.
If you want realism, it’s cheaper, faster and easier to film it as live action.
Point well taken. But then again, if we don’t push the envelope occasionally, how can we know our limits? I’ve always considered “Final Fantasy: The Spirits Within” to be a significant landmark in the genre. Certainly well beyond any other CG animated film at the time.
Don’t waste a big part of your life on something that is ultimately futile.
LOL. Too late… I’m semi-retired after too many years in IT…Frankly, I and my friends have the time and resources to pursue this project and there are other “positive factors” involved but I’m not at liberty to discuss them as of yet.
I appreciate your candor however and would like to ask, do you currently work for a studio that does animation?
Yes I do, and I teach students as well, so I never mince words when I think someone might be on the verge of wasting their time
By the way, here is a very small taste of my latest indigo render. It could end up getting us a big commercial job, so I don’t want the full size version of it anywhere on the internet! But the scaled down version shows you enough. It’s a couple of CG diamonds, one clear and one yellow. I rendered them for around 45 hours.
One problem that has occurred to me with using Indigo on a film is that Indigo’s lighting and camera are so real that you need exactly the same skillset to use them properly as people who do lighting and photography/film for a living. I don’t think you could use Indigo as your renderer unless you consulted with some professionals on this. Other renderers are easier to fudge, but with Indigo I had to use a lot of tricks from studio photography to get the result I wanted on those diamonds, and it’s still not good enough.
That’s the problem with realism - you start needing a whole bunch of real world skills on top of your CG art skills!
Very nice rendering! Well worth the time for whatever reason! I hope you land a lucrative contract with your perspective client. 
Yes I do, and I teach students as well…
Interesting that you mention teaching. One aspect of our project is an involvement with a local 501C3 non-profit who’s Mission Statement includes fostering “Graphical Arts Education” through digital medium. Our association has not been finalized as we need to cover all the legal aspects but we’re very hopeful.
I had to use a lot of tricks from studio photography…
Another interesting point concerning the “photographic skills” aspect. To be frank, while using Indigo may include some challenges, I’m more motivated to move towards Indigo than away. I guess I can’t help enjoying the problem solving aspects… I’m starting to consider the possibility of developing tools during our project that might be of use to the Indigo community at large in surmounting some of the “difficulties” in using the renderer.
Of course, there is no substitute for experience…
BTW, we have considered Yafray, Kerkythea, 3Delight and Sunflow. As well, I’m a POVray hobbyist user from 'way back. Each has it’s strengths and weaknesses and at least in the case of Yafray, fewer “issues” due to it’s close integration in Blender. I’m quite certain that either the built-in Blender renderer or Yafray would suffice in producing “prototyping” animations and scenes, and we may well go that way initially.
But we’ll always know that we could have perhaps done better…
The education aspect sounds exciting :). I will hopefully be teaching Blender to some gifted high school students later in the year. Perhaps a little class on Indigo would be a good thing to include.
Prototyping some scenes with another renderer could definitely speed things up, but I think you will find that you’ll have to hack the .xml files by hand when you transfer over to Indigo, rather than just use your materials and lighting straight from your prototype scenes. I had to do that for the diamonds. I really just used Blender to output the model data and positions, then moved the camera and lights about by hand, changed the materials, etc. I found that outputting a mesh light from Blender really didn’t work well enough, so I changed it in the xml file to a rectangle light. As things stand, you can really only achieve basic renders if you export straight from Blender.
In fact, I have found that the integration with 3DS Max is much more useful than it is with Blender. The Indigo materials in Max have all the options I ended up setting by hand in the xml files when I used Blender. But of course, it’s much better to use Blender for an educational project than Max.
Perhaps a little class on Indigo would be a good thing to include.
Raise the bar Lisa!
but I think you will find that you’ll have to hack the .xml files by hand when you transfer over to Indigo, rather than just use your materials and lighting straight from your prototype scenes.
This was one of the tools I’ve started to consider. While a seamless integration with Blender would be nice, I’m thinking that finalizing the render settings in a “scene frame modification” phase would give finer control. As the export format is XML, another “positive” for Indigo in my book, and I’m familiar with XML and its related technologies from my IT experience, I would probably write a parser/editor and automate the changes…
I really just used Blender to output the model data and positions, then moved the camera and lights about by hand, changed the materials, etc.
I’m thinking that would be our main objective as well. I would want Blender to output the scene frames with all the basic geometry, lighting, etc and then automate the “final” Indigo XML element substitutions before rendering.
But of course, it’s much better to use Blender for an educational project than Max.
Another plus. I’ve been an Open Source advocate for many years and would like to pass along the “philosophy” if I can to others. I used to teach an evening LINUX class at the local college and a number of my former students remain Open Source converts to this day.
The downside to Indigo is that it appears to be an MS-centric application and I would have loved to see it on LINUX. Oh, well… can’t have everything I guess…
I rendered my diamonds in Ubuntu Linux!! All you have to do is type “wine indigo.exe path_to_xml_file” and it works!
I have some problem with it not saving its render properly, but all I did was take a screen cap of the image on the screen. I’ll sort out what’s going wrong there at some point.
All you have to do is type “wine indigo.exe path_to_xml_file” and it works!
You just made my night Lisa! Thx a MILLION!
I will have to look into this, (I’m a Debian user) and see if I can use Expect and some shell scripting to build the farm. Now I’m thinking a OPEN MOSIX modification to the kernal for load distribution and wham! instant render farm. I built a render farm a few years back this way that distributed POVray animation files throughout an 8 node network. Worked quite well…
If I can be of any assistance in solving your “save” issue, please let me know. I’ve been a LINUX user/admin/programmer for many years…
It’s midnight here and I must be off to bed, (like I’ll be able to sleep now!) and I’ll be gone all next week on business. Thx so much for your ideas and insights this evening! I’ll look forward to conversing with you again in about a week! Take care.
Have you tought about renderman ?
I think the only way that you could use Indigo in a practical way in animation is to render out background plates for a composite, alot of background scenes are static so only need to be rendered once.
Back from another grueling week on the road…
I have considered a “Renderman compliant” engine like 3Delight for this project. Actually, I’d love to use a RIB formatted scene file because the documentation for writing Renderman style shaders is extensive… :yes:
I noticed there was a plugin for Blender that exported an .rib file but sadly the project has gone by the wayside it seems… 
i would definitly not recomend indigo for an animation project. while i love indigo and find it very good for many of my uses, its not the most animation friendly. even with a 10 computer setup, you are looking at rendertimes from 2-5 hours to get the image grainless. in an aniation project you would generaly want x-8 hours max rendertime on a single computer with the point of a renderfarm being rendering individual frames. other than that, indigo is not the friendliest towards animations in general. in theory you would have to have a seperate file for each frame and each file making a new image. this would not be so bad if it were not for the fact that indigo renders for an indefinite amount of time, so you would have to cancle the renders manualy, currently there is no way to shut the render based on mutations. even if you could shut the render based on mutations, the grains in each image would end up being different enough that you would get an odd effect. also, indigo is geared towards more technical and mechanical renderings… while it does have the ability to do sss and organic effects, the renderer doesnt necessarily lend itself towards a more artistic style.
on the other hand, i have found Gelato by nvidia to be a perfect solution in many cases. like rib you can write shaders, and even import renderman shaders into the gelato format. gelato will take advantage of the video card as well as taking advantage of the cpu, which makes it very fast. also, the blender to gelato script is very good now. if you know how to work with renderman, gelato is very easy to learn, and also easy to learn if you dont know how to work with renderman.
Sirius,
there is a skilled programmer who is still improving the Renderman (Aqsis) exporter.
LetterRip
there is a skilled programmer who is still improving the Renderman (Aqsis) exporter.
Thx for the heads up LR… I’ll do some investigating… A Renderman compliant 'tracer would be a plus along with a maintained exporter…
[EDIT] LR, are you refering to Neqsus? http://www2.stormwind-studios.com:5150/metadot/index.pl?id=2015
It’s looking, as lisae and oodmb have pointed out, that Indigo would be a poor choice indeed for the application we want to use it for…:no:
