I’ve been wanting to learn how to create CG elements and work more efficiently within in 3D space through flame, but it seems that nobody has budgets and/or time to ever utilize these skills. I’ve always wanted to take my flame skills beyond what I know, but is it worth it to take the next step in flame? I’m always finding myself in the position of clients just wanting things done fast and perfect for cheap.
Does anyone have any thoughts?
I did some 3D in Flame yesterday for some previs. I haven’t used it for 3D that much and I cheated by using free turbo squid models.
I often find myself getting to a point where I can’t progress any further. In this example I seemed to run out of shadows. As I increased the number of rectangular flats my shadows stopped working.
I was having a play with the stingray stuff and was loving the ambient occlusion but that was introducing banding on my sky dome and I couldn’t fix it.
I get the same with particles, yet I have seen people do some amazing work with Flame particles so I am sure that someone out there would do some amazing stuff with Flame in 3D. I just can’t guarantee that I can get it beyond that 70% mark. And in a client session, that troubles me.
I guess I am lucky having access to a 3D department. Maybe the necessity hasn’t been there to push me over that threshold.
I’ll follow this post with interest.
That’s quite cool! I know what you mean about getting stuck. I personally hate getting a shot that I know needs CG work and having to turn it away (I don’t have access to a 3D department).
One day I hope there’s CG integration that utilizes AI to complete 3D images in our footage. Like, yes, computer I removed the nike logo, now let’s get a solid track in 3D space and emulate fabric creases based on the previous frames of movement and interaction!
I think whilst doing some cg flame it’s cool just to know you way around some basics, it’s just not the place to do heavy lifting. It just isn’t. I’m adventurous, and try to use Substance and Particles and IBLs whenever I can. But, its hard to justify being half as good for twice the price for something that is so easily done elsewhere.
And there are plenty of companies to partner with on the cg side. There’s no reason to turn anything down for that reason anymore.
Yeah I do a fair bit of 3D in flame but unfortunately agree with Randy. It’s pretty time consuming getting somewhat close to 3D guys can get out but sometimes you gotta do what you gotta do.
The shadow issue is a biggie - related directly to gfx card from what I can tell and just turn off when comp too heavy for GPU. Shadow and AO aliasing can be turned up to improve quality. I also find I usually end up breaking action comp out to seperate passes and pre render to get over the line
A month ago, I did a demo animation for a client in Flame, and they loved the look of it so much that I decided to do the final version in Flame as well. My 3D artist did the modeling part in Houdini, and I did the animation/lighting/texturing/rendering part in Flame. We had only one week to deliver. This is the end product: Liszt Fest TVC, 2021 on Vimeo
I’ve done even more complex animations in the past with Flame when we were limited by tight deadlines, and the client was short on budget. I believe the key part is you have to have a clear understanding of the client’s needs and thoroughly think it over if it won’t reach the limitations of Flame.
I’ve tried 3D in flame. To be honest I think it’s only good for some tweaking of stuff (@kmor ’s post excepted). But for intricate stuff it’s pretty unwieldy. Also making geometries out of 3D shape is pretty limited. And even vector drawing with tracer masks is limited compared to illustrator. Flame is a Jack of all trades or image integrator as it used to say in the splash screen. That’s its strength.
I have played with Houdini but that package is full of nonsense like flame and I just can’t be bothered.
If I was to consider learning to do cg I think I would try unreal. I’ve done a few jobs with unreal and it’s nice because it have that interactivity that flame does and the ops have a similar approach to me. There’s much less of that boffin stand off ish “well that’s not scientifically correct” vibe.
I’ve done the odd bit of very basic 3d in Flame (boxes in packshots… i can texture with photos from box artwork, then add some subtle camera movement).
Anything more complex then i will model and render it in Cinema4d. If its beyond my limited skills, then send it out.
If you want to learn 3d modelling but don’t have the budget for C4d / Maya etc, then try Blender. Its been around at least 15yrs, has a very active community and i think its free (but i may be wrong). Those who use it swear by it.
I do a lot of mograph cg in flame, using 3d text, generic primitives or the occasional purchased model…in that regard, Flame is pretty good, especially for textures and lighting.
But I never really “model” anything in Flame other than extruding basic shapes…you can usually find something on TSquid to get the job done.
Remember when the refraction fx were added a couple years ago? My reel was very “glassy” that year, lol!
I like the stylised Budapest - especially the river Donau. I think they’ve used maps data. I looks fantastic!
Blender is my second option after 3d in comp have reached its limit and before asking the CG artist to take over. Evee is Blender’s real time render engine, and is amazing what it can do.
Also tried to push Flame in a 3d sense, but these days I just have blender installed on my box. It’s amazing what you can achieve quite quickly when you get into it - especially with Eevee. The 3d tracking’s not bad either. Just had a job where they needed a game like sequence for one of the shots, and managed to whip it up in a day using kitbashing. As much as I love Houdini, blender really is a “get it done” type package. You can smush things together and the viewport is pretty much WYSIWYG.
Wow @kmor I didn’t even know the depth in which you could push flame like that. You seamlessly integrated the 3D artist’s elements within the magic you created, awesome stuff!
Thank you to everybody for your extremely helpful advice, tips, and tricks! My takeaway is that there are much better programs out there that do 3D, and I would get the most out of using those programs for CG and 3D. I do want to explore the basic 3D tools a lot of people mentioned, like creating pack shots like @Lightningad mentioned - that would incredible.
The simplest way for me to do a box shot in Flame is :
Get box artwork, do screengrabs of each panel. Ensure you don’t rescale the image.
Make a note of the pixel dimension of each screengrab…this will in turn tell you the size to scale the model to.
In Flame, Action, import Fbx cube (the one with split faces). Select the main axis for the model and adjust the scaling for x,y,z to mimic the screengrab sizes (eg, 1240x650x75. Would become 124%x65%x7.5%.
Then import the screengrabs into action, and onne at a time delete the texture maps from the cube model. You’ll see the face go white, so can work out which panel needs to replace it, then attach the relevant screengrab.
You may need to tweak a few settings to get it just right, but that has worked well for lots of quick and dirty pack replacements on our tvcs.
That’s a great technique. I have to try it on my next box shot.
Now I’m going to be searching for a job that needs this to try it out!
and when you add a simple blurry squiggle as a reflection map…mind blown when the camera is animated to orbit the box!
I use a combination of blender and flame quite a lot. I got bored of the limitations of flame, and it’s helpful to learn a bit when having conversations with CG artists. It’s free, so you can install it on multiple machines and save your files in the cloud. There are tons of tutorials online so its easy to get started, and the investment of time has paid off for me.