You could do multiple light passes, render the scene’s CG passes (including albedo, shadows, ao, and all utilities) then mix that up the way you would if it was given to you as CG renders
3D shapes are great to create gobos (fake a window …)
1 Like
The beauty and chaos and the beautiful chaos of the BarberChair render in Flame amazes me just as much as the dentist’s chair that I once saw attached to the back of an open-aired lorry in Havana a few years ago, with someone on it smoking a cigar, that I did not get time to photograph but only see. I also-lutely know that I saw it and, like Stefan’s usage, it has never left my memory. As with Andy’s post, I love this R&D notion, and this is something that I love about the romance of hacking-Flame-artistry. I love seeing how people like Stefan “hack” the node structures outside of the design of the software but that the software simultaneously celebrates by dint of its design as software that asks and demands and provides creative bonuses in that stretching. Stefan explains very well in part 3 with that cirque du soleil demo the sheer power of the upstream control and the downstream monitoring in ways that I have seen but I see here in a very interesting way, and if Flame can integrate these aspects in ways that others have suggested, in terms of access to render pipelines inside of Action, then Flame stretches further that sense of housing many aspects inside itself, rather than being finishing or being a terminal aspect. If Flame harnesses, say, the cores of the upcoming M2’s/M’3s/M4’s, etc. to “parallel” the cores/nodes of Action/Batch (i.e. parallel cores, rather than serial cores) then there could be some exciting pipelines between nodes that make interactivity more fluid and less static and I can imagine that Flame could be more like a live FlaMaya that dissolves or bridges (not breaks) the barriers. The “R&D” of people artists like @Stefan point the way that the software should work to travel much further. @randy provocatively asked what a Flame artist was (today) but people like @Stefan and many on this board (I was blown way by @GPM OneFrameOfWhite) point the way in terms of stretching Flame and Flame, itself, should, in exchange, always work to stretch itself further to further stretch the strange ways that it itself finds itself being used.
Whether it integrates outside nodes inside (Matchbox, OFX) or builds its own, opening itself up to the outside of itself is the only way that it will keep its own Flame burning. Action needs the option of an output that outputs…?
1 Like
I agree with @GPM, seems ambitious if you want a general establishing shot! But you could try to get some cool shots (not too wide) on parts, areas …a pieces of train, structure of the station, light and shadows between pillars … maybe adding dust (floating particles) and a bit of light atmos (foggy, smoky) using depth data or passes?
You’ll need a powerful machine too
I think it would be a great exercise to try to pull off - and you’d learn a ton of stuff through it. (Like OFOW!) But if your client is expecting photoreal….
1 Like
Did something similar a few weeks ago. Client wanted to replace all drink glasses with branded paper cups. Pulled a model off turbosquid and altered the diffuse map to feature the brand color and logo. Lighting all came from the scene piped in as it’s own IBL. Worked pretty well when mimicking shadow changes. This wasn’t high end CG work by any means, but I was pretty excited by the results. Having a simple object like a paper cup helps. Any minor pass work was done using Object IDs. Stingray ambient occlusion and physical shaders were a huge help.
Train does sound complicated, but worth a try. Would object IDs help with breaking out areas for lighting?
2 Likes