Things are changing

[Devin Johnston]

After a month of tinkering with Unreal I've come to realize that 3d Studio is a dinosaur, the display navigation alone is archaic not to mention performance. I'm actually hating the thought of having to do anything in it now, adding objects and adjusting lights in Unreal is so much more intuitive. I'm still not very comfortable with Blueprint or Matinee but even though it's more complex being able to see your results immediately is a big positive. The only real downside is the time it takes to prep the model in Max, once Chaos comes out with their live link things are going to get so much easier.

[Ernest Burden III]

...Just as a reference, Gui Felix, from Brazil, has been doing an amazing job with Lumion, as well as 3D Cordoba from Argentina...

http://Https://www.youtube.com/watch?v=-NBuNftWw2A

 

Those really aren't that bad, though kinda bad. The photos mapped onto the back of glass windows to show lit interiors are awful. But the cars and people move pretty nicely. I guess Lumion comes with birds?

[notamondayfan]

Those really aren't that bad, though kinda bad. The photos mapped onto the back of glass windows to show lit interiors are awful. But the cars and people move pretty nicely. I guess Lumion comes with birds?

 

Must agree, it's not bad, and to be honest the rendering, lighting materials aren't half bad. The camera movements, tiling textures, tiling foliage and things like that are what's bad, which has nothing to do with the render engine.

[Jon Berntsen]

I work inhouse, and I've been periodically nervous about my job over the years because of the development in CAD program renderings, plus competition from $50 rendering outsourcing companies.

 

At our office they're using ArchiCAD, which now has that Maxon rendering engine. For our viz people that has resulted in fewer bad rush-in-between-renderings. However, the last year I've seen two things that makes me a bit more comfortable.

 

1) LEAN. It's everywhere, and it's about maximizing productivity time. Which means: Use people where they're good. Expertise. Companies need expertise as never before.

2) VR/AR experiences. These are literally games. We've got both feet in UE4 now, and just by scratching the surface of what we do when we're making an interactive button in their house, people in the office look at us like we're aliens.

 

There are additional points, and some people have pointed out that we're dealing with analyzing what to do to make an image/piece stand out. That's what we do, and that should be our main expertise.

 

Additional points:

 

- Ever talked to an architect about the best selling angle of a building? 100% chance that this angle sells the building and not the whole image.

- Still images will never get unfashionable, as people wants to climb a mountain rather than put on a pair of VR or AR glasses in mainstream ad situations.

- We should be experts in analyzing what makes or breaks images

- We must create good compositions

- Thinking creative is essential to make it stand out

- Be patient with the focus on details

- Make animations(realtime)/videos to present projects

- Game programming is needed to make good user experiences in VR and AR

- Environment/landscape/scenery design. We should all become good environment artists.

- VR/AR will be huge, and look at this as a replacement of the bad rush in-between rendrings you've been forced to do (which made you work overtime to reach deadline for the projects you were working on in the first place).

- Take initiative for changes. the VR/AR will not magically enter your office.

- Try to show your expertise in revealing composition flaws with good expertise arguments when you're taking an in-house project.

 

We still have traditional project visualizations for sales, but we're also bringing in VR/AR solutions to the big projects.

 

There's more I can think of too, but I need to get back to work to deliver a moody evening image with carefully placed composition elements.

Edited August 3, 2017 by chroma

[Devin Johnston]

Jon these are all great points!

[Ernest Burden III]

Jon these are all great points!

 

Yes, thank you.

[Rodolfo Tanno]

The bird library is still very limited. But come on, the quality is not that bad. Most clients do not need better quality than this. It is not photorealistically perfect, but general people would never notice the errors in the internal maps, or badly positioned tiles. A video like this can cost 1/4 of a video rendered in 3ds Max. It can be done at 60 fps on a home computer, it does not even need renderfarm. With a few months of study, anyone produces such videos on Lumion. In other programs, learning in the configuration of materials, lighting, cameras, animation certainly takes a few years to have quality.

I think this is a video standard that tends to grow a lot from now on, both for the agility and cost-effectiveness that customers require.

[Ernest Burden III]

...I think this is a video standard that tends to grow a lot from now on, both for the agility and cost-effectiveness that customers require.

 

Whether you run a realtime display to video or render frame-by-frame at four minutes per, you still need a good model and good 'environment' and camera work. That is where most of the cost of an animation comes in--at least when I'm pricing them. There cost of having some computers running day and night running frames is fairly low. But it lengthens the delivery time, which is often the sticking point with clients. Once they've green-lit the work they want it yesterday.

 

So my interest in workflows like Lumion is less about offering dumbed-down work, it is about offering my usual quality of work faster, and not so much less expensive. Having said that, I do not do a lot of animation because I'm too small for the scale of work clients need when you have a full package of stills and animation. It would take me months. Anything that speeds up delivery makes me more competitive with the big studios.

[Rodolfo Tanno]

Lumion 8 was released a few months ago.

The engime is even more incredible. If anyone still has doubts about the quality of the images, watch this video:

[Jon Berntsen]

The delivery time is an issue regarding animations. However, I see platform compability the biggest argument for which software to use in the coming years. We cope with rendertimes by investing in more nodes. VrayRT is not mature enough, by far. Tested it on a couple of our latest scenes, and it just crashed everytime. Stability is a huge thing.

 

The production line is expanding, in terms of more product types. You will simply loose money if you need to set up the same scene several times in different platform, or if you can't use an asset in all of your products. With datasmith and vray 4 coming, and also with 3ds max interactive as an alternative to ue4, I see no problems with using the "good old workflow" with max and vray. These softwares changes with time too, luckily, so they get new stuff implemented.

 

For us, rendering animations is still restricted to vray because we make still images in max/vray, based on a huge (and still counting) database of objects and materials. We need to use the same assets in the same looking rig for stills, animations and optimally also VR. Animation rendrings might be possible with UE4 soon, but the realism simply aren't there compared with the results that for an example Phoenix FD gives for water and fire.

 

However, I am definitely looking for a good sky/cloud simulation. We recently invested in Phoenix FD, but I'm not quite sure about its capabilities with animated cloud and sun study. Lumion seem to do this well. Why doesn't Chaos implement something easy here?

 

Obviously, we use HDRi's in domelights to light up our scenes, and that seems pretty incompatible with a sun study with realistic animated clouds.

If anyone has a great solution for this, please let me know!

[Nicolai Bongard]

 

Obviously, we use HDRi's in domelights to light up our scenes, and that seems pretty incompatible with a sun study with realistic animated clouds.

If anyone has a great solution for this, please let me know!

 

Most of our usage of sun studys is to show clients and/or government how neighbours will be affected by new buildings, and thus clouds tend to be irrelevant; you want a non-diffused sun to cast clear shadows to easily spot the affected areas. If suddenly in the middle of a sun study animation the shadows were to be less clear because of a tiny cloud passing by, most people we tend to deal with would assume it is a rendering error, not a "feature".

If you on the other hand want to show a day/night cycle in an animation, where the mood/effect is more important than having 100% accurate sun height and whatnot, you may want to check out hyperfocal designs HDRI timelapses, they should suit your workflow.

[Francisco Penaloza]

Where I used to work they use e-on software such Vue and the cloud Factory.

They created very nice sky and atmospheric effects. The render times where very long tho. I used to render time lapses of the sky only and use that with an some light when renders needs to be faster.

But what Nicolai mentioned is correct, when we deal with light studies only IES data and Su/sky system where used