Artwork for podcast Confluence
Democratizing 3d Content with Fovea
Episode 1022nd February 2024 • Confluence • Evan Troxel & Randall Stevens
00:00:00 01:10:52

Share Episode

Shownotes

Corey Rubadue of ArchVision joins the show to talk about Fovea, a tool that simplifies the process of translating and sharing 3d content across different 3d platforms and 3d file formats. In other words, it allows users to import data from various sources, process it, and then export it in a format that can be easily used in other applications. This eliminates the need for manual file conversions and reduces the risk of data loss. Fovea also supports a wide range of file formats, making it a versatile tool for 3d content creation and sharing with a user-friendly, simple drag-and-drop interface for people working on a deadline... that’s you.

EPISODE LINKS


EPISODE SPONSOR

Chaos

For a limited time, get a 20% discount on Fixed and Floating Annual License of Enscape. Head to https://chaos-enscape.com and use code RES24 at checkout to supercharge your design workflow.


Watch this episode on YouTube or Spotify.

-----

The Confluence podcast is a collaboration between TRXL and AVAIL, and is produced by TRXL Media.

Transcripts

Randall Stevens:

Welcome Corey to the Confluence podcast.

2

:

Uh, I've been looking forward to this one.

3

:

Kind of near and dear to my heart,

Corey and I have known each other

4

:

for a long time and, uh, have a

longstanding professional relationship.

5

:

And, uh, when he's in town visiting,

he stays with me at my house.

6

:

So it's, we're we're, we've been good

buddies for a long time, but, um, yeah.

7

:

Today, uh, we're looking forward Corey

to, uh, learning about your work.

8

:

on this fovea application and its

development and how this all came about.

9

:

Just as a, um, for the audience,

there's been a lot of close work that

10

:

ArchVision the company that Corey, uh,

runs, has been working closely with

11

:

Avail, we call 'em sister companies.

12

:

Uh, and, and we've been doing a lot of

integration of, of what Corey's gonna show

13

:

you directly into the avail application.

14

:

So, uh, Corey, maybe you can just

give everybody a uh, uh, kind of

15

:

an intro to what Fovea is about.

16

:

Corey Rubadue: Thanks for ha

17

:

having me on, Randall.

18

:

I appreciate it and I always

enjoyed talking with you guys, um,

19

:

about everything we're doing here.

20

:

Um, and Evan, good to

21

:

see you Um, yeah, so Fovea, as of today

is a 3D model configuration platform

22

:

that's part of our ARC vision ecosystem.

23

:

And, uh, we, uh, can basically bring

in 3D models and data sets, scan, scan

24

:

data, um, from outside sources into.

25

:

Uh, the fovea application and publish

those assets back out into, various

26

:

file formats, uh, for consumption

in many different 3D applications.

27

:

Revit being our, our key SketchUp,

rhino Blender, handful of applications.

28

:

Um, and it's, uh, the tool really designed

around, uh, democratizing 3D content

29

:

and, uh, assets and helping people to

build or firms to build their asset

30

:

libraries up in a big holistic, uh, way.

31

:

Uh, so yeah, I mean that's a, that's

the, the nutshell about fovea.

32

:

Um, I can say that, uh, when we started

this effort, we had really no plans to

33

:

build exactly what we were building.

34

:

Um, but, uh, it, it

evolved and, and based on.

35

:

Customer needs, customer requests.

36

:

And uh, also just really, uh, feeding

off from the technological relationship

37

:

between, uh, arc vision and avail and

trying to integrate in, uh, different

38

:

pieces of our technology with a avail

to make just really great customer

39

:

experiences around content, content

creation and content management.

40

:

All, um, in one, from one very

singular and holistic place, if you

41

:

will.

42

:

Randall Stevens: I think, um, you

know, with what I know, uh, uh, kind

43

:

of insight into the fovea app, the,

uh, the thing that seems, you know,

44

:

mo most, most of the time when you

would, uh, want to take a piece of

45

:

content and produce other formats.

46

:

Uh, seen that, you know, over the last

20, 25 years that I've been working

47

:

in this industry, it's usually, you

know, I would call it a dumbing down.

48

:

It's usually a l least common

denominator kind of, of these

49

:

translators into these, uh, applications.

50

:

And I know fo you know, fovea is

different in that each of those

51

:

applications and different kinds of

file formats, it translates to Corey

52

:

and his team have, uh, you know.

53

:

They understand the native

capabilities of those applications

54

:

and how to move that data into those

native applications and around.

55

:

So maybe Corey, you can talk

a little bit more about, you

56

:

know, what all went into that.

57

:

Um, you know, even just more about your

background, you know, in the computer

58

:

graphics side and, uh, how that's

played into, uh, really understanding

59

:

this problem at a deeper level.

60

:

Corey Rubadue: Yeah, sure.

61

:

So, you know, practice architecture for

years and, uh, really was quite focused on

62

:

the visualization side of, uh, practice.

63

:

And, uh, I always said we were, we had

visual information modeling, before BIM

64

:

was bim we were really trying to kind

of use, uh, visualization as a way to

65

:

describe and show building design process.

66

:

And outcomes as quickly and

as efficiently as possible.

67

:

And as part of that process,

we built some technology.

68

:

Um, we originally started working

on a product called Vray for

69

:

Rhino, Vray for SketchUp, Vray

for Revit back in:

70

:

And we got deep into the, inner works of,

of geometry materials, lights, all this

71

:

stuff really super early on, and try to

get information converting from SketchUp

72

:

into, uh, Vray file format as an example,

or Revit into, uh, the Vray file format.

73

:

And so we just, we learned a lot

about these, um, applications,

74

:

APIs to be able to take.

75

:

Material from SketchUp at the time in

:

76

:

for a material definition, um, you know,

a base color and maybe a texture map,

77

:

but we weren't getting hyper, you know,

detailed information about the materials,

78

:

like reflection values, refraction

values and stuff like that, because

79

:

SketchUp didn't have that information and

that, uh, delivering through their API.

80

:

But on the other hand, Revit did.

81

:

So we had information coming from

Revit to be able to build, really,

82

:

accurate 3D models from these

applications into, at the time it

83

:

was very strictly vra uh, VRA assets,

VRA scene files, and VRA proxy files.

84

:

so, that experience

really set the stage for.

85

:

And help me kind of set the stage, reset

the stage for RPC, and I came on board

86

:

in whatever it was, 2019 on at Avision.

87

:

You know, the main purpose for me, at

least in my mind at the time and vision

88

:

was like, Hey, there are still problems

with content out there in the world.

89

:

People don't get enough content.

90

:

People don't have enough 3D models

to access and get them in their

91

:

project and inform their designs

or, or explain their designs.

92

:

So our ARC vision just seemed to

be this platform where we could

93

:

take these rrp, this RPC technology

and ultimately just transform it

94

:

and bring it to a whole new level.

95

:

And so that experience that I had with

vray, assets and creating scene files

96

:

and trying to navigate content from one

application to another, served as that.

97

:

Kind of catalyst, if you will,

for what we could do with RPC.

98

:

And at the time when I stepped into the

ARC vision seat here, um, I really had to

99

:

dig in and understand what was, what was,

what's, what was the DNA of of an RPC.

100

:

And, uh, came, come to find out it was,

uh, as robust as it was in its early days,

101

:

there were a lot of things that were not

in the RPC that needed to be there, like

102

:

real materials for different objects.

103

:

And, uh, you know, inside of a 3D model,

it was the RPC was designed to essentially

104

:

either be image based where you had,

uh, 360 or more images inside of an RPC.

105

:

And depending on the camera, it

would actually pop up an image, uh,

106

:

and render that image at that time.

107

:

or in a 3D model, or 3D RPC.

108

:

It was a mesh.

109

:

With, uh, or meshes with

textures baked on top of it.

110

:

And the whole process of even getting

those 3D RPCs was, was pretty cumbersome.

111

:

I mean, we using three Studio Max

going through a bunch of very technical

112

:

things, uh, steps to get to, uh, and

our, uh, 3D RPC and get it out to

113

:

the, you know, to your user base.

114

:

And so, really as I stepped into this

role, you know, I try to understand like

115

:

we, we don't want architects and designers

learning all this technical stuff.

116

:

We really want to be able to kind of

come back down to the heart of what RPC

117

:

and Randall and Avision design, which

was like ease of use, drag and drop, get

118

:

everything inside of that file, and, just

make it easy for people to, to consume

119

:

that piece of content and render period.

120

:

mean we had to do a lot of dissecting of

the RPCs file structure and technology,

121

:

uh, and then rebuild on top of that.

122

:

And so, uh, looking at all these

lessons learned with Vray file formats

123

:

and, and, um, other, you know, file

for formats like FBX or OBJ and uh,

124

:

Colau files and things like that.

125

:

Uh, coupled with, uh, looking at how

mental Ray was doing things and how

126

:

Corona render and all these other

rendering engines were producing

127

:

materials, geometry, how was all this

stuff transporting around, we had a

128

:

lot of research to pull together to

try to come up and say, we want a, a

129

:

really new single source of truth of

technology underneath the RPC file format.

130

:

So.

131

:

Evan Troxel: Corey, can you explain

what an RPC is for those who don't know?

132

:

I'm sure almost everybody who's

listening to this or watching this

133

:

has seen one, but they maybe don't

tie those two things together.

134

:

What they see with RPC.

135

:

Corey Rubadue: Sure.

136

:

So RPC is a file format.

137

:

Uh, RPR stands for, rich

Content is the acronym.

138

:

And, uh, these are assets that,

um, it started out really just

139

:

entourage cars, people, trees.

140

:

Uh, the RPC families ship

with Revit as an example.

141

:

So all the planting, uh, elements

that you can get outta the box

142

:

of Revit are all RRP C files.

143

:

And, uh, yeah, simple, simple,

144

:

Evan Troxel: And, and basically what

you're seeing on the screen is like

145

:

a, a very low resolution proxy of.

146

:

When you render, it's gonna

look better than that.

147

:

Right?

148

:

But it's lightweight for the sake

of being lightweight in the model.

149

:

And often it's like you see two billboards

kind of intersecting each other with

150

:

a, with a outline around the object.

151

:

So it, it kind of represents

a tree in that example, right?

152

:

In Revit, when you place it, it has kind

of a, a base to it, and then two vertical

153

:

billboards that represent the tree.

154

:

But when you hit render,

you see a tree, right?

155

:

And that's the idea behind this,

is you have this rich content that

156

:

sits behind a very lightweight

geometry presentation in the

157

:

Corey Rubadue: Absolutely.

158

:

Yeah, that's, that's the whole idea

Randall, and I like to say it, uh, an

159

:

RPC is a proxy for a future transaction,

160

:

right?

161

:

Um, it's, it's there, it doesn't

overburden your, your CAD

162

:

package, whether it's SketchUp

or Revit or Rhino or whatever.

163

:

The RPC is in there as

a, as a placeholder.

164

:

Uh, and as soon as the

render hits that RPC.

165

:

It'll, uh, it'll pull out the geometry,

the material, DNA, if you will, and

166

:

render that in, in its photorealistic,

full photorealistic beauty.

167

:

Evan Troxel: Cool.

168

:

And content's always been the hardest

thing for architects to have to deal with.

169

:

I mean, this is a conversation

we've all had over the years, right?

170

:

Which is like we build models of

buildings, but putting the stuff in those

171

:

buildings for the visualization, for the

space planning has always been difficult.

172

:

Where do you find it?

173

:

Is it the right quality?

174

:

Is it too detailed?

175

:

Is it not detailed enough?

176

:

Is it made by the manufacturer?

177

:

Is it the right dimensions?

178

:

And this has always been hard and I,

you guys have been at the forefront of.

179

:

Content or assets or entourage, like

all different words that we could

180

:

kind of interchangeably use here.

181

:

And so that, that's where the, the

value proposition is for what you're

182

:

doing right, is democratizing the

creation of, and cataloging of, so

183

:

that firms and, you know, practicing

architects, people doing visualization,

184

:

have access to these things at their

fingertips to throw into their model.

185

:

Because then it allows you to

focus on building the model, right?

186

:

Not, not curating and

texturing and creating

187

:

Randall Stevens: Yeah, and I'll throw,

I'll just throw in 'cause I was obviously

188

:

involved with this in the earliest stages

and, you know, the, the people in the

189

:

trees, the technique that was developed,

uh, uh, kind of underlying that technology

190

:

was thrown at the tough problem of organic

things like people in trees, right?

191

:

Where, where you've got very

complex geometry, potentially

192

:

very complex materials and

textures to make that look real.

193

:

So, you know, in its early days, it's

served its purpose in helping you

194

:

know, the industry to, uh, you know.

195

:

When that first came around, there wasn't

any, there was, there really wasn't any

196

:

other good way, good way to do that.

197

:

And especially at, at,

198

:

Evan Troxel: no, you, you would load

one human or one tree model from

199

:

onyx tree, for example, back in the

day, and it would have more polygons

200

:

than your entire architectural model.

201

:

Right.

202

:

And it would just absolutely kill

back then when computing performance

203

:

was, was very limited at that point.

204

:

Right.

205

:

So, I mean, the, the idea

206

:

Randall Stevens: Yeah.

207

:

And I think the, um, um, I, you

know, I think though that the, the

208

:

DNA of that technology and which is

what made it, you know, very special

209

:

when Corey came, you know, and

brought even more, uh, I'll say broad

210

:

visualization platform knowledge to it.

211

:

Um, the underlying technology had really

concentrated on how, how do you store.

212

:

lots of different kinds of data.

213

:

Not just geometry, but material

information and texture information.

214

:

How do you package that?

215

:

That's where the RPC file format came in.

216

:

So that, so that it wasn't a bunch of

loose, you know, I call it loose content.

217

:

You didn't want JPEGs over

here that were, you know, now

218

:

not in the same path, and they

didn't render because a texture

219

:

was missing and all that.

220

:

So there was,

221

:

you know, and that was, as Corey

said, that was all about, you know,

222

:

how do you make this stuff easy?

223

:

The way you make it easy is try to

224

:

just not not let a mistake,

you know, even occur.

225

:

And then, um,

226

:

Evan Troxel: Yeah.

227

:

Randall Stevens: and then, you know,

the other piece of that I think,

228

:

you know, in the early days was, um,

that content needed to be able to

229

:

be rendered in lots of different.

230

:

Rendering engines and applications.

231

:

So there was a long history of Arc

visions, the company having, uh,

232

:

relationships with all of the rendering,

the people doing rendering engines.

233

:

So, and getting that kind of baked in

there and supported, uh, kind of natively.

234

:

And as Corey said, even, you know,

with Autodesk, I think they began

235

:

shipping that Corey, what, 2010 or

11 whenever the tree and Foliage

236

:

solution started shipping with Revit.

237

:

So it's always been kind of

in the DNA to do all that.

238

:

I think what's, you know, special

about what Corey's been doing now

239

:

is, you know, one, I'll just say

modernizing the definition of the

240

:

storage of that information inside

the file and, and those capabilities.

241

:

But then, you know, looking at,

you know, can you, um, you know,

242

:

maybe Corey, you can talk a little

bit more instead of having to have

243

:

those companies, um, support the

RPC file format in some special way.

244

:

What, uh, the team at Arc Fission's

been doing with the fovea application

245

:

is saying, look, we can store it

all this data in its highest form,

246

:

but we can spit it out into a native

form that that application wants to

247

:

without losing or dumbing down quality.

248

:

So, you know, it's, it's a

little bit of the opposite.

249

:

Instead of everybody trying to consume

one file format, it's like, look, can

250

:

we get the highest level of information

and then map that and translate it

251

:

into the native formats that each

of those applications wants to know.

252

:

So if you're trying to use it in

253

:

Inscape, it's gonna know what

to do, or if you're gonna use it

254

:

in Vray or one of these others.

255

:

So.

256

:

Evan Troxel: Give each

package what it is expecting.

257

:

Before you do that, Corey, I just,

can we just share like how many

258

:

nightmares we've had where we've.

259

:

Sent the job to the render farm

overnight or for days at a time to come

260

:

back and find that that one material,

that one diffused material or that one

261

:

bump map or whatever was not there.

262

:

Right.

263

:

Oh my gosh.

264

:

And you have to do the whole thing again.

265

:

Right.

266

:

Uh, we've, we've all been around

long enough to, to have lived that

267

:

Randall Stevens: there's

a lot of complexity there

268

:

Corey Rubadue: and that,

and that wasn't, um.

269

:

Discriminatory in any way.

270

:

Like didn't matter what the file

name was or the file size or what

271

:

the texture map was for, if it was

missing, it was missing and it broke

272

:

everything.

273

:

Right?

274

:

Evan Troxel: yep.

275

:

Or some weird character in a

file name or something like that.

276

:

It was just like the littlest

things and it was like, oh my gosh.

277

:

It was a learning experience

278

:

Randall Stevens: and it's still to

279

:

this day, right?

280

:

You can download, um, man, you know, I

see the manufacturer content all the time.

281

:

Pull down a Revit family, you

know, from a manufacturer.

282

:

Geometry will come in, right?

283

:

It's in that RFA

284

:

file.

285

:

But as soon as you hit render

and you hit the rendering engine,

286

:

it's like missing these maps.

287

:

It's like, why did you know one, maybe

they didn't even come down with that

288

:

file off, you know, as it was zipped

289

:

up or something.

290

:

So there's just, there's just still a lot

of complexity in the industry around this

291

:

and uh, you know, but maybe Corey, you

can talk more about what you're trying

292

:

to do with the fovea, you know, and

that sitting in the avail application.

293

:

Corey Rubadue: Yes.

294

:

I kind of killed a couple

birds with one stone here.

295

:

If I zoom out just a little bit and

just talk about the other part of RPC,

296

:

which is yes, there's the content, but

the, the technology that's that Randall

297

:

you were alluding to, like there's

an A-P-I-A-C plus plus API behind RPC

298

:

and originally that c plus plus API is

what our, um, our partners were using.

299

:

So, folks that, on the Revit team or, or,

um, you know, there was Lux, um, a handful

300

:

of other vendors, MicroStation, you

know, we had those companies coming and

301

:

working with us, using our APIs at that

level to build integrations into their

302

:

applications to support RPC and part of.

303

:

know, my approach with this was to,

again, even try to look at the platform,

304

:

the, the programmatic platform that

we had and figure out and try to

305

:

design something that it could become

much more friendly web-based, uh,

306

:

web API based stuff as an example.

307

:

Um, more accessible people might be

able to see, someday, see RRP C as, as

308

:

in an open source format and be able

to use Python and other languages to

309

:

be able to use our platform in general.

310

:

And so not just, you know, I wasn't,

I'm not just looking or haven't been

311

:

just looking at the content as being

democratized, but also the technology

312

:

that wraps around everything we're doing.

313

:

And part of our journey

was at the beginning was we

314

:

had to take RPC technology.

315

:

From what used to be Avision called

Avision Dashboard, which was a content

316

:

manager specifically designed for

delivering RPCs from the cloud to

317

:

users before cloud was cloud probably.

318

:

Um, and take that experience and,

uh, keep as much of it intact as

319

:

we could and bring it into avail

and integrate our content delivery.

320

:

Uh, everything from thumbnails to

the downloads into preview, being

321

:

able to preview the models, all that.

322

:

We had to build a bunch of different

technologies around our APIs and

323

:

start to democratize the c plus

plus stuff into other languages so

324

:

that, uh, we could talk with a veil.

325

:

Right.

326

:

And that started the, the really the,

in my mind, that's where we started to

327

:

have to, Build the foundation of fovea

before fovea was fovea, if you will.

328

:

Um, so the data flow, the, uh, not

just the geometry, not just the

329

:

textures, but there was meta, there's

metadata involved with all this stuff.

330

:

There's special GUID's around

the a, the assets so that we can

331

:

track them and understand where

they're going and things like that.

332

:

So ultimately, there's a lot of

engineering around making this

333

:

simple and simpler, not just from

the content consumption and use, but

334

:

really, you know, forward thinking.

335

:

Let's get a bunch of content creators,

a bunch of main building product

336

:

manufacturers, involved with our

APIs and be able to use them to even

337

:

translate and, and deliver RPCs of their

own content into Revit and Rhino and

338

:

SketchUp and wherever, as opposed to.

339

:

The broken process as an example,

Randall, what you're talking about

340

:

downloading something from the

manufacturer's site and not getting

341

:

the texture maps we can deliver.

342

:

And we have the platform now, you know,

underneath the, our ARC vision RPC

343

:

technology to be able to deliver single

source of truth assets and give people

344

:

simple APIs to access that to, um, to

be able to deliver their own content.

345

:

Randall Stevens: Corey, maybe you can,

uh, you wanna share your screen and can

346

:

you show what, uh, fovea just ends up

looking like, uh, In the application?

347

:

Corey Rubadue: Fovea comes in

three different forms right now.

348

:

Um, we've got a web app version,

iOS app, and then this version

349

:

here, which is inside of a veil,

which is our latest and greatest.

350

:

So what you're looking at

here is, uh, a veil desktop.

351

:

Uh, on the right hand side, you'll see

the FOA window here, just sitting here.

352

:

And I'm going to just burst on the

scene and do a quick show and tell of

353

:

why is this important and what we're

really, what the power of Fovea is.

354

:

I can grab a model that's part of

the Fovea publication and literally

355

:

just drag and drop it directly

into this Fovea cannabis here

356

:

that's embedded inside of a Vail.

357

:

And, this, this is mind blowing to me

even in that we're in a content manager

358

:

and we're in, an application in a content

manager where we can manipulate 3D data.

359

:

I can grab a material

on here and whatever.

360

:

Let's just say we wanted that to be

black and I can change some of the

361

:

other material parameters in here.

362

:

So we've got the full material editing

capability for the asset right here,

363

:

and I can make that maybe more shiny

and do a couple operations there.

364

:

And And then very simply,

there's a published button.

365

:

I can give this a name, and this is

where I start to put some metadata on

366

:

this, uh, asset chair, whatever, uh,

give it a furniture category, how to

367

:

tag, and, um, let's just say, uh, modern

as an example, enter, and I won't get

368

:

in the weeds on this at just at this

very moment, but this is what we were

369

:

talking about earlier, um, where the

RPC has a, uh, proxy representation.

370

:

And as part of the publishing process,

you've got, you have the opportunity to

371

:

choose different, um, proxy mesh, uh,

meth methods and what the representation

372

:

will be in your CAD package.

373

:

Uh, you can even do some, uh.

374

:

Down sampling of textures and mesh

simplification and optimization here.

375

:

So if we know this asset's going to end

up in scape or something like that, and we

376

:

need to do some decimation on this, uh, we

can, let's say, make this, uh, uh, target

377

:

mesh count down to 63, uh, hundred and

changed triangles, and I can hit publish.

378

:

Um, so we're grabbing this in this

3D data, the instructions behind

379

:

it, sending it up to our cloud

services, through our APIs, um, and

380

:

publishing that piece of content.

381

:

And, what's great about the integration

that we have with avail is that

382

:

piece of content's going up into the

cloud, processing, metadata's being

383

:

stored, and at the com point that the

processing is completed in our, in our.

384

:

Um, in our web services, we'll now push

and synchronize that piece of content

385

:

directly into, um, this, this Fovea

channel that's sitting here in avail.

386

:

So I am not just making content

and editing a, creating content

387

:

in avail through the fovea lens.

388

:

I'm actually also putting it

in a managed state right away.

389

:

Okay?

390

:

So what's important about that is,

okay, if I built the asset, uh, maybe

391

:

from my web browser and I downloaded

it to my C Drive, does anybody else in

392

:

your company have access to that asset?

393

:

Maybe if you save that from your

c drive to a network, drive a

394

:

network location, a shared location

that everybody has access to.

395

:

But in this case we're, you know, we

were very, very focused on, again, ease

396

:

of use and trying to make sure that.

397

:

This content is, is, in a managed state

for users to be able to just not even

398

:

have to think about it's, it's here.

399

:

They can search for it.

400

:

They can just drag and drop

it into Revit and go, period.

401

:

Randall Stevens: you were

talking about the avail and, and

402

:

talking about these as lenses.

403

:

Um, on the avail platform, we have

a plugin architecture to avail.

404

:

That's allowing third parties like

Avision to build on top of the avail

405

:

as a content management platform.

406

:

And we call it our plugins lenses.

407

:

So when he talks about fovea as

a lens, it's a plugin inside.

408

:

So it's, it's a provision's technology

that's running, you know, on top of

409

:

this core content management platform,

which allows them to, allows our vision

410

:

to concentrate on this high value

added capabilities without having to go

411

:

rebuild over and over again a, you know,

core content management capabilities.

412

:

So as those assets are being published,

they're coming right back into the

413

:

core application, tagged, managed,

ready to go, uh, as opposed to being

414

:

downloaded from another service, then

having to be reorganized somewhere

415

:

is kind of all one stop shop.

416

:

Corey Rubadue: Yeah.

417

:

And, and.

418

:

Just to point out that this, um,

our first kind of step into this was

419

:

actually being able to build a, a,

a, model viewer, a previewer, if you

420

:

will, that got embedded inside of

avail, uh, early, pretty early on.

421

:

And so, people want to see what they're

going to drag and drop into Revit,

422

:

a Rhino or SketchUp, beforehand.

423

:

and so you can select, you know, pieces

of content and preview that model

424

:

in our little preview window here.

425

:

And that, that actual canvas here is the

backbone of what we're using in fovea.

426

:

So this is the same 3D

canvas that we have.

427

:

Uh, in fovea, just minus, you know,

the interface, uh, for controlling the

428

:

finer details of, uh, of publishing

and, and so on and so forth.

429

:

So this was our, our first step

into building and integrating some

430

:

technology into avail, um, before

we jumped into the FOIA realm of

431

:

being able to now edit this content.

432

:

I'll zoom back out here real quick.

433

:

Just to go back to the, kind of bigger

picture of what's, what's, what's getting

434

:

produced, what's FOIA really, you know,

doing and why, and so on and so forth.

435

:

So when I published that

chair, and it goes up into

436

:

our cloud, we're spinning off.

437

:

Um, a bunch of different

file formats out of this.

438

:

Yes, we make the RPC and

we make the RPC first.

439

:

That file form, that single

source of truth fi file format

440

:

is our, you know, first stop.

441

:

The next stop along the way is we,

um, we actually rely a lot on uh,

442

:

GLTF to transport, and help us,

migrate data, uh, into some of these

443

:

other file formats that you see.

444

:

But we have, uh, direct jobs, uh, behind

the scenes where we're taking the RPC and

445

:

we're getting RPC to GLTF and RPC to data

Smith and RPC, to, uh, FBX and, and USD

446

:

and Pure Landscape Asset, for instance,

or, um, RA VA steam file as an example.

447

:

So all of these file formats and

file types are spun out of the single

448

:

source of truth of what the RPC is.

449

:

Um, we specifically and kind of touch

on what Randall was saying earlier,

450

:

like RPC contains the richest form of

data, uh, and information and, uh, uh,

451

:

textures and all that, geometry and

textures and all that inside of that RPC.

452

:

And now we can take and repurpose

any of that information into these

453

:

different file formats through

our, through our API basically.

454

:

So that's what fovea is doing.

455

:

It ultimately it'll, it, it's

just spill spits out a bunch

456

:

of different file formats.

457

:

And this is, you know, a departure from

RPC in, in the main kind of way of looking

458

:

at it, which was we really, we could have

just kept building RPCs, and relying on

459

:

our c plus plus APIs to have partners.

460

:

Build plugins or

integrations with RPC files.

461

:

Um, but really we started to lean on

this idea of this proxy as, uh, sorry,

462

:

RPC as a proxy for a future transaction.

463

:

And so the way we see this is an RPC

may come into an application like

464

:

twin Motion, and when it comes into

twin motion, it may come in as just

465

:

the, the billboard representation.

466

:

That was the preview that you

saw inside of your Revit file.

467

:

Um, but what's slick about what, uh, what

we've done is literally with a, find

468

:

and replace, we can grab a data smith

version of that same asset and swap it

469

:

out in twin motion so that you get apples

to apples, the same car person tree, or

470

:

whatever the asset was that you placed

in Revit and that you might have been

471

:

rendering with scape or V-Ray in Revit.

472

:

Um, if your team happened to switch

over to use Twin Motion and they, they

473

:

would get with a little bit of, you

know, workflow involved, they would

474

:

be able to get the same apples to

apples, the same car, the same trees,

475

:

the same people coming out with the

same materials, as, as the folks were

476

:

getting on the Vray render team inside of

477

:

Revit

478

:

Randall Stevens: maybe you can dig

479

:

into that a little bit more, Corey,

you know, just the, uh, the, the

480

:

reality that inside of a lot of firms,

the workflows, the, the modeling

481

:

workflows from design to visualization.

482

:

And back, right, back and forth.

483

:

It's not always a linear process and

everybody's going back and forth that,

484

:

um, you know, just how that understanding

drove the need to then have these

485

:

different, um, representations of the

same thing kind of readily available.

486

:

Evan Troxel: and maybe before you jump

in, Corey, I mean this is something that's

487

:

come up on, on TRXL podcast many times

and I, I just saw Ian Keough of Hypar

488

:

tweet this same exact sentiment, which

is like, we used to use design software

489

:

to document decisions that had been made.

490

:

Now we're using them to make

decisions along the way because

491

:

things are happening in real time.

492

:

And to Randall's point here, it's

like you're, you're looking at

493

:

the design in scape in real time

or in twin motion in real time.

494

:

You're visualizing what you're draw

what you're modeling and drawing and

495

:

the ramifications of those things

spatially in all the different ways.

496

:

How, how it makes you feel, how

you emotionally react to it.

497

:

And it drives the design.

498

:

It's not like the end

of the process anymore.

499

:

Right.

500

:

So like, just to reinforce what

Randall's saying, like I, I'm

501

:

seeing this, these tools are now

moving way earlier in the pipeline.

502

:

They're not just the end

of the sentence anymore.

503

:

They're are the end of the phase

right there, there is this feedback

504

:

loop that is, that is happening.

505

:

And so these things need

to be all bi-directional.

506

:

Corey Rubadue: Absolutely.

507

:

Yeah.

508

:

And so, you know, my, my

early days of working in, at

509

:

this point it was at Air St.

510

:

Gross Architects and

Planners in Baltimore.

511

:

And being in the middle of the

studio there, um, servicing

512

:

all different departments from

interiors, landscape designers, and,

513

:

and, you know, building studios.

514

:

we had to get information from the

design teams and designers into a place,

515

:

uh, in our case, it was, at that time

it was, it was three Studio Max and.

516

:

Produce a visuals, but also at the end

of it, be able to translate what we put

517

:

in the visuals back to the design teams.

518

:

So the design teams could document

what, what we put in the visuals.

519

:

Right.

520

:

And, that still exists today.

521

:

I mean, this is still happening and

Scape does a great job, but this for

522

:

folks nowadays where they can design

and visualize at the same time and they

523

:

can have things consistent and they're,

it's kind of, you know, not as, uh,

524

:

disjointed of a process as the older,

traditional, you know, design model or,

525

:

um, design documentation to a 3D modeler

specialized in max producing something.

526

:

And then the design team interacting

with what was going on in the

527

:

visualization development and Max.

528

:

And then trying to, again,

what do you do with it?

529

:

How do you get that information back?

530

:

So.

531

:

My, you know, my sentiment behind RPC

and, and want, what really drove me to

532

:

take the bull by the horns on RRP C is we

can, we can make this just this connection

533

:

between the asset and the design world,

the asset in the visualization world, and

534

:

hopefully someday be able to make that

connection back from the visualization

535

:

world back into the design world.

536

:

So an RPC in twin motion, or maybe let's

say even further down the line in the

537

:

Epic Stream, uh, maybe you're in, uh,

unreal Engine and, and with our Unreal,

538

:

uh, plugin RPC plugin for Unreal Engine

placing RPCs in inside of Unreal Engine

539

:

and, uh, ultimately be able to translate.

540

:

Back to the design team, back

to the folks that are in Revit

541

:

or Rhino or SketchUp or multiple

applications, you know, designing.

542

:

Um, which happens as we know.

543

:

Um, if I need to get, or what I

ideally will happen is you'll be able

544

:

to synchronize the RPCs that were

placed inside of Unreal Engine with

545

:

the Revit model or the Rhino model

or the SketchUp model, for instance.

546

:

All through, you know, our connections,

either APIs, um, or plugins or avail.

547

:

I mean, we'll be able to really hopefully

bring this, yes, bi-directional, but

548

:

interactive,

549

:

Randall Stevens: Uh,

Corey, I'll just jump in.

550

:

One of the favorite examples that, uh,

that I know that you've been working on

551

:

with some, some companies kind of closely.

552

:

Um, as an example, you'll have a design

team working in Revit on, uh, the ar

553

:

I'll just call it the architectural

components, but you might have a landscape

554

:

design team working primarily in Rhino,

completely separate design application.

555

:

Those.

556

:

Those models have to come together, right?

557

:

Uh, to be visualized and

vice, you know, vice versa.

558

:

The, the, the people working in

Revit wanna see the building in

559

:

context with the landscape and then

the landscape firm, terraforming and

560

:

entourage, and anything else that

might be going on on that front might

561

:

be primarily being driven outta Rhino.

562

:

Need to be able to see the context

of the building in that context.

563

:

part of what I've seen Corey and the team

being able to do is to show that, hey,

564

:

you can, you can take the entire Revit

project, right, no matter what scale, and

565

:

shove that into, say, an RPC file format.

566

:

Most people think about it as just a

chair or, or a single element, but you can

567

:

take the entire building and its textures

and materials into one file format

568

:

that then knows how to render in Rhino.

569

:

Well now you get one proxy object

that represents the building over

570

:

in Rhino that the, uh, landscape

team right is able to see.

571

:

And when they hit render, it flows into

the rendering engines and they can see it.

572

:

So, uh, maybe I'm stealing the thunder,

but Corey, maybe you can talk just a

573

:

little bit more because those are very

powerful, complex things where those teams

574

:

are respectively working with each other,

but that stuff needs to come together

575

:

and be visualized all along the way.

576

:

Corey Rubadue: Yeah, I mean the,

the, big, one of the bigger

577

:

projects that I worked on, uh,

back in the day was really heavily

578

:

reliant on being able to aggregate.

579

:

different data sets into one master

file and, um, be able to produce

580

:

a animation or a video or a film

of, of, you know, 10 30 buildings

581

:

in a urban context as an example.

582

:

And those workflows that we went

through back then were very similar

583

:

to what we were kind of being able

to do now with, with the RPC, which

584

:

is create a proxy, create a, a

container file, bring those proxies

585

:

into the container file hit render.

586

:

And voila, you've got proxies from

an outside source into the container

587

:

file, which is maybe in, like you said,

rhino, or it could be in SketchUp.

588

:

And you pretty quickly can get, uh, all,

all these data sets in and, literally

589

:

back out, in, in a very fluid way.

590

:

Evan Troxel: I had no idea that you

could shove an entire project into an

591

:

Randall Stevens: You got to, we'll

have to show you Evan, um, you

592

:

know, that we're, uh, some of this

is some of the, uh, fovea app has

593

:

integrated, um, AR capabilities into it.

594

:

So

595

:

I'm telling you, it's pretty slick.

596

:

You can take, I can be in Revit and I

can, I can hit a button, I can click

597

:

one button, shove an entire project.

598

:

I did it with this new

sample project, right?

599

:

That, uh,

600

:

that, that Harlan, uh, you know,

that we did the podcast with earlier.

601

:

Literally

602

:

within a minute you've got that entire

project shoved through fovea and I could

603

:

pull it up on my app and pull it up in ar.

604

:

And walk

605

:

through it at full scale, like

in the back parking lot and be

606

:

walking around inside that building.

607

:

And, um, it's, yeah, you can, you, you

can shove individual objects or entire

608

:

projects through this pipeline and

it, uh, and it all works seamlessly.

609

:

Evan Troxel: It's a paradigm shift

because I mean, we've always known RPC

610

:

to be these objects in the scene, right?

611

:

And so you're basically talking about

changing users' mindsets here too,

612

:

about what RPC can do and what it's

used for and the different use cases.

613

:

And so as a developer, those are the

kinds of things I assume you're also

614

:

thinking about when it comes to how you

market this and talk about it and reset

615

:

expectations around something much bigger

than anybody is kind of coming at it.

616

:

Just expecting it to be the next

small iteration on something that's

617

:

been around for quite a long time.

618

:

This episode is made possible

with support from Chaos-Enscape.

619

:

This is the perfect time

to set good intentions and

620

:

resolutions for the months ahead.

621

:

Whether you aim to solve your design

challenges faster, run your projects

622

:

more smoothly, make quicker decisions,

or share immersive walkthroughs with

623

:

the click of a button, here's some good

news: you won't need any resolutions.

624

:

Chaos Enscape has the

best 3d workflow solution.

625

:

Chaos Enscape is the industry leading

real-time visualization plugin that is

626

:

100% fast, 100% easy, and fully integrated

into your favorite design tools.

627

:

It is the trusted choice of over 500,000

monthly users across 150 countries.

628

:

Starting today, you can get a 20% discount

on fixed and floating annual licenses.

629

:

Just head to Chaos-Enscape.com

and use code RES24 at checkout.

630

:

That's Chaos-Enscape.com using

code RES24 at checkout to

631

:

supercharge your design workflow.

632

:

Thanks to Chaos Enscape so much

for their support of this episode.

633

:

And now let's get back

to the conversation.

634

:

Corey Rubadue: Yeah, so this is

for instance, that's the, um,

635

:

the Snowden Towers projects,

636

:

Evan Troxel: and

637

:

Nice.

638

:

Corey Rubadue: this is, now we're not

trying to visualize and render something.

639

:

This is the asset, right?

640

:

So now we can take this as an RPC, whether

it's, um, a bounding box of this or, um,

641

:

a simplified mesh or the old billboard

file style, uh, sorry, style of RPC.

642

:

We can, drag and drop this

inside of Revit and start, or

643

:

sorry, rhino, the source's Revit.

644

:

Bring this into Rhino or a SketchUp

and start, you know, designing

645

:

a new building on that parking

646

:

lot right there.

647

:

Evan Troxel: mm

648

:

mm

649

:

Randall Stevens: Or, or, you know,

650

:

just to emphasis on the,

the fovea that could be.

651

:

That very same source model that started

in Revit with all the materials and

652

:

textures can now be available in USD or

brought into Data Smith right into one

653

:

of the game engines or somewhere else.

654

:

So, and, uh,

655

:

but there's a connection between

all of those so you can kind of

656

:

bidirectionally, you know, get back

to the right data format that you want

657

:

right in, in whatever application that

you're wanting to, you know, work in.

658

:

Evan Troxel: So does this become

a link that, that design teams,

659

:

we all know you guys have been

in visualization forever, right?

660

:

That the design doesn't stop, right?

661

:

When you send the model to

the visualizing guys, that,

662

:

that's just the latest version.

663

:

It's not the, it's not the

the where it's gonna end.

664

:

And, and is there a way to, to

that this kind of helps solve that

665

:

problem, which is like, no, I really

need your final model on Tuesday.

666

:

That's the last day that I

can accept your final model.

667

:

Does this give people the ability

to kind of continually work up

668

:

to that deadline in, in less of a

export, import kind of a manner?

669

:

Corey Rubadue: Uh, it still,

still will be an update, kind

670

:

of, kind of a workflow, right?

671

:

I mean, someone can write out the

new, the final version of that.

672

:

I.

673

:

Building or design and, uh, the, the

user in RiNo or SketchUp or whatever

674

:

would have to update that RPC and,

and that data set inside of that

675

:

application.

676

:

So there is, there is still a, you know,

a, some workflow involved, but we're not

677

:

talking about, you know, having to redo

all the materials and rebuild all the, you

678

:

know, the lighting and so on and so forth.

679

:

gets brought in with the RPC is what

was, where the project was left off

680

:

and, and say Revit in this

681

:

Randall Stevens: But I let, let me just in

682

:

let me inject Corey, just

so to make sure, but yes, it

683

:

requires a human to do it, but it's,

I can tell you, uh, you know, having

684

:

used it in Revit, it's one button

click and you're, and that's it.

685

:

It's like in, in,

686

:

in 60 seconds, the thing

687

:

is out there and able to be

consumed on the other end.

688

:

What I'll say about, you know, what

ARC Fission's working with, avail

689

:

on part of a veil's platform is, you

know, now that's a piece of content.

690

:

Well, now imagine that the, that,

that, that piece of content is in

691

:

an application, rhino or Revit.

692

:

Part of what

693

:

avails working on is the ability,

when something has been updated,

694

:

the content gets updated.

695

:

Can it either push a notification

that to the user, Hey,

696

:

there's a new version of this

697

:

available.

698

:

Do you want to update it?

699

:

Uh, so now you can imagine these

bi-directional workflows where different

700

:

people working on different parts, and

you want that stuff to come together.

701

:

We're doing that from the

standpoint of thinking about like

702

:

a manufacturer updates a piece of

703

:

information and you want to send,

you know, you want that to be,

704

:

um, people who are using that

piece of content to know that

705

:

there's an updated version of it.

706

:

But I also think that it's laying

the groundwork for that, you know,

707

:

you could be on a nightly basis,

you know, publishing the updated

708

:

version of the Revit model for the

day, and then as everybody comes

709

:

into work the next day, they're receiving

the next updated version and, you

710

:

know, moving along with their work.

711

:

So it's not, you know, the,

the, the, the framework's there.

712

:

Right now, people are manually kind

of pushing these things, but the,

713

:

the groundwork's being laid for this

to be automated in, in new ways.

714

:

Corey Rubadue: Yeah, and I mean, I can

use an example right now, actually,

715

:

two in particular, um, NVIDIA's

Omniverse platform as an example.

716

:

It's been kind of a, a great testing bed

for us and, and we we're gonna, we'll

717

:

be talking more about this in the next

few months, but, it what that platform

718

:

has allowed us to do, because it's

open developmentally and we can develop

719

:

a lot around our current APIs and.

720

:

Do a lot of things that we can't

do in Revit specifically, or Rhino

721

:

specifically, or whatever, because

those applications might have a limit

722

:

limitations on APIs and how geometry and

information comes into those applications.

723

:

So in the case of Omniverse, for instance,

uh, we have, uh, asset store plugin that

724

:

we built for omniverse, and it's really

designed to be a, uh, number one, an

725

:

authentication source, uh, to be able

to allow, uh, a provision subscriber

726

:

to log in, authenticate themselves

inside of that platform, and then

727

:

allow them to, download their assets,

which they might make through fovea

728

:

or through our subscription content.

729

:

but what's going on is, in, in Omniverse.

730

:

If somebody brings in a Revit file

and there's an RPC in it, uh, that

731

:

what we get inside of Omniverse is

the, again, a proxy object that's

732

:

transferred from Revit, the Revit

file into the Invi Omniverse platform.

733

:

We can literally, again, select that

asset and request via API to swap out that

734

:

asset with its twin in USD file format.

735

:

So it's a targeted file format for a

targeted use case in, in that application.

736

:

Uh, but it's it's center point reference.

737

:

Center reference point is

the RPC and the RPC Goid.

738

:

Right.

739

:

That's really making that connection.

740

:

Right.

741

:

Um, and then even now.

742

:

Literally we have a development

progress meeting this afternoon.

743

:

I'll show demo this integration that

we're working on with Blender, which

744

:

again, open source, uh, platform and

able, super easy for us to get around,

745

:

uh, programmatically and build tools

around and communicate with our API

746

:

with blender's, API and literally take,

a Revit project building project like

747

:

Snowden here and get it into blender

and two, two clicks and a drag and drop.

748

:

Randall Stevens: I

749

:

don't know if you did

it, but maybe you can

750

:

open up the preview panel and just

show that, that, you know, full

751

:

model is able to be rendered, right?

752

:

Even just in the web, in the

web render, but it's full.

753

:

The entire project right, is in there

with all of its glory materials, textures.

754

:

It is back to that, like you

755

:

don't have Elvis dangling

information that you're gonna lose.

756

:

So, uh, you know, for especially

through these visualization flows, it's,

757

:

uh.

758

:

Corey Rubadue: So yeah, so just so I'm

selecting, you know, one of the exports,

759

:

if you will, or one of the translations

into RPC of the Snowden project.

760

:

And there it is in its full

glory, this full data set,

761

:

Evan Troxel: I mean, that's the biggest

problem with file formats, right?

762

:

Is there, there's inherent data loss.

763

:

Uh, and there's also just kind of

you, loss of life of brain cells every

764

:

time somebody has to open one of these

FBX or OBJ or DAE dialogue boxes and

765

:

figure out which boxes to check or not

check and, and to push these updates

766

:

and to export and to import and to

switch the axes and flip this and that.

767

:

And what you're saying here is like,

we solved all that and, and we've

768

:

actually made it like Randall, you

said one button export from Revit.

769

:

That's a big deal.

770

:

How much time has, would think back to

how much time has been spent in those

771

:

dialogue import export dialogue boxes, and

how much data has been lost through those

772

:

translation processes over time in the

architectural field alone, it's enormous.

773

:

Right?

774

:

And, and so this, this

is a big deal actually.

775

:

Corey Rubadue: Yeah, I mean, it's

a, it's an ongoing exercise, right?

776

:

I mean, we've, we've, constantly have to

evolve the communication of these assets

777

:

across these different file formats.

778

:

One renderer will treat one

texture map one way and another

779

:

render will treat it completely

the opposite way, as an example.

780

:

Right?

781

:

And we've gotta, we've gotta learn this,

whether learn it through the artistic

782

:

means that we have of having to have done

it before or our users doing it daily.

783

:

Um, but those are the, like,

the little nuances that we

784

:

gotta capture and figure out.

785

:

Okay.

786

:

How does a, transparency map, translate

differently between vray and n scape or

787

:

differently between, um, Revit and Rhino?

788

:

Um, we're trying to take all those

complexities out of the equation.

789

:

We can't hit everything with, you

know, one silver bullet, but we're,

790

:

we're progressing and our APIs are

progressing and the communication

791

:

around all that, um, is progressing.

792

:

So at the end of the day.

793

:

you know, we've got a work cut out for

us still to, to remain agnostic with

794

:

the technologies that we're building

so that we're not dependent on a

795

:

particular renderer or dependent on

a particular, CAD package or, or 3D

796

:

package, but that we can just hopefully

let people share and consume and,

797

:

uh, build

798

:

Randall Stevens: One of

799

:

the things, uh, one of the things Evan,

that I've just been excited about from

800

:

a veil standpoint of seeing this evolve

and, and, you know, now making its way

801

:

out into the market, is this idea that

inside of, inside of a firm, especially,

802

:

you know, a firm of any size, I always

refer to it as like you have people

803

:

at the edges of the network, people at

their desk working and building, right?

804

:

In the end, those are the, that's

where the, the work's being done.

805

:

But the idea that

806

:

as as assets or, or things are

created and, and the, the idea

807

:

that they can be published

808

:

centrally back into the system.

809

:

And then democratized right through,

through, Hey, I've made this in Rhino,

810

:

but you can consume it in Revit, right?

811

:

Immediately, like within

812

:

two minutes that

813

:

I've pushed this back in.

814

:

So I have this.

815

:

Um, I think that that's where this can

be at least initially, that people will

816

:

see, uh, a lot of the value out of it is,

Hey, we can, as we're creating things,

817

:

no matter which of these applications

we're creating them in, if they come

818

:

back through this hub and then they're

made available, then for anybody that

819

:

wants to now consume those, you're not,

you know, you're not having to rely

820

:

on whoever built it to go do something

and try to translate it for you.

821

:

It's like, no, this is

automatically being converted.

822

:

Not in a dumb way, but in a very

smart, you know, high level way

823

:

and keeping the integrity of that

data as it, as it moves around.

824

:

So I think there's a lot of value

to be exercised or, or found and

825

:

uncovered, uh, just through that,

uh, part of the process as well.

826

:

Evan Troxel: I think also one of the

things that this allows, through this

827

:

democratization process and through this

kind of single source of truth, but, but

828

:

multiple outlets, is that there's a lot

of tools in the toolbox and you wanna be

829

:

able to use the right one for the job.

830

:

And that doesn't mean you're gonna

use the same tool on every single job.

831

:

And it may even mean that there are

different people on the same project using

832

:

different tools to do different things.

833

:

And it's nice to have the ability to pick

the right tool for the job, but still know

834

:

that you can get all your stuff in there.

835

:

So if, if you've got a

vis person using max.

836

:

With Vray and you've got another person

doing interactive content in Unreal,

837

:

and you've got a designer sitting at

their desk using scape for day-to-Day

838

:

visualization, you can still easily

get the same content and models to all

839

:

of those people in a very simple way.

840

:

I mean that, that to me is, there's a

huge potential there to really unify

841

:

workflows, but still give choice to the

people who wanna be able to make that

842

:

choice of what tool to use for what job.

843

:

Corey Rubadue: Yeah, and Evan, I mean,

frankly, you know, having that choice

844

:

and being able to open people up to

many choices is one of the catalysts of,

845

:

you know, us breaking out of, in this

case, building all every RPC ourselves.

846

:

Like originally when this all started,

we were, we were producing every

847

:

RPC and that wasn't something that

I wanted to subscribe for and, and

848

:

for the company to, to commit to

just creating content constantly.

849

:

And so the first, you know, really

the first, you know, steps into

850

:

fovea, were like, Hey, wait a minute.

851

:

No, let's take our old avision creator

technology that we had and make it,

852

:

more accessible, cross platform, cross

device, and let people have choices of

853

:

which file formats they want to bring in.

854

:

Not just pigeonhole them into

going into three Studio Max

855

:

and doing certain workflows and

getting that data out into an RPC.

856

:

Like, okay, let's open this

up as much as we possibly can.

857

:

If you're a SketchUp subscriber

and you swear by using, um.

858

:

3D Warehouse to get your assets,

but, and, and you need to get a

859

:

chair from Warehouse into Revit.

860

:

It's two clicks from our, our setup.

861

:

Grab the file from warehouse, drag

it into fovea, publish it, drag

862

:

it from avail back into Revit, and

you've got, that chair from your,

863

:

your go-to library directly into

864

:

Revit.

865

:

Boom.

866

:

Done.

867

:

So it we're kind of, we try to

keep our blinders off and let

868

:

people use any of these options out

there to create their own content.

869

:

I don't wanna make content anymore.

870

:

So that's, you know, that was

really the one of the catalysts,

871

:

right?

872

:

Evan Troxel: It makes a lot of

sense because you're, you're almost

873

:

definitely not gonna make the

exact content that somebody needs

874

:

Randall Stevens: It's in stage five,

which is it's insatiable Right.

875

:

Journey,

876

:

Corey Rubadue: n never.

877

:

Right.

878

:

So I, I've got.

879

:

This is kind of the latest and

greatest and I'll just show,

880

:

um, what we're doing in Blender.

881

:

Again, blender, this, you know,

open source platform, easy for us to

882

:

develop around, so on and so forth.

883

:

So, we've built a publisher

for fovea directly in Blender.

884

:

So I'm logged in, uh, I can add a

file name here, so on and so forth.

885

:

I'm, I can set up my parameters

just as if I was inside of, uh, veil

886

:

and fovea like I showed earlier.

887

:

Um, but you know, what I have here is now

if I'm get a scene from my blender guy,

888

:

uh, I can do something like I can isolate

the CH chair from that scene, right?

889

:

And I can give this a name, chair,

I can give it a category furniture

890

:

chair tag, um, podcast as an example.

891

:

Again, I can choose my, my, uh, proxy

type, whichever way I want it to bring

892

:

in, uh, billboard, published a Fovea.

893

:

I'm gonna generate my preview of this.

894

:

So I've got a preview, a

thumbnail of the, of this.

895

:

So the thumbnail goes along for the ride.

896

:

Uh, I'll see this in a veil or wherever

the RPC or any of the file formats go.

897

:

And from here I can publish the Fovea.

898

:

So again, we're taking this data set,

pushing it up into the, to our services,

899

:

running the translations, all that stuff.

900

:

I can literally, now, once this is

published, go grab even a dot blend,

901

:

file a native blender file that we

are packaging up along for the ride

902

:

as the single source of truth for

the original, the original model.

903

:

And once this is all done, processing,

I can literally drag and drop

904

:

the same, uh, the blender file,

native blender file into blender.

905

:

And, Basically 360 degrees.

906

:

Uh, I can open up that file in fovea

edit, make edits and hit publish again,

907

:

and I'll get another blender file

out and I'll be able to drag and drop

908

:

that blender file into, into blender.

909

:

So in many cases, let's say full

360 bidirectional workflow, just,

910

:

and this is on an asset level, but

we can zoom, we can zoom out, and

911

:

we can do this at the, at that room

912

:

level

913

:

Evan Troxel: So you could theoretically

have somebody who's like a content manager

914

:

or a librarian, constantly be monitoring

what's coming into fovea and avail

915

:

and making changes in there to kind of

standardize things with certain metadata

916

:

or materials or looks or whatever it may

be, and publishing those to a content

917

:

library for everybody to use and they

don't have to know Blender as an example.

918

:

Corey Rubadue: So, you know, back to back

to avail and what, what people will see.

919

:

So the fovea channel is a,

is a plan level channel.

920

:

by

921

:

design there will

922

:

Randall Stevens: You mean?

923

:

You mean by that you mean company-wide

We call, we call 'em plans.

924

:

Corey Rubadue: Companywide plan.

925

:

Sorry.

926

:

Yeah, sorry.

927

:

Companywide plan.

928

:

Yeah.

929

:

And, uh, what that means is they'll,

they'll be a, a gatekeeper at the, at,

930

:

you know, at the top of this, right?

931

:

There's an administrator or

publisher that will have access.

932

:

To all of what's coming into Fovea.

933

:

And at that point, you know, you

can have folks being able to approve

934

:

pieces of content and get that

content, um, in a, in a focused

935

:

custom, uh, avail channel, if you will.

936

:

Um, so that folks can have their

own firm wide channel that's

937

:

absolutely separate from, from Fovea,

938

:

even

939

:

Randall Stevens: Yeah,

Corey's got, Corey's got

940

:

tons of stories.

941

:

Ray's told me about.

942

:

You know, somebody's

under a tight deadline.

943

:

They've got a project, there's an asset

that was available on Turbo Squid or that

944

:

they found, you know, somewhere, right?

945

:

May, maybe it was even on their

network, but it was in the wrong file.

946

:

It was in a different file format

than whatever application they're

947

:

working in, but they need it.

948

:

So for them to be able to figure

out how to get that content that

949

:

they know is right here in hand

into the right form, right, is very

950

:

frustrating, time consuming, and, and

951

:

ultimately expensive process.

952

:

So I think

953

:

there's just a ton of, you know, of, of

real un, you know, uncovering all this

954

:

value in, Hey, you can just now drop that.

955

:

I don't even have to know

anything about Blender.

956

:

All I have to know is that this

FOIA application knows what to do

957

:

with it, and then boom, here it is.

958

:

I can now

959

:

bring that into Inscape, or I can bring

that into Revit or some other place and

960

:

leverage it immediately is very powerful.

961

:

Corey Rubadue: Yeah.

962

:

Evan Troxel: Again, just how much

time that could potentially save,

963

:

because that is an iterative process

of, uh, trial by fire, right?

964

:

Which is like, okay, we're gonna import

it and see what didn't work, and we're

965

:

gonna go back and we're gonna check some

different boxes and we're gonna try again.

966

:

And okay, I, I have to learn how

to use this because I don't use

967

:

blender and I'm, I'm used to 3D

Studio Max or whatever it is.

968

:

And that is, like you said, it's

an expensive proposition because

969

:

you're paying staff by the hour or

whatever, and what a pain, like, it's

970

:

not gonna work right the first time.

971

:

It never works right the first time

unless you're an expert in that particular

972

:

file format translation process.

973

:

So, uh, taking all the guesswork

outta that is, I think there's,

974

:

that's the kind of thing that.

975

:

It's hard to explain that because firms

don't necessarily, although they bill by

976

:

the hour and they sell time for money,

they don't necessarily value their own

977

:

time in the ways that we're talking about

here because it's just, that's how it is.

978

:

Right.

979

:

Everybody just expects that's how it is

and when you can actually show that, okay,

980

:

we, we took this thing that normally takes

like literally 40 minutes for somebody

981

:

to figure out how to do it right the

first time because they don't use the

982

:

application every day and X, y, and Z

reasons, and we made it take 20 seconds.

983

:

Like that is, that's in,

that's a big deal inside of a

984

:

Randall Stevens: Yeah.

985

:

Now multiply that by however many,

anyway, the, the bigger the firm.

986

:

How many times

987

:

is that happening across

or across the year?

988

:

Evan Troxel: Right,

989

:

Corey Rubadue: can tell you guys,

you know, I just got off a call.

990

:

And, um, the, the gentleman's trying to

get assets that he scanned with his iPhone

991

:

from Poly cam into his Revit project.

992

:

And, you know, I got him access

to our iOS app beta and just did a

993

:

15 minute tour of how things work.

994

:

Dragon dropped an RPC of a poly CAM

scan from avail directly into Revit.

995

:

And, um, I, I, he, I mean,

honestly he couldn't believe

996

:

that it was that, that easy.

997

:

Um, it was, but it, I also couldn't

believe like how well the scan data was

998

:

coming in, uh, that he had captured with

poly cam and how powerful Poly Cam was.

999

:

So I'm sitting there thinking to

myself, you know, we zoomed out

:

01:04:08,073 --> 01:04:11,253

before, well before we made the iOS

app, and made a decision that we

:

01:04:11,253 --> 01:04:15,993

weren't going to get in the business

of making a a iOS scanner app.

:

01:04:16,563 --> 01:04:20,403

And that what we wanted to do was

actually embrace as many of the

:

01:04:20,403 --> 01:04:24,963

scanning apps as we possibly could in

the file formats that they deliver so

:

01:04:24,963 --> 01:04:31,383

that you can get those assets up into

fovea and into Revit, rhino, SketchUp

:

01:04:31,383 --> 01:04:33,063

or wherever as quickly as possible.

:

01:04:33,993 --> 01:04:39,573

And, you know, he happened to be a poly

cam user who's had a bunch of scans and

:

01:04:39,573 --> 01:04:45,423

he's like, I need these, this bike scan,

this bike, I want this bike in, in Rev.

:

01:04:45,423 --> 01:04:45,513

It.

:

01:04:45,987 --> 01:04:51,149

so the processing time, was probably

the, you know, the longest part

:

01:04:51,149 --> 01:04:55,739

or the most built up anticipation

and you're like, okay, well,

:

01:04:55,769 --> 01:04:56,939

well, when's it gonna get here?

:

01:04:56,939 --> 01:05:04,230

So, Once it got produced and the drag and

drop happened, he, his eyes just lit up.

:

01:05:04,230 --> 01:05:04,529

It was

:

01:05:04,620 --> 01:05:05,220

pretty awesome.

:

01:05:05,960 --> 01:05:08,055

Evan Troxel: so is that

throwing an RFA into Revit?

:

01:05:08,055 --> 01:05:08,715

Is it a family

:

01:05:08,940 --> 01:05:13,500

Corey Rubadue: So, well, technically

speaking, um, when you drag and drop

:

01:05:13,500 --> 01:05:16,110

the RPC into Revit, we build an RFA

:

01:05:17,000 --> 01:05:17,120

Evan Troxel: Hmm.

:

01:05:17,130 --> 01:05:20,131

Corey Rubadue: behind the scenes

that contains the, the link, if

:

01:05:20,131 --> 01:05:22,771

you will, to the, to the RPC.

:

01:05:23,581 --> 01:05:25,861

Um, so there's, there's an RFA involved.

:

01:05:25,866 --> 01:05:26,701

Yeah, we create it

:

01:05:27,438 --> 01:05:33,048

when we create that family, we

take metadata from the asset,

:

01:05:33,048 --> 01:05:37,518

so the tags and category of

the asset and things like that.

:

01:05:37,523 --> 01:05:43,308

And we actually try to put out as much

of that into the family as we can.

:

01:05:43,308 --> 01:05:48,078

And if it's a furniture piece as an

example, we make sure that, that the.

:

01:05:48,563 --> 01:05:54,018

FRFA that gets made is actually a Revit

family, a furniture family, and not just

:

01:05:54,018 --> 01:05:56,688

a g you know, generic, um, entourage

:

01:05:56,808 --> 01:05:57,468

family as an

:

01:05:57,468 --> 01:05:58,008

example.

:

01:05:58,278 --> 01:06:01,398

So we try to, try to do that.

:

01:06:01,398 --> 01:06:04,698

we can't do that for every Revit

category 'cause of the way things

:

01:06:04,703 --> 01:06:05,988

are in Revit at the moment.

:

01:06:05,988 --> 01:06:10,105

But, I know in the future we'll be

able to do something pretty slick, um,

:

01:06:10,255 --> 01:06:15,175

around that, uh, to be able to keep,

keep these, uh, these assets pushed into

:

01:06:15,175 --> 01:06:17,035

the right Revit category as an example.

:

01:06:17,820 --> 01:06:18,040

Oh.

:

01:06:18,485 --> 01:06:23,298

Evan Troxel: did a poly cam scan

and brought that model into rhino.

:

01:06:24,198 --> 01:06:28,338

And even then, it was the first

time I've done it and it, because

:

01:06:28,338 --> 01:06:32,568

I I Black Friday dealed the poly

cam for a year this last year.

:

01:06:32,658 --> 01:06:32,898

Right.

:

01:06:33,138 --> 01:06:34,338

And I, I wanted to play with it.

:

01:06:34,788 --> 01:06:39,228

And so I did a room scan model and

brought that into Rhino and just

:

01:06:39,228 --> 01:06:42,408

getting the scale right and everything

is still, it's still an issue, right?

:

01:06:42,408 --> 01:06:46,398

Like just getting stuff to show up,

how you expect it to just show up

:

01:06:46,458 --> 01:06:49,428

immediately is not how it actually works.

:

01:06:49,668 --> 01:06:54,228

You still need to know about the

translation and the, and which boxes

:

01:06:54,228 --> 01:06:58,098

to check and which scale to apply and

what, what units are you working in, in

:

01:06:58,098 --> 01:07:01,158

poly camm and what units are you working

in, in Rhino and all those things.

:

01:07:01,518 --> 01:07:05,928

And to make it easy is, it's a,

again, I just keep going back to this.

:

01:07:05,928 --> 01:07:06,528

It's a big deal.

:

01:07:06,533 --> 01:07:09,318

It saves so much time to just

have it work right the first time.

:

01:07:09,318 --> 01:07:11,148

That, that it's incredibly valuable.

:

01:07:11,253 --> 01:07:13,758

Corey Rubadue: Yeah, and we, I mean,

we take some of that, a lot of that

:

01:07:13,758 --> 01:07:15,294

guessing outta the equation, right?

:

01:07:15,294 --> 01:07:20,784

that poly camm scan comes in and we

in fovea, um, I mean, I can show this.

:

01:07:21,314 --> 01:07:22,604

here's, the chair, right?

:

01:07:22,604 --> 01:07:26,894

So there's that chair coming

in from, from blender.

:

01:07:26,954 --> 01:07:28,064

Uh, I can preview it.

:

01:07:28,274 --> 01:07:28,574

Um.

:

01:07:30,130 --> 01:07:32,350

That's not a slouch of a model either.

:

01:07:32,350 --> 01:07:38,350

I didn't do any decimation or, uh,

optimization of that geometry at all.

:

01:07:38,350 --> 01:07:43,360

So it, it's, you know, there's blanket

there, it's been deformed in, in the

:

01:07:43,540 --> 01:07:46,810

application, the pillow, the wrinkles,

all that, all that defamation.

:

01:07:46,810 --> 01:07:50,590

I mean, it comes at a

cost polygon triangles.

:

01:07:50,590 --> 01:07:54,452

Um, but again, we can decimate

that, um, to some degree.

:

01:07:54,512 --> 01:08:00,632

Uh, not, you know, we don't have

amazing, uh, decimation tools, um, that

:

01:08:00,632 --> 01:08:02,492

can completely rebuild all the meshes.

:

01:08:02,492 --> 01:08:06,302

But, um, but yeah, this

is, that's a chair.

:

01:08:07,232 --> 01:08:07,522

Evan Troxel: Nice.

:

01:08:07,772 --> 01:08:10,292

Corey Rubadue: And again, this is

ready to just drag and drop and

:

01:08:11,792 --> 01:08:12,302

Rhino

:

01:08:12,452 --> 01:08:12,721

Randall Stevens: So in

:

01:08:12,721 --> 01:08:17,011

that amount of time, right, you went

from, you know, maybe one format, native

:

01:08:17,011 --> 01:08:24,362

format that was completely unavailable to

you, to, I can now drop this into any of

:

01:08:24,542 --> 01:08:26,252

the applications that I'm used to using.

:

01:08:26,252 --> 01:08:26,312

So

:

01:08:27,607 --> 01:08:27,797

Corey Rubadue: Yep.

:

01:08:27,807 --> 01:08:30,482

Evan Troxel: when, when you're

publishing it from Blender in this

:

01:08:30,482 --> 01:08:36,002

case and you ingest it, fovea, fovea

is making all those other versions

:

01:08:36,006 --> 01:08:37,772

of the, the, that you showed earlier.

:

01:08:37,772 --> 01:08:43,502

So the GLTF and the the RPC and,

and all those different versions.

:

01:08:43,502 --> 01:08:44,551

Randall Stevens: all that, all that

:

01:08:44,551 --> 01:08:45,152

concurrently,

:

01:08:45,307 --> 01:08:47,551

Evan Troxel: then that's all, all

that work is done and then all I

:

01:08:47,551 --> 01:08:50,582

have to do is pick what, where I

want to send it to and it's, it's

:

01:08:50,672 --> 01:08:51,752

Randall Stevens: All 13 of those,

:

01:08:51,872 --> 01:08:54,182

all 13 of those formats

are available immediately.

:

01:08:54,301 --> 01:08:54,482

Yeah.

:

01:08:54,662 --> 01:08:55,022

Evan Troxel: Okay,

:

01:08:55,697 --> 01:08:55,937

Corey Rubadue: Yep.

:

01:08:56,042 --> 01:08:56,402

Evan Troxel: cool.

:

01:08:56,987 --> 01:09:00,287

Corey Rubadue: So again, if you've

got a blender guide who's great at

:

01:09:00,292 --> 01:09:06,069

modeling and you, your, let's say

visualization is on Unreal Engine.

:

01:09:06,770 --> 01:09:10,100

You can use our RPC plugin, an

Unreal engine, or you can import

:

01:09:10,100 --> 01:09:13,581

in a data Smith file, which is part

of the output as an example, right?

:

01:09:14,371 --> 01:09:14,946

Randall Stevens: That's very

:

01:09:15,140 --> 01:09:18,231

Corey Rubadue: So actually if

I, if I filter down through

:

01:09:18,231 --> 01:09:25,341

here, I by our application, so

application Unreal engine and chair,

:

01:09:26,691 --> 01:09:30,890

um, you know, you can see I've

got a handful of file formats

:

01:09:30,890 --> 01:09:31,791

that are pulled down here.

:

01:09:32,451 --> 01:09:35,961

I've got, uh, FBXO, B, JRPC,

and DA, you Data Smith.

:

01:09:35,961 --> 01:09:38,390

Those are all the file types

that you can import into.

:

01:09:39,411 --> 01:09:41,600

Unreal Engine, outta the box, right?

:

01:09:42,560 --> 01:09:47,571

Um, so we give you the, the, we open

up, we open up and give you choices.

:

01:09:47,906 --> 01:09:50,666

if you wanna bring in an FBX, by

all means, if that's your workflow

:

01:09:50,671 --> 01:09:53,456

and that's the way you prefer to

work, then grab the FBX and go

:

01:09:54,297 --> 01:09:54,567

Randall Stevens: Well, I know

:

01:09:54,567 --> 01:09:57,567

there's a ton more that we can get

into on this, but I think we're

:

01:09:57,567 --> 01:10:03,717

going to, uh, plan on kind of putting

a wrap on this session and then.

:

01:10:04,347 --> 01:10:08,187

Follow up with maybe a deeper dive,

Corey, assuming that you're willing

:

01:10:08,187 --> 01:10:12,627

to, uh, kind of get into a little

bit of the behind the scenes of some

:

01:10:12,632 --> 01:10:15,267

of the, um, some of the decisions.

:

01:10:15,297 --> 01:10:19,647

You know, you started out saying that

even, even where this ended up was

:

01:10:19,647 --> 01:10:24,507

not where you originally, you know,

envisioned and, and the path kind of took,

:

01:10:24,537 --> 01:10:26,637

took you there in, in different ways.

:

01:10:26,637 --> 01:10:31,107

So, uh, maybe we can end up, um, you know,

coming back to this in another session

:

01:10:31,107 --> 01:10:34,497

and digging in a little deeper about what

some of those decision processes were.

:

01:10:35,209 --> 01:10:36,784

Corey Rubadue: Absolutely my pleasure.

Follow

Links

Chapters

Video

More from YouTube