reboot activity as a 3d graphic designer

[Tom Hamelrijckx]

Hi everyone,

 

Until about seven years ago I was active as a 3d graphic designer, doing visualizations for architecture projects and giving sketchup courses. I did that for about 15 years until I was given the opportunity to start working in a carpentry workshop. I have a background in carpentry, and I just happened to start missing the manual aspect in my 3d work, so I gladly accepted that challenge. It worked very well for a couple of years, and little by little the focus shifted more to the carpentry work.

 

But then some life events made me suspend my professional activities for quite a while, actually long enough to reconsider how I will move on now. Both activities have been put to a halt completely, but they both give me energy and I would like to keep on exercising both of them part time, one having priority over the other.

 

So the idea is to first start doing 3d again, and get that part rolling smoothly. I lost some technical skills in those last years though, so I wonder how to get this started again. That's what this post is about, I could use this communities' professional opinions, insight and ideas.

 

• question 1
  

In a brainstorm with an acquainted architect that was interested in launching an internal 3d cell in his firm, we thought it might be interesting to let our mutual needs for a launch runway coincide with each other. I thought it would be good to develop a workflow in which all the current projects of the firm could get a constant 3d update as the project moves on, and gradually let that model develop according to the architects' needs. On demand, low-workload renderings could be produced to support progress briefings internally or with the client. In time I'm supposing this can get ran smoothly enough to get the renders better and better. Final renderings would be kept on outsourced anyhow. Until now I didn't get any feedback from this firm anymore, so I guess there are issues that don't convince them.

So that's my first question; what do you think might make my case more convincing? I wonder as well if this project is not too big to start for me, considering I have been out of this business for a while.

 

• question 2
  

Second question is, could I reconsider other more modest options than starting such a big project? A refresh training maybe? Or in another sector? It doesn't necessarily need to be in architecture, I think more technical applications like photogrammetry might suit just as well.

 

 

I appreciate all feedback and thank you for reading this already ;-)

Edited October 14, 2019 by Mazinger

[Scott Schroeder]

Question #1:

Most of this work has been solved by the introduction of Enscape into many places. Now the architects themselves are doing all of the in between renderings right in the software they are working in. As Enscape continues to get better and adding more content, these renderings will almost all but disappear from your traditional rendering workflow.

 

Question #2:

If you feel you are rusty, I would suggest learning on Blender. It is a fantastic piece of software that is free. From there, you can explore real time options such as Unreal / Unity.

 

Real time applications may be where you want to gain your experience. With Unity Reflect now in beta and Unreal + Datasmith, there are plenty of options that many architects are exploring. Some want VR and AR, some just want a high quality real time experience.

[Tom Hamelrijckx]

Thanks Scott that's already very valuable info.

 

Since I am in this reboot phase the purpose is to integrate a training phase in the beginning, and Blender was on my list.

Apart from what you mentioned, is there any other valuable software or skills that might be interesting and with which I can distinguish myself a bit?

 

I suppose 3d modeling and post production in PS are still indispensable skills. Or not?

 

As well, sometimes I wonder if it wouldn't be valuable to be able to produce high-res environment imagery with photogrammetry (and a drone), but it looks like that would take me too far off track. Am I wrong?

 

(PS I was using Sketchup and Vray)

[Scott Schroeder]

Absolutely, post work and 3d modeling are still very important skills. In my own personal opinion, there are 2 somewhat distinct paths in rendering. One side does everything in the rendering VFB and very little Photoshop (my current workflow). The other side does 20% in the rendering and 80% in Photoshop. Neither workflow is bad, and both have their advantages and disadvantages. However, I feel that you need to pick one side and stick with it as much as you can because the workflow can be different and may not support the other side.

 

I would for sure start to learn Substance Designer as that tool is becoming integrated into the workflow these days. Any of the tutorial series by Daniel Thiger are an excellent place to start. That may be a good area to show value to an architect/designer. With Substance Designer, you won't ever find yourself telling a client you can't really visualize that material or be left trying to make a 128x128 pixel sample tile well in your scene. Since the Substance workflow can feed most all visualization engines inside a studio, you can make yourself quite important.

 

The drone stuff is all fine and good, but throw a rock and you can get a service that does that. This is a personal opinion, but most places that need that type of work do it in house and most place that don't need it do it with cheats or Google Earth snapshots. That area may be hard to break into or you'd be faced with constant, "I need this yesterday!" requests.

[Jose Negrete]

Hi Tom,

 

I just want to add that from my experience, most firms we work with that do their own "in-between" renderings are using Enscape or Lumion. I also agree with learning blender (2.8) with its many improvements in UI and an increase in support that they're getting for development.

 

If you get into Unreal Engine and sign up for their Unreal Studio beta you will find quite a bit of training on their site including a bit of substance designer - though I can't say how it measures up to Daniel Thiger's.

[M V]

I too am a SketchUp VRay guy but also use Lumion and Enscape because this is the way of things in an architecture office these days. If I was in your situation and rebooting perse, I would take the above advice and learn Substance and Unreal Engine. You can still model in SketchUp if thats your comfort zone and just hit the pavement sunning with Unreal - you can even looking into VRay for Unreal honestly. This is where things are heading and this is where the paying clients will be. The one-off, progress renderings are now being done in house so you need to push skills far beyond that and VR is the next frontier for us.

[Tommy L]

VR has been a hot future product, just around the corner, for about 15 years. Clients will always need images, sometimes need animations, rarely need VR.

UE4 is an amazing thing, but I think its really in the same vein as Enscape. It is fast, good, but not the best. All other things being equal, the best imagery is done with Corona.

[Tom Hamelrijckx]

Great, thanks for all of these valuable tips.

I will look into the mentioned software.

 

To refocus a bit on the first question, is anybody here working inside an architecture firm doing only 3d and project visualization for that one firm? I'm curious to know about your task description. I guess the "in-between" renderings in Enscape or Lumion still need a modeling, texturing and setup phase that an architect is not going to spend his time on, no?

 

And aside all this, I remember the only less pleasant aspect of this work being the last-minute demands and the resulting stress. How do you guys handle this? It might be worth another topic maybe....

[Scott Schroeder]

I work in-house for an architecture firm. While I still do some in-between level renderings, I am mostly trying to focus on doing the marketing level renderings. As a studio, we've decided to keep that money in-house rather than outsource those later stage renderings.

 

If the Revit/SketchUp/Rhino teams are setting their models up right, there is no need for someone else to help them with the model for those design level renderings. We can support them with texture creation though as that is not a skill most architects have or are willing to learn.

 

Rendering is always going to carry last minute stress and demands, that is just part of the job at times. However, I really try to remember something that Valve Software has in their employee handbook and that is while last minute crunch time will happen when you get near a milestone or final delivery date, sustained crunch time is a sign of a failure in planning and communication. I try to make sure we have all of our ducks in a row and are on track through every project. That way if needed, it is only 2-4 hours of overtime instead of 40.

 

Perhaps that attitude for Valve Software is why I have been waiting literally half my life for Half Life 3.

[Tom Hamelrijckx]

Scott thanks again for your feedback,

 

Do I understand correct

- that the architects/designers in your firm are in the teams that you call the 'Revit/SketchUp/Rhino teams'? So they design in that software?

- that you only do texturing, rendering and postproduction? Or still more than that?

[Scott Schroeder]

Yeah, most of our design is done in Revit. We may do some early design exploration in SketchUp or Max but primary software that is used to create the full construction model is Revit.

 

While we do have Revit models, most of them are unusable as direct to rendering (Max, Unreal, Lumion, etc) models. Too many architects have fallen for the Autodesk propaganda that Revit models are just one click exports. So for the most part, I am working in a cleaned up or fully rebuilt Max model in parallel with the Revit teams. While this isn't exactly an efficient workflow, it is the workflow that needs to happen to get usable models for rendering at the marketing level quality. Marketing renderings being the ones you see in ads, here on CGArchitect, or on some other architecture publication.

 

As the Revit teams are using Enscape earlier in their process, we are starting to see the model quality become much better so maybe we are close to the one-click export, but Revit exports will probably always need to be taken through software like Max to clean it up. Revit itself is just poor at converting curved faces. What should be something with 24 verts ends up with 24,000.

 

We also do a lot of furniture modeling and custom texture creation. We'll get the specs from the interior designers and have to create 1:1 replications of them so the clients can get a true sense of what their space will look like.

[akshayarora]

Yeah, most of our design is done in Revit. We may do some early design exploration in SketchUp or Max but primary software that is used to create the full construction model is Revit.

 

While we do have Revit models, most of them are unusable as direct to rendering (Max, Unreal, Lumion, etc) models. Too many architects have fallen for the Autodesk propaganda that Revit models are just one click exports. So for the most part, I am working in a cleaned up or fully rebuilt Max model in parallel with the Revit teams. While this isn't exactly an efficient workflow, it is the workflow that needs to happen to get usable models for rendering at the marketing level quality. Marketing renderings being the ones you see in ads, here on CGArchitect, or on some other architecture publication.

 

As the Revit teams are using Enscape earlier in their process, we are starting to see the model quality become much better so maybe we are close to the one-click export, but Revit exports will probably always need to be taken through software like Max to clean it up. Revit itself is just poor at converting curved faces. What should be something with 24 verts ends up with 24,000.

 

We also do a lot of furniture modeling and custom texture creation. We'll get the specs from the interior designers and have to create 1:1 replications of them so the clients can get a true sense of what their space will look like.

 

I am little satisfied that you also need to redevelop the models in 3dsmax otherwise I was in a dilemma that we are only visualizers left which are doing this