What was your UE4 learning curve?

[Michael J. Brown]

I made a similar post over in the Software thread, but have gotten no replies yet.

 

I'm just trying to find out what peoples experiences have been who've either successfully or unsuccessfully migrated from 3dsMax/Vray to Unreal Engine 4. How steep was the learning curve? What where the biggest challenges? What workflow changes had to be made?

 

How long did it take before you were able to produce the same quality work in UE4 as you did/do with Max and Vray?

 

 

Thanks

[Scott Schroeder]

To do it properly, it is a pretty steep learning curve. I'm usually concerned with modular modeling and re-use as well as proper UV layouts. My current R&D is heavily concerned with mobile deployment and your average user actually being able to use this on their computer without having 10 Titan X's hooked up, thus the need for modular re-use and proper set ups.

 

Then come changes and how to deal with them so you are not constantly having to re-do work. So UE4 is best used when you are 99% sure the design won't change. If you are still in the design process, sticking with Max is going to save your health and sanity.

 

If you want to do the 100,001th concrete room with one chair and maybe a messy bed, then the curve isn't so bad. Just jam your unoptimized objects in there and enjoy your real time at 1 fps or try to convince your client why they need to spend 6k in video cards to run your scenes.

 

UE4 is great and is slowing gaining traction in the industry, but the old Max/Vray process smokes UE4 in every aspect of production. Sure UE4 is "real time" but you have to spend easily 10 times more time getting assets into UE4 than you save in overall render time. 1 week time for Max/Vray = 4-6 weeks for UE4 during the learning process. As you build your UE4 asset library and learn the process, you can start to save time. However, I would safely say that even with a good UE4 library and solid knowledge, 1 week in Max/Vray still is about 2 weeks in UE4.

[Michael J. Brown]

Thanks much, Scott.

 

As I had begun delving into the UE4 learning materials, the first bucket of cold water was the entire UV mapping issue. I quickly realized I was going to be deep in the hurt bag given that the work my employer wants to see done in UE4 is stuff that changes constantly and frequently with very short timelines (auto show stand design).

 

This helps me out a lot.

[Scott Schroeder]

Your best bet for things like that would be Vray RT or other GPU rendering. It's faster than stand-alone rendering and quicker than going all the way into UE4. UE4 is definitely not for the "I just want to see it in 3D first" crowd.

[RyderSK]

I subscribe to everything Scott said.

 

I had super much fun using Unreal4, and the learning curve steepness is greatly lessened by what I think is the best UI in CGI application ever, and the broadest documentation available. Learning it properly, is simply matter of short time, and good reading comprehension.

 

But real-time engine usability in commercial production of visualization is so far detached where offline rendering currently is (which almost can't be more streamlined). Asset production, scene optimization and organization, work planning, deployment of final product, etc..

 

It's not going to replace or even much supplement current pipe-lines anytime soon, or I dare to say ever. It's simply going to be new niche product for next 3-4 year. But exciting one at it :- )

[philippelamoureux]

The workflow is getting easier and easier, thanks to scripts (detaching, unwrapping, etc).

 

Here's a nice thread about Lightmass (photon mapping in unreal). https://forums.unrealengine.com/showthread.php?88952-Lets-make-Lightmass-EPIC-(and-understandable)

 

Devs are actively working with us for solutions tailored to arch viz. 1st thing is going to be portals :-)

 

I agree with Juraj for most of what he said. hopefully more than 3-4 years tho. :-P

Edited November 2, 2015 by philippelamoureux

[Michael J. Brown]

Thanks for the good replies, fellas.

 

I've made up my mind. I don't think I'll be pursuing UE4 at this time. I'm just going to stick with what I know and have invested the bulk of my years perfecting - Max + Vray. Mainly because of the frequency of design changes and the sometimes ridiculous turn-arounds in which they are expected. Just to produce still image visuals (for which changes are often made on the spot and well within the 24-hr window leading up to said deadlines) I've had to pull many near-all-nighters at the office.

 

I don't want to imagine the stress level of trying to keep the client happy by producing real-time VR content too. If I were still young and fresh out of architecture school (22 years ago), I'm sure I'd look at this glass of water differently. But I'm not trying to set the world on fire anymore. Some sacrifices are simply not worth the reward.

 

Thanks again fellas!

[RyderSK]

Don't be discourage to at least try it though :- ) It's lot of fun and you'll enjoy it, and you'll never know.

[Michael J. Brown]

Oh, I hear ya' Juraj. I'll definitely keep it on my own personal bucket list. Just not for what the office is trying to do right now.

[philippelamoureux]

I've been really focusing on Unreal for the past 6 months and every time I have to go back to max and V-ray or Corona I'm bored after a couple of minutes lol. Like Juraj said, the UI is probably one of the reason I like Unreal so much. It's very good and FAST! I can say right now that I can do better stuff in Unreal than with vray (because unreal is more intuitive and you get instant feedback). I know vray can look better in the hands of a pro,but I'm not one.) I like a WYSIWYG workflow.

 

What I suggest you is to at least keep an eye on the software, the updates, etc so if you decide to try it later on you are not going to be too overwhelmed!!

 

There is a kind of merging between architecture viz and gaming that I don't dislike, to be honest! I think becoming good with a game engine could open even more doors, if it's not for arch-viz it could be something else!

 

Just remade an Evermotion scene in unreal, it's fully playable. Runs realtime buttery smooth. Now if I had a god damn Rift! lol

 

 

I say, exciting times ahead!

Edited November 3, 2015 by philippelamoureux

[Michael J. Brown]

Good to know, Philippe (and to see). Thanks.

 

So, do you do anything in Max anymore? (modelling)

Edited November 3, 2015 by renderhaus

[philippelamoureux]

I still need to use it for modelling and I'll do a render with v-ray/corona if a client really just want stills with very high fidelity. Still try to push it towards unreal because I feel more comfortable.

[philippelamoureux]

Unreal is just getting better everyday! Look at that!

 

[Nic H]

looking foward to seeing more out of focus white beds and concrete walls from UE4 arch vis artists

always doing inventive work IMO

[Scott Schroeder]

[philippelamoureux]

looking foward to seeing more out of focus white beds and concrete walls from UE4 arch vis artists

always doing inventive work IMO

 

This is not specific to unreal arch-viz artists...kitchens and bedrooms are every-****ing-where on most forums!

[Nic H]

yes it seems to be very a very popular choice of subject!

one day i too will *hopefully* make my own kitchen where i can change the joinery colours

[beestee]

What I suggest you is to at least keep an eye on the software, the updates, etc so if you decide to try it later on you are not going to be too overwhelmed!!

 

There is a kind of merging between architecture viz and gaming that I don't dislike, to be honest! I think becoming good with a game engine could open even more doors, if it's not for arch-viz it could be something else!

 

I say, exciting times ahead!

 

I am in a similar mindset about realtime right now. I do feel that Lumion and Twinmotion are hard at work to converge this technology with architectural visualization and currently offer a more ideal toolset for architectural visualization. We have yet to reach a no-compromises convergence, but I don't feel that it is more than a few years away.

 

It would be best to stay current on the technology even if it cannot yet be integrated into your production process, although I imagine the workflows will only become more approachable when the technologies ultimately converge.

 

I understand the idea that it is unreasonable to expect clients to have video cards to run realtime visualizations, but it is something that offline renderers do not offer at all. The realtime solutions can offer stills in seconds and video output in minutes, though, which is what is already being provided by offline renderers in hours or days. Just mentioning that because I feel there are significant benefits to using this technology aside from only realtime walk-throughs.