Will SketchUp Ever Wear Big Boy Pants?
-
There was an earlier thread about SketchUp and hi-poly models and I am sitting here wondering why every modeling application out there handles hi-poly models just fine and SketchUp still does not, even after all these years.
It's really the one thing keeping SketchUp from really pushing the limits and playing with the big boys and really the only complaint I have about SketchUp. I have been a user since the At Last days and will probably always be a SketchUper but there are times where I feel like my models and my renderings would be much improved if I could get hi-poly models into SketchUp and work the way SketchUp does with the low poly stuff. Import one hi-poly sofa and my model is not 10X the file size and runs like I just took a time machine back to 2004 SketchUp in 32 bit.
Proxies and Skatter have really helped get around all of this, but it still leaves me scratching my head why nothing has been done about it and if anything will ever be done about it.
-
We're a small firm, 9 seats (excl. director, admin seats and tea-person) and transitioned to REVIT LT (we're probably going to be upgrading some of those in a bit) at the end of last year. All new machines were acquired as well, pulling out all the stops so-to-speak. Capital outlay, as you can imagine, was not small change.
I was wanting to pitch SU Pro (thus with LO; also a few essential commercial extensions) to the director as I know it can do what we need, and more, and is relatively easy to work; I even gave him some of Nick Sonder's video's to watch; but in the end it was not to be. If I compare how REVIT LT runs to how SU runs on the same hardware (i7 7700K CPU at 4.4GHz on average, 64GB RAM, 1080 GTX, SSD's, Win 10 Pro), let's just say it's a little disappointing how SU compares currently. Yes, I know about structuring your model, how you model, groups and components, choosing the right styles, etc. It's probably a little unfair comparing SU with an Autodesk product since they have spent a lot more on R&D over the years (yes I know Autodesk acquired Revit from another company in the early 2000's), their product also costs more in comparison.
I'm also a keen supporter of SU, yet I wonder about what the future holds for it. Is it destined to be a niche product when it comes to being utilized from start to end on a project, or for small firms only (mainly, I'm sure there are a few places out there using SU that could be classified as anything but small).
Every new version has improvements, thank goodness for that, but there are undeniably some areas where great improvements would help SU "hit it out of the park", or close to it.
-
it's great for small firms. Less overhead in many ways, not just money--but just to get out some presentation models is challenged as SU bogs down, and you just want to be able to bring some plants to finish the job without coming to a crawl.
-
Thanks for trying JuJu. I can't imagine giving SU the boot for Revit. Once you have a system down that you know, I simply haven't seen anything that competes. I'm still a one man show and just completed 6 custom complex homes in 5 months....all either started construction or in the building department for permit review. There is no software I know of, that could allow me to achieve that. Granted it was a very busy 5 months!
Each of these projects had between 60-80 sheets and included multiple renderings, animation, kubity model etc.
If any of you are attending basecamp, I have some significant updates to share, along with my process for detailing.
Nick
-
There are many, like Sonder pushing the limits and boundaries of what SketchUp can do. Now imagine being able to bring in a 300K poly object and have it run as smooth as importing a low poly chair from Form Fonts. Why isn't this possible, if it is for so many other modeling applications? Does it require a full rewrite of code that Trimble doesn't want to invest in (aint broke - don't fix)? Is it the limitations of OpenGL? It makes me wonder if I am holding myself back by continuing with SketchUp.
It's an amazing tool that has allowed me to find a lot of work and allowed me to design and illustrate very quickly. My struggle is when I want to render an interior and need to populate the model with hi poly objects converted through the convoluted process of getting it from OBJ or MAX into SketchUp only to make my model come to a crawl. Revit, Max, Blender, Thea Studio, Lumion and all other programs I have used can take these models and work with them. SketchUp simply cannot and I think we should be asking why.
-
I've been an SU guy from the @Last days,and having been using it in our work for years as well as my private at home stuff. We still use it at my work for initial studies and such. It's easy, fast and fun. However--when we land on a concept, we turn it over to our consultants of which we employ a lot . .. i mean a LOT. They in turn hire their consultants. . .MEP, Structural, etc.
They all speak REVIT.
I tried to get into REVIT in its infancy. . .(2.0 or something like that) didn't get it. so when SU came along it was great for me. But I need to learn revit now. I still use SU but . . BIM is the way of the world--in our world.
-
My 2 cents as a non-professional …
I too have cursed SU many, many times over the course of the years for its inability to cope with high poly models. Loudly, even.
And then the spirit of the Buddha touched me, and I thought: why not love it for what it can do, rather than hate it for what it can’t? Since then I’ve basically used SU as an archviz modeling plugin for Modo. I get far less headaches that way.
Sure, it’d be great if I could do everything inside of SU, but honestly, I’m fairly confident that’ll never happen.
I want me some consultants, too.
-
-
my 2 cents
the question can be asked in reverse also:
why are all those big boy apps not as easy/fast to work as Sketchup?
why aren't they as easy to learn?I think all apps have their pros and cons, there is no perfect app.
Sketchup has always been a fast/easy modeler, even if that comes with the price of not being able to handle high poly models.For me, proxies, components and good modeling habbits help a lot to work around those limitations
-
Consider yourself consulted with, Stinkie. (I still can't get past your nom de sketchup--even after all this time. )
Anyway. . . consult. . .consult. . .consult.
I'll Bill ya.
-
@jonfar said:
For me, proxies, components and good modeling habits help a lot to work around those limitations
Agreed, but my point we shouldn't have to and it's been proven by other applications that it is possible.
Explain how importing a 35MB obj makes your SketchUp model grow by 200+MB? Where is all that bloat coming from? My question is more along of the lines of shouldn't SketchUp, based on all the technologies out there at this moment, be able to handle hi poly models? And I am not talking about it chokes after a dozen or so, one single hi poly bed or sofa kills a file.
To me, it just seems like SketchUp showing it's age. Old coding maybe?
-
@valerostudio said:
Explain how importing a 35MB obj makes your SketchUp model grow by 200+MB? Where is all that bloat coming from?
Sketchup is saving all the textures inside the SU file which makes it bigger.
About the slowness I've read here somewhere before that it is because SU is using the inference engine all the time.
I usually hide all edges on imported highres geometry which makes it a bit faster. -
@pixero said:
hide all edges on imported highres geometry
I do this with every big model -- soften edges: smooth normals, soften coplanar, and about 40 degree angle
Also disabling "Profiles" in Styles
Helps a lot. -
I would never use a straight 35mb sofa even in Max. Vrayproxies all the way.
-
Well, while we're at it: I had a model bloating up to 78MB or so and running badly. Purging didn't help, nor "fix problems" (does it ever?). I copied everything out of it. Purged the file except scenes and styles. Put everything back in paste in place -- under 5MB!
So beware.
-
Sadly, I have to agree with what's posted here. SketchUp keeps falling further and further behind as a real 3D modeler. It's now firmly in the "not going to change" mode with all the brain power for upgrades going into how to move it to the web as a hobbyist web 3D application.
I frankly believe the old data architecture model is flawed, and much like Adobe Muse (which Adobe recently end of lifed), there just isn't anywhere to go to fix it without a huge overhaul.
I've had several lengthy conversations with a couple of the top plugin developers here, offering $1000 to develop a community distributable filleting plugin which actually works correctly with non coplanar faces with the conclusion it's just not possible to do in SketchUp because of the internal architecture. ALL OTHER 3D POLY APPS HAVE SUCH A CAPABILITY-- just not SketchUp.
From where I sit, and as a very long time (also @Last) user, the core competencies of the SU dev team now pretty much reside in learning javascript and building web interfaces. I would even posit the real 3D dev talent for SU is in the plugin community, not with the dev team. Based on this, I doubt there will ever be any catching up done other than some slick web demos, like the recent one on how to create a screw using the web based version and no plugins.
-
I've been using it since the @last days and it's still King for conceptual stuff but I have recognized like others that I need to move to Revit soon. That's just what the Arch world uses.
The other thing holding SU back is the lack of tools to deploy it in larger companies. No corporate license key for multiple installs unless you are a network license holder, plugin deployment is a nightmare and setting a common default toolbar arrangement beyond 'getting started' just isn't possible. It also frustrates me that many of the popular free plugins that we all seem to rely on have still not made it into Pro. Trimble could easily pay for the IP or license it. Having to side-load 2 dozen plugins to make it usable for us is unsustainable.
-
@stinkie said:
why not love it for what it can do, rather than hate it for what it can’t? Since then I’ve basically used SU as an archviz modeling plugin for Modo. I get far less headaches that way.
Completely this! For me, it's a tool to create architectural models for use in Blender. All my entourage and rendering happen in Blender. Every now and then I'll render a quick exterior in VfS3, but that's it. SU is just one tool in a toolbox, and it's very good at what it does.
-
Yeah ... it is indeed. I understand the frustrations, believe me, but if there's anything I learned using Modo, it's that SketchUp, limited as it is, is really good at basic architectural modeling -and fast,too.
It can be an annoying bit of kit, at times, but it's not very likely I'll ever stop using it.
-
@chippwalters said:
I've had several lengthy conversations with a couple of the top plugin developers here, offering $1000 to develop a community distributable filleting plugin which actually works correctly with non coplanar faces with the conclusion it's just not possible to do in SketchUp because of the internal architecture. ALL OTHER 3D POLY APPS HAVE SUCH A CAPABILITY-- just not SketchUp.
I think it's a little bit unfair to blame Sketchup 'internal architecture'. There are limitations in the edge/face conceptual framework of Sketchup, essentially because this paradigm does not allow to have a global grasp to the topology of a shape at once, as opposed to a NURBS based approach.
Still, advanced filleting can be done: whether it is native or in a plugin does not matter, as the algorithm does not require special performance in reality. Just a matter of work, and therefore of priority. I'll probably release my version in a few months, just need to have enough free time to concentrate on resolving a few issues to run the last mile(s).
Fredo
Advertisement