What is really possible with T-Splines.....

[PhilProcarioJr]

So some coworkers keep talking about how limited T-Splines usefulness is. To be honest it's not limitations of T-Splines or Fusion it's the person using them and their knowledge of the tools. I have even seen post here on the boards downing on T-Splines. 

So anyways a friend challenged me to a bet, he claims I can not get the orc head model from Topogun into Fusion 360 as a CAD model fully editable. The head is made up of over a million polygons. Almost 2 million. Here is what the highres model looks like.

The challenge was two fold:

1st part is to model the head using only Fusion 360 and that's what I am going to post today. 

2nd part is I can use whatever tools I have as long as I end up with a fully editable CAD model.

So anyways here is what I came up with after 3 hours in Fusion Sculpt environment using just T-Splines.

Here is a video showing that the T-Spline mesh is fully editable:

I am not happy with this at all but I think it's a good start. More to come....

Cheers

[PhilProcarioJr]

@Anonymous @O.Tan @0502880

Thanks for the kind words.

@0502880

Honestly this falls back on what I have been saying for quite some time on these boards. I could have made this entirely in Fusion but it would have taken hours to get the perfect alignment. So instead I used an old Blender trick. In blender I created a mesh grid larger then what I needed and set up some simple selections. Then I applied the cast modifier to bend the grid to the correct shape then deleted the bottom points which were not needed and extruded the points from a central location. Then I converted the mesh to curves and in the bevel object selection box I hooked up a Bezier Circle that was made to the correct wire size. That gave me a low resolution control cage then saved to .obj format. I uploaded the .obj and converted it to T-Splines made a profile to match the cage and did a revolve operation. Done and it only took 15 min including popping off the renders.

This is why I keep saying that the dev team is wasting time working on mesh tools when they should be working on T-Spline tools instead. The amount of stuff you will be able to do with mesh tools VS. T-Splines is no comparison....but I feel this is falling on deaf ears...Anyways I'm off to get some more work done.

Cheers

[TrippyLighting]

And I've been scratching my head, how this was done in Fusion 360, with the tools provided in Fusion 360.

Doing this in Blender is, well not exactly easy for a beginner either , but quite doable because it has so many tools that allow you to do  such things.

I totally agree that focussing too much developement on mesh repair tools rather than extending the T-spline tools palette is the wrong direction. T-Splines are endlessly powerful, but that power only reveals itsel if you know what can be done with Subdivision surfaces. This object is a perfect example.

Based on the feedback on this froum from the Fusion 360 team it is easy to assume that there is more development going on in the area of mesh repair ( not creation) than there is on the T-Spline and surfacing front. I personally also find this rather odd as the combination and integration of solid modling, T-Spline and surfacing (another area that needs some TLC) is the one area where Fuiosn 360 is very unique and allows you to do things that are hard to impossible to achieve with other tools.

[PhilProcarioJr]

@TrippyLighting

"And I've been scratching my head, how this was done in Fusion 360, with the tools provided in Fusion 360.

Doing this in Blender is, well not exactly easy for a beginner either"

One thing I would like to make clear is I can replicate this entirely using Fusion with no external tools (you of all people around here should know this).  So the next question is why not just use Fusion 360. this is not the first wire mesh I have made the last was done entirely in Fusion...same complexity and it took about 3-4 hours. So 3-4 hours or 15 min? As to your second statement I only have 1 hour of experience in Blender..honestly...and one hour might even be pushing it a bit longer then reality. I have been telling people that can't afford apps like Maya to just use Blender....it dawned on me am I directing them to an app that I would use? So I jumped in...

The big point here is EVERYONE should have and use multiple tools in their toolbox....you can't do everything with a screwdriver so stop asking the screwdriver company to make one that does....

Hopefully by now the dev team knows I have great respect for them and the work they do and I know that development is being driven mostly by votes on ideastation. On one hand I like the ideastation and on the other I hate it. I could post my wants all day long but I doubt very many people would even understand the benefits of what I was asking for. People are so closed minded when it comes to a different CAD way, so most of my ideas wouldn't get votes. Every time I see the road map I don't see what I am looking for..sad really but oh well I am just one voice so I'll keep quiet about development and do what I do from here on out (get work done).

Cheers

[PhilProcarioJr]

@TrippyLighting

Oh another point I would like to make..you said

"Based on the feedback on this forum from the Fusion 360 team it is easy to assume that there is more development going on in the area of mesh repair ( not creation) than there is on the T-Spline and surfacing front."

I would like to point out that it was stated two weeks ago that there is no work being done to T-Spline tools in the near future. Quoted from Colin.

"As for T-Splines, we have nothing in the short term with regards to tools, however we will be working on improving the T-Spline performance."

So my hopes for anything more then I currently have in Fusion at least for the rest of this year are nonexistent.....

All the new mesh repair tools are going to do is double the amount of questions we answer here on the forums and cause 10x the confusion for new users. The ones that will be using the tools are makers...not engineers and only a hand full of designers, why because anyone wanting CAD accuracy will know that those tools are useless. They are only good for tinkerers with rapid prototyping machines. The sad thing is I own a bunch of RPM and I have maybe a 1% use for these tools coming out. Converting 3d scans..useless because you will not have CAD accuracy so they can't be used in product development. Again only makers that tinker will see a use for them...the rest of us working on product development are left waiting...and I will bet that during this time your going to see another CAD app hit the market running with tools that engineers and designers are really after...

Just my 2 cents

[jeff_strater]

Hi @PhilProcarioJr, @TrippyLighting,

You raise good points here.  I am not replying in any official capacity here, but can give a few glimpses from inside the machine...

One of the hardest things for us in the Fusion team to do is to figure out what our priorities are, what to do next, and how to balance resources.  Don't get me wrong, it's a great problem to have.  We have a lot of passionate users, we have a decent-sized team, and a huge list of things to do.  But, our resources are finite.  So, we have to make some tough trade-offs.

While you are correct that we are introducing a mesh repair workspace, and also are not planning to do much in TSplines in the near future, those are not necessarily the same teams, so it's not exactly correct to say we are trading off TSplines for mesh, but its still a valid question.  

I'm not a product manager, but the decision to introduce mesh was driven by a couple of factors, I think.  First, there are workflows where you want to start with data that is only available in mesh form (scanned data, Sketchup data, Thingaverse designs that are in mesh form), and we've seen quite a lot of requests for working with this kind of data.  Second, we have some existing mesh technology that was relatively easy to integrate into Fusion, so it was not a high-cost project.  Third, there are some forward-looking projects that need mesh support that we think are pretty exciting.

Regarding the lack of new TSpline features, to be honest, we have not seen a lot of requests for new features in that area.  Most users seem pretty happy with the current functionality.  If that changes, and more people share your desire for more TSpline capability, then absolutely we will put more effort into those enhancements.

We can't make everyone happy, unfortunately.  We'd like to, of course.  We do take most of our strategic direction from our customers.  But, you all don't always agree on workflows and priorities .  So, we do the best we can at making decisions based on what the majority of customers want.

I don't think that satisfies your desires, unfortunately.  

Please keep these kinds of comments coming, though.  This kind of feedback is critical to have.

Jeff

[PhilProcarioJr]

@jeff_strater

First off thank you Jeff for taking the time to respond here I for one really appreciate it.

Here is my thinking on whats happening. You don't have requests for more T-Spline tools because most people have no idea what can even be done with them or why they want them. So here we are in this thread...which is exactly why I created this thread to try to drive awareness of what can be done with T-Splines and where they are lacking...Personally I don't think there is anything I can't model CAD accurately using the current tools. Now enter the problem, just because there is nothing I can't model doesn't mean modeling those HARD parts is currently fun to do or fast for that matter.

On to stl and mesh tools...the reason your team has so many request for said tools is because these people have no idea how to create things in CAD software...this is new ground now that RPM are affordable to the average hobbyist/maker. So you guys are going to add the tools and these makers are going to come to a realization very quickly...this form of model something from someone else's work is a waste of time. sure you can modify it to your needs but the reality is you will end up modeling it all over again from scratch if you need accuracy. They are taking the long road because the hill is less steep. The problem with the long road is one side of it is a cliff they can't see. I have been modeling for over 28 years now and I have been exposed to just about every form of modeling imaginable and at one point or another someone wanted me to make it work. I have worked with some of the most talented people in these fields...everything from guys working on high budget films to engineers specializing in their fields to triple AAA game makers to machinist that could make anything. Today I use my experience from all these fields, all of these talented people to do what I do. Every time someone has said you can't do this....I have found a way to do it...and I will continue to. These people (NOT ALL) that make these requests have very little experience in these fields...You may have a team dedicated to the mesh tools and that's great but where my questioning comes into play is this....someone decided we need these tools, put together a team that is now working on said tools, yet they knew they needed tools and they needed a team to do it but they have to ask on the forums what people need in that area? These are questions I would of had a ton of answers to way before I would have ever decided to put resources and man power towards development of...IMHO this was a case of we own Meshmixer and the code this will be somewhat easy to implement so lets just do it. Let me ask the team this, what good are the Meshmixer tools if once the meshes are converted they are a PITA to do anything with because the T-Spline tools they are supposed to support are lacking? If once converted you really can't do anything with them...seems to me like a lot of wasted resources.

All of these requests from customers...what are they making? Do they post work that shows their knowledge of CAD of Design or are they people claiming they want to learn but only if there is a do everything for me button? Is their end goal just to copy others hard work or are they trying to make their own creations?

Then there is a group of us that have proven our work and we help people every day down in the trenches and I see our requests shot down all the time...I ask myself why...these are the people with experience, knowledge and we pass our experiences on when we run into problems with the tools and software only to fall on deaf ears.... Sorry but this is how it feels.

Anyways, I have a great respect for you Jeff and all the Fusion dev team I have had the privilege of talking to, I just feel like the direction Fusion is taking is misguided but that's just my opinion.

Cheers

[PhilProcarioJr]

@jeff_strater

"While you are correct that we are introducing a mesh repair workspace, and also are not planning to do much in TSplines in the near future, those are not necessarily the same teams, so it's not exactly correct to say we are trading off TSplines for mesh, but its still a valid question."

This comment got me thinking...

"those are not necessarily the same teams"....I really hope these are not different teams.....the mesh tools and the T-Spline tools need to work seamlessly...hand in hand....if they don't then there will be all kinds of bottlenecks....and problems galore.... Not only that but almost every mesh modeling tool I can think of should also be in T-Splines...unless of course you guys are making some new fangled tools that don't exist out there already...

[LMD001]

@PhilProcarioJr

Hello Phil,

Stumbled on this very interesting thread.

Really amazing stuff you do with T-Splines, this shows very clearly that Sculpt is most certainly not a "gimmick". One only needs to know how to use these powerful tools.

This is high quality work!

Best regards,

Ludo

[PhilProcarioJr]

@LMD001

Thanks Ludo,

When I first started using T-Splines my initial thoughts were that these tools are very limited....but I was facing a road block with no solution. I needed to get my models from other tools into Fusion as a CAD model and T-Splines fit that bill. The more I worked with them the more I found out what they can do and it occurred to me. I don't think people really understand what can be done with them and since development revolves around customer requests I am worried about T-Splines future. So I started this thread. What is seen in this thread isn't even my best work with T-Splines, but since a lot of my work is under NDA I can't share it. Currently there are rumors floating around that a few other CAD apps are developing a full suite of Sub-D tools in their app that will convert and work seamlessly with NURBS and surface tools. If that happens T-Splines will become a rusty old tool that gets lost in the shuffle. The community really needs to become aware of the possibilities of T-Splines and support the future development of these tools. At this point I feel like AD buying T-Splines was a bad move for the development of T-Splines because they just added it to their apps and now that development seems stagnant. I hope I'm wrong but I doubt it....been in the business for 30 years and I have seen this all to many times.

"Really amazing stuff you do with T-Splines, this shows very clearly that Sculpt is most certainly not a "gimmick". One only needs to know how to use these powerful tools."

It's just my opinion but I too definitely don't think T-Splines are a gimmick, do they need work and new tools and better training and awareness? Yes definitely.....

I will resume work on the skeleton as soon as my current crop of client work is done so stay tuned.

Cheers

[PhilProcarioJr]

Back to work...

[bensbenz]

Hi Phil, I cant even begin to imagine how you have made some of this stuff, amazing work. I can agree with you when you say, many of the people are asking for feature because they don't understand how to use what they currently have. I am definelty fit the category of not knowing how to fully use the power of fusion 360. I am more in to the machining side of things, but if you ever did a YouTube video explaining how your using t-splines to create some of this I would watch very intently.

Can I ask what specs are the computer/laptop you are using fusion360 on? I need a to build a faster rig and I am curious.

Thanks!

[PhilProcarioJr]

@bensbenz

Thanks for the kind words!

I am working on training now but its kind of on hold until I figure out what the Fusion team is doing with mesh space environment and until the bugs and kinks are ironed out from the last major update...

Computer specs Intel i7 950 @ 3.07GHz CPU, 12GB of RAM, Nvidia Geforce GTX 770 with 4GB of memory and windows 10 on a solid state drive.

[TrippyLighting]

Actually at the moment its more of a mess workspace 😉

[PhilProcarioJr]

@TrippyLighting

Don't forget I warned everyone including the Dev team it was going to end up being like this but no one wanted to listen.... 

Everyone was like ohhhh wow Meshmixer in Fusion.....

[bensbenz]

Thanks for those specs. I was hopping it was something crazy which would explain why my system seems slow, but I have almost the same system, a few more GB RAM and a GTX 750. Guess I will build a 24-32 core xeon workstation and see how it goes.

[PhilProcarioJr]

@bensbenz

Your videocard has 2GB of ram? If so that's your slowdown. I have done extensive testing on both 2GB and 4GB cards and in my tests the performance increase was huge with the 4GB cards and ever larger with an 8GB card......for what it's worth I don't even use 25% of my CPU so your wasting money on Xeon processors...This is NOT a workstation app and runs like crap on workstations.

[bensbenz]

Oh very interesting. Well I was getting the xeons for faster rendering without waiting on the cloud queue. When I render especially turntables I can peg my CPU at 99%. I will get a video card with more ram and try that first, thanks for the information. 🙂

[TrippyLighting]

Get a "normal" Gaming card, not one of the Nvidia Quattro or AMD Fire GL workstation graphics cards.

Those are the reason that Fusion is not running well on "Workstations".

These Pro Graphics cards are hardware wise identical to gaming cards, but usually are a few generations behind. The difference is in the software drivers. 

PRO cards are optimized for Open GL performance, because most CAD and a lot of other DCC software uses the open GL API.

Fusion 360, however uses the Direct X API and that is something that these PRO cards are NOT optimized for.

On the mac platform Fusion 360 uses Open GL.

[PhilProcarioJr]

Ran into more issues with the surfacing tools and T-Splines saved the day....

[Anonymous]

---

@TrippyLighting wrote:

Get a "normal" Gaming card, not one of the Nvidia Quattro or AMD Fire GL workstation graphics cards.

Those are the reason that Fusion is not running well on "Workstations".

These Pro Graphics cards are hardware wise identical to gaming cards, but usually are a few generations behind. The difference is in the software drivers. 

PRO cards are optimized for Open GL performance, because most CAD and a lot of other DCC software uses the open GL API.

Fusion 360, however uses the Direct X API and that is something that these PRO cards are NOT optimized for.

On the mac platform Fusion 360 uses Open GL.

---

Good piont that! 

I need to get my cad station back from the work shop where it's being used as a machine control. It's sad! 

It's specked for SW 2015. But I have a gaming card ready to put in there as the SW machine needed less. So a 8gig card should do ok! 

At least it will stop Phil complaining about me using a Mac! Still got a long way to go until I can get nere Phils set up though! 

BTW Phil. 

Undo in topogun is cmd+z! 

Talk on Sunday evening same time! 

What aspects of that part are T Splines?