Polygon cages for smoothing

[SJD]

Hi

I'm trying to learn and get a better understanding of more advanced 3D modelling techniques, not just how to do them in FormZ but the overall concepts involved. I'm specifically interested in hard surface product visualisation.

I can see FZ has a vast array of tools and have successfully worked through many of the tutorials and I have found the videos incredibly useful. However, it's a real pity there aren't any that take a real world commercial project and show how the tools could accurately be applied to it like there are with many other 3D programs.

One approach I see a lot in other tutorials (3DS, Blender etc.) is either a controlled or free form construction of polygon cages or meshes which are then refined and smoothed to create sometimes very complex forms or subtle detailing accurately. Is this an approach you can do, or should do in FormZ?

thanks

steve

 

Edited June 3, 2020 by SJD

[SJD]

just by way of example, here's a classic Coke bottle.

The basic shape would be easy enough but it's the subtle vertical indents that are the problem.

The approach I have seen starts with a basic mesh bottle and then the relevant polygons are selected and then bevelled - the smoothed shape gives this effect. There's obviously quite a bit of work and refining involved but that seems to be the approach. I'm not sure how I could do this accurately in FZ

 

Edited June 3, 2020 by SJD

[Andrew West]

If it were me making that bottle I would start with either the Revolve tool or one of the Loft tools and then refine.  You could also start with a basic shape and then Boolean difference out the recessed areas.   With so many different tools to choose from everyone will have a different approach.  That is the beauty of Form.z and solid modeling.  I get in the habit of always going to my favorite tools and once in a while I have to remember to look through some of the others when I get stuck on something.  

Edited June 3, 2020 by Andrew West
clarify

[Andrew West]

One tool that I really wish FMZ had is what 3DS Max calls TurboSmooth.  It does a great job of taking a basic rough object and smoothing out all the edges.  All you ;have to do is make a rough approximation of anything you want which saves a lot of time.

[SJD]

cheers Andrew

using the boolean tools is typically the way I would have approached it in the past but the end results have always been fairly unsatisfactory when seen up close and also a bit hit and miss. It's the refinements that make all the difference between something amateurish and professional. I can see how amazing the tools are from the tutorials but when it comes to apply them in practical terms I find it difficult to get my head around what approach I should use. That's why I wish some of the tutorials showed them being applied to real word projects instead of random and conceptual stuff.

[SJD]

here's another wireframe of the coke bottle, the topology looks very neat and tidy

is it possible to create this kind of model in FormZ?

 

[¢hris £und]

Looks like a Sub-D model of the type Modo would produce.  But I see no reason why fZ cannot do equal or superior quality.  I might be able to give a crack at it later tonight.

 

¢£

[SJD]

hey Chris

don't make it a problem 😎 - this isn't for a project it's just me setting myself some tasks to try and get a better understanding of how and when to use the tools

tbh (and no disrespect to ADS) I'm surprised that there are no tutorials on the website that use FZ to demonstrate real world applications. I'm sure this would be a massive selling tool for them.

It's all very well showing off the tools building random fanciful structures but it's very different when you are confronted with actual objects that you have to visualise. I would have thought that this is the task that most of us using FormZ face. I have used FormZ for over 20 years and never needed to go much beyond the basics. However, times are changing and I need to broaden my skill set but I find myself looking at YT videos of 3DS, Rhino and Blender in order to get a better understanding of how to use tools in the real world.

[¢hris £und]

Steve,

No worries.  I am not familiar with 3DS, Rhino, or Blender.  Though I have to say, that I would approach the problem differently in fZ than I would in Modo.

In Modo, I would build the bottle monolithically, in fZ, I would do 1/16th of it mirror that (or 1/8th), and copy-rotate the sections stitch then apply SDS.

 

¢£

[SJD]

Hi Chris

I'm not familiar with those programs either (I've only ever used FormZ) but what I mean is I can't find any other practical tutorials for FormZ so I try and apply the modelling principles from other programs where I can. I'm not sure if it's just me that has trouble or just generally the FZ user base - there are hardly any examples in the gallery that go much beyond basic modelling techniques

I understand when you say model in a section of it and repeat and subdivide - can I ask how you might create the shape to subdivide in the first instance? I've found SubD surfaces very versatile but also incredibly unwieldy and involving a lot of trial end error. It's the generation and manipulation of the mesh cages that seems to be easier in other 3D programs.

thanks

steve

 

[SJD]

sorry, you say model in Modo - do you think Modo is better suited to this type of modelling?

Edited June 4, 2020 by SJD

[Des]

I think this is a good discussion and I'm interested how other people would approach this. I remember doing a large Fanta bottle many years ago but it was less detailed than the coke bottle. It was created using similar process described above (revolve, booleans, rounding etc.). One thing I am sure of is that any splines to be used initially must be good quality (not imported) and good information of the object (decent photos). When taking photos of an object for reference and tracing I usually use a zoom lens and take from really far back and zoomed in to reduce the perspective (narrow cone of vision). Otherwise I think FormZ could produce a beautiful model using the tools available.

Clean modelling is what interests me without breaking the bank regarding poly count, preferably smooth geometry for adjustment.

Watching with interest and if I find time I'll jump in as well..

 

Des

[SJD]

thanks Des

at least my original post wasn't too dumb then 😁

I've always thought and been led to believe that FZ is one of the best and most robust modellers out there so it's a bit frustrating that there are no practical examples of it other than either rather basic models or wild conceptual fantasy stuff. I imagine that the bottle example is bread and butter work to a 3D professional so it would be nice to see how FZ could be used to tackle some seemingly everyday tasks that should be expected of a 3D artist using a professional 3D package.

 

Edited June 4, 2020 by SJD

[SJD]

I've been looking at a few more videos on subdivision concepts and it seems a very popular approach is to create quad meshes from simple primitives or planes by extruding edges, cutting/dividing edges, adding bevels etc. It seems a logical way to approach modelling and an easier way to create complex shapes and surface detail (obviously with patience and practice). I've tried this approach with FZ but it does seems much more difficult and less reliable - am I missing something? Is it possible to apply the fundamentals of Sub-D as with other programs?

Edited June 4, 2020 by SJD

[Des]

There are some here who are very good with sub-Ds (you know who you are!) and I'm sure they could come up with a workflow for those.

However, I made my first attempt using splines and nurbs, here is the result so far. It needs more playing around with the splines (see the indents) but it shows that this approach works fine and produces a nice smooth surface. I made two splines, one for the indent, duplicated it and edited it for the bumps. Multi rotated them within 45º (using 45 as I'm thinking 8 sectors in a coke bottle?). Nurbs lofted them (selection order is important), rotate duplicated 7 times and stitched them together.

CokeBottleAttempt01.zip

Edited June 4, 2020 by Des

[SJD]

thanks Des

this looks really helpful, would you mind back saving the file to 8.6 please so I can have a look? 😎

 

[Des]

CokeBottleAttempt01v8.zip

[SJD]

cheers Des

[¢hris £und]

Steve,

Regarding Modo, Yes, and No.   Modo is pretty much a dedicated SDS modeler.  It has added other features over the years, but they seem tacked on.

Main benefit to formZ's sub-D's is they can be converted to NURBS.  You can always pull the control cage from Modo and use it in formZ if needed.    But Modo is a Polygon modeler that just hides the polygons beyond the smoothed cage.

There are some things I think that make Modo easier for this kind of work.

1. The ability to flip back and forth between Sub-D and Poly by a keystroke is great.

2. The ability to use all of its tools in both Sub-D and Poly.  (All fZ would need to do to match this is to allow it's sub-D tools work with poly models.)   I find it often easier to work out details in poly mode than in sub-D.    It is easy to get your mesh overly complicated when you can't see the controls as easy.

3. Heuristic selection.  On any specific object, if you select a segment, skip a segment select another, it can infer from this that you want to keep this selection pattern.  It isn't perfect, but I love this ability.   This is the main feature that I think would allow me to change my approach for the coke bottle.

4. Insertion of segments manually.   Something formZ has lost or changed in a way I no longer understand.  (I miss it very much!)

5. Other insertion/division tools.  In Modo, these are very fluid.

Where Modo really excels, in my opinion, is purely Organic modeling.  Currently, formZ can't match Modo here.  On the other hand,  Modo can't touch formZ anywhere else regarding Geometry.  (Its surface mapping and rendering are phenomenal.)

There are a lot of people who do Hard Shape object modeling in Modo and other similar SDS applications.  But in My opinion, it is a lot of unnecessary effort to do so.  formZ is vastly superior here.

The coke bottle is kind of an in-between object.  I don't think Modo would add enough advantage to the modeling of it to warrant a purchase.

----

In fZ,  I would create a rough bottle using the revolve tool that can be divided into 8 or 16ths.    Say with 32 or 64 sections.   Split it immediately, then work on the divots. at the edges of the section.  Once the shape is perfected, copy-rotate and stitch back together.

 

I 200% agree with Des regarding taking photos of your objects.  Get far away, and zoom in.   the more extreme on both, the "flatter" you can make the object for accurate tracing of its profile.

[SJD]

Hi Chris

a very informative response, thank you

I think it will take a little time for me to start to get grips with more advanced modelling techniques but it's something I'm determined to do whether it be with FZ or another program. Personally, I think FormZ's biggest short coming has always been the lack of decent practical tutorials to showcase its modelling power. I can see the tools in action but I find it difficult to choose an approach and then apply them efficiently to an actual project. I just don't understand the fact that ADS don't have a single video on them taking a real world complex object and showing it being reproduced as it would be in a studio - using underlays, correct sizing, ideas of how to approach certain aspects of the model etc. so they can showcase FZ in action. Contrast that to the wealth of C4D, 3DS, Blender etc. tutorials and it is little wonder FZ has been on the sidelines for all these years. I may be wrong but there also seems a bit more consistency between a lot of the other 3D packages so it's easier to relate what's being done in one program to another, FZ just seems to do things a little different.

[SJD]

Hi Chris

your items 1 - 5 look to be fairly standard across many of the main 3D programs I've been looking at, it seems that it's almost essential to work in poly mode to ensure smooth quad meshes, like-wise the division, selection and insertion tools. Is it not possible to do any of this in FormZ?

 

[johnalexander1571]

Hi SJD,

Chris is giving you good guidance. I know this doesn't help, but I did do a job where we had to model existing bottles for clients. We started with a 3d scan that was very lame and needed modeled over totally, but there was no "sizing" possible with any program that could get to that level of accuracy. For the record, we used Modo for that. form.Z could do just as well.

I tend to use the sub-tools to grow my models a row of polys at a time. I switch back to polys only when I run into a little detail area. I once had a key shortcut on the switch back and forth and it was the same as any other programs switch. With some malice of forethought, the sub creation tools are just fine in Z. Needs an add too selection in rows type of feature.

[SJD]

Hi John

yes he is and I appreciate that. I also appreciate that FormZ has a small but strong and loyal following and I don't wish to offend anyone with what is probably my ignorance. I too have used it exclusively for over 20 years and have made a fairly successful living out of it. It's just a pity that it's strength as a 3D modeller could not be made a little clearer through a few tutorials that show different approaches to creating more complex actual real world objects, large or small, not just fanciful or random models - I'm sure these would be more useful to professionals considering which 3D software they should buy. Most of the videos are over 7 years old and I'd have thought it would have been useful to do this due to FZ's small user base and lack of other online tutorials.

I too have managed to create some wonderful sculptural shapes using NURBS and Sub - D but it's a different story when you have to create something that has to conform to strict dimensions and appearance. I am trying to school myself in the basics of more complex 3D modelling irrespective of software, but find it a bit tricky when FZ does some things differently to what seem to standard industry methods. 

[¢hris £und]

Steve,

Agreed, there could be more in the way of tutorials for the contemporary version.  I think there are reasons this is the case.  fZ pro is not geared to the advanced hobbyist.  The light versions have not had enough of a marketing push.   Professionals who use fZ are just busy.

These days, I am really just a hobbyist.  Maybe, an advanced one.  I no longer use fZ professionally, at least not on a daily basis.  Funny thing is, this actually gives me more time to play with it and discover nuances I never explored back in my "pro" days.    I keep my license simply because I enjoy using it, and don't want a light version because I would be annoyed by lacking tools I am used to.

Back to SDS - no, fZ is fully capable, just not the same way as it used to. For example, I used to be able to insert a segment anywhere I wanted to on a mesh manually.  Now there is the "offset segment" and imprint tools that kind does this. In my mind, these are tools for a specific use.  I want the more generic functionality back.   Also, fZ used to have a C-mesh tool, also depreciated in favor of NURBS and SDS.  I used to use that tool a whole lot and the primary reason I got Modo was fZ dropping it. Funny thing It would make a great addition to SDS.  Roughout a cage with the C-Mesh and then SDS from there.    All of this can be worked around, it just takes a different mindset.

Additionally, Modo, etc has some specific types for SDS modeling that do improve the flow of that style of modeling methods.  As the Python SDK advances, I hope to add some of these other tools to fZ's set.

If you are looking to match specifics.  fZ is a better tool than pretty much any of the others including Modo.  3DSMax: encourages sloppy modeling behavior.  If you have ever imported other people's work, even professionals, 95% of the time they are CRAP!   It works, but from what I can tell, a person has to work overtime to make a good clean model.   I love Modo, but more for its fluid modeling style that really lends to organic modeling.    If you want to go beyond fZ in this regard, I wouldn't look back to Max or even Modo, I would look to the Alias Suits or Dassault.    While I do think fZ has the potential to match these, I don't know that it ever will.  Their primary market probably isn't pushing them in this direction.  Generally, Architects aren't looking for Class-A surfacing technologies.   Here too, I intend to add to fZ's capabilities with Python, should it allow.

Unless you have the ability to dump loads of money into one of these systems, fZ is your best bet for accuracy and strict appearances.  Yes, learning will be needed.  Same too, if Alias or Dassault is chosen.

Just as an example, the following was made 100% in fZ. (well, with some photoshop enhancements)   Granted 5/6.   Were I to do this again, I would use Modo.   On the other hand, Modo is hard to make specifics (not impossible.)

Edited June 4, 2020 by ¢hris £und

[SJD]

cheers Chris

Quote

Agreed, there could be more in the way of tutorials for the contemporary version.  I think there are reasons this is the case.  fZ pro is not geared to the advanced hobbyist.  The light versions have not had enough of a marketing push.   Professionals who use fZ are just busy.

I guess that's part of the problem, who is FormZ aimed at? I always thought it was the professional in the early days. I am/was a Mac user mid/late 90's and my employer gave me the choice of C4D, Lightwave or FZ. I plumped for FZ on account of a very impressive illustration that was on the cover of that months MacUser!

I think FZ has suffered greatly from its image over the years, it's always looked a bit 'hobbyist' to me especially as time has gone on which is a great shame because it's clearly not! The website gallery has never done it justice and that's the first thing I look at when I'm considering software. I'd have thought FZ's price tag would put the hobbyist off especially when they have Blender. As you say, the serious professional alternatives are pricey (maybe C4D is worth considering with it's monthly plan) which is why I'd have thought it was worth ADS putting in some effort to show serious users (and Luddites like me) what it's really capable of - surely a few well done serious tutorials aimed at the professional would be worth it rather than the hodgepodge of old tutorials and webinars we have at the moment?

 

Edited June 4, 2020 by SJD