Blender for the Over 60s - a Light-hearted Look at my Personal Learning Experience

Disclaimer

By ‘Learning Blender’ I don’t mean every aspect. That would be impossible. Blender is enormous and can do many things for many different creative disciplines so I haven’t become a ‘super-user’ or candidate for a job at Industrial Light and Magic. Over the last 12 months I have simply learned to take original 2D automotive package drawings and sketches and turn these into 3D surface models shaded and rendered in very simple HDRI environments. To be honest that’s enough for most automotive creative exterior and interior designers. There is joy to be had from seeing your original design in the round in a nice sunny environment. I can do this easily in Alias and VRED, but Blender is a completely different animal. I’ll also apologise in advance for my attempt to describe ‘Geometry Nodes’ and ‘Empties’ in my 'Final Observations' section. I could quote the user manual verbatim but preferred using my own words to express my understanding of these features. Modelling and rendering 3D cars in Blender is easy for me (now). The hard bit was finding out how to do it as someone not under 30 years old. This is my experience…

Introduction.

Back in December 2020 I decided to have a go at learning Blender for automotive design (Fig.1). Soon after I retired from academia I lost access to Alias and VRED and couldn’t justify buying an expensive licence for ‘hobby’ use (this was before the Learning Editions shipped). My job used to be my hobby and I still love modelling original car and vehicle designs in 3D for fun so I needed an alternative 3D ‘fix’. Blender was the natural choice as could do what I wanted, but it was also free. I knew Blender had a steep learning curve but for me it was a vertical brick wall that I kept banging my head against!

The Wall

My biggest problem was finding Blender to be quite unlike any of the 3D software I had used previously in every aspect of its conception, development, interface, and workflow. Whereas expensive commercial software has a specific industrial purpose and provides ‘accredited training programmes’ using workflow recommended by the developers, Blender is ‘open source’ and free, so the knowledge is dispersed across a vast community of disparate software developers and users (from 2D and 3D animators, to 3D asset modellers, architectural interior designers, games developers, 3D visualisation artists, digital sculptors and many more 3D creatives). Most users acquire knowledge using a shared trial and error approach (not unlike online video gaming where techniques are shared to overcome obstacles communally).

'Free' Blender learning relies upon a tacit ‘Wisdom of Crowds’ that I haven’t experienced before. Advanced Blender users of all disciplines freely upload techniques to YouTube but you have to search for specific content you might require and filter rigorously (automotive design techniques in my case) . Some online tutors are better ‘teachers’ than others. I saw some ‘YouTubers’ uploading techniques which reveal themselves as 'personal workarounds’ when I have later stumbled on more advanced users from other 3D disciplines using them properly as the developers intended. But in truth this is not unlike how people learned ALIAS in the early days.

Note! For more advanced users who really need to use Blender professionally following tried and tested workflows, ‘Blender Studio’ does offer properly structured training courses for a reasonable monthly fee delivered by a collective of existing experts in the Blender community .

The Buffet Restaurant Analogy

There is so much expert knowledge spread far and wide on the web that some entrepreneurial individuals have been able to curate the best tools and techniques for a specialist YouTube audience (like automotive design). I would compare this to a holiday hotel buffet. You can go around piling your plate with everything including stuff you don't really need and you just end up with a big pile of totally unrelated food. Or you can make a ‘proper’ meal by being selective in what you gather…and it helps if you have a chef to guide you round the restaurant (Fig.2). However, it is still important to pick at other platters and try less familiar items if you want to expand your palate. I have discerned a lot from a couple of Automotive Design specialist YouTubers, but at the same time I have learned additional interesting Blender stuff (and sometimes better techniques) from users in other disciplines.

Blender in Automotive Design

In terms of using Blender for ‘original’ car design (as opposed to just modelling existing products from blueprints) there was nothing on the web when I started in late 2020. Then in June 2021 a young designer called Berk Kaplan (veteran YouTuber demonstrating 2D automotive design techniques) created his first video modelling one of his sketches in 3D using Blender. His enthusiasm was infectious, and Berk has since become an advocate for the use of Blender for automotive design concept modelling. He has also developed his own (paid) training course which a couple of OEM studios have picked up on. Berk is extremely articulate, his content is very well-produced and I can recommend it highly. But, for me, his Blender tutorials came online a few months too late. In March 2021, a contract to teach Alias at my old University gave me Alias and VRED for another year so with no incentive to learn Blender I put it aside…

…until January of last year (2023) when I resolved to leave Autodesk software and learn Blender properly. This time I approached it with totally fresh eyes, a lot more patience, and months of absence from Alias and VRED.

Escaping the Legacy

15 Years of ALIAS muscle memory is hard to unpack so I originally tried to use NURBS technique in Blender. As a ‘mature’ designer I started my career when 2D sketching and marker rendering was progressed to 3D clay models via an accurate orthographic ‘Tape Drawing’, using black line tapes on gridded mylar film with a ‘10-line’ grid. Key lines and curves could be transferred accurately from the 2D views onto a clay model in 3D space using a gridded base and section templates (Fig.3) or a digital pointing machine like a Stiefelmayer on a steel surface plate.

When I began to use Alias for concept modelling in the early 2000s, it was easy for me to take key lines and curves from a side-view tape drawing or sketch and ‘pull’ these laterally to create a 3D wireframe network of NURBS boundaries into which surfaces could be introduced. This was how I had always worked in ALIAS and was how I expected to work in Blender. The icon-based graphical user interface (GUI) in ALIAS was also very clear and ‘relatively’ easy to learn. Also was much easier to teach the basics to large groups using the Alias GUI until advanced learners could move on to the quicker marking menus and mouse/keyboard shortcuts .

The Blender interface is very different. Blender does have a GUI (Fig.4) but relies far more on keyboard shortcuts which you really do have to memorise. Adding to the complexity, many shortcuts and menus change depending on the mode you are in or the object(s) you have selected on screen! Learning Blender is like a London cabbie taking ‘the knowledge’…I’m not joking…you have to walk lots of back streets to truly discover the city. In a year, I just got to walk around the block!

Another Blender quirk is the use of non-destructive ‘modifiers’ which create various 'visual' effects on a model until being physically ‘applied’. By comparison ALIAS modelling is ‘destructive’ and you are manipulating the real surfaces with each operation so it’s important to strategically build ‘history’ (requiring considerable expertise and years of practice). That way operations can be reversed, and a model can be changed as a design evolves. The non-destructive modifiers in Blender leave the model in a ‘flexible’ state until the designer is happy with the final result, which is fantastic for concept modelling. When the modifiers are ‘applied’ so the visual effect becomes physical. For example: using the Mirror modifier to work on one half a symmetrical model with a mirrored instance, after ‘applying’ the modifier the mirrored side becomes a physical entity. (similar to solidifying a symmetrical layer in ALIAS). The use of modifiers sounds really confusing until you understand them, but they are fundamental to working in Blender.

My Breakthrough!

?At first I had been trying to find a way to model in Blender that was similar to NURBS modelling in ALIAS. I even spent $25 on a ‘Curves to Mesh’ Add-on that I thought would allow this.

WRONG!!!

Blender hard-modelling is closer to poly modelling in MAYA (which I had never used but was aware of the techniques). Serendipitously, in summer 2017 I was introduced to Autodesk Speedform (later the basis of the Sub-D tools in ALIAS) during a 'skills refresher' at JLR under the (very patient) wing of a vastly experienced Specialist 3D Artist, Mark Reaney. In January 2023 I happened upon some old hand-written notes from the time and had a 'Eureka' moment. The familiar Speedform workflow could be emulated in Blender. I then discovered a (hidden) trick to modelling similar to placing Alias CVs, which suited my way of working. That was the turning point.

In 2020 I had no point of reference but by 2023 I had several guidance sources including current Alias Sub-D tutorials from Autodesk, as well as Blender YouTube content by Berk Kaplan and other ‘hard modelling and visualisation specialists’. Significantly though, the most useful became the Speedform training I had received at Jaguar design studios six years earlier. Who would have thought?

The Main Frustrations

As a fairly competent veteran of Alias and VRED, what put me (and probably most new users) off Blender was knowing what I wanted to achieve but then finding a sudden hole in my knowledge preventing further progress. The complexity of modifiers and ‘smoke and mirrors’ shortcuts were largely to blame. I would have to trawl through loads of YouTube videos to find the ‘next step’. I’m notoriously impatient but was forced to accept that Learning Blender would not be quick, despite what the ‘Learn Blender in Half an Hour’ videos might say.

I soon realised most ‘Automotive Design’ YouTubers just modelled existing cars from very accurate blueprints and reference photographs so they knew exactly what they were making. This is not the same as concept modelling from an original loose sketch where key lines and surfaces are being generated on the hoof from a designers imagination. It was also clear, from the terminology used, that many ‘automotive’ YouTubers were not trained car designers but competent 3D digital modellers making cars for fun or to sell as game assets. This is why Berk Kaplan’s channel is so good. Berk is a car designer teaching concept modelling and visualisation techniques as part of an ‘original design’ workflow. He understands what car designers need from Blender and has worked hard to establish the best tools and methodology, and he's now sharing knowledge with others (some of it at a price through his online course). Metaphorically he is the chef in the buffet restaurant mentioned earlier.

How to Beat Blender: You Must Advance on Many Fronts

Okay, I’m going to get a bit academic for a couple of paragraphs.

About 20 years ago two University researchers, Jan Meyer (Durham) and Ray Land (Coventry), coined the phrase ‘Threshold Concepts’ in an academic paper. (Meyer, J. H. F., & Land, R. (2003). Threshold concepts and troublesome knowledge: Linkages to ways of thinking and practising within the disciplines. In ISL10 Improving Student Learning: Theory and Practice Ten Years On (pp. 412-424). Oxford Brookes University.)

Meyer and Land defined threshold concepts as:

?"core concepts in a subject where understanding these concepts is key to transforming the way students understand a whole subject, allowing them to move on in their learning’.

End of Lecture!

In simple terms they meant there is a certain amount of knowledge that you need to have in place before you can really start to learn a particularly complex subject. This is what had been limiting my progress. I just didn’t know enough about the philosophy behind the Blender interface and the information was spread too far and thin to gather easily. Even simple things like moving pivot points and using symmetry (common in car design) seemed illogical and complex compared to Alias. The main reason is clearly that Blender was originally created for free sculpting and character animation rather than for hard modelling and creative design of manufactured products (like cars), but its potential for the latter is huge and more people are now beginning to see this. Also, Blender is open-source software that has been developed and is used by a vast community of young ‘digital natives’ from many disciplines. It seems attuned to the ‘PlayStation Generation’ who grew up immersed in technology and who can thumb-type 60 WPM on a tiny mobile phone screen unsighted. Contrast that with myself who can barely hit 40 WPM while needing to look down at a full-size keyboard!

Which explains the reason for my article’s light-hearted title. Blender is definitely not designed for ‘my generation’, nor should it be! Young designers and 3D modellers should not be held back by legacy interfaces just to make things easier for old guys like me. Software should take advantage of the generational attributes of the under 30s. Incidentally I recently watched a YouTube film about the annual Blender Congress and what struck me was the youth of the audience. Apart from Blender’s founder, Ton Roosendaal, and a few other seniors the majority of the huge audience was probably under 30. A lot of Blender YouTubers were also in attendance proving it to be a real and not just digital community.

Why was the Process of learning Blender so Tough (for an Older Guy)?

I would suggest that my generation lacks some of the foundational skills (like fast typing and video-game-derived hand-eye coordination) that make Blender so much easier for people under 30 to learn. Several advantages of this age group include:

• Being digital Natives who have grown up with technology, video gaming and using keyboards.
• Learning Keyboard Shortcuts and building muscle memory again come naturally to younger people. Graphical user interfaces are easier to learn, teach and use but are slower in operation than keyboard shortcuts if you can type very quickly.
• Collaborative online ‘communities of practice’ drive Blender learning. Young people use social and professional media to collaborate online in many ways including the sharing of knowledge and working methods, so they are totally at home with peer and self-learning.
• YouTube and online forums have opened up a world of knowledge transfer that has been quite remarkable. Blender learning benefits massively from an online community of users and developers that genuinely epitomises the phrase 'Wisdom of Crowds'.
• Young people are not afraid of trial and error. There is no right or wrong way to use Blender. Someone will discover a better technique and then share this with the online community. I find this analogous to video gaming which often includes communal sharing of methods and strategies online, and this process has automatically been adopted by the Blender Community. Try…fail…try again…fail…try again…succeed…SHARE.?
• Blender users and developers volunteer time contributing to social networks and forums (and an annual Blender Congress) to benefit the whole Blender community. As a teacher myself I can genuinely appreciate the satisfaction they derive from passing their knowledge to others.
• Younger people are entrepreneurial. In a world where ‘jobs for life’ no longer exist many expect to change careers during their lives and have an entrepreneurial approach. Some use Blender as a pastime while others have identified opportunities to make a proper living as freelance digital modellers or animators. Others just make additional income through Blender YouTube demonstrations, paid online training courses or even selling artefacts like 3D game assets or architectural models.

Final Observations

I grew to appreciate that Blender is an incredible piece of software that seamlessly provides tools for many digital creatives and permits everything from simple 2D CAD drawing through 3D soft character and hard product modelling to advanced visualisation and cinema quality animation. It even has provision for ‘hand drawn’ animation in 2D and 3D. So here are a few of the key observations I have gleaned so far on my personal learning journey to date.

• Blender is dimensionally accurate allowing 3D artifacts to be placed and scaled very precisely. It can also be used for simple engineering drawings, one reason it has found application in architecture and interior design. Quick floorplans can be turned into complex 3D building models and photo-realistic visualisations.?
• Orientation of X, Y and Z in Blender’s default set-up is a bit strange for car designers. it defaults with the Y axis as length, X axis as width and Z axis as height. I instantly rotated my Blender set-up to X = length, Y = width and Z = height to match the automotive ten-line convention used in Automotive Design and Body Engineering (Fig.5)

• It’s incredibly important to understand how the ‘Cursor’ and individual component ‘Origins’ interact. As an Alias user this was the hardest thing to grasp. The single ‘Pivot Point’ in Alias is far easier to use than the separate ‘Cursors’ and ‘Origins’ in Blender.
• You need to learn a lot of keyboard and mouse shortcuts from the outset and to develop the muscle memory to work quickly. I found this hard because I had to 'unlearn' Alias keystrokes and I am an appallingly poor typist !
• You definitely need a few specialist ‘Add-Ons’ to enhance and simplify workflow. Some are free, others are paid-for (but inexpensive and they are perpetual licences). Most add-ons can be obtained from Blender Store. One good example is ‘Machin3Tools’ (sic) which generates something akin to Marking Menus (familiar to Alias Users). Machin3Tools improved the usability of Blender massively for me. There are lots of YouTube videos describing all the most common add-ons if you decide to learn Blender.
• Modifiers, as mentioned previously, are tools that allow non-destructive changes to be made to models. In effect they impart visual changes to the model without affecting the real geometry. These include mirror/array/subdivision/booleans/bevels and many more. Modifiers are key to keeping the model ‘flexible’. Parameters can be changed at any time and their effects can be reversed, toggled-invisible or removed entirely. When a model is finished the modifiers can be permanently applied which does bake the physical geometry.
• ‘Empties’ are a bit of a curiosity at first. I’ll try to describe my understanding in the simplest terms. An empty is an object that doesn’t contain any geometry but has an 'origin' that can be translated, scaled or rotated about. However, it is possible to attach the empty to a collection of other objects.? Let’s say you had a fighter jet canopy made from various components and wanted to animate it opening. You could create an empty that created the pivoting action about a point. The other components could then be assigned to the empty. As the empty is rotated the other parts follow. This is very similar to grouping objects into ‘Nodes’ in ALIAS and manipulating them onscreen or in the Scene Block Diagram (SBD) window. I’m sure empties can do much more but haven’t used them much yet.
• Subdivision is critical to automotive modelling in Blender but it has to be used carefully. Like Alias SubD it is possible to construct a flat/simple low-poly model with a high degree of control, then toggling into a refined smooth/SubD mesh using a Modifier which automatically applies a denser mesh. Also it is important to develop understanding of edge topology and boundary flow to obtain smooth surfaces and transitions. Subdivision is another modifier that can be permanently ‘applied’ when you are happy with the model.
• Order of Operation. Sometimes modifiers have to be stacked in a particular order or you may not get the expected result. I found that Subdivide modifier is often the last in the sequence.
• Destructive and Non-destructive techniques. There are some instances where it is better to model using destructive techniques, manipulating the real geometry, rather than using the modifiers. Shutlines were one example where I did this. You can project flat shutlines (like line tapes) on the surfaces during concept development where you may which to reposition them, but for the finished model I wanted them cut into the surface to catch edge highlights in visualisations. You can only do this when you are satisfied with the quality of the model’s major surfaces because it’s hard to make changes afterwards.
• Geometry Nodes are another of Blenders most powerful tools taking advantage of Procedural Programming methods. The term Nodes in ALIAS means something different and simply describes grouped objects (or groups of grouped objects) that sometimes can only be selected in the SBD window. Blender geometry nodes are similar to a super Scene Block Diagram but each on-screen entity and its attributes are represented diagrammatically by nodes containing lists of attributes (Fig.6). Their main use is in controlling materials, shaders, lights, and environment textures which can become highly complex in animated movies. Geometry Nodes are best controlled using an add-on called ‘Node Wrangler’. I haven’t needed to use them yet, but I can see how Blender Geometry Nodes would be necessary to control a cinematic animation with many 3D elements, complex shaders, lights and environment. Procedural Programming isn't new, but I haven’t used it before. It will be one of the features I intend to explore in future.

• Although the Blender interface looks very intimidating and alien, basic automotive concept modelling probably requires less than 5% of the commands available in the interface (as did Alias). The secret is knowing which 5% and where to find them.
• My final take-away is that Blender Learning just can’t be rushed. You have to gain a broad front of basic knowledge before making real progress. It helps to make notes of operations and routines to refer back to. And you have to keep practising regularly or you will soon forget shortcuts and methodologies. You’ll also probably be watching more YouTube than football on TV if you get hooked on Blender! But the rewards are great and once you have attained the so-called ‘threshold concepts’ your Blender learning will become exponential.

In Summary.

I’m not purporting to be a Blender expert in any way and if you’ve seen my models so far that should be obvious. But as a Design Educator I’ve been able to to observe the differences in learning between expensive industry standard software like ALIAS and the open-source Blender. And as someone who has taken over a year just to make modest progress I know where and why I struggled with the latter.

Part of that, in all honesty, is because many of the skills and competences needed to work Blender are not native to me as a child of the 1960s (nor possibly children of the 1970s and 1980s). Blender is quite rightly optimized for the under-30s.

The main problem with learning Blender, unlike commercial software which comes with officially approved training programmes, is that Blender knowledge is dispersed and shared among a vast (but generous) online community. It reminds me of the strapline from the old 1990s TV SciFi series ‘The X Files’, which said “The Truth is Out There”…it definitely is, but with Blender you have to look long and hard to find it.

With no exaggeration, learning Blender has required as much research as my MA Final Project did in 2010. I have a folder of hand-written aide memoir notes and countless YouTube tutorials bookmarked. Berk Kaplan’s channel was invaluable at the beginning, but wasn't my only reference. My academic’s instinct led me to seek out other sources and viewpoints which have given independent insight and better ways to do some things. As a result, I am slowly developing my own workflow based on what I want from Blender, and I’m still learning. ?There are add-ons originally developed for video game designers that allow you to rig a car model to animate with moving wheels and suspension. This sounds a lot like Bunkspeed U-drive that I posted about a while ago and I am quite excited to try it.

So, to anyone thinking of trying Blender for automotive (or product) concept modelling but intimidated by the interface and apparent lack of tailored tuition I would recommend sticking at it. It is really hard to develop the essential ‘threshold concepts’ but there is a lot of guidance and tuition online. Once you get there your learning will accelerate.

Back in 2020 I really hated Blender because I just could not make it work (like I could make Alias work). Today, in the first week of 2024 I absolutely love Blender and what it will allow me to do in future. There is far to go and much still to learn but during the last couple of weeks, after a lot of head banging and bloodied fingertips, I managed to crawl over the top of that bloody Blender wall! (Fig.7). Everything suddenly makes sense and the software I couldn’t use at all has become really exciting. I can quickly sculpt surface transitions like using physical clay, so surfaces I wouldn’t even attempt in Alias are now quite easy to achieve. Oh, one last thing. In ‘viewport shading’ you can manipulate the model and see how highlights change on a rendered metallic surface model in real time. Imagine being able to sculpt a clay model if the metallic Dynoc film stayed attached as you went!

And again, all this capability is free. Realistically, although I’ve scaled the first wall there is still a big hill to climb, but I can now see the summit in the distance. Maybe I should use my recent learning experiences to set up a slower-paced YouTube channel. ‘Blender for the over 60s’. Actually, knowing what I do now I should probably make that ‘Blender for the over 40s’!

I wonder how many subscribers I’d get ?