“Blender for the Over 60s” Update - One Year On

Introduction.

About a year ago (9 January 2024) I wrote a long article detailing my experience as a very ‘mature’ designer trying to learn Blender. ( https://www.dhirubhai.net/pulse/blender-over-60s-light-hearted-look-my-personal-learning-clough-llwbf/?trackingId=lQ5EiROTRuODOtlI3ceRlA%3D%3D )

I jokingly called it “Blender for the Over 60s” and described getting to grips with software aimed at the under 30s, a demographic that possess very different skills and attributes to an old ‘Gen X-er’ like me. I have continued my Blender journey, and my progress has been hard won, so I felt the time is right for an update. My achievements to date may sound laughingly unimpressive to young Blender hotshots but I got here alone from scratch using just YouTube as my guide. Hopefully my progress will encourage other users of all ages to explore Blender :-D

A Watershed Moment

While writing this article (11 February 2025) Bentley Motors at Crewe placed a job advert for an ‘Expert Blender Modeller’. (Fig.1)

I found this insightful because it shows that Blender is gaining serious acceptance in mainstream automotive studios dominated by Autodesk, (and also affirms that I haven’t wasted two years teaching myself how to model cars using, to my age demographic anyway, really difficult software). One of the key job responsibilities is listed as “Leading the creation of high-quality Blender data sets for physical and virtual sign-off”. I’m unsure if this is hinting at engineering data because, in my experience, I don’t think Blender can quite manage automotive A-class surfaces (yet), but it is a step forward. Maybe the real purpose of this job is for future exploration of Blender in a commercial setting and to see what could be added. For now, I’m pondering what kind of candidate Bentley will attract because most true Expert Blender Modellers are not in the car industry (more about this later) so I’d love to see the applicant profiles.

What Motivated Me to Learn Blender?

Although I have retired from serious design and teaching, I still enjoy automotive design. After 40+ years I still get a kick out of designing cars and other vehicles, but I get even more of a buzz out of modelling these in 3D. Physical clay is not a practical medium for a ‘hobby’ designer to use at home and the digital Autodesk products I taught previously are too expensive for recreational use. ‘Free’ personal learning editions of Alias and VRED are available, but they may be time limited, and I don’t want to be left with ‘Wire’ files I can’t access in future.

My research indicated Blender was the way to go. It has so many creative applications, but my aim was to use it in automotive design concept modelling, surfacing and visualisation. An important and sometimes overlooked fact is that Blender is dimensionally accurate, so it is possible to model things with precision. Some users even 3D-print from Blender models. I’m unlikely to make physical artifacts but I still like to model digitally at ‘full size’, as I once did in Alias, so I prefer to model over a package drawing. And that’s where my Blender learning began.

The Package Drawing (A Brief Aside).

For the benefit of the unfamiliar, a package drawing is the transport design equivalent of an engineering ‘general arrangement’ drawing. It once comprised a set of full-size orthographic views and sections drawn on gridded mylar film that included all mechanical components/hard points’ as well as ergonomic manikins (usually representing the extreme smallest and largest people that had to be accommodated in a particular vehicle). In car design a 3-dimensional grid of 10cm squares called ‘ten-lines’ was used. By knowing the coordinates in terms of ten-line position in X (length), Y (width) and Z (height) from a known datum (Fig.2), parts could be accurately located in 3D space.

Individual components (e.g suspension parts) drawn separately on smaller sheets with the same numbered grid were then traced into position on a master package drawing (Fig.3).

Obviously since the 1990s, car design has been carried out largely in 3D CAD virtual space using the same ten-line coordinate system, but even today a side view package drawing with 2D manikins is a good starting point over which to correctly proportion a freehand concept rendering. This can also form the basis for creating a clay model to scale.

I liked using ALIAS in concept workflow. By quickly doing a side view sketch and importing and scaling this as an image plane, I could trace the key lines to create a planar network of NURBS curves that I could then pull laterally to make a 3D wireframe ready to surface. (fig.4).

But to ensure that the model proportions were not cheated I created a set of accurately scaled manikins (ergonomes) based on some human figures (and a dog) that were supplied as ‘assets’ with ALIAS (Fig.5).

Blender didn’t have anything like this, so my first goal was to create a set of functional ergonomic manikins to throw into my models so I could develop realistic seating and vision envelopes to work over.

Blender Manikins – My First Models.

I simply didn’t have enough knowledge of Blender to model a car at the start, so my first task was to create some ‘functional’ ergonomes. I modelled these at 2.5th percentile Chinese Female, and 97.5th percentile Dutch Male, the world’s shortest and tallest range to accommodate 95% of the global population (Fig.6). I also created 50th Percentile European Male and Female ‘average sized’ manikins to use in visualisation because the extremes can look odd (Think Richard Kiel and Blanche Ravalec in James Bond film Moonraker… Google a photo). These were low polygon articulated manikins made up of discreet body parts that could be rotated about their joints to establish seating positions in 3D. They did not have to look pretty as long as they were accurate. Modelling the manikins introduced me to many key commands, shortcuts, collections and modifiers, building up muscle memory. I even experimented with sculpting tools to give the manikins recognisable facial features. I later shaded a set of manikins so they would look better in renderings when I finally modelled a vehicle (Fig.6).

Blender can be hugely frustrating at the beginning. Coming from Alias, I wanted to run before I could even crawl so creating accurate manikins gave me an attainable goal and sense of achievement as well as a huge incentive to carry on learning. I would recommend anyone to start with small achievable objects, taking baby steps. Learning Blender requires patience!

Once I had the skill to produce reasonably well-surfaced car models to visualize in realistic HDRI environments, I wanted some properly rendered ergonomes. I downloaded free high-poly, rigged manikins from a specialist website which I scaled accurately to the same percentiles as my functional ergonomes including some 50th percentile (average) people. (Fig.7).

I tend to use the low-poly functional manikins during design development but for final visualisation I can swap in the fully rendered, rigged versions which can be placed in sophisticated poses. The final results I’m starting to get speak for themselves. (Fig 8).

Progress

This time last year I had already picked up a lot of the modelling basics and could make presentable sketch models, but the geometry (topology) of the models wasn’t actually that good. I have also mentioned that the Blender interface requires a lot of shortcuts to be memorised. Keyboard shortcuts are used far more than other software I’ve experienced before because most menus and dialogue boxes show different things depending on the mode you are in, or even what you have selected on screen at a given time. Shortcuts allow quick access to the most used commands independently of the Graphical User Interface, which speeds up workflow. The only way I could learn was by constant repetition, making notes and by ‘doing’. With time and effort things did start to stick even in my old brain but I still refer to the notes.

As a former ALIAS tutor, I appreciate why Blender Learning really lends itself to video tutorials (YouTube) but even these won’t make complete sense until you have acquired some knowledge of the basics, the ‘threshold concepts’. You really have to be motivated to learn Blender and can’t just shuffle along expecting to pick it up. Luckily, I was and still am highly motivated and the more modelling and visualisation I do the better and quicker I get.?

For now, I’m not too concerned about learning totally photorealistic visualisation, animation or exporting to Unreal Engine because those are specialised areas in their own right. I want to be able to design something on paper or tablet in 2D, then to model and render it in reasonably realistic environments with convincing shaders. It’s the digital equivalent of putting ‘Dynoc’ on a clay model to view it outside in the sunshine!

What Else Has a Year Taught Me?

First, that everyone should start learning Blender in its basic form which is totally free. Even this is very powerful and will allow you to understand all of the fundamental principles and workflows and to make sophisticated models quite quickly (if you put in the effort). If you then get serious about using Blender for design modelling there are a number of important ‘Add-ons’ that you may want to get (some free, some paid for). ?These enhance the operations of basic Blender, making the workflow quicker or streamlining some operations. In some cases, add-ons make it possible to achieve surfaces that would be difficult or impossible in basic Blender.

But I would still strongly recommend using the default version before spending money in the Blender Market. That’s what I did until I understood which add-ons would be useful to me. I then bought several that I knew would streamline my Automotive Design Modelling. Another useful tip is that Blender Market has ‘sales’ about twice a year where paid add-ons are discounted by up to 30%. That can save you quite a bit of money if you are patient. So far, I have spent around ￡50 on paid add-ons and have identified a couple of others on my wish list for a similar amount when the sales come around again. Putting things in perspective, I will then have paid about ￡100 for perpetual software that achieves similar results to commercial programmes charging tens of thousands of pounds on subscription per year! That’s not a bad deal is it?

What’s Next for Me?

I’ve now developed a workflow that I’m happy with and that closely resembles how I used to work in ALIAS and VRED (Fig.9).

In terms of projects, I have created several admittedly rough ‘learning’ models in Blender to date (and design iterations of those models) but have not yet attempted a detailed interior. My next goal will be a ‘proper’ interior for my MA Project redesign (above right). Also, I have a number of other vehicles that I previously sketch-modelled in ALIAS/VRED but which I would now like to revisit using Blender.

Regarding the Blender software itself, another important area of future learning will be a deep-dive into the procedural nodes structure. This is an incredibly powerful method for creating geometry, shaders, scenes and for manipulating objects in the background without using the usual GUI. I have scratched the surface, using simple nodes to set up visualisation, but nodes can do so much more and particularly when generating complex shaders, lighting set-ups and animation. Nodes are where the dark arts of Blender really live.

Distance Travelled

Since first trying then abandoning Blender in 2021, I have travelled a huge distance down the Blender learning path. However, it has taken nearly two years serious and concentrated effort from January 2023 until now to reach my current modest level of ability (Fig.10) because there is just so much to take in (and then retain).

I won’t pretend that it’s been easy nor that I’m an expert. Far from it! ?The interface is infuriating, so much depends on shortcuts that you have to memorise, and even online tutorials are challenging because they often assume prior knowledge and miss steps out.?

Blender is just too complex for spoon-fed step-by-step instructions. You have to help yourself to reach a certain level, (‘threshold concepts’ again). In my last article I compared it to ‘The Knowledge’ taken by London cabbies. You can study a map or satnav to get an idea how to get from A to B, but prospective London cab drivers walk or cycle every route for months or years getting to know every landmark and alternative route by heart before taking the Knowledge test. Even then roads change all the time, so cabbies are constantly learning. And being Open Source, Blender also changes and improves at a lightning pace.

Blender still has limitations for automotive surface modelling and I’m frustrated by some things I can’t yet do. It may just be that I haven’t ‘discovered’ some of the latest advanced techniques because once you go down the Blender rabbit hole there are lots of important side tunnels you can explore. Also, at the moment I’m just doing this for fun because I actually have a life and can’t spend every single day on Blender…which is almost a necessity if you want to be a real expert!

Having watched a lot of YouTube content, I know which are the ‘go-to’ instructors for the different aspects of Blender.? I’ve also observed that different content creators cover different applications of the software which suggests there isn’t a single Blender Expert who does everything. Becoming a competent Blender user is less about ‘knowing everything’ but about ‘knowing where to find things you need to know for your discipline’. I’ve had to crawl on my hands and knees to get to where I am, but the sense of achievement is incredible. With Blender you never stop learning.

Blender has also helped me rediscover the joy of automotive design. This is a 3D discipline and although I’m no longer active professionally I still like to design innovative stuff, and being able to model it myself is creatively liberating. I have no wish to experiment with generative AI because I still love the visceral human ‘art’ of design (sketching, rendering, physical sculpture and manipulating 3D digital surfaces myself) that led me into car design from being a child. And yes I still love drawing cars!

Video Snapshot

To compliment this article, I’ve created an annotated timelapse video at the end showing my personal workflow from rough 2D sketches to 3D Visualisation. The topic is my old MA project again which I have updated. I have skipped over the actual modelling of the exterior shell for now…that would have taken too long because the video shows Design Iteration number 5! (Fig.11).

This final model represents months of trial and error and steady learning to achieve surfaces that I’m reasonably happy with (quality over quantity matters to me)...but there is still much room for improvement. This design has largely served its purpose (unless I create a detailed interior). It’s now time to tackle new projects.

Summary

In summary, do I think the many months of pain have been worth it? Absolutely I do! Blender is amazing for Automotive concept modelling and visualisation and has democratised access to very high-end 3D software especially for young, digitally-native, pre-university users. As an ex-designer who just wants to model stuff in my spare time it’s perfect.

Final Observations on the Bentley Job Advert

I genuinely can’t wait to see how car manufacturers like Bentley use Blender in anger, but I question how many of the ‘Expert Blender Modellers’ they seek actually exist in the automotive industry just now. Most Blender ‘3D Digital Modellers’ specialise in hard-surface consumer products and game asset design. Many others do organic soft modelling, texturing and rigging for animated characters. Then there are 3D artists who use Blender for motion graphics and animation in advertising for example, but the industry standard software in many areas appears to be Adobe, just as Autodesk dominates the automotive design discipline.

Out of curiosity, I have searched LinkedIn Jobs and Google using keywords themed around ‘Blender Modeller’ and there don’t appear to be many such roles, nor people with a specific profile that clearly match the Bentley Motors job specification. I recently discovered that Ubisoft and Sony Interactive Entertainment are starting to use Blender in parts of the video game design process which is also very significant. Blender is currently a bit of an outlier for professional applications but that is bound to change if it is universally adopted for concept modelling in automotive OEM studios.

I’m sure many keen young automotive design graduates now have Blender skills but may lack the overall design studio experience to be considered ‘experts’ just yet. There are also Blender ‘automotive modellers’ who are great at producing and rendering existing cars from accurate blueprints (e.g. for video games, configurators and advertisements) but I would question if they can model new concepts that only exists in 2D theme renderings, as sensitively as a fully-trained Automotive Design Graduate. ?The solution must lie between these two.?

Blender has enormous potential for automotive design and if I was thirty years younger, I would definitely be steering my career in that direction (actually I wouldn’t mind dipping back into industry for a few years now to push Blender along). Amusingly, I’m probably one of the oldest ‘serious’ Automotive Blender users with prior hands-on experience of Automotive Creative Design (Exteriors and Interiors) , Clay and Hard Modelling, Alias and VRED. My background in Alias/VRED teaching also means I know how to train designers quickly to get to where I am now, based on my own experience… but I’m still no Blender ‘expert’.

For now, my journey continues…

Enjoy the Video!

?