Jason Pook's Games Design Blog

1 Comment

Client Project- Session 7/8

I have combined these two sessions due to the 7th session Myself and Phil was pulled off from doing work on the Client project personally, this was due to we are further ahead in both work and skillset so our tutor Paul wanted us to go around the room and sort everyone who is struggling with certain aspects to get up to speed. The common problem seemed to be scale, people didnt have there grid set up properly so it snapped in ue4, snapping, the units setup was generally in metres when it needed to be in generic units for the engine, and then the way modelling should be done with buildings is the modular aspect to save time and increase work flow.

I spent around 35minutes with Rinalds, he had all of the issues above but once i sat with him and showed him step to step how to go about it he took it on board very well and seemed to improve. Im hoping with the techniques i showed him that in the next group critique session he will have come on alot, even if the building modelling isnt 100% as long as its snapped and to scale thats foundations we can build on in future. I then moved on to abbie and barrie who had the same problems which all i could do was tell them the right way as i did with Rinalds. Eventually i gave dave some feedback on some windowframes and pillars he was building modular for his building and then James had a slight issue with snapping despite also going round with me and Phil.

Each on of them i linked them to a free Digital Tutors course on modular building, it not only shows you the correct process it shows you texturing methods which will be the next step for everyone once we have a initial first street block out done. Found here: Digital Tutors modular design course That concluded session 7 hopefully this wont be needed in future so it doesnt pin the likes of myself and Phil back, but it was good to give som peer to peer learning to others in the group to build on the feedback from the previous group critique.

Moving on to session 8 now, Me and Phil felt it nessacery to really push on. We had been talking over the weekend about where we are at in the project and where we should be and its not as far as we expected, this being reinforced by our tutor Paul we took session 8 as the next gear. Phil took the group talk and requested we push on, we expect all buildings that have been set to be done in a basic model by now which should be 2 each, 1 for big building projects like viktors bank. We wont check up on them until next weeks group critique were we expect finsihed products and moving on. I added that no matter how far behind you are with your current buildings im sorry but they should be done and we are moving on to the next its now addition to your own time to get that done.

With that said the majority of people took absolutely no notice and carried on with there initial buildings. Me, Phil and James all collected new buildings to be getting on with we cannot afford to fall behind for other people, but as this is a group project the lack of collaboration was dissapointing. Nevertheless i collected myself 3 new buildings, currently the Poundworld, Quicksilver and Check Centre. I finished Poundworld and Check centre in the session and Quicksilver shortly after, they are basic models with some minor details to pillars but are to scale and snapped to grid.

Here are my finished products:

Quick Silver:


Check Centre:

I will be showing these on tuesday in our scheduled group critique session to show my progress and get any constructive feedback on my work. I have put these in engine and they are to scale and snapped to grid properly. I have now contributed 5 buildings and will be collecting more shortly as soon ill have assets to focus on aswell.

1 Comment

Summer- Self initiated mini brief- Process

The brief i set myself in my previous post i am now going to discuss my own process of completing it.

Pt1. Find or Research your chosen vehicle for use of reference when modelling (Your car/ images/ sketches/ owned item etc.)

What was i going to do? there is a wide range of choice and being a huge car fan i had to go down that path. Personally owning a classic mini i thought i would go for that…but then i didnt want to jump in head first with this i could come back once i felt id improved my 3d modelling. In relation to the brief it says representation of a vehicle so i decided a remote control car would be a fairly challenging model to build to start me off with this brief, and i could also build the remote to control it.

With that in mind i set to the internet to find myself some reference images, maybe some parts i could add like a spoiler of my choice as rc cars are customisable.  SIDEtop

So this is what i found, simple shape with plenty of features to model. like the suspension springs, screws, spoiler, wheel, tire tread, antenna etc. The reason i picked these angles as when i am modelling in 3dsmax if i set up 3 planes evenly and texture the images into viewport i can use them whilst modelling to guide me in shape and size of my car. I think i will do a slight interpretation on some features due to topological difficulties but we shall see how it goes.

Pt2. Once comfortable with chosen vehicle (it has to be real, no transformers allowed!) plot yourself out a design process to tackle in 3d modelling.

I realised that doing an rc car i could cut a few corners as there is no interior and just a base so one moulded block would show a built rc car under the shell. But first things first i set myself up 3 reference image planes top, front and side. This would allow me to get the scale of my model right in at least 2 viewports for accuracy to the model.

reference setup

As you can see evenly measured planes that have been aligned together with precision, this was just a simple process of create plane to scale, clone 2 times and match, then texture each individual plane with relevant reference image. I then selected all 3 and clicked freeze selection so i wont even move them by accident, it wont be changed.

The next step was the car shell, as i dont know what the frame looks like i decided to do the main car shell then model the frame to match the shell based on what id made.

The shell started off as a simple box that i scaled to match each viewport as close as i thought i could go,car shell but obviously there are many curves and bumps in the car so a box would look right at all. By converting to an editable poly this allowed me to use the connect tool when selecting edges, this is similar to the swift loop tool but if you do it in multiple connections its more even topology. I then just grabbed vertexes and moves them to fit the shape of the model, i did this in side and front viewport so it had the reference shape as if i did just side view it would be wrong. I then deleted all polys underneathe the shape as its just a shell and i didnt need them as other objects would fill that space. But as you can see thats far too sharp and boxed to look realistic it need smooth transitions of bends and shapes.

car shell turboTo do this i added a turbosmooth modifier, which gave me the curves i wanted i upped the iterations to 2, turned on isoline display and clicked on show end result when going back to my editable poly. i used the swift loop tool to create some air intakes front and back. Obviously i wanted to show definition of the centre piece and the intakes so a trick i learnt was to use the swift loop tool whilst turbosmooth is on and the closer to the line it is the more pinch it will do to your tubrosmooth to make it a more solid shape. With being pretty happy with that shape and it being modelled to the reference images i decided to leave it for now and move on to modelling another area of the rc car. to make the topology neater when i swift looped I moved selected ringed edges and use move with edge restriction to even it and space it all out.

I proceeded to the tire, from the reference image it needed to be fairly thin with a bit of a outer ridge like a tire does. I did this initally using spline modelling techniques, i put it in the front view and used lines to draw the shape of the tire but only from top to where the rim would be not the full tire otherwise my chosen modifier next wouldnt work. I then used Lathe modifier in the X direction to create my geometry, but it wasnt in the right axis, using the sub selection of the modifier axis i used the move tool to drag it down and create a tire shape with a ring in the middle (where the rim will fit). to create the smooth inner tire bump i converted it to an editable patch and used the inner handle scaled in the Y direction to curve this out and create the tire.

The rim was created by a tube aligned to the centre of the tire and a cylinder with matching segments to the tube facing eachother so that i could use bridge to create the spokes. Then it was just tweaked with a centre cap which i modelling a screw to fit inside, it was turbosmoothed and swiftlooped to pinch the shape and create our wheel in completion.

After that i duplicated the wheel when grouped together to have 4 copies for the car aligned them and positioned them to individual locations. This allowed me to model the axel and wheel connector to the axel. I then began to model the spring suspension for the car which is clearly visable at front and back. Being a spiral shape the best thing to use was the helix shape which is essentially a spring, once aligned to my object it was just a case of making sure it looped as much as i wanted to and tighten it up around the cylinder of the suspension with height and radius. Once happy all i had to do was go to the rendering option on the modifer toolbar and tick enable in viewport and renderer, this then thickened out my helix to a 3d spring which i could adjust thickness and sides for however high poly i wanted it.

tire3 treadI then added tread to my tire, this was just cubes modelled on the surface on the tire cloned several times in a row of 2 then every other cube being deleted to create a tread, once angled the tread on the inner part of the tire it was duplicated with the axis set in the centre of the wheel using the rotate tool on snap holding shift.

To speed things up i made a list of what needed doing to finish the car. I needed to model: a frame, a inner car carriage with battery latch, spoiler joined to frame, bumper joined to frame, a car antenna, a remote with ariel. Then i needed to align everything in position relevance, make any additions i may think of, use symmetry for tires to the other side, texture.

Here is the remainder of the process as this post is getting very lengthy and images speak a thousand words, maybe not in 3d modifiers but i will explain where needed.

I modelled the frame using box modelling to create the outer edge path of the car once donei used the symmetry modifer to mirror this to the other side, and to make the support connections down the middle i just extruded inner polygons where i believed nessacery. Then the inner carriage was modelled to fit above the frame but inside the car this made sure the inside of the shell wouldnt be seen and gave depth to the model with a simple addition of a inset battery box with latch. I then made the spoiler using splines for the edges of the spoiler and the handles of converting to bezier to sculpt the edges, then applying the extrude modifer for 3d element. Final touches were needed to the connection of the wheel to the suspension to the car needed a extra part and extrusions to make connections visable. Then it was final modelling of a bumper which was beveled to create what i thought was a nice solid look to the bumper, then the antenna for the car to recieve the controllers movement. Finally the controller was modelled, using topology creation in local for some features. It was all just a case of putting shapes together for the controller and extruding some polygons for features. And that left texturing after i added my air intake on top of the car which wasnt on the reference but i liked it as a feature.

Here is the final car!

This slideshow requires JavaScript.

I chose a random colour scheme and then added some racing stripes, a racing number, some form of zig zag decal, and some racing sponsers of NOS and Falken. All textured in photoshop. When unwrapping the shell i used the relax custom setting so i had a more even texture when applying my colours this really helped in the end. Also to finish it off and complete part 3 of texturing and organising, i grouped all areas to eachother and gave every part a name so there was no Box1001 and if this was to be passed down the pipeline in a game people checking or editing would know exactly what is what. Grouping is key.


Alot went wrong dont let the process fool you, i had many a rage storm out the room when it just wasnt doing what i wanted to with the topology, capping polys and actually modelling the frame/spoiler took about 11 times as it just wasnt transitioning from one to another right with the turbo smooth applied. With experiementing on many new modifiers i hadnt used before such as lathe and symmetry it took time to adapt to using them and getting my head around how i would go about the next piece.

In the end looking at the reference image i thought well that isnt exactly what is there for me to make. But you have to remember its just a reference image its there to guide not to be precisely replicated unless specified so i was pleased with my result.

One tool which was a total life saver that i was new to was the “isolate mode” this is a little pink light bulb at the bottom which when you selected what you was editing it would hide everything else and focus purely on that object. This allowed me to rotate around the object much easier and nothing was blocking my view. Other tools i found useful: local- this allowed me to move or scale objects i maybe previously rotates and angled in the direction they was facing so the objects local. AutoGrid- when modelling something i wanted on a surface so say the control sticks for the controller if i put on auto grid it would create my shape onto the controllers face.

Overall i was very pleased with my outcome for my car. I am hoping to now take this a step further and model a proper car using the skills i have obtained here. The poly count was just over 1.7million so high poly work but there was no restrictions and i did use only what was nessacery in some cases it could have been alot higher.

Give it a go the brief i set is here.


3D Realisation- 3D Focus (9/10/11)

In this session I had to present my 3 3D artists (Tor Frick, Christopher Radsby and Paul Pepera see 3D Realisation- Artists Research for more information on them) to Gareth and Paul my tutors in 3D, I opened up by showing them everything about my 3d artists and what they have done just so they were clear on that my artists met the given task of 2 shipped games. Once I had explained to them about my leniency towards Tor Frick being my chosen sole artist to focus my design on it went down very well, they did suggest originally that I should check him out but by confirming he was my favourite pleased them as they both enjoy the work of Tor Frick. Whilst explaining my ideas of how I would adapt his work into my own, I mentioned I would like to be a level design because I like how the environments and all the visual of a level looks. This immediately caused a problem as then I was shown what a level designer was, a level designer is the person who makes the level as the basic shapes and terrain no textures at all just plain blocks. The 3d artist is the person who puts the visuals onto what they have been given with level design. So by learning this I quickly decided its definitely 3d environment artist I am looking towards doing for this brief.

“Level design is really about planning the flow of the level, making decisions about gameplay, and laying out challenges for the player to overcome. Level design as a discipline starts with written plans and drawn schematics before firing up a level editing tool.

“Pure” level design is embodied by a workflow called “white-boxing” where the level designer places blank (not necessarily white, anymore than blueprints have to be blue) geometry in the scene to establish scale and structure, places all the dynamic items and enemies around the scene, and then environment artists go through and replace all the stand-in geometry with polished assets.” – jhocking.


This is level design.

So from finding out my misunderstanding of 3d environment artist and 3d level designer I focused on what I was going to do with my work. Then I was sent on my way by my tutors as I needed to now focus on what I was going to be designing for my brief. I had thought up a few ideas from going into extensive research on Tor Frick, but wasn’t sure if I was thinking too big for my first ever 3d task. So until next week it was pure design focus for me. I was also told that I would need to produce a physical model for whatever I was going to be designing in 3d, and that there were multiple ways of going about it: clay, cardboard, paper etc. I have no experience of modelling things apart from with Lego bricks as a child so this will be a learning curve for me im sure. Regardless of that my ideas will not be toned down just because I may struggle on the physical side of modelling.