A Foggy start

 A fog hangs in the air. The morning wears a drab cloak of biting cold. I know because I had to pop out to get food items for my eldest son's food-tech lesson. I do wish he'd remember to ask for these things sooner—not the night before he needs them!

Asking him to be better organised feels like travelling back in time to tell my younger self the same thing,  no doubt with equally disappointing results.

Start of Week (SOW)

Last week I kept the Medium Rig build going consistently. I often find the transition from Large to Medium scale troublesome because a model isn't in a great state. I need to do more "squint tests" to evaluate composition and layout before starting on the medium, then the small details.

I got almost nothing done on the Halo: Raptor terrain work.

Almost no training!

This week

I must make progress this week. I think Brannon is expecting a surface treatment, but I have a couple more stages to complete on the mesh. The cliff treatment still needs some testing and experimentation to see if I can apply outcroppings.

Medium Rig Crit 2

We're not too far off. I really want to do a good job with the materials, so the model gets much of its small detail from the surface treatment. Those Hitem3D cargo cans lead the way; they have small details at the right scale, but I won't apply that detail density to the rest of the vehicle—it would be far too cluttered.

I don't want to do too much to the aft section, just break up some of thse blank spaces, just a bit.

A final step will be to finish the detailing on the MFI rig. It should be quite uniform, but I'll add more plating, radiator panels and wiring in the inside surface.

Test shots!

I added some post-effects to see how the design looks. Not bad! It certainly looks a bit less synthetic. The lighting is extreme, and the final renders will be a bit less blown out and distorted, but it's not bad for three minutes' work on an unfinished model.

Training

Complete an existing course. Maybe a terrain/landscape-based one. 

I should do a refresher on things I will need for the terrain work, such as Nature Generator and or Geo-Scatter.

Start of Week, Start of November

 Start Of Week

Last week I did some good work on terrain sculpting, pushing the Halo: Raptor work slightly forward. I need to continue with the work this week.

I also continued with the personal project/Orbital Elements vehicle/Commerical Product effort with the Medium Rig (hold on, didn't it begin as a light rig?). 

This week, I will focus on completing the Medium Rig. My worry is the CG artist's adage: Completing 90% of a design leaves about 90% of the work time remaining. Polishing and tweaking takes much longer than laying the foundations.

Last week, I completely failed to do any training. I look back over the last few weeks, maybe a few months and see a big motivation and creativity leak happening for ... reasons. I think that things have improved since the half-term break, but we'll need to see how this term pans out.

I have been a lot busier with busy work; more frequent grocery shopping, more food prep, more cooking, and more house chores. It's all good, except it has cost creative time.

November Plans

Along with weekly plans, let's try monthly plans (SOM). 

Halo: Raptor. Complete the Terrain effort so that by the end of this week I have a solid base mesh to tweak and polish. The aim has been to create a complete service road model that can be chopped up and turned into the individual scene elements. That suddenly sounds like a tall order!

Orbital Elements: I'm aiming to write an hour per day. I am just not managing that. I will not beat myself up about the failure. Thank you, Aimee McNee, for that advice!

I need to develop my terrain creation skills outside of True Terrain; it doesn't look like Richard or Zach can get much done on a new version. No fix seems to be coming soon. <wipes tears away> 
To do this, I will complete a hefty course and go back to complete or repeat courses I already have.

Clean up time

Next run: tidy up all the mid-level details. Next run: Tiny details.

Fly-past days, weeks and...whoosh, here goes October

 Raptor Progress

Slow but steady. All the terrain-sculpting fun has gone into developing the service road cliffs. I'm happy with the progress! 

Cliff section sculpt

This element is really big, about 16 million polygons. Really, this is a miracle of Blender 5, which always seems to laugh off polycounts, at least until it crashes. On the subject of crashing, I've been getting a lot, but all on the older machine. This could be CPU erosion, and I need to start limiting more cores. We'll keep an eye on that. I use the secondary (4090) machine for sculpting because it has my Wacom tablet attached and all my lovely rock-sculpting brushes installed.

Quick test on instancing

Take the giant terrain element and create instances, then arrange them into a big scene that doesn't swallow all your system memory thanks to instancing!

Camera view

Yes, you can see the repeated elements, but this is a test. In anger, you could add more clutter to help hide repeating patterns.

Other stuff

This felt a bit Blender-schooly, so I didn't do any additional training. I will at the weekend. I wanted to collect some notes on the recent phone-camera addon. It's time I created something worth showing.

Other other stuff

The Medium Rig started to gather dust, so I jumped back on it yesterday.

The model is in serious need of a clean-up, but I am holding off until I complete the first pass on all the elements. Later today, I will complete the propulsion section and add a comms array, possibly on some clusters attached to the MFI ring. Some of the details need improving. I really want to do a better-than-usual job of the materials.

A tale of two terrain generators

 I've somehow fallen into a new workflow where I use both World Creator (WC) and Gaea to create a terrain. I use WC's ability to quickly build and sculpt a basic terrain from primitive landforms. I then import the heightmap into Gaea in order to use its superior node-based filter system.

Then I can take the Gaea output and import it back into WC to make more tweaks...and presumably ping-pong between the two applications until I'm in a happy place.

World Creator base terrain

Imported into Blender 

Into the Octoberverse

 Month ten, already! Wow!

Last week I kept thinking, developing, and sketching for Orbital Elements, so I don't have much to log in this blog. This week, I'll be back here catching up.

I like how close this design comes to the first OE space vehicle. Most big ships are built on a super-strong stem, made out of space elevator material, which has mind-blowingly high tensile and compressive strength.

I need to account for heat management - that MF ring generates a lot of thermal energy, so there needs to be a big radiator array or some novel heat sink system -- maybe huge tanks of salt.

Playing with the idea that smaller vehicles still need really big MF rings. Maybe they fold out/inflate, or they just don't appear in the vehicle sketch.

Blender time! Astronaut created using Hitem3D

Start of Week

Last week was all Orbital Elements. This week, I'll confine this work to my evenings out -- Mondays and Thursdays, when I walk with Noah to Army Cadets.

Halo: Raptor

Time to appraise where I am and what I need to do. Have I become fixated on the details and accuracy? I feel like I've been led to a task that is more open-ended and can easily become impractical. I need more Gaea expertise, but more importantly, I need to really understand how terrains are engineered. 

The service road's walls are very steep, and it seems they need to be enhanced manually after the terrain is converted to a regular mesh.

A practice service road-like terrain

OE Vehicle project

As a side project that works in both streams, I'll create a highly detailed OE vehicle to showcase the new design rules.

Proof-of-workflow test

A point of inflection

 Halo: Raptor

The news of Microsoft's closure of Halo Studio and its transfer of the IP to Activision has, unsurprisingly, plunged the project into fear and uncertainty. I think Raptor will likely continue, with an adjusted aim point, since Microsoft can no longer provide a path to a web series production. I might even prefer a worse scenario, where they develop a whole new IP. Clearly, Halo has a huge following, and anything Halo automatically generates interest. This automatic interest comes at the cost of creative shackles; you're very constrained in what is a pretty derivative (Aliens blended with golden-age sci-fi) universe. A fresh start could open so many interesting new doorways, without significantly changing the story we intend to tell.

We will see!

I'll keep punching out iterations of the service road terrain, although I am struggling to get the fine detail I crave. I'll keep trying different approaches.

WC-Gaea-Blender workflow practice. Something quirky happened...

Reduced displacement looks better, but it isn't what I was aiming for.

MSVAs

I'm enjoying the concept creation phase of the work. Sketching and thinking, usually fueled by coffee and occasionally cake. The Modular Space Vehicle Assets line is tied directly to the Orbital Elements (OE) universe. Killing two birds with one stone, if you will. This means that OE provides design constraints, while the Blender modelling end provides OE with some ship designs, if not a ship kitbashing system.

A mid-sized space ship vehicle (In OE, we don't call them ships)

I still need to take a decent of Noah to use with Hitem3D for his Attack on Titan commemorative render.

A Blender sketch I made while trying to persuade my 11-year-old that it was bedtime. 

Tyr's day

5.2.2 due

Today, another Blender update should drop: 5.2.2, which I imagine fixes any serious bugs uncovered since 5.2's release. I was surprised to learn that Blender 6 is entering Alpha. Blender 6!!!

Start of Week (missing from yesterday's post)

Halo: Raptor

I am still beavering away on the test version of the Gaea-generated service road area. This week I will work out a better strategy for adding lots and lots of undulations around the service road canyon. Right now, all my tests are too sparse and too flat. Getting more density might require a more procedural approach, rather than manually painting the terrain.

Modular Space Vehicle Asset Set (MSVAS)

Allow a couple of hours (this week) to plan the next steps. I think that the first thing I need is a design language for the space vehicles, which are part of my 2288A.D. Orbital Elements universe. Once I know what a 2288-era vehicle looks like, I can start planning how many engine pods, how many cargo pods—that sort of thing.

The big inspiration right now is the Project: Hail Mary space vehicle, which looks like a real thing, and yet looks cool. <I just grabbed a link to the original design by Paul Chadeisson, which was on his LinkedIn account, but found that my account has my old email address as primary. So I tried to change my primary to a working address, but got stuck in a loop trying to update it...>

Not this, but you get the idea

Blender School

I need to do more training/learning/practice. There are so many things I could or should be doing. So many that it's now impossible to choose what to do. Nature Generator? More rock sculpting? A newer addon that I've hardly touched or even forgotten about? Come up with a training timetable for the month, then jolly well stick to it.

Therapy

You guessed it. Terrains...endless, almost pointless, terrains

Workflow: World Creator --> Gaea --> Blender 

This is becoming a bit of an old chestnut—a bowl-shaped terrain you can plonk the camera in the centre and do a 360-degree pan without the terrain suddenly vanishing.

Wider (20mm)

They're not bad, but I really need to get the masks in True Terrain matching again. Maybe I should beg Richard to take a look at it when he has time?

Whoosh. Another week flashes before my eyes

 It's flown! Here we are on Friday, seemingly moments after Monday's close. This might be the power of routine. Compared to the Summer break, everything runs on rails of routine, so you stop noticing the flow of time.

This week

Progress on the Halo terrain. I need to start a third iteration of the canyon/service road element. This time, I need to spend much longer filling in the land above the canyon with peaks and undulations. My main beef with the previous two attempts is that they don't include enough terrain detail.

Cheeky to show (NDA) but it's rejected, right? So it illustrates the lack of detail around the canyon.

Train travel

Clara and I took the train to Worksop because our broken-down car has been towed there for repair. We dropped off the service documentation at the garage and lugged most of our luggage back. While waiting for the train, I snapped photos of track maintenance engineers because their safety packs look really cool, maybe a bit Sci-fi.

Rough source!

It struck me that the figures would be a good test of Hitem3D. Exactly how usable would outputs from quite poor sources be? A test worth doing, I thought. I  wasn't expecting this to work at all -- look, his feet are essentially missing!

So that worked out pretty amazingly...

The first engineer with the best photo is the best by far, and I think it shows. That said, they are collectively very useful for background elements. I'm impressed! Also, a bit scared of what this does to the 3D industry. There's going to be a whole lot less work going forward. 

Fluent back in the house

The hard surface modelling tool, which wasn't initially working in Blender 5.2 looks to be fixed.

Quick test element: some tiny issues on the mesh, as usual.

Apparently, I bought this add-on in the opening weeks of the Ukraine invasion. Since then, I  have used it occasionally. Mostly, I get into a mess and need to learn how to make the most of the mesh/normal repair tools. Occasionally, it's done an amazing job.

Back to work, again

 Halo: Raptor

The director reminded me of a test I said that I'd run, using Gaea to rebuild the terrain element for the service road. It should be possible; the question will be how to do it for the quickest and best results. I must confess, I've not touched Gaea since my rush to proficiency. I'm also hoping that we see Gaea 3.0 appear in the next few weeks. I'll have this test completed by the end of the week.

More Hitem3D

The daily free credits have stopped, but I've continued to generate content using Hitem3D. I took one of my heads and placed it on one of the bodies I'd created in last week's free-credits splurge.

V3 and V2.1 What a difference! Twice the resolution now, apparently.

I also started to create an asset library for the latest high-resolution figures.

I used decimation to create three variants of each figure: at 10k+, 500k+, and 1M+ resolution.

The men! I'll add the women once I've generated a couple more.

While there's a super improvement in the overall resolution, there also seems to be better accuracy. Hands are less prone to weirdness. This could be an issue with using generative AI source images vs high-resolution photographs of real people. There's room for improvement, but we're really only a year into this particular implementation. Where will we be in 5 years? Photorealistic (or near enough) output, maybe?

Start of Week

 Here we are, burning through the summer holidays. It feels like we're only just starting, when in fact we are more than halfway through. Time to roll up the sleeves and get to work. 

Last week's progress: I get a pass! I was on holiday in Europe. This week will come with high expectations. Fully rested and raring to go? Well, maybe.

Aims

Halo: Raptor

New instructions from the new Environmental art lead. I'm to return to the opening planet scene and redo it following a different artistic vision. Marvellous! Who doesn't like throwing away two months of work? Well, this is the job; it's how it works. If you are too precious about work product, you are probably not cut out for commercial film/content production. That said, the fact that they didn't think that I "knocked it out of the park" with the last run-through is not a great sign. The new directions give me concern; they are driven by a series of Midjourney frames that, in my humble opinion, are overcooked -- like much of Midjourney's output. They are eye-poppingly attractive, but it just looks like gloss over substance: a striking image that doesn't support looking too closely at the details.

Training

Where to start? So much to cover. I feel like I need revision on everything, but I need to start somewhere. I'll draw up a priority list. It could be that I focus on environmental design because of the project, although I would like to get back to basics for some stuff. 

I have a ton of addons that I have not yet touched. I need to install them, read up on use and give them some practice.

I need to build a good camera rig. Souternshotty does a great tutorial on his setup. I will follow that again.

Physical Addons has their Atmosphere 2. It looks like a shameless copy of TrueVFX's True Sky3. I don't make such comments lightly, but given that Atmosphere 2 includes an Earth model that supports planetary atmospheres, rather than making that functionality part of their PCO: Planets addon, seems rather twisted. All said, I will be picking up a copy after the current sale -- which seems weird but its because the sale offers a small discount, but I can get a big discount as an owner of their first version, only during Superhive sales; you can't choose the bigger discount, it's the sale discount or nothing, until the end of the sale.

Sell something!

I'm looking at putting a package of scanned objects together and selling them really cheap. Rocks, mostly. I need to process the scans:

1. Make the objects watertight -- most need a bottom.
2. Extract the texture maps from the object and store them in a consistent file structure.
3. Turn them into assets and create an asset library. I may need to read up on the best way to do this so that when shared it works as expected.

Plan B will be a series of planet maps, created from an improved workflow that comes out of next week's Halo project work.

It seems like the new lead thinks my textures were not crisp enough. That was feedback on my 32k x 16k master file!! Planets are really big; when you start getting close in, you need monster-sized files. I'm already using files that are an order of magnitude larger than most of the production team because I have a workstation-class machine to work on.

I see pixelation! More resolution!!!

I see a 21k x 10k planet map. Oh, that's pretty big. Smaller filesize!!!

Raptor - quick request

 Gabe asked me to help make scale adjustments to the previs version of a starship's bridge. Not a lot of creativity, but a good exercise. It was clear that the source model was exported from either a game engine or had at least been optimised into a triangle-based model. A horrible starting point if you want to make edits! It took a couple of hours to cut the model in half, then turn it back into quads for adjustment, then separate so that some parts could be turned hidden.

The director approved the changes but requested adjustments to some railings because there needed to be more space for characters to move around a command table. Then, while I went on my holiday, Gabe remodelled the command table a bit.

The NDA prevents me from showing anything off, but it should be pretty cool.

End of Term

 It's the final week. My final day is next Monday, but the boys will end their school year at lunchtime on Friday. 

My sister, Tracy and her youngest child have been visiting, so I've mostly put down tools and spent time with them. They're off to spend some time in the Peak District, so I will be catching up over the week.

High Bradfield on Sunday

Start of Week

Last week was not very productive. No progress made on Halo: Raptor. There was a warning from the director that we needed to create some quick-turnaround previs environments to meet the needs of mocap capture that will be happening soon. I've not received any new work requests yet. I'll soldier on with the shot002 work.

Personal project

Monorail system. Quick and fun. Happy with the results, so far. I need to finish the rail system and build a cab.

On a whim, I decided that the cab would be more sci-fi than on Chiba's actual monorail system.

Blender 5.2

The shenanigans with crashing to the desktop after installing a bunch of add-ons really threw me off my stride. I have held back on installing any more addons. I'll possibly wait until tomorrow's official release, with a final batch of bug fixes, before having another go.

Onwards into Quintilus (or Septimus)

 Harsh week. Back pain. Chronic sleep deprivation. Stresses and strains. We endure!

Diabetes

I'm off the medication, handling my insulin resistance through diet and exercise alone! Yay! Now to stay on the straight and narrow, so that my next HbA1c test continues to show me below the pre-diabetes range.

Halo: Raptor

I chatted with the director about my progress with Gaea. We're looking at using Gaea for my next scene. In the meantime, I will be doing some quick test terrains to see if I can muster the right type and quality.

I also need to finish the city. For this, I need to grab the terrain beneath the city and create a copy of its topology so I can properly map the roads onto it. Currently, much of the city's surface level is sunk below the hills.

Now, let's see if I can break out of my disconbobulation and get some productivity.

Time to get back to work

 

I feel like I have put one foot through the doorway into the Gaea terrain creation. Its node-based system and impressive selection of erosion systems make it a preferred choice for complex terrain. Not in every single case, but in general.

Spot art for a Facebook group I belong to:

True Sky 3 and some old models

Start of Week 

Last week's aims:
Complete Gabe's challenge - tick! I conceded and moved on to the next aim: learn Gaea. That's another tick! Also, refresh my Random Flow skills. Tick!

This week

• Re-engage with the Halo: Raptor scene 002 work. There's a conversation happening with the env team.
• Take another look at Random Flow -- there's a new major release out. I'd better get it installed and take a look.
• I desperately need to complete a small project. A fully-fledged but simple-ish scene.