November has come and gone in what felt like record timing. The Steam Autumn/Fall sale is underway and I threw in Airship Dragoon for the Steam Awards category "Labor of Love", due to updating it after nine years on Valve's platform. More on that available on Steam.
November was a blur of getting the dreaded lurgy - which seemed to take forever and half a million hot toddies to clear, and testing, testing, testing ...
Testing many, many things ...
Such as trying to lower the polycount on Monsters Loot Swag models. Previously I had used an automated plugin to reduce polycount for Level Of Detail, however this made a horrible mess of UV mapping. So I decided to try Blender's own decimation tools such as "UnSubDivide" - which would surely reduce edgeloops by simply unsubdividing the mesh ... yeah, well, actually it didn't ...
 |
| Unsubdivide edgeloops ... unsubdivides right? |
Almost like it does one half of the body fine, then gets bored and just compresses the edgeloops and swaps to the other half of the body.
I am not a fan of retopology because I do not understand why it cannot be automated perfectly - especially on a model with perfectly equal edgeloops. This is the exact sort of soul destroying donkey work which should be easily automated and technically is automated if you want it to be a shyte ...
I did try a number of other programs - which I believe cool kids call "apps" these days on the grounds any word over four letters is too difficult or energy consuming to utter. One of these was MeshLab which seemed to do a fairly decent job at reducing 25K to 7K, though I still felt should have saved the UV map better. Or I could just manually retopo, but as I have avoided doing that for years I don't really want to start now ... especially as getting mesh deformation to not be shyte seems to be even more of a pain in the arse ... see below.
In addition to polycounts there has also been a whole host of messing around with the newer aspects of the engine which has been the focus of the Monsters Loot Swag BETA test build. This has been a lot of testing, fixing, streamlining and bug reporting, all of which are slowly but surely going into improving the engine and game build, though at the moment there is nothing too concrete to show as a game update.
But back to low poly.
 |
| Low poly Lara with some good low poly deformation techniques |
I started making a low poly character, in a somewhat stylized plane based style. Nigh on 30 years ago I actually sculpted this type of head in concrete with a metal frame made of metal coat hangers... which proved to be a mistake when attempting to move it anywhere as I had not made it hollow, and so it weighed a ton.
Low poly mesh deformation is somewhat more of a pain in the butt than normal. Got a bad deform when bend bones? No problems, slap more edgeloops in and jiggle them about a bit. Except this cannot be done when you are modeling a plane based low poly humanoid.
 |
| Back of the elbow needs to come out somewhat |
The simpler meshes get, the more horrendous deformation becomes, and the more difficult retaining versus collapsing polys are.
This is all being done for two purposes, one is for a potential low poly with hard edges/planes style future project, and the second is for more practise in reducing topology whilst keeping good movement deformation for current work.
In other news, it's cold and damp. The Pentalon Javelins - that's a type of winter spud that is harvestable at Christmas, and not a 1950s RAF experimental supersonic jet fighter; though it really sounds like it should have been - appear to have died a death, going the way of this year's strawberry crop under the chronic wetness. I've been eyeing up designs of a padded gambeson to try and survive winter and also padded armour may help survive the GREAT HAPPENING when the lights go out and society crumbles.
December approaches ... and don't even get me started on Space Year 2024.
