Sunday 30 September 2018

Friar Tuckish

Here we have yet another of Reaper's plastic Bones miniatures, from one or other of their Kickstarters. I don't know the SKU.

He's a stout fellow, in every sense of the word, but I get the feeling he might be a bit of a bastard if you cross him.

This guy is painted as a more or less mediaeval European mendicant friar, though the weapon is one the Japanese call tetsubo, an iron-bound club-staff. It was notoriously used by some of the militant Japanese Buddhist monkish sects for killing people without (theoretically, though probably not actually) spilling their blood. There was one sect who carried their sacred bits and pieces in a casket into battle, and threw it into the enemy lines to inspire the monks to a fanatical frenzy to get it back. They were, let's face it, completely barmy.

Thursday 27 September 2018

Pit Howler

This is the Motleyverse Kickstarter freebie STL that I posted about a few days ago. It's been printed at 0.08mm in no-name PLA.

I've replaced the file's original thick disc base with one of my own 25mm paving bases.

It's not a stellar print; the thing's front-left paw is completely missing, as Cura decided it didn't need any supports... which it clearly did, and the mouth detail is a bit messy.

The bony spikes on its back ended up looking a bit more like candy corn than I had intended.

Ah well, it will just have to do, as I can't be bothered taking any more trouble with it.

Tortoises – Finishing Comparison

Clickupon to enlargenate
I decided to paint the Tortoise I printed most recently, the one printed with tree supports rather than zig-zag verticals, without any surface finishing at all apart from string removal.

Both models have been printed with identical quality parameters: 0.08mm layer height, 10% triangular infill, and printed fairly slowly. The one on the right, the first, has had many of its surfaces scraped down to remove as much of the printing layer lines as I reasonably could. The one on the left, the newer model, has not, but has just had paint applied straight over the surfaces as printed.

There is certainly a visible difference between the two models when seen in close-up, but with paint on, at tabletop distances, the variation in surface quality is pretty much invisible. For purely gaming purposes, I'm doubtful that the extra time and effort involved in smoothing down visible layer lines is really all that worth while.

Tuesday 25 September 2018


 This is the Tortoise, a British WWII super-heavy assault gun. Half a dozen prototypes were built, but its main intended purpose was smashing through fortifications like the Siegfried Line, and by the time it was ready, there was no longer any need for it.

This model is 1:100 scale (15mm), and was printed in PLA+ on my Ender-3 FDM printer.

eSun PLA+, 0.08mm layer height

I printed a second one, using Cura's tree supports, just to see how it would do. Apart from the stringing which seems to be an inescapable feature of this filament (brand unknown), it printed fine for the most part.

However, I cannot figure out at all how Cura's tree support placement algorithm works. You can see, under the gun barrel and along the bottom of the side-skirts, supports are clustered together in dense groups, and then there are long bare gaps.

With my printer in its largely stock setup, I can just bridge those gaps, but not completely cleanly. If the same number of supports had been spaced evenly along the length of the area to be supported, it would have printed much better.

Now, I'm impressed with anyone who can figure out this sort of programming wizardry to begin with, but I don't think it can be denied that the algorithm could do with a bit of refinement. It would be even better if I could manually add or delete supports in Cura; it's possible in another slicer app (Solidify3d), but that costs quite a large number of yankeebucks, so I'm unlikely to get a copy unless we win the Lotto.

Sunday 23 September 2018

Magic Missile d4

Here's another design for a 4-sided die, this time marked from 2 – 5 for use with the D&D Magic Missile spell.

The file is on Thingiverse at

Romanesque D4

These are some 4-sided dice I designed a while ago and just got around to printing. I've stripped them off their supports, but haven't done any more cleanup than that so far.

I wanted them to look Roman-ish, so apart from using Roman numerals, I've tried to echo elements of Roman architecture in their shape.

They seem to roll adequately randomly, from what I've seen so far. I'm working now on refining some "baked in" supports (see below) so that I don't have to rely on Cura's occasionally inexplicable support placement.

They're a fraction more foot-friendly than the caltrop-tetrahedral d4s, but they've still got some pretty stabby points on them

There's a printable STL file available for download at . The fin-like structures at the bottom are supports which will need to be stripped away after printing. I had to enable "Print thin walls" in Cura as they're only 0.4mm thick, and with my 0.4mm nozzle Cura wouldn't slice them reliably without that setting enabled.

Friday 21 September 2018

Ganging Up

When I'm printing things like the Printable Scenery ruins I've mentioned before, I first create and re-size the assemblies I want from the individual component STL files, and arrange them into a compact group, using Blender. Then I export a single STL of the whole group.

When that STL goes into Cura to be translated into a gcode printer file, I'll fill up the platen with as many copies of it as I can fit, so I'm printing multiples at once rather than having to restart the print job over and over. Laying out several models for the same print run is called "ganging up", it's a very old term that comes from 2d printing.

Of course, with an FDM printer this increases the printing time in a 1:1 ratio; twice the models means twice the printing time. It doesn't really save me any actual time, just the faffing about between print runs. On the other hand, a DLP resin printer, which uses an LCD screen to project and expose each layer sequentially, the amount of stuff on the platen has no effect at all on the print speed; it runs at a set rate per millimetre of height, so it pays off, in terms of printing efficiency, to cram things together.

The down-side to ganging up on the platen is that if for any reason the print fails, then you've lost  at least twice as many models and wasted twice as much filament (or resin).

In this picture, I've got two iterations of the STL I exported from Blender. That pretty much fills up my platen, though I could possibly have put some other smaller models in the corners. Not this time.

More Corporate Gouging

Well, bugger.

It appears that PayPal now want me to pay a seven dollar fee on each and every single transaction I make using them. That seems rather excessive to me.

They used to be a convenience, but I guess now it's time to look for other means of transacting business online.

Wednesday 19 September 2018

Paint Test — 3d-Printed Ruins

This is my first painting test for the ruinous bits and pieces I've been printing, from elements by Printable Scenery.

My focus has been on speed of production, to get them on to the gaming table as fast as possible, rather than creating gorgeous diorama-quality miniature models. To that end, they've been painted pretty much entirely by means of washes and dry-brushing over some zenithal priming.

At this standard of painting, they will suit my needs fine. I'm hopeful of having a table-covering mass of urban ruins within the relatively near future.

I have been tossing up whether or not to mount the individual pieces on some sort of base. They are sturdy enough as is to take some fairly careless handling, so they don't actually need basing. So any basing would be more for aesthetics, rather than for structural stability.

Having some debris scattered inside the ruinous shells is appealing, as without it I suspect they'll look more like obstacles on a paintball range than an actual town that has had the shit knocked out of it. The trick will be to make it look enough of a mess without compromising the ability to move troops around inside.


I added an internal base to a complete building assembly, and as I suspected it does look much better.

It does add considerably to the preparation time, but it doesn't need to be done at the same time as the painting of the shell. I can do the basing progressively as I have the time.

Thursday 13 September 2018

The Heartbreak of Change

Experimental Chaos Blob Thingy
Blender 2.8 is peeping over the horizon, and I've been playing with its Sculpt mode in an Alpha daily release.

I can see that, if I can get used to it, the UI could be an improvement. However, many of the keyboard shortcuts I've come to know no longer work, and finding stuff that used to be right there on the tool-bar is proving to be rather a trial.

The new render engine, EeVee, looks like it will be amazingly useful, but again things have changed enough that when it comes to basic lighting and rendering I'm completely lost now. I couldn't even figure out how to grab and move a light; I had to transform it using the tool-bar. I couldn't figure out how to light and render my Thingy at all, and ended up just taking a screen-grab of it. On the plus side, it's a much better screen grab than I would have got before. I think.

More fiddling with it, and
another screen-grab
I don't think I'll persist with it until it's actually in Release Candidate status. Things are changing around enough that time spent relearning how to operate the program will most likely be largely wasted until things are set in stone. Once it is released though, then all the multitude of YouTube mavens can set to work teaching us Ignorant Ones how the New and Shiny is better than the Old and Busted.

Printable Scenery Scenery

Printed at 0.2mm in PLA. This model took a couple of hours.
There's a company called Printable Scenery who sell... well, printable scenery. For tabletop gaming.

One of the many sets of models they offer is the one this came from (actually a combination of two of the pieces from the set) — their Modular Town Ruins.

They're designed with sockets to be able to clip together with OpenLock clips, but I prefer to just combine the models during printing to create a single piece. Being able to mix and match bits on the table is fine and dandy, but I'm just as happy to print a whole new ruin if I need one in a specific configuration. There's also the issue that I wargame primarily in 15mm, so the models need to be rescaled from their base size to fit in with models and figures of that size — the clips might still work, but it's an added complication I don't care for.

I've resized these two bits to 60% in Cura, and arranged them to print together. They seem to fit pretty nicely with the 15mm figures I've photographed them with. I could maybe have taken them down to 50%, and I'll give that a go, but I suspect they'd then start looking a little small relative to the figures and what-not.

You get a good few different bits for your nine yankeebucks, and with these I should be pretty well set for 20th century building ruins for the foreseeable future.

The main feature that's missing is floors, for creating two-storey sections. However, that's not something that would be terribly difficult to create myself, if need be.


15mm figures from Battlefront, PaK36 and crew from PSC.
Shown here to the right is a more ambitious assembly from the same collection of pieces.

This one took about eight hours to print, which seems like a long time, but I can put something like this on the go in the evening and have it waiting for me to pick up off the printer when I wake up.

One thing that I've noticed from this exercise is that one does have to take note of which side of a wall segment is an inside or outside face; the piece in the top left, for example, should probably have been the other way around so that the fireplaces were actually inside the building, where they could have done some good when it was whole and inhabited. However, many of the pieces are fairly ambiguous, and you could get away with putting them either way around.

Wednesday 12 September 2018

Filament Reel Roller

I wanted to get the weight of the filament reel off the frame of my printer, so I made this proof-of-concept reel holder, using Wood™, and the Power of Magnetism to hold it to the filing cabinet next to my workbench. I originally intended to use it as a test-bed, and make a better-finished one at a later date, and that might still happen, but I suspect inertia will take hold and I'll end up just using this one forever.

It works very well, except that I failed to take into account that although the magnets I inset into the back of the frame cling very strongly to the steel cabinet, they slide very easily on its smooth surface. At the moment I have it propped up on a cardboard tube, but I'll screw a little tab to its top to hook over the top of the cabinet, and that should secure it without any other outside aid.

The 3d-printed spool runs very freely on a pair of skateboard bearings, and they should take a load off the extruder gear. I haven't noticed any issues with the reel just running on a fixed axle, as is supplied with the printer, but there's no harm in relieving any potential stresses when one can.

I built the reel-holder so that it could also be set down on the table, and the reel mounted on top, on what is now the front edge. Whether that will ever be necessary, who can say, but the option is there if I ever need it.

I suppose I could have 3d-printed the whole thing, but that would have taken considerably longer than it took to knock this up out of scrap wood.

Monday 10 September 2018

Half-Finished Howler

This is a free giveaway model from a game Kickstarter, the Motleyverse. It's called a Pit Howler.

It's an OK creature design, but the model is ridiculously over-engineered for printing. In its downloaded form, it weighs in at over 800,000 faces, and about 45 MB. About five minutes work on the .STL by me in 3d Builder reduced that to 66,914 faces and 3.14 MB without appreciably affecting the level of detail that would be printed.

3d Builder also identified, and fixed, a large number of structural issues with the original model. It's not impossible that it would print properly without fixing those issues, but it's a matter for concern that they were still there in a commercially released model.

I'm not one to look a gift-horse in the mouth, and this model was free after all. But the issues with it do not inspire me with confidence for the rest of the models in the Kickstarter.

Wednesday 5 September 2018

Stone Bridge

I finally got around to printing and painting the stone bridge I designed a couple of weeks ago. The STL is at, this is the version with the cobbled road-bed.

I printed it at 0.2mm in eSun PLA+. I got a bit of underextrusion here and there; I suspect the filament might have been getting a bit moist and/or dirty.

The bridge is intended for 15mm wargaming, and in that scale it's just large enough to take a truck or a light tank. Certainly nothing larger than a Panzer II.

Monday 3 September 2018


Here's another 3d print from a model by Duncan Louca. This one's called the Fleshwalker, and it's huge — I've printed this model at 50% of the base size.

I printed it at 0.08mm in some grey no-name PLA, and mounted it on a 50mm fender washer.

Sunday 2 September 2018

Bare-Nekkid Hill Giant

Here's one of Duncan Louca's hill giants, all finished up.

It's been re-scaled to 66% of the base model size, and printed at 0.08mm in eSun PLA+.

I'm never really very comfortable with painting vast areas of bare flesh, but overall I think it turned out OK.

I designed some round bases, intended to be combined in a slicer with a baseless figure to be printed as one, though of course they could also be printed separately and be glued together. The links in the image captions will take you to the Thingiverse page where the files are located.

So far, the diameters available are 25mm, 36mm, 40mm and 50mm. They're available either with a simple generic undulating texture, as shown to the left, or with a rough paving texture, like that on the right.

The average thickness is about 2mm, I guess.

I neither know, nor much care, what the official critter-size bases are in D&D 3 to 5, but if that actually matters to you it should be a simple exercise to resize one of the existing bases to the exact size you need.

Note that in Cura, if you're combining a base with a figure model, you will have to go into the program preferences and turn off "Automatically drop model to baseplate", or else any figure you put on top of the base will be buried up to its ankles. Be sure to zoom right in and make sure that the soles of the model's feet (or pseudopods or whatever) overlap with the surface of the base so that they do actually print as one piece.

Saturday 1 September 2018

Unusually Tough Supports

With Cura's tree supports
This is one of Duncan Louca's models, a hill giant, printed in eSun black PLA+ at 0.08mm. It was about an eight hour print.
After support removal

I re-scaled it to 66% of its base size, as it was originally much too large for my existing 28mm roleplaying miniatures, and I added a 36mm diameter base. It's still taller than an official D&D hill giant should be, but at a true-scale 10' tall, they just don't look giantish enough to me. This guy is more like 12' or 13'.

I used Cura's experimental "tree" supports, which I favour for miniatures like this because they tend to leave less mess behind than line or zig-zag supports, and normally they're easier to remove. In this case though, removal was a real chore. The trees are usually hollow, but these ones were filled with zig-zag infill, which made them very, very strong. Too strong. I've clearly checked a box or changed a parameter some time in the past and neglected to change it back, and I have no idea now what that parameter might have been. There are a lot of options in Cura.