All posts by Corey Burger

How to run gaslands (or any game) by webcam

The second wave of COVID-19 is here and with it, new restrictions on in-person events like gaming. While in British Columbia we don’t have a strict lockdown, we are limited to our “core bubble” aka our immediate household. Which means no in-person gaming. Our little gaming group (myself, Brian and our friend Sean) took a break, we wanted to get gaming again and preferably avoid fully-online solutions like TTS. Enter the webcam.

Attempt 1: Ancient webcam + gotomeeting

Our first attempt used an ancient webcam that spits out low resolution image, plus using Gotomeeting. I shared my screen, which had the default Windows camera app up on this. It worked, but was very difficult to see.

We were also challenged with how to track stats – gear, hazards, damage, etc. Brian tracked it on his computer and I tracked gear phase and hazards on the gaming board. Neither really worked – we couldn’t see what Brian was tracking and due to the webcam, nobody could see the hazards I’d placed on the board.

Attempt 2: Phone as webcam + OBS + Discord

Having tried the old webcam and realized it didn’t work, plus the need to track stats, I decided to download and play with OBS Studio. I figured I could overlay a Google sheet to track stats, plus bring in the camera feed.

To deal with the poor image quality, I downloaded #LiveDroid, an Android app that streams your phone’s camera over a local server. To help with low light, I used my new ring light that my wife had bought me for Christmas.

At the end of the third game. My phone was in the light ring holder, hence why OBS is showing a blank image right now.

Setting up Open Broadcast Studio

OBS sources

OBS is pretty easy to setup – this was my first attempt at using it and it ran flawlessly. I had four different sources: cam link from phone (as a browser source), read-only Google sheet (as another browser link), white image for a background and the Esquimalt Thunderdome branding image. Note that I didn’t run any audio through OBS, as we were talking via Discord.

For stats via Google sheet, I first setup the needed columns and then cropped the source down using a Crop/Pad Effects Filter to only show the key part of the stats. This method is easy to change on the fly, which we did when I forgot to add Gear Phase to the initial sheet.

Streaming from your phone

I downloaded #LiveDroid, a free Android app that streams your camera to a browser on your local network. You will need to set the resolution for both the camera and in OBS. I set the phone to be 1920×1280 and the OBS source to 1280×960. For me, Camera 0 was my rear camera.

I also tried DroidCam, which has an OBS plugin, but that stuck an ugly watermark on the image.

Connecting it all together

Finally, I shared the OBS window out using a virtual webcam. On the bottom left of the OBS window, it is labelled Start Virtual Camera. We then started a video call in Discord and I changed my webcam source to OBS Virtual Webcam.

If you’re the presenter, don’t be alarmed when your webcam looks like this:

Discord helpfully flips the webcam image, which works great for your face. Not so much for this.

But for everybody else, the combination works well. This was Brian’s view of the final hit of the game.

Modifying Gaslands to play via webcam

For those playing remotely, the hardest challenge is spatial sense. A lot of Gaslands is about choosing the right template. Also, play will be slower, so plan for that.

Adjusting the template rules

The touch it, use it and no pre-measuring rules from Gaslands assume you have a good view of the table, which remote players really don’t. So we changed the template rules a bit. They could ask me about what a template would look like and sometimes I would even lay it down after they’d described what they wanted to do.

Keeping the cars and game simple

In order to keep the game from bogging down, we decided to institute a few simplifications to the game:

  1. Everybody drives the same car, in our case a car with a single HMG forward facing & a mine layer aft
  2. No sponsors
  3. Easy scenario – we ran both a modified flag tag and arena of death

Have a record keeper (and track stats locally and remotely)

The person moving the cars and running the feed doesn’t have the time to deal with record keeping that others can see. So we ended up tracking the key stats both on the game board (so I could see them) and in the Google sheet. So Brian, sitting at his computer, could track the stats for everybody, leaving me free to move cars and keep the game moving

Be ruthless about whipping people along

In order to complete a game, you’ll need the person who is moving the cars to be a ruthless GM. They need to keep the game moving via ensuring each player moves, constantly asking all the key questions like “How many dice do you want to roll”.

What’s next?

Overall, the second game ran really well. OBS made it a lot simpler for everybody to see all the stats.

One of the key problems we ran into was seeing the dice. Instead of an online dice roller, I rolled everybody’s dice in person. For the next game, I plan on using the crappy old webcam as a dice cam – set it up vertically over a dice rolling box.

I also need to play with the colour balance, contrast and saturation of the webcam. Unsurprisingly, all the various pieces of software being used tried to do some sort of correction and not all of it worked.

Where to get the software

OBS is free and runs on most major OSes, while #LiveDroid is Android-only.

Links:

Esquimalt Thunderdome – a Gaslands race track

I love a big terrain project and a few of us here got into Gaslands recently. So I decided to create a racetrack. A big one, at 5′ by 3′. So off I set to do that.

Initially, I wanted it to be portable, so I did some experiments and found that acrylic caulk will peel off of wax paper. I ended up regretting this decision, but I am going to use the technique for a future project.

Racetrack laid out on wax paper
Racetrack laid out on wax paper

The track is 36′ wide and 60′ long, with the track itself being 8′ wide. All of these widths are approximate. The cars on the track are either older Hot Wheels or Lledo.

Once the caulk was added, I added quite a bit of texture with a rough caulk knife that I had roughed up on a previous project (oops). Here was another error I made – the wax paper was not fully taped in the middle, I ended up with ridges.

Fresh caulk on the racetrack
Fresh caulk on the racetrack

One note about drying acrylic – it loses a lot of water, so you’ll want something underneath your mat like another drop cloth to catch that water.

Once dry, I inked it using and then heavily dry brushed it with a variety of browns. Overall it works, but I was aiming for a slightly different brown than what I ended up with.

Painted race track
Painted race track

We used it a few times like this, but it really didn’t work that well. The edges of the mat were thin and a tear started developing in one curve. So onto a canvas mat it went, with more acrylic caulk to stick it down. Should have done that first…

Now that it had a permanent home, I decided to add some concrete pads with thicker grey caulk, mostly as an experiment. It worked well, I would do this again.

Then it was time to flock the rest of the mat. I decided a dead grass theme fit the post-apocalyptic Gaslands well. The dog decided to come and leave his mark on the mat (good thing acrylic caulk is non toxic, as it ended up on his pads too). And thus Crater Dog Paw 1 came to be.

Once the mat was fully flocked, it looks pretty good. This is the 85% stage done project – I need to cut the mat down to size and paint the concrete pads. I may also add some more bushes or other flock.

BSC 2018: Correcting some too thin errors

As I am fairly new to 3D printing, I am learning a tonne about what not to do. One of the biggest issues I have discovered is that things that look good in CAD can look absolutely terrible once printed, given the resolution of the printing or the limitations of FDM 3D printing. Today I wanted to talk about three different, but similar errors I made: making a part too thin to print correctly.

Back corner of the cab

First up – the back corner of the cab. As the backside of the cab is curved, it thins quite a bit at the very back corners.

Back corner of cab in blue. See the thin joint at the very end

When you bring this into Cura to slice it for printing, the problem becomes obvious.

Back corner of the cab, showig that when printed, it will only be two layers thick (~0.8mm)

The solution to this is to thicken the back wall of the cab, which I did by adding a flat piece to it:

Back corner corrected with additional piece (in red)

Once you bring that into Cura, you see that the narrowest part is now at least 1mm thick, so the piece shouldn’t be so weak.

Thicker corner sliced in Cura

Fenders

The next piece I tackled was the fenders. They were originally 0.25mm thick, which means that they were just over one layer thick when printed at 0.2mm and only two layers thick at 0.1mm. This meant they basically didn’t print at all.

Fender dimensions

Fender sliced, showing no overlap

The solution to this problem is two-fold: thicken the fender up to 0.5mm and also only print at 0.1mm (considered Fine quality).

New fender, now 100% thicker!

New sliced fender, showing overlap

Top of the windshield

This is actually entirely my error. I made the windscreen and frame around it taller than the sides or back of the cab. This meant when I cut up the model to slice, the top of the windscreen disappeared. Oops.

 

Top of the windscreen, showing the height difference

 

Top of the cab when sliced. That grey area is supposed to have yellow lines for printing

BSC 2018: First print of 1956 PANG

The new printer board for my 3D printer finally showed up, so I got to printing the first print of my 1956 PANG. It was less than a full success. As you can see in the pictures below, I have a bunch of work to do.

Printed truck on build plate. Printed without the side compartments. Excuse my messy garage

As can be seen on the build plate, a bunch of details didn’t come out right – the biggest of which was the top of the cab for some reason completely failed to print correctly. I also accidentally selected brim intead of skirt, so I had lots of cleanign to do.

Front & side view. Lledo Model A, Copplestone Chinese and Infinity Ariadna for scale

The sides of the truck aren’t very smooth, and a bunch of the finer details simply didn’t print.

Side rear view of the truck
Side rear view of the truck with Lledo Model A, Infinity Ariadna and Copplestone Chinese for scale

Front and side view of the truck

So what next? First of all,my printer needs some upgrades:

  • A part cooling fan (the Ender 2 lacks it by default)
  • Belt tighteners (and the belts replaced with metal-core of some kind)
  • A more modern Marlin (the firmware that runs the printer)
  • A lot more tuning

On the truck, the first thing I am going to do is shrink it by 25 to 30%. It is too large for what I want. Then I need to some work on the model itself

  • Thicken some of the walls so that they print, especially the fenders, which didn’t print at all
  • Clean up some of the smaller details, possibly removing them for now
  • Re-cut up the cab so it prints correctly

Anyway, I have my work cut out for me!

BSC 2018: PANG print layout

I finally got around to laying out the 1956 PANG for the first printing. As you can see from the pictures below, I am trying to remove the need for supports which will mark the surface of the final print.

Final draft for reference
Final draft for reference

Laid out in TinkerCAD for printing
Laid out in TinkerCAD for printing

I chose not to break out the rear section or the bins for initial printing. I likely will in the final version, as it will speed up print time and make it cleaner.

Cura print layout
Cura layout, with approx. print time. Note that my printer (Ender 2) has a fairly small print bed, so this will fit on it

 

One final note: I am certain that there will be tweaks needs to be this and I am missing things, namely an interior and side mirrors to start.

BSC 2018: 1956 PANG first printable draft

Build Something 2018 continues and I have been pretty quiet for the last week. Part of this is because I am waiting on the warrantied printer board, so motivation is low. Also been busy with other things. But today I kicked myself into gear and go the first printable draft done of the 1956 PANG. Take a look below (the red rectangle is 32mm high and is there for reference):

Front view of 1956 PANG truck
Front View

Front angled view of 1956 PANG
Front angled view

Rear view of the 1956 PANG
Rear view of the 1956 PANG

Side view of the 1956 PANG
Side view of the 1956 PANG

Rear angled view of the 1956 PANG
Rear angled view of the 1956 PANG

 

BSC 2018: 1956 PANG continued

Well, Build Something 2018 is well underway at Lead Adventure. You can follow all the entries, including at least two other 3D printed things, over on the subforum.

And what of our 1956 PANG tricycle truck? Yesterday saw a lot of work on the back end.  As of two days ago it looked like this:

I really didn’t like the aspect ratio of the truck – height, etc. So I completely rebuilt the back end so it was taller and thus looked narrower:

Early work on 1956 Pang

Today I started working on the front end, which is a lot harder. It has a lot of curves with curves, so it going to be a real challenge to do well. I am not overly happy with the look (too wide and squat) so might rebuild the front. Here it is in exploded view.

Front end of 1956 PANG, exploded view on right
Front end of 1956 PANG, exploded view on right

What I love about modelling this way is that if you don’t like something, it is trivial to rebuild it. Play with the ratio, etc.

An evening in the air at Trumpeter Salute 2012

Trumpeter Salute, the best annual miniatures convention on the west coast of Canada (sorry, Gottacon. Your relentless focus on GW turns me off) is happening again this weekend and I am again reminded of why I come to conventions: to be inspired.

This evening found both Brian and I playing aerial games, although he had considerably less success with his World War One 1/72 than I did in Leviathans, where I fairly handily destroyed a British aerial cruiser and destroyer. Our third member, recovering GW player Sean, was evolving. Or something involving tails. I didn’t ask.

And the spending has begun. I picked up some Saxons for my eventual Dux Brittanium, Sean some ECW/30 Years War and Brian also managed to spend a great deal of money. And we haven’t even hit Pulp Miniatures yet.

But now, pictures:

Leviathan cruiser on cruiser action

German World War One aircraft, 1/72 scale

British World War Two aircraft, 1/300 scale

More photos tomorrow, including of Brian’s RCW game.

When the Chinese met the train, a RCW game

In a cavalcade of firsts this Sunday, my (mostly) freshly-painted Chinese met their match in our new armed train, all being played on our new gaming mat, some 6′ by 9′ of fully flocked canvas drop cloth.

The game, a preview of Brian’s Trumpeter Salute convention game, was  the defense of a small Russian town by a combined White and Chinese force. Thankfully for the Whites, they had managed to scrounge up a field gun plus a machine gun. The wealthy Chinese warlord Wu Do also brought one along to the party. Facing against them were a large Red force, including the newly-built armed train and the armoured car.

Opening shot. Red armed train the back, Whites in the foreground. Chinese are out of shot to the left.

Due to a lot of luck, the Chinese ended up being the central troops in the game. They had lots of officers and their cards kept coming up, so they quickly occupied the central building and started taking fire. Another squad rushed up a hill, only to discover what happens when a HMG opens up on you when you cannot respond.

The maddest moment of the game came when Sean’s Reds cavalry attempted to run down the White field gun. The gun was right in the middle of town, which meant that his close assault was rudely interrupted by two squads of Chinese, including the Warlord’s aggressive Household troops. This did not end well for the Red cavalry, which were slaughtered to a man (and horse).

Red cavalry attempt to run down the White gun. Chinese infantry and Wu Do in the foreground.

By the end the train had done its job and had cleared a route into the town. We called it when it became clear that the Chinese would retreat (having no interest in defending this particular muddy bit of Russia to the death) and the Whites were too weak to hold back the largely untouched Red infantry (even if they couldn’t move due to bad card luck.)

(As usual, we used Through the Mud and Blood from Too Fat Lardies. An excellent rule set all ’round)