Fish! Update - Part 2

In this second part of my update on our collaborative text adventure livestream experiment, I explain how we're producing the experiences, both technically and logistically. Ultimately we'd love for other groups of people to do this as well, and I’ll cover this in part 3.

How are we doing this?

In Part 1 of this update, I explained the 'why?' of all this, but I imagine there are people who are interested in the 'how?' as well. Here's a very brief breakdown of what we are actually doing in practice:

  1. Firstly, I am running the game and streaming to YouTube from my desktop computer, while all the other players join the same Zoom call from a variety of devices.

  2. We are playing a cracked version of the original Commodore Amiga release, running in the WinUAE emulator set to mimic a stock Amiga 500. We're actually using a slightly out of date version of WinUAE as we had flickering text display issues with the latest release. The version we use is 4.40 and it's been rock solid so far.

  3. We're also mapping the game as we go, using Trizbort which is an excellent mapping tool designed specifically for text adventure gamers and designers. There will be a wider discussion about live mapping and the best tools to use in part 3, but currently we're using the latest native Windows version of Trizbort, and I have set all the map colours to 'dark mode', as the game itself uses light text on a dark background and the contrast would be too glaring when switching between the game and map if I'd left the defaults.

  4. The streaming software is StreamLabs, configured with a custom scene and all the graphics, screen captures, overlays and audio inputs, and we stream to a scheduled YouTube id, which is set up in advance on my DeadFleshRetro channel. If anyone is really interested in all the settings, do get in touch and let me know - streaming is always a bit arcane, and the settings I have appear to be working ok and that's good enough for me :)

  5. As I mentioned, we’re using Zoom for the video call between all the live players. This works pretty well I think, although not everyone has a professional recording setup and I have had a few audio editing issues - I’ll talk about this some more below. 

  6. I am running all this on a fairly beefy desktop machine with two main monitors - the left one has the game and map, the right one the streaming software and chat monitoring - and I also run a small tertiary screen underneath the two main ones which runs the Zoom call in speaker view, which is what is shown on stream. In addition, I have an iPad set up which is also connected to the Zoom call, but has the view set to gallery mode so I can see everyone who’s joined the call (or arrives late, or leaves and rejoins for whatever reason).

  7. The players are all watching both the Zoom call, and the Livestream. There is of course a fairly significant lag issue, sometimes as much as 15 seconds between me typing into the game and the players seeing it via the livestream. There is an option in the streaming software to present the stream as a ‘virtual webcam’ and show it in the Zoom call, which might reduce the lag a little bit, however I can only join the call as either myself, or the stream, so I would have to show my own webcam separately which would mean that the conversation between us would be subject to lag as well as the game!! So, unless we figure out a better way of setting everything up we’re just going to have to deal with the lag and at least with a text adventure latency isn’t too critical. I do occasionally have to reduce reaction pauses for the podcast though, but this isn’t too onerous. The agreement we have is that when new text appears, I start reading it immediately, and as I’m reading it the stream the players see catches up and they can skip to the text I’ve then gotten to, to read along. Occasionally listeners might hear players telling me to “just read it!”, this is because I’ve reacted to some new text before they’ve been able to see it and I should just read it out so we can all react to it together. (Does any of this make sense? These are the logistics we’re trying to iron out 🤷😃).

  8. As well as player reaction time being affected by lag, it is also affected by the need to keep Zoom on mute for most of the time so that player background audio, (police sirens, dogs barking, etc.), doesn’t interfere with the recording. However, players need to remember to unmute before reacting to something (especially something funny), otherwise we don’t get their laughter at all. This has made some of the jokes appear as if they are falling flat - on screen I can see them laughing. We’re not sure whether unmuting with push-to-talk is really going to become second nature - it hasn’t yet. It’s quite unnatural.

  9. I should probably also say something about switching between the game in WinUAE and the map in Trizbort. As long as they are next to each other in my open apps list, I use Alt+Esc to switch to the previous app in the list, and Alt+Shift+Esc to switch to the next app in the list. These are standard Windows shortcuts. I just have to remember which is which (and I did get extremely confused during episode 07 where I showed a screenshot of the game’s boot sequence and it threw the app order out of whack!).

  10. Finally, I edit the audio for the podcast version using LumaFusion on an iPad Pro, which is really designed for video editing but works just as well for audio of course and it’s a fantastic product. I think it’s quicker than using a desktop app, as the Apple Pencil is very precise and quick to select, cut, extend, reduce and move chunks of sound along the timeline. I probably spend too much time ‘fixing’ the conversations at times - it takes me about 4-5 hours to edit approximately 1 hour of stream audio into a 45-50 minute podcast. I then upload it to Spotify for Podcasters, add thumbnails (created in Canva), and have set it to syndicate to all major podcast platforms via RSS. 

Anyway, that’s about it in a nutshell. The logistics are fairly straightforward, and it does appear to be working pretty well. There are definitely still improvements to be made though, of course.

In the next part, which I’ll post tomorrow, I’m going to reflect on some things we’ve learned from the experience so far, and discuss how we think other poeple could get involved as well!

Meanwhile, please check out the livestreams and the podcast, and do join us live if you’d like to help us figure out how to stop those dastardly anarchists, the Seven Deadly Fins!!

Previous
Previous

Fish! Update - Part 3

Next
Next

Fish! Update - Part 1