The Mozak Community reconstructed a phenomenal 60 neurons in 2018. What an incredible year it’s been. Thank you to all our phenomenal players! Let’s see what we can accomplish together in 2019!!!
(01/29/2019) Hi players, in our latest updated we have introduced quests in the game. These are tasks that you can complete during your tracing session. At the moment they are primarily used to introduce new users to the game during the tutorials but hope to make it more engaging for more experienced users in future updates. We also have a handful of bug fixes for some long standing issues. View all the updates here in the changelog.
(01/07/2019) Hi players, our first release of the new year has some improvements to the Connect-the-Dots tools to simplify it a bit more. We’ve remove the notion of having to click on a node again to ‘commit’ a trace. In addition, ending a segment on an existing node will no longer have that existing node be selected. We hope this changes make using Connect-the-Dots a bit smoother by reducing the number of clicks needed to select/unselect nodes. There were also a few bug fixes, all the changes in this update are in the changelog here. Happy tracing!
(12/24/2018) Hi players, the latest update was mostly modifications to the code to prepare for big new feature updates in the future. However, there were a few bugs fixes and ui tweaks you can view in our changelog. Thank you for playing!
(12/11/2018) Hi players, we have released recently released an update to the game. The primary changes involved overhauling the tutorial to more gradually introduce some of the tools and have better positioned tooltips and making comments have a solid fill so they are easier to see and click on. See the full changelog here.
Connect-the-Dots got a user experience makeover! In preparation for releasing Mozak on touch devices (coming shortly), we needed an alternative to the right-click button, and make the interaction more streamlined. There is no longer a notion of “finalizing the trace”. You simply add an new edge from the current node by clicking on a new point.
We also hope that the new user interface will encourage people to grow the traces from existing traces rather than starting a new branch out of nowhere. You can still start a brand new branch by manually unselecting a current node highlighted orange, and then clicking on a new location. As an added bonus, now that right click is not used, you can rotate while in CtD mode!
See the changelog here for all the updates. If you have suggestions to change how new interface work please let us know!
If you had given up on Virtual Finger in the past because it didn’t do what you intended, or created some weird traces, you should give it another try!
We invented a brand new algorithm that is super fast, does the right thing when user makes small trace errors, and handles intelligently erroneous traces in empty space. It should work on any connected trace in the volume. It will intentionally not trace over empty space because it is specifically looking for continuous portions of the neuron. For traces over neuronal signal, it is very accurate even when a person’s trace is not. In fact, we think you can now trace even with your toe. We should probably call it Virtual Toe. Now, it’s still new, so we may have missed something. We challenge you to find any connected trace where Virtual Toe creates a weird trace. Special callout to all who find a confirmed bug.
Rendering performance is also drastically improved, allowing the game to work on computers with lesser graphics capabilities. Rendering the view is now done in two separate layers, where the volumetric image (3D image) is rendered independently from the “geometry” (edges, nodes, hotspot boundaries, comment boxes, etc.) One nice consequence of this is that we no longer need draw the geometry at 1/3 resolution when we render volume at 1/3 of resolution (e.g. during rotation). Previously, that meant that the entire view became blurry while zooming around, including geometry and text. Now dropping down to a 1/3 resolution is almost unnoticeable. Moving the mouse around, hovering over nodes and comments, virtual finger, deleting nodes and branches and any manipulations that do not require moving the camera (zoom, rotate, pan) will now be much more responsive.
We also found a rendering issue that was washing out the neuron signal. You will notice that all volumes now have 20% more contrast in the image. I bet you’ll be able to see more faint signal now than before. Play with the contrast slider, making it brighter will bring out more signal than in the past.