For the last two weeks I have been working on facial animations and sounds. And it’s so complicated that I have to change the strategy to something smarter and thus change the scope of this update.

Don’t worry, there are already plenty of changes you’ll like, so it won’t be “I added a save file to the game. By the way, my game used to have save files” kind of update ;).

Generally speaking, the animations are one of the hardest parts of the game and they are done by professional animators. Working with the professional animator is a no-go for me right now as I do not have enough funds to do so.

It is difficult to find good animators. Even harder when you are on a budget.

When it comes to the face, it is probably the hardest part to animate. Now there is another problem - the voice acting looking good with model. I want the game to feel better than “open mouth then talk”.

So to have custom animations and custom voice acting I would have to hire the professional.

I asked three different VA actors if they could give me a quote for recorded facial animations with VA…

And it’s expensive. WOW, what a surprise! Also, each new scene/VA line will require extra money, so after the first batch I will have to pay even more in the future. Another challenge is that there is model incompatibility between the one in the game and the one that VA might use to record. This will require more re-targeting and probably more software in the stack (which I will have to learn!) or even extra software licenses.

To sum up:

• Animations are hard
• Facial expressions are even harder
• VA is not cheap -> using soundpacks (I bought two) might be a good idea for scenes like ASMR or single sounds, but they are not good for the tailored experience our game requires.
• The VA actress with recorded facial expressions is expensive and it will slow down the development.
• Using external VA also adds the problem of re-targeting.
• Using external VA with facial expressions will slow down the development (I work full time and have a mortgage, so I don’t have much time).

Solution:

Wow you are still here? So I found the solution. It’s pretty cheap and should work. Of course I won’t know until I try it :).

The solution is to use a face recording and then run the voice through a voice curation program. It is not a silver bullet, but it has the following advantages:

• It’s affordable - the ARKit that works with the unreal engine requires an Apple iPhone with TrueDepth Camera. I should be able to buy a used one (unfortunately only Apple has this type of camera in phones :/ - so I have to buy IPhone) - estimated cost is about 250$.
• It allows rendering and testing in real time!
• Anyone can be an actor
• It will speed up development considerably
• Oh, you want different VA in the game? No problem, it will take about a few minutes to retarget the sounds through a different VA profile/model.

It has the following disadvantages:

• It might not work :)
• Some cost

Like always you can check the status of the development on the Trello board: https://trello.com/b/irQ9RiFE/domjoi-kanban