#comp770

Today was testing day, so we got people from other projects to come help test our vr experience (also Stefan even though he had seen it already).

Zhe testing on my android phone, good visuals but bad sound.

Patrick testing on Kent’s android with Kent’s headphones, bad visuals but good sound.

Here’s the people that tested it, the time they took and any feedback they had: - Zhe, approx 2:30, DNF because my phone died whoops - Stefan, 7:32, got stranded on the top of the hill before he reached the fern - Liam, 2:38 - Ashton, 4:02, he suggested instead of teleporting having the player slowly move and stop when they get to the object and have a voiceover explain the information - Yemon, approx 7:15 - Patrick, 3:48

And here’s the survey results:

I died laughing reading these, it shows that while they did retain some knowledge, it wasn’t impactful enough as their repsonses are...limited.

There were 5 models in total. The fern, kiwi, white kiwi, fantail and pou whenua. Due to the teleportation, it takes forever to move around the map (which is also sort of the point) so I expected that not all of them would be discovered.

I was hoping to get results about it being calming so I’m pleased there was a bit of that, but I do understand the not being able to read the signs as they may have been a bit too out of reach.

Very pleased with these answers, and I hope Kent is pleased too since he was the wizard behind the sound.

I put in this question knowing what the results would be but I had to confirm it. It was difficult trying to find what was the best method for teleporting.

Would have liked more responses but simply ran out of time. Even with a few people I feel it does give a good indication as to what others would think.

With the VR assignment due this Friday, I’d be lying if I said I wasn’t panicking.

Low Poly Fern

Made two attempts at a low poly fern; on the left is triangle planes, exported as obj and textured in unity, on the right is rectangle planes with image as texture and exported as fbx into unity.

Had a small breakdown when Kent imported it into unity and it lost its texture, so I spent a good 2 hours trying to fix it. Turns out that as a fbx it still had its texture, but it was embedded so I had to “extract embedded textures” in the asset inspector in order for it to get its proper texture.

Low Poly Fantail

I know we are supposed to be focusing on functionality over “prettiness” but its the visuals that I’m better at, I’d rather not have to do the scripting but since our team is little I kinda have to do a little bit at least. I feel like I've hardly done any work.

My search history is full of “unity how to...” and I still don’t have a clue what I’m doing.

We only were just able to start testing on my android phone this week, and the first one crashed so we only managed to get it it work when all of the trees we painted onto the terrain were deleted. Kent had to painstakingly place low poly trees around the environment, but its not that “forest” more like a field. Then the sound wasn’t working. Turns out the volume was off. End me.

Feels like we’re going backwards down mount everest on a bobsled that’s spontaneously combusted.

Kiwi Model

I got this kiwi model from online, the kiwi is based off Manukura the white kiwi.

Quickly realized that whenver I tried importing it into unity it would lose the texture and was time consuming trying to then redo the texturing in unity. The problem was because I was exporting as an obj, and I needed to export as an fbx so that I could also export media (texturing).

Kent finished the teleporting function, just a standard stare at something, timer on the ray and then it puts you in that spot.

Here’s a pretty photo Kent took from in unity:

Fern

Environment

Kent is currently designing the environment + pathways, this was just a practice.

Pou Whenua Model

Here is my terrible attempt at making a pou whenua (basically a maori totem pole) thats based off the one in Orewa. I read the story behind it too, although it was quite sad. Her husband was a sailor and died so she would look out at the water, and later died of a broken heart.

Still don’t have the student account so I can’t work directly with Kent, but I got the google cardboard off him so I can try test things out on the uni computers and my computer at home.

2D vs 3D sounds

2D is used for music and background noise

3D for anything that is directly a part of the environment

Questions to test

If the character is teleporting will the sound immediately cut, fade out and then back in again, or somehow blend between locations?

Should we implement the sound getting louder if there are creatures on the ground so the user ducks down to hear it better?

Do we use faint background music, and if so do we make it diegetic (like coming from a radio) or non diegetic (coming from the environment)?

We can increase the roll-off of the object that makes sound so that the user isn’t overwhelmed by so many sounds at once.

Have to consider how many sounds we use due to using a phone, need to not overdo it.

Utilizing the spatial sound design (position, direction and shape) if we have birds so the sound comes from the front of the bird and you can tell its coming from above the user when the user walks under it, as well as water. Ambisonic for wind and tree sounds.

Personal Reflection

Sound is such an important part of our VR so it’ll be a main focus to start testing as soon as possible.

Holiday Work

My role in my team for VR is mainly doing the 3D models, so I made a test 3D model of a bracken fern over the holidays. As usual, I am usually kept occupied over the holidays (as well as hardly being motivated) so I didn’t get anything else done other than the singular leaf.

General Principles of VR

1. Be aware of the limitations of VR The main thing I overlook because of my lack of awareness of what is possible in VR. It is better to do something little well than doing something well...little. 2. If something looks interactive, it has to be interactive Definitely do-able for our experience - making the interactive models more detailed and keep any other models basic. Making sure the menu works as well. 3. Always keep comfort in mind 4. Be gentle with your difficulty curve and pacing 5. Work around having no camera control 6. Focus on player movement 7. Choose your locomotion mechanics carefully 8. Be consistent with level metrics 9. Use positional audio to keep players immersed 10. Choose your technology carefully 11. Think multiplatform from day one 12. Always think about your performance targets I take this as no MegaScans for our experience, due to the high resolution of them and its better to have something simple and comfortable than intricate models. 13. Don't underestimate testing Self explanatory - we will test as often as we can.

(Dealessandri, 2020)

Next step is to “block out” the environment in unity to test the functionality first, rather than the designs of the 3d models as it would be higher risk to having problems further down the line.

References: