First impressions

Finally, the box is here!! :smiley:

Now, while the first setup was a little bumpy, I am sure all this will get better quickly. Here is what I found:

  • my one controller came sort of broken: the button ring that gets shifted was open (glue pad did not stick) - some repair pads would be appreciated!
  • the app drains my battery in no time (Galaxy Note 4), the device gets very hot, and the frame rate is very low (I see a series of still images)
  • I tested mini golf in my hallway. I put 8 markers on the floor, very similar to the setup video, and after mapping and starting the game, the play area was orthogonal to the marker layout (so I couldnā€™t reach certain areas)
  • my phone (Note 4) is symmetric, while the box is not; the lens adapter is where the box lid tries to be (seems to work though) - symmetric phones should have been considered at least, no?
  • the headphone jack is covered by card board (is there audio with mini golf?)

EDIT: In minigolf, the left controller completely disappears and only the bat is visible, while the right controller is always represented as a plain white controller model.

This all sounds rather negative, but still I am thrilled what you can do AR-wise with such little money. Great work, guys!

Ole

Hi Ole,

Thanks for reporting your thoughts, and the positive spin at the end! :slight_smile:

To start with the one that is fixable right now - the orientation of the golf course. When you tap the screen to build the map, the side of the pointcodes that youā€™re standing on at that point determines the ā€œfrontā€ of the map. For mini-golf, youā€™ll need to be standing on one of the long sides to get it aligned the way you want.

We originally thought weā€™d fix the x-axis to be the longer side of the bounding rectangle for the codes, but that still gives 2 potential solutions. Also table-top maps are often roughly square, so that had a 4-way ambiguity. Deciding the front based on the side the user is standing when they tap to build the map gets rid of any ambiguity and is actually easy for users to reason about, at least once they know how it works.

Other stuff:

Sorry to hear that. We have seen a few of these reports; I think it might be caused by temperature variation in storage and shipping weakening the glue a bit. When weā€™ve seen this happen the glue has still been sticky enough to just press the ends back together, and it seems to hold when kept indoors.

Is that not the case for you? If not Iā€™d just recommend sticking it with any standard household glue - in manufacture they used an applicator device for the glue strips so theyā€™re not separate pads that we could include spares for unfortunately. For the next batch weā€™ll just try and source some stronger glue to hopefully fix the issue for good.

Yes thereā€™s definitely work to do here still, weā€™re confident thereā€™s big performance improvements we can make. Stay tuned for updates on this one.

Saying that ZapBox is always going to push devices pretty hard - most phones will get pretty warm and use a chunk of battery in a session with any VR app, and with ZapBox we also want to run the camera at 720p@60 FPS and search all of those pixels (over 55 million a second!) looking for pointcodes.

Do you mean the bottom right section at the ā€œfrontā€ of the box interferes with the lens adapter? We considered devices with central cameras in the design - the cutout is below the camera on most devices, but youā€™re right with the lens adapter attached that increases the space needed.

The cutout was a balance between rigidity of the cardboard and how universal it was between devices, but we might reconsider for the next batch. As itā€™s made of cardboard you can always cut a larger hole if required for your device!

We added the wings at the edges as a safety measure to prevent phones sliding out if they werenā€™t held tightly in place by the foam strips. Unfortunately the range of devices means anywhere we positioned those would have covered up some port on some device, so instead Iā€™d recommend you to cut a specific hole for access to your headphone jack on your device.

I canā€™t actually remember if thereā€™s audio in mini-golf, but if so it will just be incidental. We donā€™t yet support spatialized audio so headphones wonā€™t bring a huge benefit for now (outside of privacy).

There are definitely some content bugs in the mini-golf game which Iā€™ll discuss with the content team. Weā€™re intending the left and right controllers to behave the same.

At one point the controller transitioned from a club to the white controller model when the world codes went out-of-view, and it seems thereā€™s still some echos of that behaviour present at least on one of the controllers.

Hi Simon,

Thanks a lot for your detailed answer!

You know, for me this is just a great idea and fun thing to play with and share with my friends and colleagues. I am not a potential commercial adopter. I simply was very eager to receive my box in the beginning. But then I received a LOT of emails from the campaign that were not at all for me (all this dev stuff): initially I would think each time ā€œcool, news about the arrival of my boxā€, and then ā€œoh no, just another load of text that is not for meā€. So when the box finally arrived, my enthusiasm already had adapted a little, so that the thing sat in my kitchen for a while until I unpacked it. Now I think I deserve to be a bit disappointed about all the things that were broken or do not work so well. Donā€™t you think?

Which is where the wall of my hallway isā€¦ :wink:

How about letting the user choose? By rotating the final map? Because the mapping procedure requires certain camera perspectives with many markers in the image, which cannot be done from just arbitrary view points - at least not in my case. Or by rotating the appā€™s orientation in the beginning?

Wouldnā€™t work initially as the adhesive tape was crumpled together. I was able to flatten it out, though, and make it stick again. Seems to hold for now. Just wanted to pass this on to your QC.

I understand that, and I admit that my phone is 3 years old. But a Note 4 isnā€™t a weak phone as such. So I am looking forward to the improvements! :slight_smile:

Yes. The cardboard lid slightly sits on the lens adapter. Not on the lens, though, and so it is working altogether.

I know this is something that likely is impossible to solve for all models out there. Makes individual modifications necessary. Just wanted to let you know. Or a larger re-design of the wing-thing: maybe smaller wings and not only in the bottom area? Anyhow, I think the foam stripes are doing a particularly well job of holding the phone in place!

Seems not, I checked that my volume was up, by now.

Or audibility, as the phone speaker is sitting inside some cardboard enclosureā€¦ :slight_smile:

Simon, thanks again for you extensive comments and helpful support!

Hi Ole,

I didnā€™t mean to sound like I was diminishing your disappointment - thatā€™s of course a feeling youā€™re completely entitled to. Weā€™ll still try to turn that around with some future updates.

On map orientation the longer-term plan is to have a pause mode where the user can move and rotate the map, that would mean content developers donā€™t have to add that functionality to every experience. For now even with a narrow hallway you should be able to tap to build the map with the phone over towards one of those long sides - you donā€™t need to be completely outside all of the codes, itā€™s just which side of the centre point you are on when you do that final tap.

Can you share a photo of the lens adapter position? Iā€™m still a little unsure of the issue from your description.

Hi Simon,

And I donā€™t want to sound unappreciative. The ZapBox is a great thing. Period.

Iā€™ll play around with the mapping and see if I can achieve what you are suggesting. Seems to me I didnā€™t quite get your first comment on this. If it is actually only about the last tap, then this should be fairly easy. I think this pause mode would really be a benefit, though.

A photo of the lens adapter positionā€¦ well, here it is:

But maybe this is by design, or just not an issue at all.

Thanks, Simon!

Speaking of photos, here is one of my repaired controller:

Regarding my snappy ā€œQCā€ remark, I think the hinge of the button did not get pre-flexed enough during assembly. So the sticky pad holding it down came loose and this creates pressure on the button marker ring. This looks pretty my identical with the second controller. So if you will be shipping those cardboard models again, it might be worth taking a look here.

Just want to add that I am pretty impressed with the purely cardboard-based button design as such!

Thanks for all that.

I think the front looks fine with the adapter - weā€™d expect the clip to be under the cardboard like that on a lot of devices, but the foam pads that hold the phone in place should mean it fits under without getting caught.

Weā€™ll definitely aim to solve the trigger glue issue for any future batches - even just a change in manufacturing to leave the glue overnight to strengthen before inserting the plastic spring would probably do it, but weā€™re also looking at a design where the trigger piece would be glued on the inside of the shaft which gives a cleaner look and should be less susceptible to this issue.

Still once repaired we havenā€™t had ours completely popping open again, and the action still works pretty well (I just push down on the hinge section before each use). I definitely agree we can do better than that for the next batch though.

Thanks for info and open words, Simon!