Web AR instant tracking for Wall Mural


Hello everyone, we have been asked to create a web ar activation for a mural that is 20’ wide x 8’ tall. Will we be able to create this inside of Studio, and what are the parameters when it comes to being able to tracking it?


Hello! Conveniently, this is the exact type of project I tend to work on. I just recently finished a demo for a 18’ x 8’ wall, so very similar to yours! I built it entirely within Studio, so you should be just fine to use Studio as well.

As always, take a peek at the documentation for what makes a good tracking image before getting too far into the project.

Once you have that underway, you will need to ensure that the image you want to train is in RGB colorspace, and if your original file is at full scale (meaning your image is 20’ x 8’), you will likely have to scale down your .JPG for Studio. If you import your tracking image and it seems to have a checkerboard, chances are the file is too big. I believe mine ended up being a 50% scaled down version of artwork that was created at 10% the actual scale of the mural. Everything still tracked just fine on the real mural, so you can get your tracking image file size to be quite small and still be fine.

Hope that helps!


I am currently working on this type of project. Some of my findings are use a cropped image of the actual mural (not artwork used to create it), photograph it in a more neutral lighting scenario, and use a guide for the user to line up the artwork (image tracking usually needs the image framed in about 25% of the screen to start tracking).

I ran into some initial struggles as my artwork is large 100’x20’ and didn’t understand why the tracking wasn’t working. In my initial tests I was too far away. So I tried a smaller portion of the mural to improve tracking, but that required users to get even closer. Using what I outlined above allowed me to get something that works pretty consistently. I also added that when the image is lost the user gets the guide again to reinitiate tracking.

Good luck,


Branching off of this, there’s a subsymbol that you can add in relatively easily that lets users know they need to “find” the image again.


This has all been very helpful, thank you everyone!