While this website’s primarily aimed at reviewing 360 cameras, VR headsets and related news from the tech world, we also thought it would be interesting and helpful to include a few how-to articles on things related to 360 video and photography. So, for this article we’ll be looking at how to create 360 panoramas, and then exploring how to turn these into interactive ones you can view using a VR headset.
Step 1: Take photos
There’s the easy way and the hard way. If you’re using an android smartphone then you’ll be able to create one using the ‘photosphere’ mode on the built-in camera app. After tapping the camera icon and selecting ‘photosphere’ you’ll be asked to slowly move the phone around until you’ve taken photos in all directions around you, 360 x 360 on both the horizontal and vertical axes. You won’t have to worry whether you’ve missed a bit, as you’ll see an arrow icon on the screen that will tell you which direction to move the phone until it’s done. Then the app will automatically stitch all these images into a single panorama or ‘photosphere’.
However, while using your smartphone app is the easiest option, you’ll frequently find that there’s a trade-off in terms of quality over convenience. Yes, it’s the quickest method to create a photosphere, but anything that moves frequently gets blurred, the exposure often isn’t quite right, and most camera phone’s aren’t great in low light. To get a better quality photosphere you’ll need to use a proper camera and slowly pan around taking photos until you’ve captured a full 360 x 360 without any gaps. This is much more difficult to do as you won’t have an arrow icon guiding your movements. A lot of people use a tripod for this but I just steadily move the camera around taking photos, with the first image facing downwards, then gradually moving my body upwards taking more photos, then slowly turning, repeating the process (down to up), until I’ve turned a full 360 on the spot. I get a lot of funny looks from passers-by doing this, but the results are usually pretty good.
Tip: bear in mind that in order to properly stitch the images together to create a single photosphere there will need to be some overlap in the photos. And this means that if you’re trying to take one in a fast-changing environment then you’re going to have problems when they’re stitched together, as you might find half of a person missing, an arm elongated, and other odd results. To avoid (or at least minimize this) you should look around first and if you see someone walking past then wait until they’re further away. The same applies to cars and other traffic that might distort your panorama. To see what can go wrong when making a panorama have a look at this compilation of alarming fails.
Step 2: stitching the images together
If you used the photosphere app on your smartphone you don’t need to worry about this bit as the app will have done this for you. However, if you’re using a dedicated camera (e.g. a DSLR) then you’re now faced with the additional task of joining all the images together. There are a number of photo-stitching programs out there, including Kolor Autopano Giga, but my personal favourite is Microsoft’s Image Composition Editor (ICE). As with all these programs, you add all the images that make up the 360 scene, and then the software automatically stitches them together. You’ll then be offered a number of options to tweak the output, either to minimise (or maximize, if you like) the geometrical distortion. As you’re planning on turning the resulting panorama into an interactive photosphere you should choose settings that minimize distortion though.
Step 3: finishing touches to the panorama/photosphere
By now you should have a 360 image. However, you might notice some problems with your panorama. For example, you might have missed a spot in the sky so it’s blacked out, or someone walked past you when you were midway through shooting and they’re all warped in the resulting panorama.
That’s not great, and it’ll diminish the quality of the final image as people will inevitably focus on that. The only way around this is to open the stitched image in whatever photo editing software you normally use (photoshop, or GIMP, for example) and manually correct it. In photoshop the best tool at your disposal is the ‘Content-Aware’ function. For example, to get rid of any black areas or small lines between badly joined parts of the panorama you should select those areas using the Polygonal lasso tool. Then press Shift+F5 to bring up the Content-Aware window where you’ll see several options. Just make sure ‘Content-aware’ is selected from the drop-down box with the blending mode set to ‘normal’ and opacity at 100%.
Click ‘ok’ and photoshop will fill the selected area as best it can by matching it to the surrounding image. This usually works for me, but if it doesn’t then the only way around it is to use the clone stamp and alter that part of the image. This can be extremely tedious to do, but I find it’s worth it to get rid of artefacts and blank areas caused by the photo-stitching process.
After tidying it up you should have a finished panorama with no blacked-out areas, like this.
Step 4: turn the 360 panorama into an interactive one
So, you’ve taken lots of images, stitched them together, and tweaked the resulting panorama a bit so it’s all ready. All you need to do now is upload it to a site that understands it’s a full 360 image and can make it interactive by allowing viewers to pan around it as a 360 image instead of a flat two-dimensional panorama.
The most common method is to upload your image to Google Street View. But Google have a lot of requirements you need to meet for your image to be allowed. Here’s a link on how to do this.
The alternative option is to upload your photosphere to another site. Personally, I’ve always had problems with Google Street View, so I use this site: http://viewer.spherecast.org/ It’s free, they don’t seem to have a limit on image resolution (while Google’s limit is 8,000 x 4,000px) or any other stringent requirements, and you don’t need to register. Simply upload your file, and you’ll be able to share it. You can pan around the image on your desktop, but for a more interactive experience you should open it using your smartphone or tablet as you can then pan around the image using your devices’ gyro. And the best thing about the site is that you can view your image in 3D using Google Cardboard (or whatever VR headset you have). By simply clicking the google icon on the bottom left-hand side of the screen, followed by the full-screen button on the bottom RHS of the screen, you can then view it in VR!
Here are some examples of photospheres I’ve made and uploaded to the site.
- Times Square at night
- Bethesda Terrace in Central Park
- Camden Town in Central London
- Somerset House in London during London Fashion Week