How to create a 360-degree panorama and upload to Google Photos for display

Litchi is able to automatically create a 360-degree panorama by stitching together a series of photos and then displaying that 360-degree panorama within the Litchi app (IOS Only). However, there are a number of reasons why one may wish to process the series of images manually. These reasons include:

  • To improve the stitching quality
  • To perform color grading of the sky and foreground
  • To display and share the 360-degree panorama in Google Photos

The above steps can be complicated and difficult to perform. This guide will provide some guidance to performing those steps with an emphasis on getting the 360-panorama to be properly recognized and displayed as a panorama in Google Photos.

Improve the stitching quality

Litchi’s image stitcher works quite well. However, sometimes the stitching produces unacceptable results. For example, the following panorama of Bear Butte in South Dakota had an unfortunate seam passing right through Bear Butte.

To correct this, it might be necessary to use another stitching program. A few are available. I use Microsoft’s Image Composite Editor (ICE). Stitching the individual images in ICE produced superior results.

Perform separate sky and foreground color-grading

Depending on the lighting conditions, a panorama might include a very bright sky with a relatively dark foreground. For example, the following photo of the open-pit mine in Lead, SD combines a light sky with a dark ground.

To correct this, it may be necessary to perform some image work using Photoshop or similar. I duplicate an image like this into two layers then adjust one for optimum sky and the other for optimum foreground. Then, I combine the two with a smooth blend across the horizon.

Prepare the image for display in Google Photos

This last step can be the trickiest part. Some photo display applications like Google Photos will automatically detect and display 360-degree panoramic images using an interactive panoramic viewer. However, Google does not make it clear at all how it automatically detects a 360-degree panorama. The question of how to make Google Photos detect and properly display a 360 pano has been asked hundreds of times in Google’s support, but never with an acceptable answer. Google’s responses are always along the lines of “try clearing your cache”, “try a different browser”, “try the app”, or “try a different computer”. Clearly, Google’s support people have no idea how their own software works.

Here are the steps required to make Google Photos automatically recognize and properly display a 360-degree panorama.

  1. The aspect ratio of the image is important. A 360-degree panoramic must be twice as wide as it is high. For example, if the image is 2000 pixels wide, it must be 1000 pixels high. If the image aspect ratio is not 2:1, it will not display as a panoramic image in Google Photos. After stitching the images, the result must be brought into an editing program (I use Photoshop) and the canvas size must be made 2:1 by adding some sky. I normally copy the entire width of the sky and then flip vertically and stretch to fit. One can spend more time blending the flipped sky with the existing sky but this step is not entirely necessary when viewing the image as a 360-degree pano. The details of how to do so are beyond the scope of this tutorial.

  2. Google Photos has an upper limit on the size (in pixels) of images uploaded to their site. This limit is 100MP. Don’t confuse this with the image size on disk (this is mega-PIXELS, not mega-BYTES). This is simply the number of pixels in the X-direction multiplied by the number of pixels in the Y-direction. To get the optimum size, we know that x * y <= 100. We also know that x = 2 * y. Solving for Y yields 7000. Therefore, the optimum size for a 360-degree panorama uploaded to Google Photos is 14000x7000. This will be a 98MP image. If necessary, scale the down to 14000x7000.

  3. Google requires that the EXIF data contained within the image properly specifies that the image is a panorama. To edit the EXIF data within the image, one must use an EXIF data editor. A few are available such as EXIFTool (for advanced users) or “EXIF Pilot”. I use “EXIF Pilot” to set the EXIF/XMP data as shown in the image below.

Capture

Once the above steps have been completed, the 360-degree panoramic image may be uploaded to Google Photos where it will be properly detected as a 360-degree panorama and displayed accordingly. Here is a 360-degree panoramic photo (displayed as a flat image) that has gone through the steps outlined above. Notice that Google Photos placed a circular arrow in the upper-right corner of the image indicating that it has been detected as a 360-degree pano and will be displayed properly if selected.

Capture

I hope this saves someone some time. It took me a while to research this and figure out all of the details to make this work.

3 Likes

Nicely done! Would another option be:

When shooting a panorama, do not select the automatic option.

  1. First go into the normal camera settings select AEB to bracket each shot.

  2. Then select Pano from the main menu, and choose the settings icon below the folder and globe.

  3. The number of rows and columns by default gives you the same coverage 22 images. You can always add more. I found that shooting over water can cause some stitching issues, so in that case I add 2 Nadirs. I also have a .5 second delay added before each shot to steady the position.

  4. When you press Start, you will have taken 3 shots at each location.

I know that instead of 22 shots normally taken in its auto mode, you will have 66, but you can pick and choose the best exposures for highlights and shadows. You could even combine them into a HDR.

1 Like

Yes! There are countless ways to create a pano and get good results.

My focus in the tutorial above was:

  1. What can be done when the automatic stitching does not produce the results you want
  2. How to get a 360 pano to properly display in Google Photos

I have not tried using AEB during a pano and then selecting the best of each shot prior to stitching. Have you found any difficulty when using two shots with different exposures adjacent to each other?

1 Like

495d6a5f9d87f407efafd97ad3690c1644fd986e
I don’t know the actual size of your original image but this is how it can be done without converting it to 2 to 1 ratio. It works in Google Photos and Facebook.

Very good. Thanks. That does work. The only drawback is that it leaves a big black hole in the sky.

1 Like

Yes it does. Not as noticeable on Facebook as in Google Photos.

1 Like

You should try PTGui as it solves many of the problems you encountered in your tutorial. It has advanced stitching options and gets it right the first time. It lets you adjust all the photos as if you were in Photoshop, except it’s way easier. It automatically generates sky based on what the rest of the sky looks like, so the part where your drone camera couldn’t point up looks pretty natural. It properly tags the file with the right metadata.

Basically, you could squash your whole tutorial into: “Use PTGui”. :laughing:

14 posts were split to a new topic: How to fly to a point then take a panorama