360˚ Video Top Tips: Episode #002

Welcome to episode two of our 360˚ Video Top Tips series. The aim of this series is to provide useful tips to teach you how to shoot 360 degree video to the best of your ability. 360˚ video and immersive content creators can apply these tips and tricks to any future productions. In this series we talk about lighting and exposure for beginners, and at a more intermediate/advanced level, automated stitching algorithms in post-production.

Level: Beginner

Topic: Lighting / Exposure


It’s a common misconception that you cannot and therefore should not control the lighting of a scene during a 360˚ video shoot. This misconception cannot be further from the truth and unfortunately lives on because it justifies less expenditure to the production.

In most cases you should consider using practical light sources such as ceiling fittings, desk lamps, and almost any other home or office fixture. If the existing practical light sources aren’t able to fulfil your creative intentions, an often unconsidered tip is to replace the bulbs for optimum luminance and to add dimmers to maximise the creative control you have of the light.

If your scene requires the use of medium to large lighting set-ups, similar to traditional production requirements, you should consider fragmenting your scene to contain most, if not all, of the obtrusive lights within a single region of the 360˚ sphere. Don’t forget, it’s always wise to consider using practical props such as furnishings to hide the lights from the camera’s view.

If you can’t sufficiently hide the lights, the following tips may be of interest:

If your camera is moving, it’s advised to keep the lights as high as possible from the horizon line whilst still being fit for purpose to make clone-stamping and/or 3D projection as simple as possible.

If your camera is static, it’s advised that you block your scene and direct the action to avoid crossing between the camera and the physical location of the light. Such a precaution will allow you to capture ‘clean plates’ (videos or photos from the exact same camera position which do not include the items you’d like to remove) to be taken after you’ve moved the lights. You can also hide your director, production crew, and client in the same portion of the scene that will be replaced by the clean plate.

It’s also very important to bracket exposures while capturing static scenes, clean plates, and when using 360 degree video cameras with insufficient dynamic range for your scene.

Level: Intermediate / Advanced

Topic: Automated stitching algorithms


If you use professional hardware which includes proprietary post-production tools developed specifically for the automated post-processing of the RAW footage captured by your 360 degree video camera, you should spend a sufficient amount of time during the project’s research & development phase to ensure you’ve identified the limitations of the algorithm(s). It is also important to exercise extreme caution during production to ensure your footage adheres to a set of rules which will be specific to your project and serve to minimise the risk of post-processing errors.

Post-processing errors are in most cases only identifiable after your selects have been made, your edit has been assembled, and your client has approved the chosen shots for fine-stitching. This is why it’s so important to take all considerations seriously during development and production to avoid increasing your costs to manually repair imperfect footage.  

Examples of automated post-processing tools include the industry leading Jaunt Cloud Services stitching / rendering tool, Nokia’s (offline) OZO Creator software, and Google’s Jump Assembler / toolset.

The primary items which often interfere with the algorithms include: consistent geometry (lines, fences, wires, patterns, shapes etc.), high-contrast (caused by overexposure or underexposure), and high-frequency items such as plaid or striped clothing. A general rule of thumb is if there’s a risk of the moire effect occurring, there’s most probably going to be an algorithm processing risk.