image: wallpaperflare.com
My last Touchstone article was the first of a series on a few easy and (relatively) cheap ways to make your livestream experience memorable (for the right reasons!). Perhaps paradoxically, the best way to improve the technology side of a livestream – sorry, technology can’t improve the content – is to improve the audio quality.
So, your paradox detector may oscillate even more wildly when I tell you that the second most important contributor to video quality is … lighting! That’s right, still nothing to do with actual livestream systems! Let me quote from and article by Paul Alan Clifford, self-styled Pastor to Techies and ChurchTech consultant, from an article he wrote for churchproduction.com’s blog.
What’s the problem?
It is often difficult to explain why it is that lighting is needed even in environments where we can see perfectly well. Our eyes “see” far better than any video camera ever made. They have superior low-light performance, superior colour palette, superior focusing ability and superior resolution. Our eyes work so well, that it seems inconsequential to architects and lighting designers to include features in our buildings that are problematical (at best) in terms of delivering light to the things we want to look at. Large windows directly behind the pulpit and top-down vertically directed lighting are 2 of the most common issues in this regard, but most of the time, our eyes automatically compensate for this, and we don’t notice – we even think the scene looks good!
Cameras, on the other hand, will “look” at that same scene and potentially struggle to produce a usable image. Let’s look at how a camera works. (I’m going to walk through a typical electronic digital camera, but in terms of lighting issues, a film camera has the same types of issues, though there are more options to mitigate these in the processing stage).
So, at its simplest, a camera is a device that controls the admission of light into a hole in a box, to illuminate a sensor/recording device. In proactice, we enhance this by adding lenses and other modifying devices to give us more control over the image we end up producing. With a digital camera, the light entering through the lens strikes an image sensor. The signal output by the image sensor is processed within the camera to create image data, which is stored in a memory. For motion, multiple images are compiles into a video stream and transmitted or stored.
image: panasonic.com
In the camera, the image sensor is the equivalent of the human eye. It has the role of converting the light directed on it through the lens to the electrical signals that are processed by the image processing engine. Sensors inherently have dynamic range limitations – too much light will wash out the picture; too little light results in little or no image being recorded. Light can be regulated to some degree by the lens aperture, (the size of the opening in the camera’s lens), which can be adjusted to allow more or less light in. However, Cameras always require more light to produce a given image compared to the naked eye. Even at its most open setting, a lens will not allow enough light in to produce an adequate image for livestreaming.
Camera Manufacturers have traditionally responded to this issue by using the Image Processing Engine in the camera to attempt to compensate for the sensor shortcomings as best it can. In reality, it is the quality of the Image Processing Engine that makes the difference between low-quality and high-quality cameras. The Image Processing Engine is a dedicated, special purpose computer that processes each image in the stream pixel by pixel. The workload for this computer increases as resolution, colour depth or frame rate increase – there are simply more pixels to process in a fixed timeframe. So if the scene the camera is capturing doesn’t have enough light reflecting back to the camera, the Image Processing Engine has to work even harder and may just run out of processing capacity all together, resulting in flat, sub-optimal images. Note that it is the reflected light we need – there may be plenty of light in the environment as far as the eye is concerned. The camera and the light source need to be on the same side of the scene for this to work. A window behind the scene will easily overpower the light reflected from the scene, and overhead lights will light the top of the scene and cast shadows down the scene which will turn up as dark patches in the video.
Solutions
The obvious (and technically preferred) solution is to put more light on the scene, and I will give some thoughts on that in a moment. However, I promised you easy and (relatively) cheap solutions. So here are some ideas that may help without actually biting the funding bullet and putting in a theatrical lighting system.
- Reduce the quality of the video output of your cameras (less resolution, colour depth, frame rate) to reduce the processing load on the Image Processing Engine.
- If you are able to relocate any of your existing lights, see if you can move them closer to the scene, and reduce the angle to the camera.
- Control natural lighting with blinds or curtains or similar – usually even indirect sunlight will overpower any other lighting you have. Also, if the sun goes behind a cloud in the middle of your livestream, you may suddenly find that what was working well now doesn’t work at all.
Theatrical Lighting
Don’t get carried away – just because I said Theatrical Lighting doesn’t necessarily mean a full-on coloured light extravaganza (even though that’s so cool!!) Theatrical lighting can be optimised to deliver lots of white light, targeted to where you need it to get great video. Unfortunately, putting in lighting from scratch often means leaving the realm of relatively cheap solution. Having said that, the advent of LED lighting systems has reduced the price of lighting dramatically and you should investigate fully before assuming you can’t afford it. Some things to remember, though;
- Check your building’s electrical capacity – again, LED lighting has reduced the requirement for power significantly, but still check.
- Structural capacity – if you are considering mounting lights from walls or ceilings, be safe and get it checked beforehand. Don’t forget that the Building Code will require secondary seismic restraint, so things don’t fall on you in an earthquake.
- In a perfect world, the objective is to create a perfectly even wash of light over the full scene. This tends to require lots of light fittings, which is why 3-point lighting was developed. This identifies the areas of most interest in the scene and illuminates them with Key light (main light), Fill light (fills in the shadows from the key light) and Top light (sometimes called back light – this highlights the top and side of the subject so it visually “pops” out of the background). There are lots and lots of good videos and articles on the web on how to do 3-point lighting.
Have fun experimenting – I hope there is something here that helps to improve your livestreams and videos.
Peter Lane is Principal Consultant at System Design & Communication Services and has over 30 years of experience with Technology systems. We invite your questions, suggestions and ideas for articles. These can be submitted either through the editor or by email to dct@dct.org.nz. We also operate a website focused on building a community of people interested in improving how we can use technology in churches, located at www.dct.org.nz.