Produce a Professional-Quality Webcast

The demand for live video streaming or webcasting is growing rapidly. Frost & Sullivan is predicting a compound annual growth rate of 24.8% between 2010 and 2017. Growth of this magnitude means constant innovation and changes, which challenges a webcast producer who tries to support legacy, current, and future workflows.

Webcasting can be an add-on to existing efforts, but increasingly efforts are created primarily for webcast. In many ways, live video production is more complicated and less forgiving than productions that afford the luxury of postproduction, be it editing, color correction, audio mastering, or the addition of graphics and titles. With live production, there are no drafts, reshoots, second chances, and opportunities to "fix it in post." Broadcasters have known this for decades with live televised video. Now a new breed of live video producers, who broadcast live over the internet, are experiencing just how challenging "live" can be.

In its most basic form, a webcast requires a video camera, microphone, hardware, software, streaming servers, and a website page where the webcast video can be embedded and viewed. This might sound complicated, but all these elements are present in a simple webcam chat using popular chat, social media, lifecasting, or streaming video websites. On a professional level, the webcam chat turns into a full-blown production, and the equipment workflow becomes especially challenging when a webcaster is working with other vendors and different technologies that don't always play well together.

The key to any successful production workflow is, well, the workflow. You have to ensure that the different standards and protocols work together, which is easier said than done.

Video Cameras and Lenses

Video camera type and the number of cameras used on the shoot can often influence requirements down the workflow, so let's start here. As with all video productions, video camera selection is an exercise in balancing advantages and trade-offs. Traditional prosumer/professional camcorders and ENG video cameras have three sensors (either CCD or CMOS), and each sensor is relatively small when compared to the new breed of large sensor video cameras. The small sensors, typically ranging from 1/3" to 1/2", yield a noisier image but offer the benefits of both a servo zoom lens and a greater depth of field.

Some movie producers frown on the use of camera zoom to change focal length and prefer to move the camera with devices such as steadicams, jibs and cranes, dollies, and sliders. But webcasting is a different medium from film production, and especially when producing with a single camera, lens zooming is very common. Whereas camcorders and shoulder-mount ENG cameras have servo zoom ability, traditional large-sensor video cameras are paired with manual zoom lenses that can be grouped in two categories: DSLR lenses and Cine lenses. DSLR lenses are designed for use on photo cameras, and consequently the zoom action is usually not as smooth as on an ENG video lens or a Cine Zoom lens. DSLR lenses are also typically limited to 3x and (rarely) 4x zoom in fixed aperture models. Although there are nonfixed aperture models available with longer zoom range, they cannot hold an aperture throughout their entire zoom range. Also, the iris stops down as the focal length increases, which isn't desirable for live video. DSLR lenses are also not typically parfocal, which means they cannot hold their focus when zooming in and out.

Cine Zoom lenses offer smooth focal length changes (zooming), fixed aperture (measured in the more accurate and consistent between-lenses T-stops), and have longer zoom ranges. However, they cost between $20,000 and $100,000. One alternative to expensive and heavy Cine Zoom lenses is the Sony E-mount system, which has a new servo power zoom lens that can hold an f/5.6 aperture throughout its zoom range and maintain focus. The other alternative is to use more video cameras in order to cover additional focal lengths.

Codecs and Outputs

So if zoom lenses are so challenging and expensive on large-sensor video cameras, why do webcasters bother? The reason is that a large-sensor camera produces a cleaner image than a small-sensor video camera. Webcasts, more than most other video mediums, benefit from a low-noise video signal because webcasts are broadcast at dramatically lower bitrates than the internal recordings of the video cameras that are supplying the live video feed. Whereas a 720p HD webcast might have a 1Mbps bitrate, the internal recording could be anywhere from 24Mbps using AVCHD to 240Mbps using ProRes HQ. A lower-bitrate encode can fall apart, including visible macro-blocks and an extreme loss of detail, when paired with a noisy video image.

While on the subject of codecs, the internal camera codec matters little for a webcast, as this recording is relegated to backup duty. The live video output is what is used in a webcast. Modern HD connections include HD-SDI, HDMI, and Component. HDV over FireWire is no longer relevant, as that codec has been replaced with AVCHD and anamorphic codecs have more undesirable video artifacts than a full-raster recording. Component is an analog signal and requires three video cables, which means it's found on very few video switchers. HDMI is a consumer connection and interconnects every device along its chain, which makes it very unstable when all but a straight connection are used. HD-SDI is the professional connection standard for video switchers and webcast encoders. It allows longer cable length runs, and its bayonet mount makes for a much more secure connection than HDMI, although the signal quality is the same between these two digital formats.

Switchers and Scalers

Using a video switcher allows the use of multiple video inputs, meaning multiple video camera inputs as well as computer inputs. Pairing video and computer signal outputs with video switcher inputs avoids the use of converters. When this isn't possible, webcasters use video converters. In addition to converting video signal connection types, some converters have built-in scalers, distribution amplifiers, aspect ratio converters, and even noise reduction. Converters are available in small battery powered units up to large rack-mount units.

Capturing Clean Audio

Clean audio is just as important in a webcast as clean video. A strong signal-to-noise ratio is the starting point, and the easiest way to accomplish this is proper microphone selection and placement. Microphones with directional polar pickup patterns are best for blocking out extraneous noise, but they are more sensitive to level changes if a presenter turns his head from side to side, which is common in panel discussions or when he is trying to engage a live audience.

Wireless microphones allow the presenter to move freely. If a lavalier microphone is used, it can potentially capture more consistent signal levels than a handheld microphone if a presenter moves away from it. They can be more prone to interference and dropouts, but a diversity receiver with pilot tone and scanning for available frequencies can reduce this problem. Another consideration when webcasting and using separate audio and video mixers is audio delay. A 1-3 frame difference is hardly noticeable, but anything more and the lip sync difference becomes obvious. An audio delay device should be part of the workflow if the video signal delay is noticeable and if the video switcher does not have a built-in audio delay function.

Some webcast workflows can allow audio inputs directly into the video camera, which embeds the audio in its digital HDMI or HD-SDI outputs, but some video switchers strip this audio. Adding a soundboard is best practice because it allows multiple audio inputs and more control over balancing audio across stereo channels.

Microphone polar pickup patterns