Streaming Media

Streaming Media on Facebook Streaming Media on Twitter Streaming Media on LinkedIn
 

Review: Livestream Mevo

If you frequently produce live or on-demand videos for social media sites, Livestream Mevo can help you produce more polished and engaging videos. While there are some rough edges in version 1.0, and there is a learning curve for operation, Mevo is an essential tool for any organization seeking to leverage the power of video in its social media marketing.

Mevo encodes all live streams at 720p30 at up to 5 Mbps, with the data rate automatically adjusted to match the connection speed. The unit can stream directly to your accounts at Livestream and Facebook Live.

There is currently no generic RTMP interface, though Livestream is considering adding one. If you’re looking for a device to use with Ustream, YouTube Live, DaCast or other services, you’re out of luck.

The unit supposedly can record at up to 20 Mbps, again at 720p30, though all videos I recorded flatlined at 10 Mbps. You can stream and record simultaneously, but in that case, the unit will store the streaming file at the lower bitrate.

Neither the Mevo nor the Boost has an external microphone connection, but you can substitute in higher quality audio by routing it through either your iPhone’s microphone/speaker port or Lightning Connector. Livestream has an excellent series of video tutorials detailing various aspects of Mevo operation on its website, and the tutorial on using external audio sources warns that not all sources are compatible.

This proved true for me, as all four mics/mic systems that I tried failed. Livestream has a short list of compatible devices on its website, and you’d be well advised to choose a known-compatible model. As you can see in the Audio Mixer in the Mevo app in Figure 4 (below), you can choose which audio source to include with the video (Mevo or iPhone), and set levels.

Figure 4. Here’s where you choose the audio source (Mevo or iPhone), and set levels. The blue box around my face is facial recognition.

Software Operation

Though I live in the self-proclaimed world capital of old-time mountain music, there isn’t always a concert to be had when you need one. So I ran my first series of tests on the next best thing, a single-camera shoot I had produced two summers ago for local group Loose Strings.

Figure 5 (below) shows the main production interface. One characteristic I didn’t like was the fisheye effect shown in full-screen output; the top and sides of the Mac monitor in Figure 5 are not curved. However, this effect is apparent only when the complete frame is visible, so it’s not that distracting in practice. The picture-in-picture on the upper right shows the video actually being streamed or captured, which is the complete frame at this point. The three boxes around the faces on the left show facial recognition, and these are all stored shots I could access by pressing the six-button icon on the lower right. This opens the “grid” view that shows all shots created by facial recognition or by manual selection.

Figure 5. The main production interface shows the fisheye effect. Click the image to see it at full size.

In manual mode, you can click on any point in the frame and cut to that “shot.” You see this in Figure 2, where I’m focusing on the two band members on the left. You can resize and move this box as desired with typical gesture controls, and it retains the 16:9 aspect ratio. You can also zoom in, but not past the actual 1280x720 pixels in the box. Operationally, you touch the three circle icons on the lower right to access camera controls, such as those shown in Figure 3 and Figure 4, and you touch the double arrow icon on the lower right in Figure 2 to jump back to full frame.

Running in Auto Mode

I ran my next series of tests in auto mode, capturing my daughter Eleanor sitting with me on a stoop on the side porch of our home. The lighting was unfortunate, as we were in the shade, with the sun behind our heads and shining into the camera lens, producing the slight streaking shown in Figure 6 (below).

Figure 6. Testing in auto mode

I tested primarily in auto mode, letting the camera perform all switching. The camera didn’t do a bad job, but moved constantly, like a novice cameraperson trying to achieve perfect framing but never quite succeeding. I found it irritating, but fans of the TV show The Office will probably feel right at home. Livestream could definitely improve the algorithm here to steady the camera, and not follow minor head movements quite so vigorously.