Telestream Wirecast 5: The Definitive Review
Beyond these new features, Wirecast’s interface is relatively unchanged. In particular, the shot editor, where you perform a wide variety of activities from combining multiple inputs into a vertical shot to choosing and creating titles, was not changed at all. The Source and Output Settings windows have been altered slightly, generally for the better, but otherwise the biggest change is that Telestream now provides information such as memory, CPU and bandwidth usage, and warning indicators on the front of the interface, which is very helpful (see the top of Figure 1).
Output
The breadth of Wirecast’s output capabilities has always been a key product strength; in particular, it’s one of the few products that can send a stream to multiple streaming services and servers from one live source. This can be a huge time convenience and cost savings for producers serving multiple outputs.
One of the headline features in Wirecast 5 is support for the x264 codec, which delivers both encoding efficiency and higher quality, though both will vary according to the stream configuration and x264 preset used. For example, I tested three different output configurations using both the MainConcept codec supported in Wirecast 5 and previous versions, and the new x264 codec. In the first comparison, 1280x720x29.97 @ 2.5Mbps, there was minimal difference in quality. This is a configuration that might be used when sending a single high-quality stream to a service such as YouTube Live or Brightcove that re-encodes that stream into multiple lower-resolution, lower data rate streams for adaptive delivery. In that use, I wouldn’t expect to see any difference in quality between x264 and MainConcept.
At the other end of the spectrum, many producers encode on location for direct distribution without re-encoding. To test this type of use, I encoded at the admittedly aggressive 1280x720x29.97 @ 800Kbps and more reasonable 640x360x29.97 @ 600Kbps rates (all data rates video only). In both those configurations, particularly the first, x264 was noticeably sharper and clearer. In all tests, using the default setting of Very Fast Encoding (3), the x264 codec used about 60 percent of the CPU resources required for the MainConcept codec. Even if you ignore the other new features, the quality and CPU efficiency delivered by the x264 codec is reason enough to upgrade.
One of the benefits of the x264 codec are the profiles shown in Figure 4, which let you trade off quality for CPU usage. To try to isolate the optimal quality/performance setting for my test computer, I ran several encoding trials at the 720p@800 configuration using the Very Fast, Medium, and Slower presets. For the record, I performed my performance tests on an HP Z400 driven by a 6/12-core (with HTT) single Xeon CPU running 64-bit Windows 7 with 12GB of RAM. Figure 5 shows the CPU usage in the Windows Performance Manager for all three tests. To put this in perspective, CPU usage while encoding with the Medium preset is very close to the CPU usage of the MainConcept codec producing a similarly configured stream.
Figure 4. The quality and efficiency of the x264 codec was reason enough to upgrade.
Figure 5. CPU usage in 720p@800Kbps encodes using the designated presets
While the 55 percent to 60 percent CPU usage measured when encoding with the Slower preset is acceptable, I wouldn’t want to go much higher when producing a real live event. Interestingly, however, when analyzing the files, I noticed that Wirecast dropped frames when producing the Slower file, even though there was plenty of CPU overhead to spare. My tests also revealed that the Medium preset produced much better quality than the Very Fast preset (Figure 6) with very reasonable CPU usage, as shown in Figure 5.
Figure 6. Different frame quality using the designated x264 presets
Of course, these results will vary by computer and CPU. In general, if you’re encoding for delivery, as opposed to encoding for re-encoding into adaptive streams (as with YouTube and Brightcove), you want the best possible quality at the lowest possible bitrate. Using the x264 codec, I would run some test encodes to determine the highest quality preset that can produce video without dropped frames at a reasonable CPU usage, say around 50 percent to 60 percent. On the Z400, which I do use for some live productions, I would definitely use the Medium as opposed to the Very Fast preset.
As an aside, before comparing quality, I (of course) verified that the streams I was comparing hit the target data rate. In all tests, the files produced by the x264 codec were between 12 percent to 15 percent lower than the target. So I boosted the target until I had files of equivalent data rates for comparison purposes.
In these circumstances, coming in lower than the target is always better than higher because while a lower data rate could impact quality, a higher data rate could prevent the stream from leaving the building or complicate delivery to viewers on lower speed connections. Still, once you choose the target bandwidth for your streams, you want your encoding tool to hit it. Going forward, as I use Wirecast in my productions, I’ll choose a target data rate, then run some tests to disk and adjust the data rate configuration in Wirecast as needed to hit my target.
The last notable new feature in Wirecast is the ability to output a program feed to Blackmagic Design Intensity or DeckLink hardware. In my tests, I output a crisp, clear HD-SDI signal out of the DeckLink 4K Extreme, which I fed into another DeckLink on a different computer. This feature will be invaluable for any producers who need to output their Wirecast productions into a broadcast or similar environment or who need to use a third party encoder to produce streams outside of Wirecast. While the Blackmagic cards were the only products supported initially, Telestream expects that other companies will use Wirecast’s Output SDK to provide similar functionality.
Summary
Overall, in terms of usability and functionality, Wirecast 5 is a significant advancement over previous versions. I ran several longish trials, including a 12-hour broadcast to YouTube Live, and found the program very stable on two computers, the aforementioned Z400 and an 8-core Mac Pro. So, while your results may vary, I feel comfortable recommending that Wirecast 4 users upgrade immediately, and I will certainly do so myself. Note that while you can install Wirecast 5 on Macs without uninstalling previous versions, on Windows you’ll have to cut the cord and uninstall.
This article appears in January/February 2014 issue of Streaming Media magazine as "Review: Telestream’s Wirecast 5."
Related Articles
We've upgraded our testing methodology to accurately reflect what broadcasters and other high-volume video publishers ask of their encoding systems. Our first subjects: Elemental Server and Telestream Vantage Lightspeed Server.
05 Dec 2014
The live streaming tool takes a big step forward with the ability to create instant replays, automate broadcasts with playlists, and much more.
05 Nov 2014
Creating multiplatform and multi-device architectures is easier, since the Telestream transcoding platform supports additional formats.
26 Aug 2014
Lynx customers will enjoy elastic private cloud transcoding when the service debuts sometime later this year.
26 Mar 2014
Debuting the first commercial product using the x265 codec, Telestream credits the work of the global open source community.
04 Feb 2014
Live streaming on a notebook can be the best of all possible worlds, but how powerful a machine do you need, and how do the various live streaming software programs stack up? We look at tools from Adobe, Kulabyte, Microsoft, and Telestream.
18 Feb 2011
There's an old Hollywood joke with the punch line, "but I really want to direct." I'm not sure what the actual joke is, or if it's funny, but the reason I bring it up is that if you talk to writers, most would say, "but I really want to be onTV."
03 May 2010
Companies and Suppliers Mentioned