Review: Inlet Technologies Spinnaker S3000

Another winning real-time encoder from Inlet.
Price: $9,995
For more information, contact:
Inlet Technologies
www.inlethd.com

When I first got the weighty Federal Express shipment for the Inlet Technologies Spinnaker S3000 real-time encoder, I figured something that weighed that much was sure to be huge. I was surprised, then, when I opened the box to find what looked like a shrunken "pizza box" server.

The distinctively blue unit is the standard width for mounting in a 19" rack (1 rack unit [RU]) but only 18.2" deep. The depth, about 3"–6" shorter than the standard 1 RU server, means that this unit can fit in a standard audio-video rack that is 20" deep, rather than requiring a server rack, which is ideal for the type of place this device is used.

Packing all the server features into a smaller-than-a-server box means that it’s got some heft to it—my review unit weighed 35 pounds, and I figured I could use it to replace my weight machine—and it tends to be a bit noisy. All joking aside, though, Inlet has been able to pack this unit with features, including multiple Gigabit Ethernet ports, a front-panel, multiline display that signals device status as well as alarms, and a full complement of analog and digital video and audio connectors.

Product Positioning
Inlet positions the Spinnaker line as its real-time encoding solution, as opposed to its ultra high-quality Fathom line, where encodes take longer than real time to accomplish in return for what many consider the highest-quality encodes on the market. Many content owners use the Fathom to transcode HD content for Blu-ray and HD DVD, and broadcasters can use Spinnaker to simultaneously stream their live programming feeds out to multiple bitrates and multiple formats.

Inputs
Inlet understands that quality input aids in quality output, honed over years of its founding engineers’ work at Osprey, which ViewCast later acquired, so Spinnaker offers multiple analog and digital inputs: the standard definition (SD) serial digital interface (SDI) connectors allow input of up to 270Mbps component digital video along with six channels of AES/EBU digital audio, while the analog inputs can be component, S-video, or composite video and balanced (XLR) audio. The AES/EBU audio is extracted, or de-embedded, from the SDI video signal and is paired in stereo channels 1&2, 3&4, and 5&6, respectively. The six-output audio scenario cuts a good line between today’s standard AES/EBU four-channel de-embedders and the next round of broadcast de-embedders that work with four stereo pairs or eight channels. If analog video is chosen, one stereo balanced pair of channels is available, so SDI audio/analog video cannot be mixed and matched for 5.1 surround options.

Encoding
Spinnaker uses a web interface to access encoding and status settings. The interface is accessed by connecting to a portion of the network that any of the three Ethernet ports are on. Two gigabit Ethernet ports, designed for serving up the encoded video, are side-by-side with a 100Mbps maintenance port. The maintenance port, normally put onto a separate part of the local area network (LAN), is in this case completely unrestricting, meaning that it can also be used to serve up video streams in a pinch—although that’s not recommended.

As is typical with Inlet’s products, audio and video settings can be extremely detailed, delving into most of the compression parameters that the codec’s creator reveals in an API. This is a great feature for those familiar with the intricacies of codecs such as VC-1 or On2 VP6 (Flash 8), but it can be quite overwhelming for the average user. I’m not suggesting that they hide parts of the API—after all, the Inlet team is one of the best when it comes to giving full control over codec parameters—but it would be helpful, if the product is aimed at the mass encoding market, for Inlet’s engineers to modify the user interface to put the key information front and center. While they’re at it, they could write the user documentation so it reads less like an engineering document. Some of the more esoteric information isn’t quite complete, so users learn by rote to set a setting at 250; if the documentation could explain why a particular setting should be 250 instead of 15 or 2000, it would serve to make the product accessible to those who focus on content more than engineering.

So back to settings and presets. Once all the settings are selected, a preset can be saved and named for future use. This is helpful for those who have tweaked a few key settings and are trying out the differences between encodes, but presets are far more helpful for the average user who is responsible for day-in, day-out encoding.

Encoding sessions can be scheduled using either custom settings or presets, and the schedule can also be overridden by a manual start/stop encoding process. Given the way Inlet approaches presets, though—saving presets after all settings have been completed rather than creating a preset and then adding to it—changing parameters or creating presets cannot be accomplished during an encoding session.