Encoding.com: Streaming Media's Comprehensive Hands-On Review
In contrast, Amazon charges based on the duration of the encoded file, which is the more typical schema used by cloud encoding services. Specifically, SD files (less than 720p) cost 1.5 cents per minute and HD files (more than 720p) cost 3 cents per minute. This is in the U.S. eastern region, with per-minute prices varying slightly in other regions. Elemental uses a third model based on computing time consumed, which is a very common approach for other cloud services. With Elemental, pricing varies based on a number of factors, including monthly commitments for encoding time and how quickly you spin additional virtual machines up and down.
Obviously, this disparity in pricing models makes it difficult to assess comparative pricing for the various services. Encoded minutes and GB of I/O are the simple ones. For Elemental, you’d have to run some test encodes using a typical file and estimate the encoding time necessary to produce your monthly run rate.
Quality for Encoding.com
One area where there wasn’t a lot of disparity was quality; all three services produced excellent results (see Figure 3). Perhaps this isn’t surprising, since both Amazon and Encoding.com use the x264 codec, while Elemental has been competing against x264-based products since its appliance-based product was launched.
Figure 3. Quality comparisons—no real differences here
Specifically, I compared the quality of my standard 720p test file encoded to 800Kbps using 2-pass VBR, High Profile, and standardized settings for key frame (300 frames), B-frames (three), reference frames (five), and other parameters. After ensuring that the data rate was within 5% of the target, I then compared the files side-by-side during real-time playback and input the files into Adobe Premiere Pro so that I could compare the frames. In all tests, the results were visually identical.
As part of my standard test suite, I compared the actual data rate for the 11 output files to the target data rate. With my shorter (92-second) HD test file, Encoding.com was substantially lower than the target, ranging from a 16% difference at the highest quality files to a maximum of 25% for some lower quality iterations. I ran some quick checks on the 52-minute files encoded by the service, and saw that some were 25% to 30% under the target data rate as well. This is neither serious nor unusual; just be sure to check the data rate of the files you’re encoding and adjust the data rate if necessary to achieve your actual target.
On the plus side, Encoding.com produced very consistent data rates when encoding in constant bitrate encoding (CBR) mode, which many producers prefer when encoding for adaptive streaming or streaming to mobile and other constrained bandwidth scenarios. This is shown in Figure 4, which is a screen shot of the discontinued Inlet Semaphore product showing a data rate graph of my standard HD test file produced by Amazon Elastic Transcoder (in black) and Encoding.com (in faint blue). I couldn’t find a CBR switch in Amazon’s encoder. If it’s there, let me know, and I’ll rerun this test for the next cloud review.
Figure 4. Comparing average bitrates from Amazon Elastic Transcoder and Encoding.com
Even without the comparison to Amazon Elastic Transcoder, the Encoding.com results are impressive; by design, this test file has very low motion footage at the start and very high motion at the end. Notwithstanding the changes in motion and the resulting encoding complexity, the data rate stayed very consistent, not only in this file, but in all the files in the 11 file adaptive streaming group.
I also tested deinterlacing quality by encoding a special interlaced 4:3 DV file containing scenes with multiple deinterlacing challenges such as screen-based text and hard-to-compress backgrounds such as rattan curtains. Encoding.com produced very good results in all scenes, which is obviously important for producers encoding older interlaced footage. Overall, beyond the data rate disparities, Encoding.com was a very polished performer across the board, particularly when it came to performance.
Vid.ly for Universal Playback
Once you’ve got your file encoded, the next challenge is making it play on the variety of computers and devices that producers typically target. Encoding.com has a service for that called Vid.ly, which encodes your video into multiple flavors for various target platforms with device detection that sends the proper stream to each player, including HLS streams to iOS and compatible devices.
You can secure your videos by specifying start and end dates, by limiting the originating IP addresses, or with token-based authentication (see Figure 5). Each Vid.ly video generates a 300 dpi QR code you can integrate into your marketing efforts, and you can monetize your videos via integration with FreeWheel.
Figure 5. The Vid.ly control console
To test Vid.ly, I uploaded my standard HD test file to the service, generating the vanity URL vid.ly/jansdemo, which should be available should you care to try it. In my tests, the file played well on my iPhone 4S, iPad, and Toshiba Thrive (an Android tablet) and was flawless on current versions of all browsers, with WebM sent to Firefox and Chrome and H.264 files sent to Safari and Internet Explorer. While the player is neither as easily configurable nor as functional as players from online video platforms such as Brightcove or Kaltura, it’s certainly competent, enabling comments, links to Twitter, Facebook, LinkedIn, and Delicious, and the ability to generate an embed code.
I did experience some problems with older browser versions, including Internet Explorer 6, which distorted the video, and Safari 5.1.7, which wouldn’t play the video at all. Still, for most small companies, these results are vastly superior to what they could achieve creating their own player and playback schema. However, if there are any older platforms that your videos absolutely must play on, you should definitely test those before going all-in with Vid.ly.
You’ll pay extra for the Vid.ly service, with charges outlined on the Encoding.com website. While not perfect, I think it’s definitely worth a look from companies that don’t have the budget to (or attempt to) supply their own universal playback strategies.
This article appears in the August/September 2013 issue of Streaming Media magazine as "Review: Encoding.com."
The cloud video encoder plans to boost its engineering and sales teams, and continue its collaborations with Harmonic.
With formats expanding and more processing power required, transcoding was an early cloud test cast. Now, more workflow elements are going to the cloud.
Brightcove customers should look to this new service for inexpensive real-time event transcoding, but several rough edges mar the experience.
Elemental is a strong choice for high-volume video publishers or those already using Elemental hardware, but it doesn't come at a bargain price.
Will the major Hollywood studios warm up to cloud encoding? Encoding.com plans to disrupt the market.
Power customers will no longer have to incorporate the Vid.ly API, and can use Encoding.com's graphic interface.
Customers using the Wowza Media Server will find it easier to prepare their files for any-screen delivery.
Companies and Suppliers Mentioned