By Mark Tennent in Reader
Posted in Uncategorized on June 3, 2007 at 11:40 am
The new Elgato Turbo.264 is a £51 dedicated video compression device in a tiny, 85 x 30 x 10mm black box. Running under Mac OS X it plugs into any USB 2 port, or via the short USB extension cable supplied. Also supplied in the packaging is the software on CD.

How was it tested?
Compressing videos is a time consuming task, even with the Turbo.264. It has been tested over a period of a week on two systems, a G5 and MacBook, both with dual 2GHz CPU’s.
As well as full feature films and TV programs recorded in EyeTV, a short 10 minute segment of a recording was made as a reference movie and saved in DV format totalling 1.9GB. This was from the recent Cat Stevens/Yusuf Islam concert on BBC, recorded in EyeTV and chosen because it is a particularly well lit and quality program, with excellent stereo sound, lots of close-ups and plenty of movement.
In addition, short, home-made MPEG2 videos from a Sony Cybershot camera were also compressed, as were avi’s, wmv’s and other formats of videos QuickTime can read. A final test was made by compressing all the films using VisualHub, a software-only compression system based around ffmpeg that uses different codecs - x264 instead of QuickTime’s H264.
What is it like to use?
The software for the Turbo.264 is in true plug and play tradition and like Elgato’s other offerings, well designed and easy to use. It ran trouble-free for all the testing. The software has one window where videos for compressing are dragged and dropped, batches can be queued. Each can be set to one of four compression levels: iPod high, iPod standard, Sony PSP and Apple TV. Output is to PAL or NTSC formats. A small window shows the video stepping through as it is compressed, with conversion speed and time remaining, which was accurate. Another small window can be opened to see the original movie’s format, size and fps.

Currently, output is restricted to 800×600 for this version of the Turbo.264 due to the limits of its chip. According to Lars Farber, Elgato’s product marketing manager, future versions of the hardware will be able to produce HD content and anyway, as Lars explained, there isn’t a great deal of HD content in Europe yet. In reality Turbo.264 produces high-quality viewing at 800×600.
The Turbo.264 is automatically detected by any application using QuickTime such as iMovie and Final Cut Pro. New export options appear in their menus. When the Turbo.264 is active, the export progress bar changes colour and little red blobs appear under it to indicate the Turbo.264 is working.
The results
The Turbo.264 certainly speeds up video conversion with movies recorded at 25 fps being compressed at up to 80 fps on the MacBook. Strange things did result however. When compressing the reference movie the G5 ran at a maximum of 16 fps and used 40% of the G5’s CPU’s whereas compressing the same movie on the MacBook saw 22fps and only 30% of the CPU’s being used. This was even when loading and saving the movie over a 802.11g wireless network from its storage on the G5. An increase in speed that was completely unexpected as the Turbo.264 is supposed to run independent of the main CPU.
Normally when compressing video, the fans on the G5 and MacBook spin up and the machines get hot. With the Turbo.264 running this did not happen and it even allowed the Macs to compress videos using VisualHub at the same time as the Turbo.264. By accident, one time this was the same videos, muxed Mpeg2’s from a Sony Cybershot. The Turbo easily out-performed VisualHub, knocking several minutes off the compression time.
The output options from Turbo.264 are set with no opportunity to make changes. The reference movie compressed to Apple TV standard in 15 minutes, from 1.9GB to only 223MB, with no obvious deterioration in sound or video quality. By contrast, Visualhub compressed the same video in 21 minutes but did make a far smaller 95mb file. Sound and visual quality were identical to Turbo.264’s.
It appears that a feature of all QuickTime-based conversions is for the exported films to become slightly darker than the originals and as Turbo.264 uses QuickTime codecs, the same darkening resulted. Visualhub did not do this and its output movies were the same lightness as the original. QuickTime playback does have controls to adjust colour balance, brightness and contrast if the movie is a little too dark for personal taste.
Compressing full length feature films still take a long time, even with the Turbo.264 on the MacBook. Visualhub takes up to several hours to compress a 90 minute film, with Turbo.264 it was actually possible to do the conversions in less time than the film to to run. A wide-screen avi recorded at 25 fps compressed using Turbo.264 at 60 fps and as a result was twice as quick as it took the movie to run. A 4.95Gb Mpeg1 muxed, 25fps format film compressed at 74fps in 65 minutes.
Conclusion
The Turbo.264 is a great little gadget for anyone doing a lot of video compression and especially Apple TV owners who want to export EyeTV recordings to iTunes. Owners of older Macs, G5’s as well as G4’s, will see much faster compression times so considering its relatively cheap price the Turbo.264 would make an excellent purchase. For Intel Macs, the speed increases are not so dramatic but still enough to make it worth considering the Turbo.264 for its ease of use alone.
During testing the Turbo.264 and its software worked faultlessly and released the host Mac for other tasks, including other software-based video conversions. Elgato have a great track record for providing support and regular upgrades to their products. I know from personal experience that questions emailed to them gets almost instant responses depending on time-zone differences.
Make a comment
Tag cloud
Archives
- November 2008
- October 2008
- September 2008
- August 2008
- July 2008
- June 2008
- May 2008
- April 2008
- March 2008
- February 2008
- January 2008
- December 2007
- November 2007
- October 2007
- September 2007
- August 2007
- July 2007
- June 2007
- May 2007
- April 2007
- March 2007
- February 2007
- January 2007
- December 2006
- November 2006
- October 2006
- September 2006
- August 2006
Most commented posts
Highest Rated Blog Posts
- The difference between a Parapedal and a Pedalflanger (100%)
- CrushFTP 4 (100%)
- CD Ripping Rip Off (100%)
- Another Ripping Rip off (100%)
- That'll fool 'em (100%)
- A bad Tool always blames the browser (100%)
- Reasonably priced car hit by star (100%)
- 10.5.3 and Time Machine (100%)
- Carbon Dating (100%)
- Rough Wiis do shake the darlings (100%)

