Mobile Web based games may have taken a big step towards the mainstream today. HTML5 development company appMobi benchmarked its new directCanvas to accelerate graphics and found that on average, iOS devices performed at least 10-times better than before.

AppMobi may have also figured out how to deal with the HTML5 sound problem. One of the biggest headaches with the HTML5 spec is that it can only play one sound at a time. AppMobi is now talking a “multi-channel sound” approach that will give developers the chance to layer sounds on top of each other for a more dynamic audio experience. See the results of the tests below.

AppMobi used ImpactJS and Box2D created two identical apps to use HTML5 and JavaScript to simulate an animated game with certain numbers of on-screen entities. One app uses the standard HTML5 Canvas while the other uses directCanvas. Frame rate was then tested by how many redraws there were per second. The results of this test are not absolute but do show the trend of how each Canvas solution performs.

The company claims it is seeing acceleration rates between 728% and 1525% on the spectrum of iDevices from the 3GS to the iPhone 4S running the mobiUs browser extension. The iPhone 4S without directCanvas running 50 entities on the screen ran at three frames per second. With directCanvas it ran at 38 frames per second. Those numbers were pretty standard across the iDevice spectrum. See the chart below.

The test apps had a maximum of 40 frames per second. Each test was run with no other apps running or background tasks. See the company’s blog post for full results.

While the acceleration of graphics within games will be an important innovation in HTML5 development, perhaps the more important aspect here is the ability to run multiple sounds at once through directCanvas. This should help game developers create sounds that are synced with game events but the uses for HTML5 layered sound go way beyond the needs and wants of game developers.

What do you think of appMobi’s breakthroughs? Do you trust the numbers the company has come up with? What would you change about the conditions of the test? Let us know in the comments.