So, some comparison points:
* Dirac is able to achieve very high quality at accordingly high bitrates (it can even be run losslessly); Theora has a maximum quality which is not quite unconditionally visually lossless.
* Dirac is more cpu intensive than Theora.
* Dirac tools are somewhat less mature (In particular; there exists a decently performing Theora decoder in Java, which means that a majority of internet users can video Theora today with no software install)
* With the encoders existing today Theora clearly out-performs Dirac for low to moderate bitrates.
Which codec you should use today depends on your application. If you are archiving video, building a unencumbered-bluray, or squeezing production grade 1080 HD video into 270mbit SD channels, Dirac is what you want. If you are webcasting with under a couple of megabits per second, Theora is what you want not only will it have better quality at the applicable bitrates, but you have more playback options. (Though you might also wish to offer a high quality dirac download option
For example I recently put a Theora video online at http://www.celt-codec.org/presentations/ At 150kbit/sec Theora is perfectly adequate for this application while the slides are not always readable with the current Dirac 'Schroedinger' encoder at that bitrate. Yet a full resolution version at several megabits per second looks better in Dirac than Theora. Unless you're using Firefox 3.5beta the video is only playing for you on that page because the page is able to fall back from the the video tag to a java based decoder, something which isn't yet possible for Dirac. (And might never be reasonable due to CPU use?) I'm planning on putting up a Dirac download version
but I'm still playing around with the encoder. (
and it's a fact of life that no one manages to do multiple format parallel distribution well, myself included)
The gap between Theora and Dirac at the lower bitrate end of the spectrum may persist indefinitely as thats really what theora was designed for
future improvements to Dirac encoders may close the gap but it is difficult to speculate about that as that isn't an active area of development for Dirac today.
If I had to place a bet I'd guess that Dirac would continue to be used for "artefact free and better" video for the foreseeable future; the use of which will increase over time as IP transit costs decrease and bandwidth becomes more plentiful, while Theora will be obsoleted in a number of years by a replacement or an extension
It's only natural for lower bitrate distribution formats to have a shorter life compared to archive grade formats, and there is more to be gained from small improvements in formats used where every kilobit counts.
Copyright © 2018, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds