Back To Listings RSS Print

Apple’s new iFrame codec implementation

Some perspective on this puzzling new format

By Matthew Jeppsen | October 25, 2009


So recently Apple released specs on a new video format they are calling iFrame. What is it, and how will it affect you, the video professional? As all consumer codecs seem to trickle into the professional space in one way or another, I expect the same to happen with this format. So here's some info to provide perspective in case you are called upon to deal with the codec by clients. Read on...

First of all, what is it? Well, iFrame is first and foremost a confusing name choice. PVC's Allan T©pper has a good post on this.

Technically speaking, iFrame is not a new codec, it is a implementation of an existing proven video format, MPEG-4. It's a wrapper, a collection of settings. iFrame is an i-frame only a spec of H.264 that also uses AAC audio Update: (I cannot locate an official reference that iFrame is i-frame only, so until we can get an official spec from Apple it's safe to assume that iFrame also includes P and B frames, and is not much like AVC-Intra after all). iFrame is locked at 960x540, apparently at 30p, so resolution is 1/4 of 1080P HD. It appears to be an option that offers better-than-DV resolution that takes the proven efficiencies and advantages of H.264, but gives you something that's a little more edit-friendly. Clearly Apple is thinking this will be implemented on consumer solid-state camcorders, and they've already added iMovie support. One way to think of it is like AVC-Intra, only aimed squarely at the consumer editing space. This post over at Discrete Cosine sums the editing advantages up nicely.

For a little more vitriol towards Apple's new format, head on over to occasional-FreshDV-contributor Eugenia's blog and check out how SHE would have done it if she was Apple. It's an interesting counterpoint. My personal opinion is that this was intended for the $150 camera Youtube generation, where a 540p source is perfectly acceptable. Update: Consider also that Apple has a huge stake in the mobile market, and a 540p file looks rather nice on an iPhone. It's also a lot less processor intensive to edit than a comparable 720p or 1080p source. What do you think?

Editor's Choice
PVC Exclusive
From our Sponsors

Share This

Back To Listings RSS Print

Get articles like this in your inbox: Sign Up


Allan T: | October, 25, 2009


Thanks for the cross-post!
I am curious to know how you know that the Apple/Sanyo iFrame format actually uses i-frame encoding (as opposed to long-GOP encoding). I don’t know for sure at this point, but I’m more inclined to believe that it was more likely to be progressive long-GOP. I wouldn’t be at all surprised if the marketing people at Apple and/or Sanyo didn’t have any idea what i-frame encoding is, or what the HTML IFRAME command or implementation is. I haven’t found any reference to i-frame encoding in any of the Apple or Sanyo specs so far. Have you? I would really like to know for sure.

Allan T

Eugenia: | October, 25, 2009

Thanks. I think that 540p is not enough for the Youtube generation either though. Now that Youtube supports HD, requires a minimum of 720p to re-encode in HD. I don’t think that consumers can be fooled with a non-HD re-encoding for too long.

Jim Hines: | October, 26, 2009

Eugenia _ I found your blog as a result of this article. You’re all right and you do good work. Rock on.

Matt Jeppsen: | October, 26, 2009

Hey Allan, thanks for asking about the i-frame thing. I’ve gone back and double checked, and now I’m not so certain it’s an i-frame only implementation of MPEG-4. I’ve updated the post accordingly.

Bjarkovic: | October, 29, 2009

Wouldn’t being rid of the long-GOP encoding potentially make the codec better suited for editing and post? I still shudder when I think of my first HDV edit. A music video which took FOREVER and then some. I know today’s machines are more powerful, but especially if it’s geared toward consumers/prosumers, you’d think that the specs shouldn’t have to be too extreme. Any thoughts?

Allan T: | October, 29, 2009


Yes, it certainly would. However, doing so either means either a higher bit rate or subsampling (throwing away pixels). Please see my article WHEN TO EDIT NATIVE, WHEN HYBRID, WHEN PURE I-FRAME, AND WHY ( for the historical examples about that.

Allan T

Bjarkovic: | October, 30, 2009


Thanks for the link. I actually had an interesting conversation with a Sony rep at IBC a couple of years ago, having unintentionally waded into a debate about codecs. All present agreed that the all codecs are good at one thing. Being kind to the camera or kind to the editing process. The space-friendly long-GOP formats were really handy for the cameras, taking up less space and delivering good images, but were hell on editors. At the same time nobody was really pushing the post production-friendly codecs because they weren’t a viable option in the field. I think ProRes may have changed the way we think about them, even if some people may not like it. If only for having united so many formats, it’s been a godsend for most of us.


Please login or register to comment