Any thoughts on proposed MOX open source audio-video interchange file format?

S

Shenan

Well-known member
I haven't seen this discussed anywhere here. Apparently there was a discussion on CML, but I don't usually read that.

http://provideocoalition.com/mchristiansen/story/mox-and-open-source-video

Everything that I've read so far sounds pretty interesting and seems to make sense. The only thing I wonder about is if it will result in much larger files than ProRes or DNxHD for similar quality, and if that would hinder adoption. I would love to hear what other folks here, especially the more technically minded, think about it.
 
Summary:

A programmer and visual effects artist named Brendan Bolles is seeking $20,000 to finish an open audio and video file format called MOX, meant to be traded easily among editing programs and computers (Windows, Mac, and Linux). Instead of reinventing the wheel, he's just bringing together a subset of all the open stuff out there:

Container format: MXF
Video codecs: Dirac, OpenEXR, DPX, PNG, JPEG
Audio codecs: FLAC, Opus, PCM

(fundraising page)


My take:

I like it! I like Linux and hope to use it more. This project could finally bring about a standard version of MJPEG, for example (which believe it or not was never formally specified). MJPEG, to me, is a nice balance of quality, interoperability, and file size --- sort of like ProRes. ProRes and MJPEG both use DCT, and I think that MJPEG could do ProRes quality --- although I'm not sure if it would be quite as small in file size. But it's a little-known fact that JPEG actually supports 10- and 12-bit color and 4:2:2.

It kind of sounds like a lightly compressed sister of CinemaDNG, for those who would rather use something that looks 95% as good but takes 1/10 the space.
 
Last edited:
I'm backing MOX, I think it's something we do need. As a great example of why I came across this article this morning:
http://provideocoalition.com/ryoung/story/apple-prores-4444-xq-in-creative-cloud

Basically even Apple on Mac OS isn't letting advanced versions of Quicktime ProRes out!

On Windows Quicktime is pretty much completely broken. 32 to 64 bit bridging of 4k frames is insanity. AVI works great on Windows but has nearly no Mac support and no camera support at all. Apple seems determined to cripple Quicktime on Windows. We need a cross platform, open container format. Remember MOX isn't a codec. It's a container format and a small set of supported codecs (all based on open things).

I really do hope it works out. I'm actually hoping it gets funded so I can try writing a plugin for eyeon Fusion to support it. Maybe even a DirectShow filter (which I suspect will be very important). I'll admit I am worried about how great it will be at playback. Getting EXR frames to decompress fast enough to be able to use them in a media player and an NLE may be harder than it looks. Same with DPX, PNG, JPG. Dirac should be workable, since that's what it was designed for. But the still image formats may prove a bit harder.

I've done work on code like this and I suspect there will be some major headaches along the way. But right now with Apple holding all the keys to Quicktime and using it as a weapon against Windows and Linux platforms we really do need to do something. Even if MOX initially fails (to get funding, to get to a usable project) perhaps it will help wake up some corporate interest. Having something like it seems like it would be a win for just about every company in the post production business except Apple.
 
Thank you for your thoughts, combat and Matt, and for your summary, combat! Regarding a DirectShow filter, does that make it available as a general codec in Windows? I know that some NLEs, like Vegas, make use of whatever codecs you have available to the Windows OS, so by doing something like that it would become available instantly to Vegas and possibly other NLEs. That's one of the first things I hoped for when I read about the project, but I figure that even if he doesn't do it himself someone else will!
 
Having support for it in DirectShow I believe will allow Windows Media Player to play the files. My hope is that it will allow Zoom Player PRO (my choice for media playback) to play the files, without having to write some sort of custom plugin. But since it's not a codec I don't think it will enable things like Vegas necessarily. But Vegas has some support for DirectShow, which might add up to enough so that the new file format is available. To be honest I'm not sure. I just know I need the ability to double click and have MOX files open in Zoom Player PRO. I think I can get that with some kind of DirectShow filter support.

It's tricky because MOX isn't a codec, it's not a choice that appears when making an AVI or a Quicktime file, it's a whole new file format. Getting support in various applications will almost always require some sort of new code. For applications with advanced plugin support that new code can just be added on. But that won't work with lots of things.
 
Thanks again, Matt. I supported the project, so today I received an invitation to join the private google group for it. There are some good discussions going on. Getting a little over my head technically, but I see some smart people having some smart discussions about the details of implementation. I'm a bit surprised that there hasn't been more discussion about this in other places, like here and in a couple of computer graphics forums that I also frequent.
 
It's very hard to support something like this on most of the better forums/communities, like DVXUser for example. It not clear if pushing someone's Kickstarter/Indiegogo project isn't just spamming. Even when it's something like this, that may very well play a huge role in what it means to do post production on video in the future.

I'm very encouraged by the conversation on the google group. If you look at some of the file formats that define our digital lives, WAV, JPG, PNG, MP3, even GIF had some form of corporate/academic backing in order to get going. None of them happened over night. All of them actually required a fair amount of work. But now they (and others like them) form the basis of a vast economic market. A decade ago there would be working taskforces and other standards bodies that would do work like this, but lately things have changed. Almost all the standards are now coming from high-level corporate sponsorship and then being setup as a standard. It's because it takes money to develop this kind of specification and the middleware/software to go with it.

This crowdfunding of a standards process is a very interesting take on this kind of problem. No clue how well it's going to work out just yet, even if everything goes perfect this is a multi-year kind of effort. But we do need something like this, so I'm rooting for it.
 
Agreed with everything you said. The only thing I would add is that at least he isn't starting from scratch - he recognizes that that would be an impossible effort at this (money/time) scale. I think that his goal of putting together existing codecs and container formats in a predictable and easily reproducible fashion is reasonable and likely attainable.
 
Yeah, so that means his initial targets will be:

1. Open-source specification and library
2. Premiere plug-in
3. After-Effects plug-in
4. Nuke plug-in
5. Integration into as many other programs as possible, especially open-source ones like FFMPEG, etc

Let's wish him luck and see how it goes!
 
You must log in or register to reply here.
Back
Top