-3

Not looking for opinions. I'm searching for data.

As it is now, I want to become a webM evangelist. However, I assume there are some actual technical reasons why mp4 is preferred over webm in the bigger picture. I want to know them so I can be accurate in my assessments.

I'm working on developing a broadcast video messaging graphics engine (think chyron) and using the Chromium engine like OBS does for messaging. So far the results have been excellent.

One of the best features I've found is using webm for video. I should note I am using small (640x480 max) videos as graphics that are on top off a larger full HD video.

Not only does it seem to have a better compression:quality ratio than mp4 for my use case, the most important thing is that it has full alpha support, which allows for excellent layering of video objects on top of each other in the HTML DOM, in real time, with no noticable performance hits.

Aside from it's predecessor, FLV, I can't think of another high quality, high compression codec that also supports alpha. I feel like you are stuck using pro-res 4444 or the ancient animation codec to reliable distribute video with an alpha.

So, that said, are there technical reasons why webM isn't more adopted than mp4?

I already know the obvious, that there is dedicated hardware to decode mp4. But, is there any technicality that would prevent a hardware webM decoder? I really want to understand more what the benefits of mp4 are over webM, which i assume is why it is more widely used than webM.

Thanks!

marc_s
  • 732,580
  • 175
  • 1,330
  • 1,459
Dani
  • 1,220
  • 7
  • 22

1 Answers1

2

You are confused. mp4 and webm (and flv) are containers, but you are comparing the attributes of the codecs they contain. For the most part, mp4 can contain every codec webm can, and dozens more. It’s also decades older, making it more accessible and more supported.

If you are trying to compare avc to vp9, the different is availability of h264 decoders in devices is much greater. Vp9 is also often decoded in software, reducing battery life. And for encoding, avc is much (much much) faster, hence much cheaper. So the only justification is if you can encode vp9 once, and send the same copy to thousands of viewers to make up the cost in reduced bandwidth. If only a couple hundred people download the video, vp9 cost more than it saved.

Finally, change is expensive and time consuming. There must be clear and justifiable reasons to do so.

szatmary
  • 29,969
  • 8
  • 44
  • 57
  • thanks for the reply. Not confused, just using the vernacular. While technically correct, avc and mp4 are basically synonyms at this point, I've never seen a web encoded mp4 that is anything other than avc. As I asked in my question, is there any reason why VP9 couldn't be decoded on a dedicated IC chip? As dynamically generated video becomes more prevalent, the need for distributed video to include an alpha channel becomes increasingly relevant, does it not? – Dani Jun 04 '19 at 21:48
  • Mp4 and avc are incorrectly synonymous. I have done vp9 in mp4 on a large streaming site, and it is becoming very common. Yes you COULD make a vp9 hardware decoder. But h264 has a 10 year lead and hundreds of million of devices in the field already. And as far as alpha becoming increasingly relevant, I have not had any customer requests for that. – szatmary Jun 04 '19 at 21:53