Bitmovin raises $30M Series B for ‘next-gen’ online video software | Hacker News Reader

Bitmovin raises $30M Series B for ‘next-gen’ online video software

techcrunch.com

64 pointsjonballant24 comments

Thx folks, it has been a awesome ride so far. That's only possibly due to the great Bitmovin team. (I'm one of the cofounders)

Thanks for leaving a comment here :)

Now that I have your ear, may I make a suggestion? Please be a little more specific about what Bitmovin does on your website. Right now the first thing I read is:

    "Software to Solve Complex Video Problems.
    Bitmovin API based products help developers around the world
    solve the most complex video problems with cloud native software that runs anywhere"

That is a very broad statement. I would consider myself quite technical but I don't have deep domain knowledge in video coding and infrastructure. Quite honestly I can't get much information out of this.

What kind video problems are you solving (except that they're "most complex")? What is "cloud native software"? How do you define "runs anywhere"? Will it run on my Smartphone / PC / Car / Container / Toaster?

I just want to know what your company does [1]

[1] https://news.ycombinator.com/item?id=15170182

some of the problems that they are solving are

- inserting ads between videos(configurable)

- serving videos for different screen size(if you send same video to all device, device has to do resizing in addition to decoding)

- serving videos in different network conditions(HLS/DASH) where based on bandwidth lower/higher bitrate video is requested from the server.

- seamlessly serving videos by switching video sources even for encrypted content.

- In HLS/DASH, player is the complex(intelligent) part as it senses the bandwith and send requests to server, they are giving readymade player implementation through SDK's

Most of those problems are already solved by open source tools and standard on video streaming services like amazon, netflix, hbogo etc.

There are mature open source clients that will do most of those things automatically out of the box. For example shaka player from Google.

Better not to depend on client for functionality.

Don’t repeat the “JS required” situation in video, or next thing you know you’re rederiving Flash.

Congrats! It's very encouraging to see video tech with a solid customer base, and good investor support.

Which factors do you think contributed most to your success? Was YC helpful in opening doors for you? How important is the right kind of investor?

Thanks!

For us it was really important to find people that help us along the way, and that definitely includes all our investors, as well as Ycombinator. We're engineers and we definitely learned so much about building a product and a go to market strategy during our time at YCombinator. Having a technology alone is often not enough.

We also raised money form some silicon valley veterans and angels, and I use them as a resource for advice and expertise. One of the was the first CTO of Cisco for example, and I always learn so much in talking with him. Pretty cool to see how one can apply things from those companies also today.

But most importantly, it's the team. you need a great to come to this point, otherwise you struggle quite early. I'm really proud to work with our folks, they really do all the work.

Thank you for the tips - and best of luck to you guys!

Congrats, but I never understood why they don't build their own storage and CDN solution. Infrastructure is hard, especially operating a video cdn, but focusing on a video player, that is easily replaced with another solution or an encoding service that can be build inhouse way cheaper are (for me personally) no selling points.

But apparently that's enough :)

The premise for MPEG-DASH is that infrastructure is easy: Just small files transferred over HTTP. You can use every CDN which supports HTTP for that.

Live encoding is surprisingly hard: You have to get a live input signal instead of a file and transcode it on-the-fly to DASH with minimal latency meaning:

- encoders in the vicinity of the input signal,

- multiple input signals,

- efficiently transport a high data rate from a venue (with often a subpar connection) to your encoders,

- support many different input signal formats (i.e. camera output signals),

- support very high data rates etc.

Some companies do it to a varying degree but not that many like Wowza, Elemental (bought by AWS and put into AWS Elemental Delta), Adobe, Bitmovin, …).

Yes, but they are putting together a huge sales team, their customers are using their video player and a video encoding service. That means they also have the need for video storage, a video CDN, etc. So they waste a lot of potential.

Building a encoding service in the cloud era, even for live video, isn't really that hard. The hard part is that every customer has different needs and different source files, if you build it inhouse you are only building it for one customer, yourself.

Commodity storage and delivery prices have been declining steadily. Building something that is scalable but also profitable, is hard.

Still, selling is the hardest part, in my opinion. Source: run a global video cloud.

So you are exactly confirming what I am saying, they have or are building the sales team anyways.

One thing I didn't make clear, that may explain why somebody just wouldn't build their own storage/delivery: when certain services become a commodity, and long-term commitments become an exception rather than the norm - it's easier to shift the pricing pressure and uncertainty for the commodity side to your providers. You specifically don't want to "build your own" in a market like that, but instead focus on specialized, high-margin services. One exception: when you can predict/control the demand (e.g. you are a Netflix), then you build your own, and still leverage providers for capacity on tap.

The one part of live encoding that needs innovating is RTMP for ingest. It's crazy to me that this lousy protocol is the first step for most live video.

You can already use webrtc for that.

Funny how they already support the AV1 encoding but neither the encoding is finished nor the hardware implementation is done.

It's always interesting to see how local startups succeed internationally (especially the ones founded in such rural areas as Carinthia).

Also I just recently was at a talk with one of their engineers and learned how they scaled their monitoring services; quite fascinating...

Keep up the great work (and see you at the next OpenTechTable).

Hopefully they start offering a free tier soon (dev-type). Always nice to have, and by letting devs use it in personal projects, they often carry the API experience and preference to their jobs.

Pied Piper has got some serious competition!

haha, yes :-)

Interesting if 'Bit' prefix has helped?

I had the same thought that with the latest Crypto craze the "Bit" prefix probably helps them stand out, but as the founders are the co-creators of MPEG-DASH which is used by Netflix, they are well known enough to not need the benefit of the crypto "bump". Plus the company was founded back in 2013.

Yeah the company was founded 5 years ago, with the same name.

Reply on news.ycombinator.com