Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Leonardo Chiariglione – Co-founder of MPEG

leonardo.chiariglione.org

Post reply on HN
Print view View original

41–50 of 209 posts

Re: Leonardo Chiariglione – Co-founder of MPEG

#41

Post by mike_hearn »

pornel wrote: post #22
mike_hearn wrote: post #4

Earlier quoted context omitted.

Possibly, nothing. Codec development is slow and expensive. Free codecs only came along at all because Google decided to subsidize development but that became possible only 15 years or so after MPEG was born, and it's hardly a robust strategy. Plus free codecs were often built by acquiring companies that had previously been using IP licensing as a business model rather than from-scratch development.

IP law, especially defence against submarine patents, makes codec development expensive. In the early days of MPEG codec development was difficult, because most computers weren't capable of encoding video, and the field was in its infancy. However, by the end of '00s computers were fast enough for anybody to do video encoding R&D, and there was a ton of research to build upon. At that point MPEG's role changed from b…

IP law and the need for extremely smart people with a rare set of narrow skills. It's not like codec development magically happens for free if you ignore patents.

The point is, if there had been no incentives to develop codecs, there would have been no MPEG. Other people would have stepped into the void and sometimes did, e.g. RealVideo, but without legal IP protection the codecs would just have been entirely undocumented and heavily obfuscated, relying on tamper-proofed ASICs much faster.

Re: Leonardo Chiariglione – Co-founder of MPEG

#42

Post by mike_hearn »

Taek wrote: post #23
mike_hearn wrote: post #4

Earlier quoted context omitted.

Possibly, nothing. Codec development is slow and expensive. Free codecs only came along at all because Google decided to subsidize development but that became possible only 15 years or so after MPEG was born, and it's hardly a robust strategy. Plus free codecs were often built by acquiring companies that had previously been using IP licensing as a business model rather than from-scratch development.

I avoided a career in codecs after spending about a year in college learning about them. The patent minefield meant I couldn't meaningfully build incremental improvements on what existed, and the idea of dilligently dancing around existing patents and then releasing something which intentionally lacked state-of-the-art ideas wasn't compelling. Codec development is slow and expensive becuase you can't just release a n…

Well, a career in codec development means you'd have done it as a job, and so you'd have been angling for a job at the kind of places that enter into the patent pools and contribute to the standards.

Re: Leonardo Chiariglione – Co-founder of MPEG

#43

Post by mike_hearn »

newsclues wrote: post #8
mike_hearn wrote: post #4

Earlier quoted context omitted.

Possibly, nothing. Codec development is slow and expensive. Free codecs only came along at all because Google decided to subsidize development but that became possible only 15 years or so after MPEG was born, and it's hardly a robust strategy. Plus free codecs were often built by acquiring companies that had previously been using IP licensing as a business model rather than from-scratch development.

This is the sort of project that should be developed and released via open source from academia. Audio and video codecs, document formats like PDF, are all foundational to computing and modern life from government to business, so there is a great incentive to make it all open, and free.

Universities love patent licensing. I don't think academia is the solution you're looking for.

Re: Leonardo Chiariglione – Co-founder of MPEG

#44

Post by cornholio »

pornel wrote: post #22
mike_hearn wrote: post #4

Earlier quoted context omitted.

Possibly, nothing. Codec development is slow and expensive. Free codecs only came along at all because Google decided to subsidize development but that became possible only 15 years or so after MPEG was born, and it's hardly a robust strategy. Plus free codecs were often built by acquiring companies that had previously been using IP licensing as a business model rather than from-scratch development.

IP law, especially defence against submarine patents, makes codec development expensive. In the early days of MPEG codec development was difficult, because most computers weren't capable of encoding video, and the field was in its infancy. However, by the end of '00s computers were fast enough for anybody to do video encoding R&D, and there was a ton of research to build upon. At that point MPEG's role changed from b…

That's unnecessarily harsh. Patent pools exist to promote collaboration in a world with aggressive IP legislation, they are an answer to a specific environment and they incentivize participants to share their IP at a reasonable price to third parties. The incentive being that you will be left out of the pool, the other members will work around your patents while not licensing their own patents to you, so your own IP is now worthless since you can't work around theirs.

As long as IP law continues in the same form, the alternative to that is completely closed agreements among major companies that will push their own proprietary formats and aggressively enforce their patents.

The fair world where everyone is free to create a new thing, improve upon the frontier codecs, and get a fair reward for their efforts, is simply a fantasy without patent law reform. In the current geopolitical climate, it's very very unlikely for nations where these developments traditionally happened, such as US and western Europe, to weaken their IP laws.

Re: Leonardo Chiariglione – Co-founder of MPEG

#45

Post by unlord »

thinkingQueen wrote: post #19
rwmj wrote: post #13

Earlier quoted context omitted.

Who would write a web server? Who would write Curl? Who would write a whole operating system to compete with Microsoft when that would take thousands of engineers being paid $100,000s per year? People don't understand that these companies have huge R&D budgets! (The answer is that most of the work would be done by companies who have an interest in video distribution - eg. Google - but don't profit directly by selling…

Are you really saying that patents are preventing people from writing the next great video codec? If it were that simple, it would’ve already happened. We’re not talking about a software project that you can just hack together, compile, and see if it works. We’re talking about rigorous performance and complexity evaluations, subjective testing, and massive coordination with hardware manufacturers—from chips to displa…

> People don’t develop video codecs for fun like they do with software. And the reason is that it’s almost impossible to do without support from the industry.

As someone who lead an open source team (of majority volunteers) for nearly a decade at Mozilla, I can tell you that people do work on video codecs for fun, see https://github.com/xiph/daala

Working with fine people from Xiph.Org and the IETF (and later AOM) on royalty free formats Theora, Opus, Daala and AV1 was by far the most fun, interesting and fulfilling work I've had as professional engineer.

Re: Leonardo Chiariglione – Co-founder of MPEG

#46

Post by mike_hearn »

rwmj wrote: post #13
thinkingQueen wrote: post #9

Earlier quoted context omitted.

Who would develop those codecs? A good video coding engineer costs about 100-300k USD a year. The really good ones even more. You need a lot of them. JVET has an attendance of about 350 such engineers each meeting (four times a year). Not to mention the computer clusters to run all the coding sims, thousands and thousands of CPUs are needed per research team. People who are outside the video coding industry do not un…

Who would write a web server? Who would write Curl? Who would write a whole operating system to compete with Microsoft when that would take thousands of engineers being paid $100,000s per year? People don't understand that these companies have huge R&D budgets! (The answer is that most of the work would be done by companies who have an interest in video distribution - eg. Google - but don't profit directly by selling…

Google funding free stuff is not a real social mechanism. It's not something you can point to and say that's how society should work in general.

Our industry has come to take Google's enormous corporate generosity for granted, but there was zero need for it to be as helpful to open computing as it has been. It would have been just as successful with YouTube if Chrome was entirely closed source and they paid for video codec licensing, or if they developed entirely closed codecs just for their own use. In fact nearly all Google's codebase is closed source and it hasn't held them back at all.

Google did give a lot away though, and for that we should be very grateful. They not only released a ton of useful code and algorithms for free, they also inspired a culture where other companies also do that sometimes (e.g. Llama). But we should also recognize that relying on the benevolence of 2-3 idealistic billionaires with a browser fetish is a very time and place specific one-off, it's not a thing that can be demanded or generalized.

In general, R&D is costly and requires incentives. Patent pools aren't perfect, but they do work well enough to always be defining the state-of-the-art and establish global standards too (digital TV, DVDs, streaming.... all patent pool based mechanisms).

Re: Leonardo Chiariglione – Co-founder of MPEG

#47

Post by tdullien »

karel-3d wrote: post #15

I... don't understand how AI related to video codecs. Maybe because I don't understand either video codecs or AI on a deeper level.

Every predictor is a compressor, every compressor is a predictor.

If you're interested in this, it's a good idea reading about the Hutter prize (https://en.wikipedia.org/wiki/Hutter_Prize) and going from there.

In general, lossless compression works by predicting the next (letter/token/frame) and then encoding the difference from the prediction in the data stream succinctly. The better you predict, the less you need to encode, the better you compress.

The flip side of this is that all fields of compression have a lot to gain from progress in AI.

Re: Leonardo Chiariglione – Co-founder of MPEG

#48

Post by yxhuvud »

mike_hearn wrote: post #43
newsclues wrote: post #8

Earlier quoted context omitted.

This is the sort of project that should be developed and released via open source from academia. Audio and video codecs, document formats like PDF, are all foundational to computing and modern life from government to business, so there is a great incentive to make it all open, and free.

Universities love patent licensing. I don't think academia is the solution you're looking for.

The solution to that is to remove the ability to patent codecs.

Re: Leonardo Chiariglione – Co-founder of MPEG

#49

Post by scott_w »

thinkingQueen wrote: post #19
rwmj wrote: post #13

Earlier quoted context omitted.

Who would write a web server? Who would write Curl? Who would write a whole operating system to compete with Microsoft when that would take thousands of engineers being paid $100,000s per year? People don't understand that these companies have huge R&D budgets! (The answer is that most of the work would be done by companies who have an interest in video distribution - eg. Google - but don't profit directly by selling…

Are you really saying that patents are preventing people from writing the next great video codec? If it were that simple, it would’ve already happened. We’re not talking about a software project that you can just hack together, compile, and see if it works. We’re talking about rigorous performance and complexity evaluations, subjective testing, and massive coordination with hardware manufacturers—from chips to displa…

> Are you really saying that patents are preventing people from writing the next great video codec?

Yes, that’s exactly what people are saying.

People are also saying that companies aren’t writing video codecs.

In both cases, they can be sued for patent infringement if they do.

Re: Leonardo Chiariglione – Co-founder of MPEG

#50

Post by mike_hearn »

roenxi wrote: post #32
thinkingQueen wrote: post #14

Earlier quoted context omitted.

That’s hardly true. Nvidia’s tech is covered by patents and licenses. Why else would it be worth 4.5 trillion dollars? The top AI companies use very restrictive licenses. I think it’s actually the other way around and AI industry will actually end up following the video coding industry when it comes to patents, royalties, licenses etc.

Because they make and sell a lot of hardware. I'm sure they do have a lot of patents and licences, but if all that disappeared today it'd be years to decades before anyone could compete with them. Even just getting a foot in the door in TSMC's queue of customers would be hard. Their valuation can likely be justified based on their manufacturing position alone. There is literally no-one else who can do what they do, l…

The top AI companies don't release their best models under any license. They're not even distributed at all. If you did steal the weights out from underneath Anthropic they would take you to court and probably win. Putting software you develop exclusively behind a network interface is a form of ultra-restrictive DRM. Yes, some places are currently trying to buy mindshare by releasing free models and that's fantastic, thank you, but they can only do that because investors believe the ROI from proprietary firewalled models will more than fund it.

NVIDIA's advantage over AMD is largely in the drivers and CUDA i.e. their software. If it weren't for IP law or if NVIDIA had foolishly made their software fully open source, AMD could have just forked their PTX compiler and NVIDIAs advantage would never have been established. In turn that'd have meant they wouldn't have any special privileges at TSMC.

Post reply on HN