Late...

[AnthonyFlores]

Fasincating analysis Thomas! If your numbers are even halfway accurate, it not only helps to explain why things are so delayed, but also why Red is doing something with the Epic (and Scarlet) that truly has never been done -- or ever will be done -- by a camera company, in history.

What other camera company -- much less owner -- would invest hundreds of millions of dollars out of pocket to gamble on developing a radically new technology like this???

Wow, I don't mean to flatter ... but this truly does take "epic" balls.

I know it's been said before on this thread and many times elsewhere, but like others I can't help but think we're witnesses something unprecedented in the history of cameras, cinema and even business at large.

It's just now sinking in for me ... :blushing:

Thanks Thomas, I hope more people get the implications of your post.

Anthony

 

[Stefan Antonescu]

The skill and commitment of the Red development team is no doubt keeping the cost way below this estimate, but at this point I can safely assume that Mr. Jannard has laid a sizable fraction of his personal fortune (see article on World's richest people http://www.forbes.com/lists/2006/10/S37W.html ) on the table to make this happen.

Great analysis Thomas.

IMHO, Jim would invest 100% of his fortune if needed to make this happen.

 

[Peter Moretti]

I think we all have to have a new appreciation for what Sony, Arri, et al have done in the digital cinema arena thus far.

Cleary it's not as easy or as cheap as many have speculated to make a digital movie camera and workflow.

 

[Jacek Zakowicz]

raise

raise

I'm sure Mr J will appreciate this analysis when Graeme and Deanan will come for a raise with the printout:001_tongue:

 

[Michael Hastings]

Fasincating analysis Thomas! If your numbers are even halfway accurate,

Anthony

With all due respect, I am highly skeptical that Thomas' numbers are even halfway accurate.

First of all, as "mission critical" as we may think our productions are - they aren't truly "mission critical" like that of an airliner, or missile system, or other defense systems where hundreds, thousands, or even millions of people can die if there is a glitch - we can in fact survive a "crash" on set - and that fact alone keeps "normal" programming probably a tenth of the cost per SLOC.

And although I'm sure there is a tremendous amount of new custom code, there is also probably a substantial percentage of the code that is available off the shelf, or maybe even already built in to many layers of the ASIC as there are numerous functions that are similar to many platforms. Video processing is already highly advanced and used in millions of cameras, TVs, set top boxes, etc. for which there are numerous code libraries and chip designs already available - and again these are much easier to integrate when a camera reboot, or picture artifact is the biggest "mission critical" danger.

I have no doubt that there is a very expensive engineering team, and that Epic will be the finest, most advanced camera our industry has ever seen but I'm not sure we should to put it in the same category as the Manhattan Project or Apollo program.

 

[Lee Fingersh]

You're right, Michael.

I spent some time at a prominent defense contractor in the area where they write mission critical control software (in this case, jet engine controllers). It's quite the process. They were averaging 1 bug per 4000 lines of new code, which is astoundingly low. The production rate of code was extremely low, but the code was extremely reliable, extremely well tested, and in many cases proven to be reliable. 1 hour per line of code seems about right based on what they told me.

On the other hand, I've also spent time in commercial software production places. Production of finished code was at a rate of 30,000 to 60,000 lines per person per quarter, or about 100 times faster than the mission-critical code. This was at a time where libraries were far less developed than they are now.

Like you, I suspect that RED is in the second category, and they are relying on libraries far more than defense contractors do for mission critical software. Nothing wrong with that - apply the tools appropriate to the task.

 

[Gavin Greenwalt]

Probably a fraction in development but even half of the mission critical would be affordable over several product lines I imagine.

10,000+ REDOnes sold. Scarlet will probably sell 10x that many. 100,000 x $10,000 would be $1.0B in revenue for the 'inexpensive' scarlet. Not sure what the cost of manufacturing is but that's ignoring the S35 scarlet, FF35 scarlet and Epic line which all share a lot of the development expenses I imagine for the firmware and chip design.

 

[brian hanson]

Probably a fraction in development but even half of the mission critical would be affordable over several product lines I imagine.

10,000+ REDOnes sold Scarlet will probably sell 10x that many. 100,000 x $10,000 would be $1.0B in revenue for the 'inexpensive' scarlet. Not sure what the cost of manufacturing is but that's ignoring the S35 scarlet, FF35 scarlet and Epic line which all share a lot of the development expenses I imagine for the firmware and chip design.

let's not forget about accessories with these modular cameras

 

[Gavin Greenwalt]

let's not forget about accessories with these modular cameras

Well I was assuming accessories. $5k + all your official RED branded T-shirts, hats and module merchandise.