Feature request: Red Code Web

Well, like most of you, I have been dreaming of features I would like to see on my Red One. I didn't make it in time for the contest, but I don’t need a contest to have me think of things I’d like to see in my future camera. So, here's my idea:

If the camera could be an on-set web server, either wireless or wired, where a script supervisor could use “Red Code Remote” software, running on a laptop, to post scene information to the camera (Via HTTP Post) which would then be added to the next shot as metadata, that would eventually be handed over to a NLE, pre-logging all of the shots. This onset pre-logging would save tremendous amounts of time in post.

Once the shot has been tagged, users of the remote software should be able to update the shots still on the attached hard drive with more details that may come up after the shot is done (director notes, quality, etc).

Ideally, this “Red Code Web” server would also stream an MP4 video stream in an IP broadcast to whoever is attached to the “Red Net”. This would allow for N number of participants to record on their own laptop/tablet computers an on-set, NLE view, of the footage without video cables.

The remote shot logging software should be able to store locally DV quality footage if necessary for on-set playback to check continuity or performances, without having to go to the camera. Down the line, software that included the ability to store pre-vis shots and replace them with real footage (that is connected via the metadata to the actual footage) would be an amazing addition to the workflow.

So, the script supervisor logs and tags, and everyone else getting the feed (director, producers) get a fully tagged version that they can record onto their own playback system so they can replay, or refer back to prior shots without slowing everyone else down .

Of course, there is the obvious desire to remote manage and control the camera (to allow an additional set of eyes on the monitors, remote set of hands on the dials or remote triggering) so, please include those features in with this as well.

In Summary: On Camera web server, maybe with it’s own DHCP/equiv server and only accept traffic from locally, camera assigned, IP addresses (to avoid some idiot hooking his camera to the internet allowing for the first RED Virus) an HTTPPost page to post XML requests the camera and streaming to an IP broadcast channel, a UDP broadcast of triggered events (camera start, camera stop, etc) Eventually, a Red “SDK” could be made public to allow 3rd parties to create cool set management applications.

Anybody else think this is a good idea?
 
Last edited:
It makes no sense to put into the camera what doesn't need to be in the camera.

All you need for the camera is an API and a suitable port. The camera settings can then be set and monitored from the port via the API or directly from the camera controls and on board menu. Period.

Everything you're talking about can be done and hosted from an external computer connected to one or more ports running whatever applications you want. If you want to see what the viewfinder sees then you take the same approach.

I like the Red price point exactly where it is.
 
"Thom Steinhoff
Director of the first RED feature in Sundance 2007
Red #516"

What if my feature makes it into Sundance too, how could we prove who was first?

hehe j/k.
 
craigbowman said:
All you need for the camera is an API and a suitable port. The camera settings can then be set and monitored from the port via the API or directly from the camera controls and on board menu. Period.

If you agree it needs a control port, why go proprietary and not a TCP/IP port? If you agree it needs an API, why not HTTP, which is an open standard with thousands of developers already knowing how to communicate with it rather than some potentailly proprietary standard?

Red, to me, is about breaking the chains of proprietary technology and thinking outside the box of conventional restrictions. Shouldn't the comm method be a part of it?

While a TCP video feed may be "blue sky" dreaming, controlling the camera via HTTP is not. You get so much for free as far as developing robust communication protocols, bus control, etc, that it is hard to ignore it as a technology option. ALso, there are lots and lots of low cost HTTP server chips available--and since you have to choose a communication chip to empower the control port--why not HTTP?

I, too, like the price where it is--and it is so well publicised I find it hard to believe they will add features that would bump it--so I think we'll be fine.
 
tonacidigital said:
What if my feature makes it into Sundance too, how could we prove who was first?

Then one of us will be forced to change our title to "Director of the first RED feature to WIN Sundance"
 
ThomSteinhoff said:
If you agree it needs a control port, why go proprietary and not a TCP/IP port? If you agree it needs an API, why not HTTP, which is an open standard with thousands of developers already knowing how to communicate with it rather than some potentailly proprietary standard?

Red, to me, is about breaking the chains of proprietary technology and thinking outside the box of conventional restrictions. Shouldn't the comm method be a part of it?

While a TCP video feed may be "blue sky" dreaming, controlling the camera via HTTP is not. You get so much for free as far as developing robust communication protocols, bus control, etc, that it is hard to ignore it as a technology option. ALso, there are lots and lots of low cost HTTP server chips available--and since you have to choose a communication chip to empower the control port--why not HTTP?

I, too, like the price where it is--and it is so well publicised I find it hard to believe they will add features that would bump it--so I think we'll be fine.

HTTP is NOT an API. HTTP is not exactly the fastest car in the garage either. Nor is it the network hardware of the port either. Its one of the higher level transport protocols with rides on TCP/IP. Just as my RJ45 style port uses ethernet to carry communications and layered onto that is TCP/IP if I select it as so.

There is no use for putting a socket server in Red either. Too much stinkin' overhead. The method I described will get you IP communications with Red without trying to jam everybody and his flying car into Red either!

Hmmmmm.....flying car....
 
No, it's not an API, but it is a very capable, industry standard, transport that you can use to write an API on.

I have been a programmer for 20+ years, a CTO of a worldwide, public company, for 5, and I am very familiar with HTTPPost and what it can and can't do and how fast it can be as a transport--plenty fast for the metadata control/data gathering I was suggesting.

Seems to me I read somewhere else on this forum of a wireless 802.11 capabillities written by a member of the Red team--which would suggest some form of TCP/IP communication already.

As for the flying car, I'm okay with it--as long is it is an option. What if I want to drive? And who would use a flying car underwater?

Flying car... Too much overhead.
 
ThomSteinhoff said:
No, it's not an API, but it is a very capable, industry standard, transport that you can use to write an API on.

I have been a programmer for 20+ years, a CTO of a worldwide, public company, for 5, and I am very familiar with HTTPPost and what it can and can't do and how fast it can be as a transport--plenty fast for the metadata control/data gathering I was suggesting.

Seems to me I read somewhere else on this forum of a wireless 802.11 capabillities written by a member of the Red team--which would suggest some form of TCP/IP communication already.

As for the flying car, I'm okay with it--as long is it is an option. What if I want to drive? And who would use a flying car underwater?

Flying car... Too much overhead.

When you start equating terms like API and HTTP which are clearly different in their usage, then you must expect correction. Protocols are fast when they run on fast systems. Speed also means heat, and I understand that is going to be a big concern to Red for the main tasks.

In the long run only time will tell how Red will address camera communications. For maximum flexibility so everyone can have their cake and eat it too we will see if "Occum's Razor" rules the day.

By the way nice to meet you. I've been developing and programming systems for almost thirty years now myself. I've only agreed once in the past to CTO for a private company.
 
Back
Top