Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

RED-Key (security proposal for RED cameras)

RED-Key (security proposal for RED cameras)

  • Total voters
    146
Photocon said:
The point is not if the collision is hard or simple, or the amount of time it takes. The point is that by using client side software you make this risk significantly larger than if you generated the SD keys behind Reds website. If I can RE the software and figure out what auth scheme they are using and even have a copy of the camera-specific key, Im thousands of miles ahead in the race to crack any cameras password. Your right it could take billions of collisions to figure it out. But it could take 2. The chances for any one collision being the right one are exactly the same, and your chances of finding the right one increase with every wrong one you discard. And you can use side channel attacks to help determine how close you are.
Click to expand...

You can't. You keep treating this like the RED one is a PC and You are hooked in and monitoring the firmware response. This is not the case. And You can't be serious that someone will actually go and take his chances to guess the password. By the time You have tried even 1% of the possible combinations, we will all be shooting 3D (i mean a true 3D with Z-depth - not just stereo) in 28K at 1000 FPS uncompressed to a PETA-BYTE CF card that costs $50,- USD...

Photocon said:
By simply putting the SD key generation behind a website you mitigate this significant risk and make cracking any auth scheme orders of magnitude harder to crack.

Another way to put it: Ive got a box in a field 100 yards away from a burglar, and it contains $17,500. I put a lock on it. Do I want to give the burglar half a key and a lockpick, or do I want to blindfold & spin the burglar?
Click to expand...

Quite the contrary - it will be much easier to hack into a website. Students are getting through DOD security systems for Pete's sake...

I am glad we disagree on this...

:) Peter
 
NO! As emphatically as could possibly be expressed, no.

Why? like everything it's a matter of risk and reward, or cost benefit if you will. I think any reasonable look at the numbers of people affected both positively (those that have their camera's stolen, but yet smile due to the knowledge that it won't now work), and those that are affected negatively (being on a shoot and having a dead camera and having to expend the time and energy to get a working camera), would have to at least consider that more people will be hurt than helped. I think maybe 5 out of 5,000 cameras are stolen.

In terms of the opportunity costs I'd rather RED save the engineering hours and incremental camera cost and offer some type of insurance. I don't like the idea of RED expending energy on anything that doesn't produce pretty pictures or improve how we work with them.

Second, lets be honest if theft becomes a problem and this is implemented this will be defeated. People who are serious about getting your stuff will get it and they don't have to be as elegant with defeating a countermeasure as the designer of that counter measure. I don't believe that even if the perfect theft countermeasure was implemented that any fewer cameras would be stolen. The ignorant thieves will steal cameras anyway, the sophisticated ones will defeat the countermeasures. Once your camera is gone it's gone.

Which brings me to my last point. Without being too judgmental. Watch your stuff. You've purchased something that at minimum costs tens of thousands of dollars, beyond that it may be the thing that feeds your kids. Be smart.
 
Peter Majtan said:
You can't. You keep treating this like the RED one is a PC and You are hooked in and monitoring the firmware response. This is not the case. And You can't be serious that someone will actually go and take his chances to guess the password. By the time You have tried even 1% of the possible combinations, we will all be shooting 3D (i mean a true 3D with Z-depth - not just stereo) in 28K at 1000 FPS uncompressed to a PETA-BYTE CF card that costs $50,- USD...
Click to expand...

Again, I dont have to guess the password. I dont care about the password you set. I just have to get a collision, which is significantly easier.

In addition. the Red One IS a PC with a lens attached. It would be quite easy to monitor for side channel attacks, using only the monitor out ports. Stuff like this does happen, right now.

Peter Majtan said:
Quite the contrary - it will be much easier to hack into a website. Students are getting through DOD security systems for Pete's sake...
Click to expand...

But with a website you can watch them do it. In addition, the key generation doesnt have to be on the web facing server. I never said that the web solution is flawless. However, the web solution is far easier to achieve with Red's limited time, easier to deploy, reduces risk and does not inconvenience end users.
 
What if the owner of the camera has the choice to enable security when firmware is updated. Being that we don't already have clocks and extra sd ports, how about revisiting the idea of firmware "build upgrade" security. If I chose to "security enable" my camera at the next firmware upgrade (assuming that such a feature existed), at every boot up I would be advised on the screen "SECURITY ENABLED". If my camera was ever stolen and reported to Red, Red would place my serial number (PIN info) info in the next firmware build. At the next attempt of firmware upgrade, the camera would lock-up and announce "Factory service required". Just as upgrading from Build 15 to Build 16 was non-reversable, if I choose to security enable my camera, there would be no backward firmware update available, therefore older firmware could not be used. If I ever wanted to disable my "security feature" once enabled, I would be required to return the camera to Red, provide proof of ownership and of course no existing report of theft to Red of that serial number. For Red to place a camera on the "stolen list" in the firmware upgrade, registered owner varification would be required, most probably accompanied with a police report.

The idea of camera pin number in metedata sounds good also. Then Red Alert!, Redcine, and other software that reads R3D files could also flash an advisory somewhere on the screen should that version of Post software contain a list of the stolen cameras.

Such a security scheme could be up and running soon should Red choose to adopt it. Anyone who wanted to buy a used Red camera would only have to download the latest version of firmware and upload it to verify legitamacy.

As one who had all of my film equipment stolen many years ago, I can say that it was no fun at all. With equipment theft, there is always a loss of money, to a greater or lesser degree depending on insurance. The loss of time is unrecoverable. Any theft deterent is appreciated, and any way to provide a trail back to the thieves is devine.
 
The idea is good. But I think this puts RED too much into the process. I mean, they would have to work for us. When the camera has to be sent in, to implement the process etc.

What I think might be a little easier is just blocking the docoding.
As of now, RED controls all the decoding of the footage.
I am not so much a fan of it, but right now, if a certain serial number in the metadata would be found, it could be blocked in the decoding/converting/importing process very simply.

This way RED would be taken into te process just for the stop-registration for that certain ser-no, no more.
 
augenmensch aka therock said:
The idea is good. But I think this puts RED too much into the process. I mean, they would have to work for us. When the camera has to be sent in, to implement the process etc.
Click to expand...

The implementation would be entire in firmware, they would simply have to add it to a new firmware release (I say 'simply' but obvious the process of actually building this may not be simple).

What I think might be a little easier is just blocking the docoding.
As of now, RED controls all the decoding of the footage.
I am not so much a fan of it, but right now, if a certain serial number in the metadata would be found, it could be blocked in the decoding/converting/importing process very simply.
Click to expand...

That means RED would have to release a new version of their decoding tools (as would every third-party developer who implements the R3D SDK) everytime a RED camera were stolen or recovered - and it would require that people actually use that version, which would be unnecessary. Also it would mean that people who already had footage shot with that camera would be unable to process it - so if someone stole RED #23 and it was disabled I would be unable to process the R3D footage from the short film I shot last year.
 
Augenmench,
In the "firmware build upgrade scheme" mentioned above, Red would only be in the process if the camera was reported stolen. The owner could respond to prompts during the firmware upgrade upload process to either activate security or not. Maybe I wasn't too clear on that.

All the best,
Vance
 
First, I have to say I really really like the idea Peter. Great job, appears to be well thoughtout.

That being said - short of personalizing the camera to begin with(by adding owner's info, adding this info to metadata in RAW file) I doubt that this type of security would be too highly regarded by actual owner-operators. While I agree security is great - if I'm the only one using this camera - I doubt I'd go through the effort to actually complicate my workflow or make the boot up process any more difficult for myself.

It seems that this would most likely be in high demand for rental houses and owners who rent/loan thier equipment out - instead of a # of boots (as stated earlier, it's too ambiguous) it sould be set to a certain timeframe - so if you rent it for a week the rental house can set the "return date" as it's lockout date. So even if you do get the bright idea to just simply not return the merch - it's useless to you untill you bring it back to the rental house, OR the rental house gives you a new "key" for extra time.

Just my 2 cents.

BTW: Just to throw my opinion on something - just because digital security measures can be bypassed, doesn't mean that they shouldn't be employed. Barbed wire fences can be rendered utterly useless with wiresnips - but - most people wont be getting on the other side of the fence.
 
Just a thought that crossed my mind at 3am today, I bought a pay-as-you-go phone sim card for a friend for Christmas, (approx $1) to go in one of my old phones. Orange, my provider in the UK are able to disable stolen phones remotely by sending a text message or signal, making them completely useless. Would it be a option to have a sim card in each camera, telephone no unknown to the user but tied to the serial no. Every time the camera is turned on it picks up a cell phone network and therefore its location can be traced as well. If the camera is stolen / not returned from a rental on time, Red can text the camera for next time it is turned on, complete lock down of body ensues and has to be returned physically to Red for a new sim card/ number and reload of software. If this can be done in a $10 phone why not in a camera?
Any thoughts or have I missed something completely obvious?
 
Actually the way cell network works is that a phone is disabled by locking out it IMEI (unique serial number basically).

The problem is in using any soft of networked security with something like RED that it needs to default to working in cases where network connection can't be established (in the middle of Mongolia, for example) which means to steal a camera and make it usable all you have to do is disable the network connection, which would be as simple as removing the SIM card.
 
Thankyou for that Dylan but if the camera was fired up in the morning with default setting to find a network wherever you are in the world and then afterwards to be able to turn off this setting in order to film with no interference at some point the signal will catch up with the camera - Unless it spends the rest of its life in a nuclear bunker. There must be a way of setting a bit of software where if it receives a text/ code all operations stop. The imei number disability function kills the phone so it cannot be hacked to another network. If the sim was 'welded' in to a circuit board and removal would kill it as long as the camera was rendered unusable what would be the difference? Could the signal be set to 'fry' a red unique piece of hardware - prompting the return? Obviously with my limited knowledge I will not be designing the system, but something cheap and simple needs to be done especially as another has gone in Las Vegas recently. I look forward to hearing your reply.
 
OK, lets say a camera gets locked out in error on a $250,000 a day shoot, who is going pay, RED? I doubt it, sounds like a very silly idea that can never work with a camera intended for high end productions.
 
Ok gents, thankyou for your replies, I am merely publicly brainstorming, insurance and a nearby rental house the only way to go then. I hope it never happens to any of us and we never have an insurance company that wriggles on payment. My first lesson camera assisting many years ago was never put the camera down/ turn your back on it. Park the van tailgate hard up against a wall. Thefts will always happen though.
 
I have no idea why security in the camera would be a bad idea. The problem is making sure it is a simple as possible and didn't interfere with renters business, etc... so perhaps back to the lo-jack idea where you don't even know it is there. I knew a guy with a Varicam back in its peak and he claimed to have installed lo-jack in it somewhere...

is that at all a possibility or just BS? Could it fit? Could it cause interference? Does it only activate when you ask for it's location?
 
I have not read all the posts, but I did read Peter's idea. I think it's major cool he's thrown something up there for discussion, and it's a fine concept. I, however am not in favor of it. It just introduces to many steps and possible problems for me to think it's worth it. I would prefer what other people have said in the past, things like owner's name and address in the meta data. Or something as simple as a password to log onto the system. There could be a master password held for each camera at RED, (in case the user forgets theirs) and the user password. But even then if they fail to enter it correctly I would NOT want the camera to shut down, but just flash a text warning at the beginning of every boot that the password has NOT been entered correctly and to verify ownership of RED serial ##### with red.com.

It's been my experience that the only trouble crooks are willing to take is the actual stealing.. If they have to do a lot of workarounds to get something to work, they pass it by. And a big text warning saying the camera might be stolen is certainly something that will keep people from wanting to buy a "hot" RED.

My thoughts.. Still think Peter has a fine idea as well.

Jay
 
Reward Program

Reward Program

Not a very high tech idea, but just to add to the mix:

A cash reward for information on RED thieves is something that would highly deter thieves. It would also apply to those purchasing stolen RED cameras.
 
Ok then what about an ownership passport - like a car licencing form/ pink slip that has your name on and your build update code on, buy a second hand camera, register the ownership with red and they get a new code to their name. No passport, no code, no updates. Very hard to sell on with no paper work and suspiciously old build on.
 
Having some type of security would be nice. Maybe a simple Lojack or GPS tracking device. Anything that needs a connection to the net is trouble, IMO. I picked up a GPS for a shoot last week, and only figured out once I was up in the mountains that the pre-bundled software required a net connection to "activate" the detailed TOPO maps. Needless to say I was extremely pissed off. There should never be a net connection needed for anything security related. Although embedding some type of secret code into the firmware, just to catch the lowest hanging fruit (ie, the most stupid criminals), through an IP trace, might be nice. I'm not sure what is required, in terms of evidence, though, to get the police to trace IPs for criminal warrants.

My guess is that 95% of the people who steal Reds are just opportunists who probably have no idea what the camera even is, let alone the resources or brainpower to defeat even simple security measures. If you stashed a lowjack deep inside the brains of the camera, I bet that would catch some thieves.
 
You must log in or register to reply here.
Back
Top