Virus

Hi, maybe it's my impression, but i don't see anyone concerned with

the subject of DSLR virus. Since now, a camera is a computer, the

same problems arise, right ? What happens if my computer is infected

and my camera is connected for image transfer ? Worse yet, since

there is now a wireless option for the 1DsMII, 20D and others, the

can is open.

 

Just a thought . . .

Very interesting thought.It's not made by microstuffed so would anyone actually bother?LOL (yes,more seriously,other computers get viruses too...)

<BR>I guess it would be quite easy to design a firewall(?*) for it as it's a pretty simple program with very straightforward tasks isn't it?

<P>*Purely guessing about the firewall thing,i'm no computer expert by any stretch of the imagination!

The camera is not like a computer where any program can potentially download itself in the background and write to whatever. The only way to get a virus is if you, personally, willfully, install a firmwire update. I am not sure what the exact procedure is (depends on each camera) but that would usually include writing a file to the memory card in the right format (ok, maybe a virus can do that) and then you restarting the camera. You would then get a message from the camera or some sort of menu to say whether you want to do the update or not. By then you will have spotted that something is wrong so you should be looking at what is in the memory card.

 

If you blidnly do that without wondering why the camera suddelny wants to update itself you deserve what you get! :-) Unless you get an infected version of one of the firmwire hacks that will appear shortly. But then you shouldn't complain.

 

Wi-Fi is no different. You still have to do the final bit of installing the virus by pressing some button on the camera.

I'd say it is potentially (remotely) possible to subvert a camera in either of the two methods you suggest. I guess the motivation would be the pleasure of trashing someones camera or stealing their photos.

 

If the intent was to steal photos, an attacker would be *much* more likely to attack your PC (which would be a prerequisite for the cable attack) or the WEP wireless transfer or the FTP server at the other end of the transfer.

What would be the point? A digital camera doesn't run Windows or one of the other commonly used operating systems, so you need specific virus code for each camera type. It might be doable for some individual cameras but what would be the gain for the virus-maker? A digital camera is used for 2-3 years after introduction, and some of that time would have to be spent in writing the code.

"What would be the point? A digital camera doesn't run Windows or one of the other commonly used operating systems, so you need specific virus code for each camera type."

 

Not only that, but the power of the computer virus is contagion: one infected computer infecting another, which in turn infects another. Since digicams aren't networked devices, the interest in the virus-writing community (those misguided, sexually deprived miscreants) is nill.

Nope. No concern whatsoever.

My digital camera never even comes into direct contact with my computer... I use a memory card reader. Anyways back on topic, the chance of this is nil.

As they said, a virus is a self replicating program, which is transmitted hidden in files from one 'computer' to another.

 

Unless you know of a bunch of guys networking their camera's or swapping memory cards about, uploading other peoples files using their cameras, I'd say the risk is non-existant. There is no gain, and no way of ensuring the replication.

Cameras ARE getting networked. 1Ds II, 1D II, 20D will have the wireless file transmitter capability, so do the Nikon pro bodies. If this is not networking, then what is? Wireless control of the camera from your laptop, wireless firmware upgrades etc WILL happen sooner or later.

 

 

However the danger is probably not viruses nor worms but targeted attacks. Imagine, for example..

 

 

... A man-in-the middle attack on the image upload at a major sports event. Zillion AP/Reuters/SI/... photographers covering the game, and YOU capturing all their imagines on YOUR laptop on the side of the field.

 

 

... Or simply blowing up all cameras of a competing news agency at the same event...

 

 

... Or maybe something less evident. Maybe a small firmware tweak that leaves the camera functional but affects the picture quality of a rival photographer just a bit, ensureing that YOU get the front-cover shot, not him? A small back-focusing issue, maybe, unnoticeable on the field but definitely there on the enlargements? Or sudden digital noise at ISO 100? Much-slower-than-normal AI servo? Wrong AF point highlighted on the viewfinder?

 

 

And so on...

Face it: camera is a computer - which so far has been living on an isolated hardware island but which probably won't do well at all when exposed to the dangers of the network world. Limited user interface, no focus on security and a push from market(ing) for "more features now!" makes the issue worse. The virgin wireless perimeter of the cameras will be breached, this year or next one, and the camera first hacked will probably be the most professional one (having most features and thus most weaknesses).

 

In a nutshell: Cameras are getting (wirelessly) networked, now. Photography is a commercial activity and a target for commercial attacks. Therefore, network insecurity & attacks on image capturing systems/setups WILL become a real pain, now or later.

Kaur, your scenarios strike me as very unrealistic, in that they combine minimal financial incentive with the near certainty of a prison sentence.

 

Malicious viruses written merely to cause random havok are in fact the main danger. They've already started attacking cell phones. As noted above, however, there's not much incentive because the potential audience is very small compared to the number of PC users, or the number of cell phone users.

Yes, but there are always going to be characters who have nothing better to do, and I'd bet you'll be seeing a little trouble on and off.

Mark & Vincent, business crime is not performed by "characters who have nothing better to do", it is done by hired thugs. So is computer crime, with the difference that computer thugs can be found more easily, they remain even more anonymous to you and charge less. And prison sentence is not a serious threat to them, e-criminals who have got jail time (for stealing business information, identity theft, blackmailing, or simply messing up someone's services) are really really rare. Law enforcement systems are simply not up to the task yet, they cannot catch the crooks and even if they could, the courts would not understand the crime. This will change, of course.

 

 

I agree that viruses are not yet a serious threat to such spare and loosely-connected devices as cameras. Unless you consider the ubiquitous bluetoothed mobile phones as "cameras"... well as a SLR user you should not :)

Connecting a camera to a computer -- regardless whether with a cable or wirelessly -- is not networking. 802.11x devices don't have MAC addresses like Ethernet devices do; besides, connected digicams (and memory cards in readers, for that matter) look like storage devices on the host computer system.

Interestingly enough, the Digital Rebel (for example) is running an embedded version of DOS from Datalight. In addition, there is a Linux-based program floating around that will allow you to hook the DR to the computer's USB port, and mount the internal flash memory as DOS drive letters (Datalight DOS has a remote execution debug mode, which was left enabled in the Digital Rebels). The camera program, called 'camera.exe' is found on the B: drive, while there is an AUTOEXEC.BAT file, among others, on the A: drive, which is apparently the protected un-writeable (normally) area of the internal flash. When a card is inserted and the camera powered on, the AUTOEXEC.BAT checks to see if a new firmware file is on the card, and asks the user if they wish to install it. If so, a new camera.exe is unpacked from the flash card, and written to the internal flash B: drive, and the camera is re-started. If there is no card inserted, or if no firmware file is found on the card, the AUTOEXEC.BAT just jumps to the B:Camera.exe on startup.

 

Now, normally, the only way to write to the B: drive is to use the flash update firwmare which is on the protected A: drive, and normally invoked by the AUTOEXEC.BAT. But, a 'trojan horse' type of code that might be embedded in a new copy of the firmware (obtained from an unofficial source), might open up the possibility to update the internal flash without asking the user's permission, by calling the firmware update code directly from within code running in the camera.exe file on the B: drive. At this point, a 'backdoor' would be running in the camera, which could be leveraged to do all kinds of interesting things.

 

The saving throw is that the firmware loader code itself, and the DOS startup code and OS are contained in the protected area of the internal flash. Normally this area can only be written to by unlocking the area, either by moving a jumper on the circuit board, or by using a JTAG connector. If so, then supposedly a really hosed camera can always be recovered by inserting new (clean) firmware and restarting the camera.

 

On the other hand, if (say for convenience at the service centers, etc.) the protection can be disabled via some software sequence (this is common with EEPROMs on motherboard, for instance), then the protected area itself can ultimately be patched or overwritten. In that case, the only way to 'clean' the camera would be to re-load it from scratch, using the manufacturing jig or service bench pod to do so.

 

So, while it is unlikely that we will be seeing things like this in the near future, they are not entirely out of the question, and eventually some camera manufacturer, striving for maximum manufacturing and servicing flexibility will make some dumb mistake, allowing those that have a lot of time on their hands to figure out how to create a mess.

. . . which is how others have hacked to drebel firmware to enable some hidden digic features. . . .

... and yet, still provides no path for virus contagion ...