I resolved to replicate the projects that used IR LEDs. Towards this end, I lined the hood of a hoodie with 10 high-intensity IR LEDs all the while documenting the project to share with others. In fact, I was super-excited to share this project with others when I was done with it. So, modeling my project after those that came before me, I inserted the LEDs, wired them up, plugged it in and took a picture of myself with a camera modified to view IR light. When I flipped the camera around to look at the picture, the first thing I noticed was that it hadn't worked. Instead of creating a halo of light in place of my head, I looked like I was wearing a sweatshirt lined with Christmas lights.
I was ready to write the whole thing off as a fake, but I figured that perhaps the project required some more thorough testing and that is precisely what I did.
Update: Please stop telling me I used the wrong kind of IR LEDs, I powered them incorrectly, I used the wrong kind of camera (I didn't), the LEDs are meant for pulsing or that you are simply smarter than me (I don't care).
Most of you are missing a key point. The simple fact of the matter is that the viewing angle (even a high viewing angle) will not protect you if you turn your head slightly away from the camera. Assuming that you will not always know where the camera is, the prior devices that my work is modeled after won't do much to help you.
In short, it is of my opinion, that we have all probably been misled.
Cryptography that only works maybe sometimes if the hardware is old enough and Jupiter is perfectly aligned with Saturn is ineffective cryptography. The same could be said for personal-protection camera blockers. If the device is incapable of hiding your face from every IR camera at every angle, at all times, then it doesn't work. Heck, I would settle for a device that can hide you from all IR-enabled security cameras at night, but I even doubt this is possible using current methods.
Like I said, if anyone can take my experiment and make it work without altering it too much... take 12 5mm IR LEDs with no more than 12v of power and make it work at all viewing angles (without the addition of optics), I would be actually very pleased to be proven wrong. However, I am suspicious as to whether anyone can do this given these parameters.
Step 1: Establish a Method
So, I thought of the reasons my project failed:
1) The LEDs were spaced too far apart
2) The viewing angles made some of the LEDs seem to be turned off
3) There was only 5v of power
4) My camera's resolution was too high
5) I took the picture from too close
And then I looked at some of the other projects that seemed to work. The first one I examined was by aKaMaKaVeLy. After my initial test, this one seemed highly suspect for a number of reasons. First off, it used a 9v battery and no resistor. That alone made it seem fishy. Next, the LEDs were spaced out around the hat at about the same distance mine were around the hood. Last, throughout the video, everything was done as video except for the two quick shots of it working that were still images. From this project I took away very little.
I next examined the project made by URA / FILOART. Their project seemed slightly more believable. They used 12 LEDs powered by 12v of electricity that were centered on a headband. I figured if any project was going to work, it would have to be theirs.
So, I resolved to buy a host of IR LEDs from the Electronic Goldmine and conduct tests with 8 LEDs (the minimum number aKaMaKaVeLy claimed were needed) and have them be centered on the forehead like URA / FILOART had done. I would power it at 12v and take a shot from the front and the side. A successful test would block the camera from all directions with 8 LEDs centered on the forehead powered using 12v (with a 220 ohm resistor).
Step 2: Conducting the Test
I placed these LEDs in a breadboard and then took video of myself holding them to my forehead while powered by 12v.
Some were brighter than others, but none worked as intended. None of them were effective from the side-profile view.
From the initial test, I concluded there was no way it could work.
Step 3: Other Considerations
I considered that perhaps distance to the camera played a factor. I moved the camera back 10 feet and tried again. From this distance I couldn't make out my face too well with or without the LEDs. The LEDs made me look as though I was wearing a headlamp and as far as I was concerned I could see my face.
I decided that perhaps my hacked near IR camera was of too high a resolution (even in 640x480 video mode) to get an accurate gauge. I brought the setup to work where I tested it on the night vision security camera in the stairwell. This time at about 20 feet I got the same results at 10 feet and once turned sideways had no protection whatsoever. I tried this both with the lights on and off. When the lights were on, it almost made no difference at all.
Lastly, I remembered my earlier discussion with Dan who felt that this could work if all the LEDs were pointed directly at one's own face. Pointing high-intensity IR LEDs directly at your face can't be a good idea, but I closed my eyes and tried it anyway. There was little noticeable difference.
Step 4: Conclusion
However, I suppose that if you cover your forehead in super-bright IR LEDs you could work something out. Although, at that point, why bother? Just wear a mask.