This is a super fun hack that’s been around for ages — but now with cheap full 1080P HD camera availability, it’s probably a good time to make your own infrared camera!
It’s actually a very easy modification to perform. All cameras are capable of “seeing” infrared light, but for standard photography and video, you don’t want to see the infrared light. So most sensors just have an infrared filter in front of the sensor, to block out any excess infrared light. If you remove it … you have a converted infrared camera.
The following video shows exactly how to modify a camera to do this. It is a bit misleading though as it labels it as a thermal camera; and while it is seeing “infrared”, it’s not actually full thermal infrared, like a FLIR or Seek Thermal can see — it’s a mixture of visible and near infrared light. You will be able to see some hot things glowing through the camera, but not to the same degree as a real thermal imaging device.
The coolest part is a lot of black plastic is transparent to infrared light, meaning you can see through it with the camera!
https://www.youtube.com/watch?v=IdJgxNVVrs4
Compared to six years ago, this hacked infrared camera is pretty good! And if you really want a true thermal camera — you can still build your own!
Converting cheap cameras to see near IR was exciting 10 years ago. It would be at least marginally interesting if he bothered to install visible light filter. Just because he used more modern, higher resolution camera does not make it hack worth mentioning.
Thing is, that apart from resolution the sensitivity went up as well, you can find CMOS sensors from Sony that will match gen II night vision devices in terms of sensitivity and unlike a gen II amplifier, will have far greater resolution, will not be destroyed when exposed to normal daylight AND are able to colors when you put the filter back on.
Any suggestions for a sensor that would be easy to use in a DIY project?
citation needed
Ehhh, I can comment here. So for as long as I have been interested in night vision, one thing has always bothered me: misinformation. Yes, digital sensors are getting better but the only sensors right now that can best a real intensifier tube are EMCCD (electron multiplying charge coupled device) chips. Lux ratings on those Sony chips are misleading. They may be sensitive down to moonlight levels, but they require moonlight to provide a usable image. Generation 2 IIT takes a multichannel plate and multiplies the incoming electrons, which increases gain several orders of magnitude even under starlight conditions. There is a reason they can be destroyed by excess light, and it’s because of the high gain from the MCP. This technology has been around for decades and is extremely power efficient, using around 20ma of power at 3v. Nothing can truly compare. Cost is high but it’s because the damn thing just works and it isn’t high demand. When you step into the third gen tubes, you get a massive increase in performance over gen 2. Btw, EMCCD cameras, which are equivalent to intensified tube tech costs upwards of $10,000. Toshiba makes a few. Very cool stuff, especially for astronomy since you get full color.
I ruined a gen 2 tube. Left the lens cap off. Left it on the floor.
Forgot to shut it off first! Stupid mistake that won’t ever happen again. I want to fix it but tube is from Russia, and out of stock.
And those digital ones are ridiculously noisy, image wise that is.
Hey, don’t fret. So basically what you are experiencing is ion poisoning of the photocathode. Intensifier tubes have a finite lifespan of around 15,000 to 20,000 hours, due to the photocathode failing. When secondary electrons are created at the multichannel plate, negatively charged ions are created, and these are attracted back to the photocathode. Ions are bad. They slowly burrow into the cesium or whatever material your cathode is and prevent free electron holes. When the electron holes are blocked, photoelectricity can not occur and thus you get black spots. These are inactive, or dead, areas.
Now, there is a trick to fixing ion poisoning. You probably won’t find this info anywhere because it’s closely shared in the field. Get a shoebox, cover the inside in aluminum foil. The idea is *LIGHT TIGHT*!! Put fresh batteries in your scope, power it on and leave it for several days in the box. You eat the batteries dead. What happens is when there is no photoelectricity, the photocathode builds up a residual charge, very high voltage. The ions will be driven out of the photocathode and travel back towards the anode. Usually it will strike the MCP and become neutralized, as the MCP is its own circuit. Some of the photocathode, if not all of it, should be recovered.
Hope this helps!
WOW! Thanks man! Low priority project but will do.
Yeah, Google didn’t try to help me that time.
The edge of a negative makes a decent visible light filter IIRC.
I have a great-grandfather who says he remembers negatives.
He might be mistaken though.
Incidentally, it’s suppose to be the end of the film that you use, or you use a piece of a floppy disk which also allows IR to pass through.
It may be difficult to find today, but that’s what I used many years ago to block visible light for my ghetto NIR camera. I used two layers of exposed color negative, glued in front of a cylinder from paper blackened by printing entire black page on a laser printer, fitting snugly on the camera lens. The filter completely blocked the visible light while being pretty transparent for NIR. Then I used my soldering iron at several temperature settings to measure lowest bound on spectral sensitivity curve of the sensor by comparing the NIR brightness of the heater with theoretical radiation spectra computed from black-body model. It matched very well the theoretical lowest spectral bound for silicon sensors.
the guts of a 3.5″ floppy disk also work great for a visible light blocking filter. This is what i used to make my FITR multitouch.
Finding a 3.5″ floppy may be today even more problematic than getting a piece of exposed film
I agree and disagree..
As someone who has spent some time this week looking at IR object recognition, this “hack” although old, did interest me.
I love the quantity of content HAD produces
We’re just not going to mention that many synthetic fabrics are also transparent to IR, right?
Shhhh… don’t tell people that….
( ͡° ͜ʖ ͡°)
How did you get in here?
Transparent? Not so much, somewhat translucent but usually still opaque. I have a Canon 5D Mark II that has been modified for full spectrum (UV to NIR) and I use it with IR filters, this is one image from a shoot this weekend, she was wearing a black jacket, shorts, and top hat:
https://www.flickr.com/gp/67292116@N00/328DZZ
That is a nice hack..
My first infrared dSLR was my Pentax featured here: http://hackaday.com/2011/04/22/dslr-infrared-camera-conversion/
It was retired and I had my 5DII modified after I bought a 5DIII.
I was wondering, but…
It seems it doesnt work with all….
Well the hack is old but I have to admit some of the applications are new to me. Making the numbers on ICs more readable is good but the IR clear plastic is cool, makes me think someone should take a phone with an IR modded camera to the next CES and see if they can get a “peek inside” some new stuff.
The “IR clear” plastic is clear only because the device houses an IR receiver and the plastic is dark plexiglass. That box with the “Harmony” label is the receiver for the Logitech Harmony universal remote.
So yeah, it is IR transparent, duh …
If you are seeing hear with a NIR camera then it is just about glowing in visible light.
If you do modify a camera for NIR you should use a filter to get the best image quality. Visible light focuses at a different location in reference to the sensor than infrared, so when you are shooting in NIR you are getting out of focus visible light making things fuzzy. If you look at a SLR lens you will see a red mark on the focus scale, this is the IR offset to get focus.
Older SLR lenses did. Modern DSLR lenses do not unless you go to the high end glass.
“The coolest part is a lot of black plastic is transparent to infrared light, meaning you can see through it with the camera!” — I didn’t know that, that’s pretty cool.
If you haven’t already done so have a go at converting a camera to IR, lot’s of interesting things to see. The fact that cola is also transparent for example.
I would really like to, but I don’t have any available camera that I’d be willing to risk ruining. I could order something from eBay, but you never know with eBay-cameras what resolution the actual CMOS has, like e.g. they may sell the camera as a 1080p camera, but in reality the CMOS is just 640×360 and they just resample the picture up.
(Disclaimer: I don’t really trust Chinamans on eBay)
Never noticed those dome cameras they put EVERYWHERE have domes you can’t see through with the naked eye?
Or did you not know those were cameras?
Of course I knew they were cameras, I just figured it was something similar to how one-way mirrors work and thus never gave it a second thought. I’ve never gotten to actually look inside one, you know.
back in the old days (640×480 cameras were new) I used a piece of film negative that was black to make my IR filter. doubt thats easy to come by these days though.
I use Lee 87C plastic gels, much more rugged than wratten gels.
A piece of floppy disc will remove the visible light and pass the IR.
Haha. The photo album in the attic or the basement? The one that is falling apart and nobody knows who half the people in it are. We always kept the negatives from the one-hour-photo place.
I used to ask for developed film leader at the supermarket photo counter to use for IR receiver filters. They were usually happy to give out their trash by the bagful, including used one time use flash cameras and (now rare) plastic film canisters.
My Uniden dash cam has a user selectable IR filter, you flip it one way or the other to use the built-in IR leds.
Playing the video shows “This video is private”. Is this just a German thing or did the user revoke the video?
I am in the US, same issue here
IMHO revoked, it was playable an hour ago here in CZE…
plays OK here in .sk
Now here too. Probably fixed. :)
Is it me or is the ‘video’ on top annoying? it is way to distracting while reading the article. Please stop those animating GIFs from the 90s
If you close your eyes it will go away.
Thanks it is less annoying now, also the other nuisances from the site are gone!!
Annoyances like me?
I’m sorry, crappy day. They bother me too!
I’ve done this a couple of times with normal webcams (like 10€ webcams from some local store). It’s a very neat trick, and extremely useful if you develop stuff with IR-sensors.
A good way to add an IR-pass filter is to use exposed film (as in the one found in analog cameras). Apparently, it will block outr visible light and let IR through.
Couldn’t have had better timing. I have the same camera, and was having issues reading some SMD IC’s. I’ve seen these hacks before, and they’re usually a little more involved. Thank you for the post.
The first time I did this venerable hack was in, uh, 1992. I was in 8th grade, and took my Fisher Price Pixel Vision XLs IR filter off, and went to American Science and Surplus in Chicago and got a big IR filter which I put over a 200watt floodlight.
It actually worked, a little, and the highlight of it’s use was during a full-campus capture the flag at the private science academy lock-in.
I have a box of those cheap dashcams, got them from Aliexpress for $9.00 a piece (shipped free)
once you get over the fact that they are natively 640×480 rather than 1080p, and the poor mjpeg compression fills a 16g card in a couple of hours (at 480p), they work just fine.
they also work as a webcam via USB, so I have one on my 3d printer, and I am always looking for new things to do with them.
Would this be useful for looking for overheating components – like what technicians use to troubleshoot electrical short circuits in electronic equipment?