Laser light kills sensor (Canon)

<p>There is a ongoing discussion regarding the above subject based on<br>

<a href="

on an local site. Any one with a little negativ experience in this respect.<br>

I have never heard above previously, but if correct then one have to find the analog camera</p>

<p>I was shooting behind-the-scenes (television) on the set of Jennifer Lopez' musicvideo, "Waiting for Tonight," back in 1999. After a few minutes of shooting, a laser, similar to the one shown in the above link, hit my lens, and my EVF went solid white. The laser fried the sensor on the Sony BVW600 Betacam I was operating (then, a $60,000 broadcast video camera), requiring a complete chip-block replacement (optical block + CCDs), costing over $5,000 in repairs. Having experienced that first-hand, I wouldn't expose any of my personal DSLRs to that kind of environment.</p>

<p>Even a small 5mW laser pointer can do serious damage let alone a club laser usually spec'd at tens of mW, as high as hundreds of mW for large concerts. </p>

<p>You would think that the transient nature of sweeping lasers won't do much harm, but it's the power density of one wavelength concentrated in a small area that makes it damaging. I'm surprised to see the laser setup in a line of sight over the dance floor; it's presumably not in the U.S. where such practices are regulated. </p>

<blockquote>

<p>I'm surprised to see the laser setup in a line of sight over the dance floor; it's presumably not in the U.S. where such practices are regulated.</p>

</blockquote>

<p>I've been to shows where the same kind of laser displays were used here in the U.S. I'm sure this kind of thing is regulated, as you say, so presumably the levels of light used are eye-safe (or companies would be sued out of existence), but that doesn't mean they're camera-safe...<br>

<br /><br>

I'm not sure if it's because of better light-gathering ability (the human eye has a maximum aperture of <a href="http://en.wikipedia.org/wiki/F-number#Human_eye">about f/3.2</a>) or because camera sensors are more sensitive to damage, but sadly cameras and laser shows don't mix unless you've arranged with the laser operator a location where you can shoot safely.</p>

<p>Hi Michael </p>

<p>I've never worked in the area of sensor damage, but my impression is that the physical mechanism behind such damage is usually thermal. If a CW laser beam is being rapidly swept around a dance floor, the residence time on any particular lens will likely be very small, and hence the total energy deposited will be small. This is not like intentional damage from a pulsed laser. </p>

<p>Do you have any sense of how common such damage is among photographers and the mechanism behind it?</p>

<p>Tom </p>

<p>Tom,</p>

<p>I'm afraid I don't. My only knowledge of laser damage to image sensors is restricted to what I was able to find doing some searching. I did find a few journal articles talking about laser damage, but they always seemed to be talking about pulsed lasers (and often CCD rather than CMOS). This makes sense, since their focus was on using sensors for research in a lab. I've done some limited work imaging a laser beam in a lab, but it was heavily attenuated (and spread); we were very careful to keep from damaging the sensor.</p>

<p>The fact that the videos I found showed entire rows of pixels damaged is very curious, and I'm sure if I understood the design of CMOS image sensors better that it would be a giveaway. (Perhaps some component related to reading out the row is destroyed by the high voltage associated with intense light, or brief heating might cause adjacent rows to short out; I found that this is possible for CCD sensors with pulsed lasers.) Unfortunately, I can't do more than speculate. (If anyone wants a fairly technical description of image sensors, however, I found <a href="http://www.olympusmicro.com/primer/digitalimaging/cmosimagesensors.html">this page</a> on Olympus's website.)</p>

<p>Hi Michael - My recollection is similar to the results of your search in that almost all the lab studies that I remember were always done with lasers with pulse widths in the microsecond range and faster. For thermal damage mechanisms, this guaranteed that the energy deposition time less than the thermal diffusion time out of the top-most tens of microns of the device so the heating of the active layers was maximized and heating of the substrate was minimized. </p>

<p>Anyway, if you find out anything, let me know. This is an interesting topic.</p>

<p>Regards,</p>

<p>Tom</p>

<p>I thank all of you for your comments/explanations - nice one still can learn something and in this connection do not be too worried taking photos at nightclubs.</p>

<p>Yes, the direct beam from a non-pulsed, continuous-wave laser can harm camera sensors and video projectors. This is for both low-power pointers, and high-powered concert lasers.</p>

<p>The more concentrated the laser light, and the longer it is on the sensor, the more likely it is to cause damage. By "concentrated", this means both the raw power of the beam and how tight the beam is when it enters the lens. A 5 milliwatt laser pointer held close-up onto one spot on the lens could cause similar damage to, say, a 2 watt concert beam that momentarily passes over a lens 100 yards away.</p>

<p>In our experience at the International Laser Display Association, camera sensors and video projectors are much more sensitive than the human eye. Our Members are responsible for laser eye safety, but unfortunately we cannot be responsible for camera sensor safety. Certainly when we work on TV shows we watch out for pro equipment. However, in today's world where it seems everyone at a concert is shooting or filming with some type of camera, it is impossible for us to watch out for every camera.</p>

<p>In the U.S., it is technically illegal under FDA regulations to directly scan audiences with laser light over 5 mW. However, some U.S. shows do this anyway, illegally, because the chance of FDA enforcement is low and the chance of eye damage is even lower. Outside the U.S., audience scanning is commonplace.</p>

<p>Pulsed lasers are, in general, much more likely to harm human eyes and sensors. This is one reason ILDA bans the use of pulsed lasers for audience scanning. But continuous-wave lasers, under the right (or wrong...) conditions could also damage sensors -- even from a single sweep of a beam into a lens.</p>

<p>Finally, let me re-emphasize that it is the DIRECT beam from a laser, going into the lens, that could cause sensor damage. Merely photographing laser beams and graphics is not harmful. If you are in a show that does not use audience scanning, feel free to take pictures. But if beams are coming into the audience (and be careful -- they may come on as a surprise effect), don't aim your camera towards the laser sources (the laser projectors). Avoid any direct beams onto, or into the lens.</p>

<p>More info is at ILDA's website, <a href="http://www.laserist.org/camera-sensor-damage.htm">http://www.laserist.org/camera-sensor-damage.htm</a></p>

<p>-- Patrick Murphy, Executive Director, International Laser Display Association</p>

<p>Hi Patrick - Thank you for contributing to this thread, particularly, your experience and practical advice derived from real-world experience.</p>

<p>By any chance, do you know if the physical damage mechanism for CW visible lasers on CCD camera sensors is essentially always thermal? If so, that could provide further insight into situations which might be particularly dangerous. For example, if a CW laser is scanning the audience with a quasi-Lissajous pattern then if you happen to be at one of the turn-around points (ie, where the velocity of the beam is lowest), your camera is much more likely to experience damage than compared to being somewhere in the middle of the scan pattern where the velocity of the beam is highest.</p>

<p>Sincerely,</p>

<p>Tom M</p>

<p>The sensor damage mechanism is probably always thermal, but I am not an expert. With eye damage for visible light, there can be some photochemical damage from deeper blue and violet wavelengths. Since sensors have been damaged even with red light, my guess is it is thermal.</p>

<p>As regards dangerous situations, the key is the raw laser power, how much area that is spread over (the width of the beam at the eye/lens/target), and the time on target. As a very rough approximation, if a laser beam scans over you and the light is uncomfortably bright or creates an afterimage, the light is possibly over the Maximum Permissible Exposure limit for eye safety. That does not mean that eye damage has or will occur -- there are some extra safety factors built in to the MPE. But I would miminize exposure by looking down or away when the beam comes towards me, or simply block the direct beam by putting up a thumb or fist between my eye and the projector.</p>

<p>For scanned patterns, there is a 4 to 10 times safety factor compared to the same beam non-scanned (static). But the exact power reduction also depends on the scan pattern. As you point out, any time the beam changes direction or otherwise dwells on a point, this will increase the exposure compared to a smooth scan. The size of the scan also matters. If a lens intercepts half of a very tight circle, more light energy hits the sensor than if the lens intercepts a small part of a very large circle.</p>

<p>Color matters as well, for RGB projectors. These use three lasers to create white light. White light thus uses maximum power (all 3 lasers combined), compared with red-only, green-only, or blue-only.</p>

<p>The variables are so complex that I would not worry too much over getting hit by the center of a scan vs. the corners. The best practice is to simply keep direct laser light from entering a camera lens or video projector lens. (Of course, if there is a shutter in front of the sensor you would be OK, but for people using Live View or whatever, just keep the direct beam from the lens.)</p>

<p>One laserist had a show where 12 Barco R18 projectors ($100,000 each) were mounted up high on truss. Before the show, he carefully aligned all scans so as to miss the projectors. But during the show, the rigging person didn't get the projectors to the pre-show height, and the DLP chips were damaged. I don't know the repair cost for the DLP modules but I do know the lamp module alone on a Barco is about $10,000!</p>

<p>-- Patrick </p>

 

<p>This is interesting because I remember when I had a D1, I shone a laser pointer straight onto the sensor (I might have even taken off the lens!). I will try to find those frames if I still have them. I don't recall any damage or issues.</p>

<p>Karim, when things are not moving (ie, as in your situation), the probability of damage is actually greater with the lens on because the same power is focused onto a vastly smaller area, so that the local temperature increase can be much higher than if non-focused.</p>

<p>Tom M.</p>