Industry News: Sony officially warns about CIS laser damage

Aug 7, 2018
356
306
In night clubs in Germany and on music festivals we had this planes of laser light that were moving through the crowds and some of them were vertical and therfore also hit the eyes for a short moment. However if the time is very short, that should have the same effect like a low energy. For example the stove can be hundreds of degrees hot, but if you only touch it for a 1/100 of a second, you might not burn your skin. I think for that reason lidar scanners are pulsed.

In my university we had a laser that produced a pulse of 100,000,000,000,000 watts (14 zeros!), but only for 1/1,000,000,000,000 seconds. So it only used the same energy as a 100 Watt laser for one second. Of course 100 Watts is still a lot for a laser, but that example shows that short pulses can reduce the overall energy by a lot and still achieve a goal.
 

Ruined

EOS R
Aug 22, 2013
932
60
As was pointed out before, it is not a mirrorless issue. In a DSLR, the laser just gets into your eye instead of onto the sensor, which may actually be worse depending on the laser and how much you value your eyesight.
Unless it's an illegal high powered laser you are not going to damage your eyesight from looking at a laser for a brief period of time. If it is too bright you will close your eye just like when you look at the sun, maybe see a floater for a couple of minutes and that will be the worst of it.

It is a mirrorless issue, because the sensor is exposed at all times you are using the camera unlike a DSLR via OVF. The mirrorless camera has no ability to "look away" because it's viewfinder cannot function without sensor being exposed at all times. Time is what does the damage, not just power alone. DSLR also susceptible in Live View, but with DSLR you can elect not to use Live View in an environment with lasers (like club etc). Or you can also risk your sensor with live view on the DSLR as another option- with mirrorless tho that's the *only* option.
 
Last edited:

randfee

I'm New Here
Aug 27, 2020
16
6
This is mainly a mirrorless camera problem because of the length of time the sensor remains exposed - which is basically all the time you have the EVF or Live View active. DSLR sensor obviously can be at risk as well in Live View, but far lower risk via OVF since sensor is only exposed for a fraction of a second instead of continuously.

Sony likely made this statement to cement future warranty claim denials.
hm, I don't accept that as an excuse ... that makes no sense when talking about LIDAR and show-Lasers. Both should be scanned around at high angular speeds and therefore the local cumulative energy/dose should remain small, no matter of exposure time of the sensor, since your eyes (!!!) are also "always open" like a mirrorless camera. Silicon has great heat conductance so if the energy is small enough, the laser could just continuously irradiate the sensor and it won't get damaged! As soon as the intensity is so high that heat damage occurs, the laser power is definitely much too high!!

Does anyone know if the sensor's readout electronics might actually get permanently damaged from overcurrent generated by photo-effect (absorption of laser in silicon). The failure pattern (total rows and columns failing) would suggest to me, that thermal damage has taken out the conducting electrodes and not the readout electronics (!) - but I'm no expert on failure patterns of the chips.
 

randfee

I'm New Here
Aug 27, 2020
16
6
In night clubs in Germany and on music festivals we had this planes of laser light that were moving through the crowds and some of them were vertical and therfore also hit the eyes for a short moment. However if the time is very short, that should have the same effect like a low energy. For example the stove can be hundreds of degrees hot, but if you only touch it for a 1/100 of a second, you might not burn your skin. I think for that reason lidar scanners are pulsed.
those planes are usually created by diffractive optics or scanners and there is safety mechanisms that open the interlock circuit immediately. A passive element splitting or reshaping the beam is safer because it inherently lowers the intensity peaks. Also, the higher power lasers are not allowed to shine into the crowds, at least not here (germany, europe).
In my university we had a laser that produced a pulse of 100,000,000,000,000 watts (14 zeros!), but only for 1/1,000,000,000,000 seconds. So it only used the same energy as a 100 Watt laser for one second. Of course 100 Watts is still a lot for a laser, but that example shows that short pulses can reduce the overall energy by a lot and still achieve a goal.
yep... femtosecond lasers. I have several of those around.
The show-lasers are definitely cw (continuous, not pulsed) so I'm still confused as to how these could potentially damage the silicon chip while being eye-safe. I doubt it VERY much, these lasers must be illegal then or have super large beam size which, together with a large entrance pupil (long focal length, large aperture) lens reaches the silicon chips' damage threshold. Again, I doubt the lasers are legal if this happens VERY much!

As for LIDAR. Many time of flight systems now use few nanosecond (or shorter) pulses... so it is possible that the damage thresholds for silicon are much lower... Still, I would think that no laser system is legal (considered eye-safe) in a developed country that could damage a camera chip this simply. Again, super large entrance pupil optical system obviously increases the intensity and for (ultra) short pulsed beam sources the damage threshold is much lower anyways, but I still see no window of legality for a laser killing ordinary camera setups while being declared officially eye safe.
Again... imagine somebody looking at a LIDAR based car through binoculars.... no way will cars be allowed to be equipped with lasers that could blind you in EITHER case, naked eye or binocs! ... only in a banana republic!
 

Kit Lens Jockey

EOS R
CR Pro
Nov 12, 2016
889
606
Is there any consensus about what classes or output wattages of lasers are dangerous for camera sensors? I'm going to be shooting in an area that will have some of those "holiday laser light" projectors in it. Those seem to be limited to about 1mW of intensity per beam, so I would think I'm ok. But I just don't want to fry the sensor on a $4000 camera by using it in the vicinity of a $40 holiday laser light gun.

 

randfee

I'm New Here
Aug 27, 2020
16
6
there can't be a power limit like this. If I take pulse lasers, 1mW means not much as this could mean 1mJ pulse energy at 1Hz.
Suffice to say, this type of laser is unlikely to be irradiating the your typical environment, but with lidar, it is not necessarily continuous and therefore at least I can't make such a statement.

IMHO, eye safety (taking into account that people use binoculars, telescopes etc) would mean, that one can't damage a typical (silicon sensor) camera. But looking at the evidence... some of the LIDAR manufacturers and MANY of the laser-show-beam source providers do not appear to be taking their power and intensity levels seriously.

I'm really interested to learn if lasers that kill silicon image sensors can physically be truly legal! I would have the means to test and find out everyday, but I'm not going to irradiate my own camera and we're just too busy for me to tell one of my students to go purchase some used cams off ebay and do a series of tests. If I ever decide to allocate some play time to this, I'll post the results here!
It would make for a nice paper that would probably get cited tons though... hm :p