Reddit: We have seen image sensor damage from a direct laser beam hit, but it turns out that a diffusely reflected laser light can cause damage too. In this short video, a tattoo removal laser cases multiple spot damage to Sony A7SII full-frame sensor:
Via: SonyAlphaRumors, DPReview
The CFA is a kind of tattoo, no ? :)-
ReplyDeleteI guess the color means photodiode itself rather than any sharing structure dead.
DeleteThe majority of damaged pixels don't seem light sensitive anymore. So it looks like a stuck-at-white kind of defect. If the laser would only damage the green pixels to yield such a stuck-at-white defect you'd still expect a color shift when moving the cam over the scenery. That doesn't seem to happen, so it appears here, green pixels are damaged in a different way than red or blue. But they do all seem damaged.
ReplyDeleteWhat I found interesting is that I also believe to see some blinking pixels in the left upper corner of the first cluster defect.
not all White pixels are green, the white/yellow ones must be also R and B. it depends on the laser wavelength.
ReplyDeleteLaser is the new Chuck Norris. Switch on a laser and all cameras within a mile stop functioning.
ReplyDeleteWould be interesting to see how Sonys deadpixel detection handles that
ReplyDeleteThe problem is that the lasers used to remove tattoo are very strong and very concentrated.
ReplyDeleteFor example: for a laser of Nd,Yaq with a pulse energy of 1J/cm2 and duration of 10us (may be much shorter), the irradiance at the object is 10^9W/m2. Which is around one million times higher than the direct sun reflected over a white surface. This very strong power density probably will produce a degradation of the microlens, because these are of a organic resin and the same happens for the CFA. Even the photodiodes can die, but is more probable that the organic material degrades first
I think it is the energy deposited that counts, so pulse duration and rep rate is important as well. Also, these lasers are designed to interact with organic material (and ablate it in some cases). Hopefully everyone is wearing safety goggles or glasses during the procedure. Just forgot to put them on the camera too.
DeleteHow is the microlens melting theory supposed to explain that all pixels appear green? Also melting is related to radiant flux/temperature not irradiance/electric field. Usually damage thresholds for irradiance stem from issues like ablation. But also here, I wouldn't see why all pixels appear green. It would be nice if someone could tear down the imager and actually check what happened with the device - CFA, mircolens, diode, insulation, BEOL...
Delete... chipworks?
The color of the laser seems green. Is one of the used most frequently. If this is the case, the color filter green will be the filter with better transmitance and then the other color filters will absorb most of the energy. This means that all colors filters but the green will be heated and probably destroyed.
DeleteIf all color filters are damaged but the green, then all senses signal will be green, which may explain the result. This can be verified if the damaged pixels go to black when no light is incoming
Maybe. But a piece is missing that explains why the green pixels always stick at the same intensity. Possibly the Blue and Red color filters get knocked out as you say. And in the green pixels maybe excessive damage is created to the lattice causing the pixel to saturate by dark current. Would indeed by nice to see a tear down to confirm.
DeleteIt doesn't make sense to speculate about the intensity or colour of the artefacts. The footage doesn't show the original sensor output (raw data), but the output of the image processing pipeline including all operations like AWB, tone mapping,...
Delete-dkf