As someone mentioned in comments to Invisage patents post, there is a 6+ min Invisage video from the DEMO event showing its QuantumFilm sensor working, albeit B&W only. In the video Jess Lee says that the pixel size is 1.1um. Here it is:
Why does this lack the shock value of the day SONY announced Mavica or the day Foveon announced a pixel that read out all three colors? The black and white image is nice, but it is somehow lacking the credibility that sent upper management to accost their engineers on these previous occasions.
I lost an evening with my family responding to an emergency request from top execs in another time zone. An evening I will never get back! Well, some CCD guys might say that is karma. I say it should have been a worthier cause.
Are the images taken by Invisage's camera, or do they show an image that simulates the 4x increased sensitivity ? It is hard to believe that the image shown (of the guy in the bar) is made by the new camera under low light levels : not a single white spot, not a single defect, no FPN .... Someone mentioned already earlier : this is a similar story we heard already over a decade ago about a-Si, and I fully agree with this statement. If you have to make an electrical contact between the photoconverting material and the silicon you have a problem ! That is one of the reasons why a 4T-PPD (no contact in the diode) is so much better than a 3T pixel (with contact in the diode) in CMOS. A.T.
I talked to InVisage for their Demo debut then later heard them present at Image Sensors Europe conference this week in London. The last slide was a color image. It wasn't stunningly awesome, but it worked. It's hard to say how much of the issue was due to the hotel's really bad projectors.
@Anonymous: I'm not expert enough to comment on how well it works, but InVisage uses an electric field to draw the electrons down into the silicon. They say it doesn't use much power.
There was a lot of chatter about InVisage at the conference, but everybody I spoke to was cautious--proof in the pudding and all that--but also because there was a lot InVisage didn't reveal at this stage in its development.
Well, it's hard to believe that there is a serious science behind this show. I hope that this could really be a breakthrough in imaging device, but ...
Aside from arrogant marketing hype based on erroneous comparisons and leftout details (e.g. CFA transmission) I believe there is serious science in quantum dot films. I think Ted Sargeant has done some interesting work on QD films and it is easy to get excited when you see that the properties of your films enter the realm of practical application. But, being in the realm is not the same as being competitive.
The devil is in the details, as they say like: -making contacts -fill factor -dark current -longterm stability -uniformity and FPN -blinking effects -mechanical aspects (e.g. delamination) -low noise readout circuits
I suspect Invisage will have trouble with all these details.
And even once competitive, one has to argue compelling advantage for technology insertion.
Lastly there are just the normal business considerations: -Yield, COGS and margin -Can you deliver reliably? -What is the stability of the supply-chain?
The top 3 companies are collectively delivering millions of sensors per day. It is a hard club to break into, even if you are a potentially big player like Magnachip, Dong-bu, and lots of others who have dropped off along the way.
On the other hand, I am only 90% sure we will be asking what happened to Invisage 2-3 years from now. It is 10% probable that they solve enough problems to become a niche player, and maybe 3% probable that they become a top-20 player.
Anyway, as I said before, I would not invest my own money here. But they raised $30M and with such an amazing albeit outrageous PR campaign, they will likely be able to raise another round.
Hey, at least there is an image. The Men in Black get credit for that.
ReplyDeletethey have a slick talk. looks to me like they went to charm school.
ReplyDeleteWhy does this lack the shock value of the day SONY announced Mavica or the day Foveon announced a pixel that read out all three colors? The black and white image is nice, but it is somehow lacking the credibility that sent upper management to accost their engineers on these previous occasions.
ReplyDeleteI lost an evening with my family responding to an emergency request from top execs in another time zone. An evening I will never get back! Well, some CCD guys might say that is karma. I say it should have been a worthier cause.
ReplyDeleteAre the images taken by Invisage's camera, or do they show an image that simulates the 4x increased sensitivity ? It is hard to believe that the image shown (of the guy in the bar) is made by the new camera under low light levels : not a single white spot, not a single defect, no FPN ....
ReplyDeleteSomeone mentioned already earlier : this is a similar story we heard already over a decade ago about a-Si, and I fully agree with this statement. If you have to make an electrical contact between the photoconverting material and the silicon you have a problem ! That is one of the reasons why a 4T-PPD (no contact in the diode) is so much better than a 3T pixel (with contact in the diode) in CMOS. A.T.
I talked to InVisage for their Demo debut then later heard them present at Image Sensors Europe conference this week in London. The last slide was a color image. It wasn't stunningly awesome, but it worked. It's hard to say how much of the issue was due to the hotel's really bad projectors.
ReplyDelete@Anonymous: I'm not expert enough to comment on how well it works, but InVisage uses an electric field to draw the electrons down into the silicon. They say it doesn't use much power.
There was a lot of chatter about InVisage at the conference, but everybody I spoke to was cautious--proof in the pudding and all that--but also because there was a lot InVisage didn't reveal at this stage in its development.
Well, it's hard to believe that there is a serious science behind this show. I hope that this could really be a breakthrough in imaging device, but ...
ReplyDeleteAside from arrogant marketing hype based on erroneous comparisons and leftout details (e.g. CFA transmission) I believe there is serious science in quantum dot films. I think Ted Sargeant has done some interesting work on QD films and it is easy to get excited when you see that the properties of your films enter the realm of practical application. But, being in the realm is not the same as being competitive.
ReplyDeleteThe devil is in the details, as they say like:
-making contacts
-fill factor
-dark current
-longterm stability
-uniformity and FPN
-blinking effects
-mechanical aspects (e.g. delamination)
-low noise readout circuits
I suspect Invisage will have trouble with all these details.
And even once competitive, one has to argue compelling advantage for technology insertion.
Lastly there are just the normal business considerations:
-Yield, COGS and margin
-Can you deliver reliably?
-What is the stability of the supply-chain?
The top 3 companies are collectively delivering millions of sensors per day. It is a hard club to break into, even if you are a potentially big player like Magnachip, Dong-bu, and lots of others who have dropped off along the way.
On the other hand, I am only 90% sure we will be asking what happened to Invisage 2-3 years from now. It is 10% probable that they solve enough problems to become a niche player, and maybe 3% probable that they become a top-20 player.
Anyway, as I said before, I would not invest my own money here. But they raised $30M and with such an amazing albeit outrageous PR campaign, they will likely be able to raise another round.