First of all, the use of the unities (V/(lux·sec)) isn't adequate, because the output in volts or in digital number always can be amplifier arbitrarily. Is much more convenient to express the sensitivity in terms of e-/(W/m^2), although this is basically equivalent to the QE.
Other important point when sensitivity is expresed in (V/(lux·sec)) is that this depends strongly from the wavelength. For sure that for wavelenghs that are almost invisible for the human eye, the sensitivity in luxes (that is relative to the human sensitivity) tends to infinity. Then, this data hasn't sense if the wavelength isn't specified.
Is probable that the 20000mV/lux·sec is a bug and probably must say 2000mV/lux·sec, that is a good data, but isn't crazy.
The most important parameters are the QE, the noise level and the FWC. The sensitivity in V/lux·sec is only marketing
Of course, this came the first time from FillFactory datasheet. They used 650nm wavelength as reference. The sensitivity was given with V/Lux*s at the beginning and you can find the wavelength at the end of the datasheet ... very smart!
what is the motivation of 'upright' field of view (1752x2336)? I think most 'landscape' type of sensors come from their 'photography' type of application. I think for microscopy or 'machine vision' square is best in many cases because this fills the usable area of the optics best. But what is the benefit of a 'upright' field of view?
How can we get 20000mV/Lux*s with 4.2um pixel pitch??
ReplyDeleteFirst of all, the use of the unities (V/(lux·sec)) isn't adequate, because the output in volts or in digital number always can be amplifier arbitrarily. Is much more convenient to express the sensitivity in terms of e-/(W/m^2), although this is basically equivalent to the QE.
DeleteOther important point when sensitivity is expresed in (V/(lux·sec)) is that this depends strongly from the wavelength. For sure that for wavelenghs that are almost invisible for the human eye, the sensitivity in luxes (that is relative to the human sensitivity) tends to infinity. Then, this data hasn't sense if the wavelength isn't specified.
Is probable that the 20000mV/lux·sec is a bug and probably must say 2000mV/lux·sec, that is a good data, but isn't crazy.
The most important parameters are the QE, the noise level and the FWC. The sensitivity in V/lux·sec is only marketing
Of course, this came the first time from FillFactory datasheet. They used 650nm wavelength as reference. The sensitivity was given with V/Lux*s at the beginning and you can find the wavelength at the end of the datasheet ... very smart!
Deleteflyers are listed on their homepage... https://www.smartsenstech.com/en/gs_products
ReplyDeleteThose flyers list significantly lower sensitivity values. Also "greater than 99% shutter efficiency" is not really inspiring confidence.
DeleteIn their flyer, shutter efficiency has been given as 1:10000. Not bad.
Deletewhat is the motivation of 'upright' field of view (1752x2336)? I think most 'landscape' type of sensors come from their 'photography' type of application. I think for microscopy or 'machine vision' square is best in many cases because this fills the usable area of the optics best. But what is the benefit of a 'upright' field of view?
ReplyDelete