I purchased some longer focal length lens from M12. I really wanted to use the PT-2016 a 16 mm F2.0 board lens. Unfortunately, it didn’t have a IR cutoff filter, so the colors are not good to say the least. I emailed support at M12 asking if any board filters were available with a build-in IR cutoff filter, and they told me they were not. I noticed the board lens that came with PIXY had a reddish coating on the rear lens element, which I assume is some type of IR adsorbing coating. Are you aware of any longer focal length board lens with an IR cutoff coating?
Hmm we’ll have to do a bit of searching for this. I’ll get back to you with what I find.
Scott
Hi George,
We couldn’t find anything specific to your request. You might try looking “here”:http://www.electronics123.com/new-products/?search=lens+filter if you haven’t already.
Good luck!
Scott
Thanks. The longest f.l. they have that includes an IR cutoff filter is 6 mm. 6 mm is too short for my application. One though I had was to buy this:
IR cutoff filter and place it behind the lens, but in front of the camera. Would doing that cause any problems or issues?
I wouldn’t play with mounting an IR cut filter myself, but that’s my preference. I’m sure there are other lenses out there, longer ones, although 6mm is pretty long for the size of the sensor (1/4"). 6mm on Pixy isn’t like 6mm on a 35mm camera. What field of view are you shooting for?
So I think this is correct— the horizontal FOV for Pixy = 2*atan(2.14/f) where f is the focal length in millimeters.
Here are some other places to buy lenses:
http://www.edmundoptics.com/imaging/imaging-lenses/micro-video-lenses/infinite-conjugate-micro-video-imaging-lenses/2196?showall#products
Hi Rich,
regarding your equation for the FOV 2atan(2.14/f), I thought that the FOV is calculated through 2atan([sensordimension]/2*f) and since the width and heigth of a 1/4’’ are 3,2mm and 2,4 mm respectively, the results I got are 60° of horizontal FOV and 46° of vertical FOV. However, on practice I have it again calculated and in fact they are 75° horizontal and 47° vertical exactly like in the Wiki. Now the question: Are the measures of the sensor 3,2mm and 2,4mm or I am wrong? How do you obtain 2.14 in this equation?
Thanks in advance!
Roberto
Hello Robert,
Pixy’s lens is 2.8mm (where did you get 2.4? We want to make sure this isn’t in our documentation!)
Sensor sizes (1/4") are diagonal, so the width is less than 1/4". The 2.14 is just a constant that makes the calculation easier. We calculated the 2.14, and it takes into account the sensor width, etc.
Edward
Hi Edward, you have misunderstood me. I know that the focal length is 2.8 mm, the 2.4 mm is the height of the sensor (all the sensor whose diagonal is 1/4’’ have 2.4 mm of height and 3.2 mm of width, or at least that is what Google tells me). I also know that 2.14 takes into account the sensor dimensions, because like i wrote before, for calculating i.e. the horizontal FOV the equation is: 2atan([sensorwidth]/2f) (being f=2.8).
But this equation (using sensorwidth=3.2) gives a result of 60° and not 75° like in the documentation.
However, for the vertical FOV: 2atan([sensorheight]/2f) (being sensorheight=2.4mm) gives a result of 46° and that matches with the documentation.
So what I want to know is how this 2.14 related with the width or better, which are the measures of the sensor. (I look into the datasheet of the sensor and it is not there)
I hope not bother you with that and thanks in advance!
Roberto
Hello Robert,
3.2mm width is not accurate. 1/4" is the diagonal, so the width is closer to 4mm. I’m not sure what you need. You want to know the derivation of 2.14? 2.14 was determined by measuring it. We took Pixy’s current lens, did the FOV measurement and backed out the constant.
Edward
Hi Edward,
by taking 1/4’’ I get 5.08mm of width (taking into account a size of 4:3), so I do not know where this 4 mm come from…
At the end I just want to know the exact measures of the sensor (I could measure myself but I am afraid to scratch the sensor)
Thanks!
Hello Robert,
I don’t know the exact dimensions of the active region, sorry 
The aspect ratio is 1280/800, or 1.6 (not 4/3 or 1.33)
Hope this helps!
Edward
Hello Robert,
I don’t know the exact dimensions of the active region, sorry The aspect ratio is 1280/800, or 1.6 (not 4/3 or 1.33)
Hope this helps!
Edward
According to the OV9715 datasheet:
Width 3888 micrometers
Height 2430 micrometers
Thank you both,
I thought that these were not the sensor dimensions since they did not match (through maths) with the field of view of the specifications. I will try it again
Thanks again for the help!
Robert
Did anyone ever come to a conclusion about this?
OV9715 datasheet:
Width 3888 micrometers (3.888mm)
Height 2430 micrometers (2.430mm)
This results in field of view with a 2.8mm lens:
69.54 degrees
46.91 degrees
(using the standard formula 2atan([sensordimension]/2f) )
In order to get the stated FOV in the wiki of alpha_w=75deg and alpha_h=47deg, I had to put the sensor height and width at:
d,w 4.297 mm
d,h 2.43 mm
This is very important info to know for me as I am estimating object distance using these numbers. Thanks!
Hello GB,
These are the correct equations. You are estimating object distance based on width of the detecting object? It you’re interested in accurate numbers, you should consider calibrating f by using known values for (1) width of object (2) distance to object and (3) width of object (number of pixels) on the sensor.
Hope this helps! and let us know what you find.
Edward
Thanks, Edward.
For number of pixels on the sensor, do you use
width = 640
height = 400
OR do you use
width = 319
height = 199
which are the max X and max Y range reported by the CMUcam, respectively?
Hi GB,
Use the 320x200 numbers for resolution. This is the effective resolution for the color connected components algorithm that Pixy uses.
Edward