@Kawboy12R and @crc2004 have brought up in a few places (e.g., Two cams with 12mm lens, but VERY different FOV. Why?) an issue with calculating HAoV from imager sizes and focal lengths.
I wanted to address the question, share some feedback and see what others have to say. I'll start first with the general problem and then into the specific issue raised.
The general problem is that manufacturer specified AoV is frequently different from what the theoretically calculation is when entering focal length (e.g., 2.8mm, 4mm, etc.) and imager size (1.3", 1./2.7", etc.).
The best way we have found to solve this is to enter in the actual AoVs from the manufacturer datasheet, e.g., we have 5,300+ (and growing every month) models in our IPVM Camera Calculator V3 While it's time consuming and costly, it provides the best solution. One key downside is that new models are always coming out and, while we are happy to add new ones in, we can't get to everyone.
Now, the specific issue with imager size calculation that @Kawboy12R has rasised . To keep it simple, historically, one just chooses a size (1.3", 1./2.7", etc.).) Problem is these are not the actual physical sizes (discussed here Image sensor format - Wikipedia). To make things more complex, imagers now have different aspect ratios and those differences will impact the potential AoV. In particular, I believe this is the issue with 1/2.7, 1/2.8, etc.
What To Do?
The solution that 90%+ of users (at least with IPVM) have done is to pick from the 5,300+ models and get the AoV listed from the manufacturer. This eliminates needing to find or enter or figure out imager sizes, etc.
Even if we address the imager size issue (and I am not sure since that depends on us tracking not only the size but the sensor model, which would increase complexity), many times the AoV of the model will be significantly different, depending on how the manufacturers uses/crops the sensor, the lens choice, etc.
My gut feel is to either remove the imager size input and focus people on either entering the model or finding the HAoV on their own or to move the imager size / focal length input to a more advanced section that explains the issues.
What do you think?
I wanted to address the question, share some feedback and see what others have to say. I'll start first with the general problem and then into the specific issue raised.
The general problem is that manufacturer specified AoV is frequently different from what the theoretically calculation is when entering focal length (e.g., 2.8mm, 4mm, etc.) and imager size (1.3", 1./2.7", etc.).
The best way we have found to solve this is to enter in the actual AoVs from the manufacturer datasheet, e.g., we have 5,300+ (and growing every month) models in our IPVM Camera Calculator V3 While it's time consuming and costly, it provides the best solution. One key downside is that new models are always coming out and, while we are happy to add new ones in, we can't get to everyone.
Now, the specific issue with imager size calculation that @Kawboy12R has rasised . To keep it simple, historically, one just chooses a size (1.3", 1./2.7", etc.).) Problem is these are not the actual physical sizes (discussed here Image sensor format - Wikipedia). To make things more complex, imagers now have different aspect ratios and those differences will impact the potential AoV. In particular, I believe this is the issue with 1/2.7, 1/2.8, etc.
What To Do?
The solution that 90%+ of users (at least with IPVM) have done is to pick from the 5,300+ models and get the AoV listed from the manufacturer. This eliminates needing to find or enter or figure out imager sizes, etc.
Even if we address the imager size issue (and I am not sure since that depends on us tracking not only the size but the sensor model, which would increase complexity), many times the AoV of the model will be significantly different, depending on how the manufacturers uses/crops the sensor, the lens choice, etc.
My gut feel is to either remove the imager size input and focus people on either entering the model or finding the HAoV on their own or to move the imager size / focal length input to a more advanced section that explains the issues.
What do you think?