FOV or angle of view on the DS-2CD2342WD-I
- Thread starter tomhanman
- Start date
These specs from vendors are notoriously inaccurate, and there's really only one way to be sure - measure them yourself (or use someone else's actual measurements). The easiest way is to measure the width of a full-screen item and the distance to the item measured.
I have a post over at cam-it that shows how to do it with a ruler and rod and gives examples, but you can do it with a fence, a wall, or anything that's easy to measure:
http://www.cam-it.org/index.php?topic=5424.0
The problem with working this backward to give your lens size, as CamFan found, is that it doesn't give your actual lens size, only your virtual lens size. For instance, a given 4mm lens will have a particular image circle size at its focal plane, while a different 4mm lens will have a different image circle size. If you order 3 different 4mm lenses for 1/3" sensors, each will give different results and will appear to be different lens sizes. Generally, the image circle is a good bit bigger than the sensor it's designed for; an image from a 1/3" lens will often cover most or all of a 1/2" sensor, based on my tests.
Add in different sensor sizes, and the problem's worse. A 1/3" 4mm lens will give a different FOV on 1/3", 1/2.8", and 1/4" sensors, and not all 1/3" (or whatever) sensors are the same size or sample the same number of pixels from the sensor.
It's also good to know that a 1/3" sensor isn't actually 1/3" diagonally. The specs for these sensors will often list them as "1/3" class" or similar. The sizing comes from the days when imagers were in vacuum tubes, and the size was the outside diameter of the glass tube. Nobody's willing to give the actual sensor size, as it would make them look smaller than the competition, and their marketing would suffer.
You can ballpark the actual size by multiplying the marketing size by 0.7, but to get the precise size, you need the image area dimensions from the spec sheet for the sensor.
I have a post over at cam-it that shows how to do it with a ruler and rod and gives examples, but you can do it with a fence, a wall, or anything that's easy to measure:
http://www.cam-it.org/index.php?topic=5424.0
The problem with working this backward to give your lens size, as CamFan found, is that it doesn't give your actual lens size, only your virtual lens size. For instance, a given 4mm lens will have a particular image circle size at its focal plane, while a different 4mm lens will have a different image circle size. If you order 3 different 4mm lenses for 1/3" sensors, each will give different results and will appear to be different lens sizes. Generally, the image circle is a good bit bigger than the sensor it's designed for; an image from a 1/3" lens will often cover most or all of a 1/2" sensor, based on my tests.
Add in different sensor sizes, and the problem's worse. A 1/3" 4mm lens will give a different FOV on 1/3", 1/2.8", and 1/4" sensors, and not all 1/3" (or whatever) sensors are the same size or sample the same number of pixels from the sensor.
It's also good to know that a 1/3" sensor isn't actually 1/3" diagonally. The specs for these sensors will often list them as "1/3" class" or similar. The sizing comes from the days when imagers were in vacuum tubes, and the size was the outside diameter of the glass tube. Nobody's willing to give the actual sensor size, as it would make them look smaller than the competition, and their marketing would suffer.
You can ballpark the actual size by multiplying the marketing size by 0.7, but to get the precise size, you need the image area dimensions from the spec sheet for the sensor.
@MaxIcon
I have read your work back in time when you did it and I fully agree with your measuring methodology and all your technical explanations. I also fully understand the optics and sensors characteristics. However very few people have repeated the work you have done with their cameras like you did. Also few people report properly and accurately the source of their information.
Therefore my question to @tomhanman still stands.
I have read your work back in time when you did it and I fully agree with your measuring methodology and all your technical explanations. I also fully understand the optics and sensors characteristics. However very few people have repeated the work you have done with their cameras like you did. Also few people report properly and accurately the source of their information.
Therefore my question to @tomhanman still stands.
Last edited by a moderator:
A big issue was that Hikvision actually released the wrong FoVs in the spec sheet. Many suppliers are still using the old spec sheet with the wrong info.
Hi, if you read the thread, you would see that I set out trying to get reliable FOV angles from the web and this forum... and failed!
So, I purchased the three of these cams, one in each of 2.8mm, 4mm and 6mm lens length (plus a similar variable zoom camera for fun).
I set them each up in a workshop, measured out some big triangles, then used a bit of trigonometry to calculate the angles.
I avoided all this crap about trying to work things out via the conceptual size of sensors or any of that nonsense, which doesn't work, and from the start just sounds like a waste of time. This is simple measurement of what exists rather than theoretical calculation of what should (if we were dealing with Leica but we're not) exist.
So assuming these things are all produced the same (big assumption for anything coming from over there) and being clear that these figures are for the full 4MP 16/9 resolution then I'd say the figures I gave above are accurate (+/- 1°) horizontal FOV angles for DS-2CD2342WD-I cameras labelled with 2.8, 4 and 6mm lenses.
Enjoy!
So, I purchased the three of these cams, one in each of 2.8mm, 4mm and 6mm lens length (plus a similar variable zoom camera for fun).
I set them each up in a workshop, measured out some big triangles, then used a bit of trigonometry to calculate the angles.
I avoided all this crap about trying to work things out via the conceptual size of sensors or any of that nonsense, which doesn't work, and from the start just sounds like a waste of time. This is simple measurement of what exists rather than theoretical calculation of what should (if we were dealing with Leica but we're not) exist.
So assuming these things are all produced the same (big assumption for anything coming from over there) and being clear that these figures are for the full 4MP 16/9 resolution then I'd say the figures I gave above are accurate (+/- 1°) horizontal FOV angles for DS-2CD2342WD-I cameras labelled with 2.8, 4 and 6mm lenses.
Enjoy!
@tomhanman Thanks for your answer. Therefore it is clear that you measured the angles through simple trigonometry of the visual dimensions of whatever you measured. I wish you could also measure the Vertical FOV angle. Personally I am planning an installation with cameras at the corridor mode (optically) to maximise the view but I also need the Vertical FOV angle which will become the horizontal angle in the corridor mode. Anyway thanks after all for your effort.
Last edited by a moderator:
Hello Aster1x,
Sorry, but I stuck them up on walls just after taking the measurements. Can you not calculate the vertical angles by simply using the 16:9 ratio on the horizontal angles measured? It kind of has to work assuming that things are going to be seen in their correct proportions. So I'm guesstimating:
Lens Hor° Ver°
2.8° 105 59
4.0° 81 46
6.0° 55 31
The lens has equal optical effect in all directions, it is only the crop of the sensor that dictates the FOV, and as the sensor is 16:9, then I guess this has to work... But it is still theoretical, and I ain't Einstein!!
Sorry, but I stuck them up on walls just after taking the measurements. Can you not calculate the vertical angles by simply using the 16:9 ratio on the horizontal angles measured? It kind of has to work assuming that things are going to be seen in their correct proportions. So I'm guesstimating:
Lens Hor° Ver°
2.8° 105 59
4.0° 81 46
6.0° 55 31
The lens has equal optical effect in all directions, it is only the crop of the sensor that dictates the FOV, and as the sensor is 16:9, then I guess this has to work... But it is still theoretical, and I ain't Einstein!!
@tomhanman You assume that the sensor is 16:9 and unfortunately we do not know the sensor maker and model. Additionally the sensor vendors prefer to keep the top and bottom few lines blanked during frame refreshing. That means that even if the physical sensor may 16:9 accurately, the pixel elements are not square and the final optical result is almost never 16:9.
So as we all now, assumption is the mother of all f@#$%!^&#*@ ups.
Thanks any way for your help. Soon I will be getting a 4mm camera and hopefully I will tear it down and report back my findings.
So as we all now, assumption is the mother of all f@#$%!^&#*@ ups.
Thanks any way for your help. Soon I will be getting a 4mm camera and hopefully I will tear it down and report back my findings.
Last edited by a moderator:
Kawboy12R
Known around here
- Nov 18, 2014
- 1,763
- 609
I would trust Tom's measurements and reasoning to be well within the tolerances of most end users' ability to judge correctly their own needed field of view. If you need more accuracy than what might amount to probably at most a degree or two of variance then you shouldn't be buying off-the-rack fixed lenses because the odds of them matching your precise needs correctly are basically nil. Buy varifocals instead.
Assumption is bad, agreed! But as the image is displayed in perfect 16:9 format, and a round thing appears round, that means whatever shaped pixels they are they representing the world mostly accurately. And if that is the case then the vertical angle of view has to be locked to the horizontal otherwise every burglar I film would either look like a dwarf, or a giant. As far as I'm concerned the sensor can be the size of the Isle of Man, but the pixels used are pretty much spot on - maybe 8 pixels narrower than it should be. But that isn't getting close to even half a degree. So I stand by my guesstimate... (in order to be knocked down when you get yours).
Hi Tomhanman
I am planning to buy cameras with Hofotech. Their cams are OEM. You camera is running well? Shall I go ahead with them as a supplier?
I am planning to buy cameras with Hofotech. Their cams are OEM. You camera is running well? Shall I go ahead with them as a supplier?
PSPCommOp
Getting the hang of it
Thanks, I've purchased the 4mm and was looking at a few 2.8's for other areas. This helps greatly.
Hi Argupta, my cameras are all working perfectly. They are international versions in international packaging (red and white boxes). I have done no firmware mods, I have no flicker/screen white/black problems. All I can say is that I am happy, and that I will be purchasing from them again. Get in contact with Cola at Hofotech and see if you feel happy...
Glad to have been of assistance PSPcommOp
Glad to have been of assistance PSPcommOp
n4gry
Getting the hang of it
Thanks for the advice. I have placed an order with them for a sample for a 4 megapixel 2142. Your previous comment made me order with them. They have unbelievable prices which made me skeptical. Thanks though once again! I am satisfied now
Nobody came back to say if they were happy with Hofotech? I hate looking like an advertiser or a muppet. I hope to hear good stories. I'm just about to make an order of ten cameras to go with the 4 I already ordered and the NVR. Everything is working well, except the HDMI output from the NVR which got fried by a lightening strike that took out the monitor as well, but VGA still works fine. Looking forward to lots of cat5 wiring!
Hi
Sorry I didn't get back. Yes I ordered from Hofotech. Yes I am really happy with their service. I ordered the 2342 DOME 4 MP from them, just one piece, only to try the product and their service. The camera works excellent, no problem at all. It is english firmware. I use it for live viewing only.
You have messaged at a great time coincidentally because I am about to place my second order with them for
DS-2CD2142FWD-IW 4mm 1 PCS
DS-2CD2T42WD-I8 6mm
DS-2CD2T42WD-I8 4mm
DS-7608NI-K2/8P
Very good pricing. I am spending almost 50% less the amount of money that I would have spent had I purchased this from my local dealer of Hikvision.
Thank you @tomhanman , because of your thumbs up for Hofo, I ordered from them and I am really happy.
Which cameras are you going to order and what NVR?
Sorry I didn't get back. Yes I ordered from Hofotech. Yes I am really happy with their service. I ordered the 2342 DOME 4 MP from them, just one piece, only to try the product and their service. The camera works excellent, no problem at all. It is english firmware. I use it for live viewing only.
You have messaged at a great time coincidentally because I am about to place my second order with them for
DS-2CD2142FWD-IW 4mm 1 PCS
DS-2CD2T42WD-I8 6mm
DS-2CD2T42WD-I8 4mm
DS-7608NI-K2/8P
Very good pricing. I am spending almost 50% less the amount of money that I would have spent had I purchased this from my local dealer of Hikvision.
Thank you @tomhanman , because of your thumbs up for Hofo, I ordered from them and I am really happy.
Which cameras are you going to order and what NVR?