Actual lighting requirements for Dahua Starlight 2MP cams

S

SecuritySeeker

Pulling my weight
Oct 5, 2018
266
157
Netherlands
Building of our new home is progressing nicely and I'm currently looking into what kind of streetlights I want to install on the side facade of our home (circled red):

New%20home%20security%20-%20outside%20model.png
New%20home%20security%20-%20outside%20street%20plan,%20camera%20suggestions%20-%20small.png


The main area I want to illuminate is the path right next to these inbetween the house and the parking area (there'll be a metal fence separating the the path and parking space). I suppose it wouldn't hurt to also illuminate the parking area but my cameras will be trained on the path. I'm not sure if I'll install a camera to watch our car, current car is 20 years old so...

I will have mounting points for lighting (230VAC) and cameras (CAT6) in the following locations:

New%20home%20security%20-%20outside%20lighting%20and%20cameras%20-%20small.png

Cameras will probably be Dahua Starlight 2MPs. I'm wondering how many Lumens of light I should install to get the best images from these. Because this is a purely residential area I don't want to cause any irritation by installing lights that are too bright or shine too wide (the block of houses labeled 20 might be bothered by glare from the lamp). I was thinking of something like this:

30W SMD Street Lamp With Photo Cell Sensor Grey Body Natural White
Color Temperature 4000K

Luminous flux 2400 lm (30W)
Color Rendering Index CRI >70
Beam angle 120 °

These come in varies varieties:
Beam angle 100 ° or 120 °
Color temperature 3000, 4000 or 6000K
Lumens 2400 or 4000
Photo Sensor yes/no

I'm not sure if all combinations are available but I was wondering if anyone could comment on the best beam angle, color temperature and amount of lumens for this application.

I'm not sure if these lights will be on their own private electrical circuit group so I might need the built-in photo sensor to have them switch on automatically at night (I realize they won't switch on exactly at the same time if they have their own sensor but that doesn't bother me).

I'm thinking (without the benefit of experience) that 3000K might be more pleasant for the neighbours and anyone passing there (including ourselves). I'm also thinking that 100° is probably wide enough. I really have no idea how many Lumens I should choose. Like I said I'm likely to install 2MP starlights and I'm guessing that those would provide a good image with the 2400lm lights, or not? Any ideas?
 
  • Like
Reactions: mat200 and BigFoot
Remember to also consider getting a camera focused on where you will be parking your car, or where your guests may park.
 
mat200 said:
Remember to also consider getting a camera focused on where you will be parking your car, or where your guests may park.
Click to expand...

Thanks for the comment. I've thought about it but like I said:

SecuritySeeker said:
I'm not sure if I'll install a camera to watch our car, current car is 20 years old so...
Click to expand...

I'm also not yet sure if it will even be possibly physically to get a good view on where our car will be parked (our spot would be right about where the one car is parked in the red ellipse). However, the houses in block 20 and those parking places will not be built for another 2-3 years and in the mean time we'll have to park our car at a very different location out of view from our home so for now it's a moot point really.

The parking area for guests is also a bit further away and it will almost certainly be out of view from our home and propery as well plus they will have to pick a random place from the available parking spaces for guests. Apart from that it is very much illegal over here to aim private (security ) cameras on public places. We can get away with filming the yellow highlighted area because it's actually our property, even if some neighbours will have right of way and will be get filmed by our cams.
 
Two perspectives with 3400 lumen spotlights. (Both cameras as 3231E-Z CVI cams, very equivalent to the 5231 IP series)

In this one the camera stays in color mode thanks to (2) 1200 lumen coach lights, and average street lighting. See the cats trip the 3400 lumen spot. Camera is about 20-25 ft to the cats.

*Remember to switch youtube player to 1080p



This one is in B&W/IR mode. Dock is pitch black at night, cant see it from porch. Notice how the 3400 lumen spot brightens the immediate area of the dock. Dock is 33 ft from camera, which is only slightly zoomed. (added bonus see how quickly it reacts switching to color and back)

 
  • Like
Reactions: japjoe7, looney2ns, dark current and 2 others
To be somewhat technical about it, lumen is not really the number that answers the question.

Lumen is essentially just the light output of the lamp. What is really needed to answer the question definitively is the lux illumination level in/around the target area.

The lux level is harder to ascertain because it depends on the light output (lumens) and how that light is spread over a given area.

For example, just using the 2400 lumens, the lux level could be very high if the 2400 lumens is focused/controlled in a tight beam or could be very low in the case of a bare lamp with the output going all directions.

Depending on the supplier of your lighting, they may have data that gives lux levels at various mounting height/installation circumstances for the specific fixtures you're considering.
 
dark current said:
Even if you don't use them now, setting a few boxes and at least having pull line in place would be wise for the parking area, or a few other possible camera & light locations. It will be much easier during construction than after.
Click to expand...

I have done what I could in that regard, having those extra conduits installed was hideously expensive.
 
For example of my previous post regarding lumens/lux.

Here is an example of illumination profile. This gives foot-candles but conversion to lux is roughly x11.

(edit: for clarity this is just an example I picked at random, it is not intended to represent the performance of the specific lighting mentioned in this thread.)
Screenshot_20190117-091445_Drive.jpg
 
Last edited:
  • Like
Reactions: bigredfish
Whoaru99 said:
To be somewhat technical about it, lumen is not really the number that answers the question.
Click to expand...

I realize that and that's why I included the picture of how high they are mounted and why I asked for comments on both the beam angle and lumens.

Lumen is essentially just the light output of the lamp. What is really needed to answer the question definitively is the lux illumination level in/around the target area.
Click to expand...

I kinda knew that but your comment made me realize I could calculate the lux illumination level myself. So I did a quick Google and found this:

Lumens-Lux Calculator

I has several calculators including a Lumens to Lux calculator that takes Beam angle, Distance and Lumens as input and has Lux as the output, so if I enter something like:

Beam Angle: 100
Distance: 5 meters
Lumens: 2400

it returns 42.77 lux. If I look at the specs for the IPC-HDW5231R-ZE they say:

Minimum Illumination

0.006Lux/F1.6 ( Color,1/3s,30IRE)
0.05Lux/F1.6 ( Color,1/30s,30IRE)
0Lux/F1.6 (IR on)

So at first sight it would seem that 42.77 Lux is plenty since it can do 30fps at 0.05 Lux. However it also specifies 30IRE. I'm not sure exactly how to interpret that but if this website is correct an IRE of 100 is a perfect image whereas 30IRE basically means that the sensor's output is 30/100 of it's maximum output, ie. only the bottom 30% of the sensor response is used or to put it yet another way the dynamic range of the output is 30/100 of what it could do under better lighting conditions.

I also don't know the shape of the curve that describes the maximum sensor output with increasing lux input into the camera It might be exponential or something so it might take a lot of extra Lux to get a significantly better IRE (according to that same website an IRE of 50 would be suitable for security purposes).

I do get the impression that with 42.77 Lux the Starlight would perform stellar though, although it would be nice to know what is needed to get 30fps with 100IRE from those cams.

As for the street lighting level itself for persons passing through, this post suggests that 20 lux is "good for mixed vehicle and pedestrian traffic with separate footways" so I'm guessing that 42 Lux will be quite tolerable and sufficient for this application.

I just noticed that this Practical Guide for Outdoor Lighting suggests that "When white light is required, warm white LEDs (<3000 K),which are now available on the market, are recommended", something I was already thinking might be the most pleasant to use. Unfortunately the brand I found only seems to carry those without photo sensors.

Edit: @Whoareu99 I see our posts crossed, I was still editing mine when you posted
 
Last edited:
I'm wondering if I'm interpreting the minimum lux specs correctly. I've got a feeling it refers to the light directly hitting the sensor or perhaps the lens, not the amount of light hitting the scene being viewed.
 
SecuritySeeker said:
I'm wondering if I'm interpreting the minimum lux specs correctly. I've got a feeling it refers to the light directly hitting the sensor or perhaps the lens, not the amount of light hitting the scene being viewed.
Click to expand...
The absolute minimum lux ratings on the cameras is worthless. At those light levels, the camera mfgs basically just decided that there's something barely discernible from the typical noise you see in the dark.

Don't Trust Lux Ratings
MP Camera Low Light Shootout 2011
 
tangent said:
The absolute minimum lux ratings on the cameras is worthless.
Click to expand...

Yeah, that appears to be the case. As I did some more reading yesterday (I also found the two sites you link to) I slowly started to reach that same conclusion. It seems that the only proper way to evaluate a camera's low light performance would be to watch some (standardized) footage in properly specified and standardized lighting conditions with movement in the scene. In principle it should be possible to express that in numbers but there's no standard to do that and as a result it's too easy to manipulate.

I guess what it comes down to is that I'll just have to order some streetlights and see if the Dahua starlights I plan to order will perform properly in this particular situation, i.e. if I can identify someone running past them. I think I'll start with 2400 Lumens at 3000K which is probably going to be the most pleasant and acceptable lighting for us and our neighbors. Worst case I can always switch them to IR if I don't want to increase the amount of lighting but then I guess I wouldn't really need starlights and could just as well go with some 4MP or higher cams.

The lighting is just as much to provide a (feeling of) security for anyone walking there as it is for the cameras to produce some nice color images at night. I very much expect any potential burglars to wear hoodies or at least baseball caps but they probably won't like the idea of being on camera in a properly lit area anyway. Probably the only way to identify them is if they case the house in advance without the hoodie etc. so it's probably a good idea to have a generous recording retention time. In addition to acting as a deterrence the cameras are also to help verify a break-in if the alarm system gets triggered.

I appreciate all the reviews done by people here and all the footage that is uploaded. It would be nice if people reviewing cams could measure the actual lux of the scene they're filming and put that info in the review. But like I said, I'm happy with all the effort put in by people to upload videos and do reviews in the first place so thank you all.
 
I thought about measuring lux and providing pictures for a reply since I am presently using two cams just on the light provided by the City. One cam is full color Starlight with no IR and the other is a non-Starlight with the IR turned off.

But, I have only a cell phone app to do it and a bit of reading suggests they are not too accurate in general. Even worse, are not consistent readings between different apps or different brands and models of phone.
 
Instead of a meter I would take a picture using a SLR camera of the scene and report the ASA, shutter, and F/no. Unfortunately my SLR seems to have died.

I ended up using Dahua
IPC-HFW4239T-ASE 2MP cameras out front where the neighborhood lights provide enough light. In the back and side yards I use
IPC-HDBW5231R-ZE 2MP cameras in IR mode. I originally bought one of each and tried them out using a 2 x 4 pole in a bucket as discussed elsewhere around here.

The 4239s are fixed focal length with the focal lengths picked while testing with the 5231. The 4239 is F/1.0 while the 5231 is F/1.4. This lets in twice as much light. This is at the widest focal length setting of 2.7 mm on the 5231. Most lenses get slower when zoomed to a longer focal length. I expect an additional one stop advantage for my 4239s with the 6 mm lenses. Two stops lets in four times as much light.
 
You must log in or register to reply here.
Top Bottom