DS-2CD2325FWD-I 2MP Ultra-Low Light Network Turret Camera

What kind of performance increase (at night) would the i8 bullet have over the turret?
 
What kind of performance increase (at night) would the i8 bullet have over the turret?
In my test of the 4MP DS-2CD2T42WD-I8 16mm at 1am in the morning, no street lights, no other lighting, it could get a marginal id shot at about 66 feet. It has much stronger IR than the 4MP DS-2CD2345-I 4mm turret, which can only see to about 20' before the IR is too weak (at about 40 feet its just black, it can't see anything).

The new ultra low light models should perform better than the 4MP models. The IR range might be more important when using a longer lens, but even with a shorter lens its useful to be able to see if something is out there or not to have a longer IR range.
 
Here's my 6mm i8. From the camera to the front of my neighbors house is 50'. The camera is about 10' off the ground.
home_IP Camera1_home_20170322083803_34594.jpg home_IP Camera1_home_20170322004433_87205.jpg
 
The IR range might be more important when using a longer lens, but even with a shorter lens its useful to be able to see if something is out there or not to have a longer IR range.
Thank you for your review of other low light cams by Hikvision. I wonder if this cam's IR will wash out, or over expose, a person's face. Any chance of snapshot, or video, of person's face with motion at night?
 
2MP is the Sony Exmor-R Starvis IMX291
3MP is the Sony Exmor-R Starvis IMX124

I just did a bit of Googling and it appears Starvis and Exmor-R are two related sensor ranges?
STARVIS Image Sensor for Industry | Sony Semiconductor Solutions Corporation
ExmorR, Exmor Image Sensor for Industry | Sony Semiconductor Solutions Corporation

The 3MP IMX124 appears to be under the Starvis range, there are a few videos on Youtube from cameras using that sensor, but footage without another camera to compare the conditions doesn't really help much.

There's a page talking about the IMX291 here IMX290LQR, IMX291LQR | Sony Semiconductor Solutions

I'm in the process of setting up my 2MP Dahua Starlight IPC-HDW5231R-Z next to a 4MP Hikvision DS-2CD2342WD-I to compare day and night. I'll be locking both in day (colour) mode and not using IR as I've got 3 x 10 Watt LED flood lights that come on automatically at night. I want a camera that can give best results in those conditions.
 
Exmor R mean Back-illuminated CMOS Image Sensor
Starvis mean Back-illuminated CMOS Image Sensor with a sensitivity of 2000 mV or more per 1 µm2

The best lowlight-camera I know is the Hikvision Darkfighter with IMX185, could be quivalent to the Dahua UltraSmart IPC-HFW8232E-Z..

IMX185 is old, no starvis, no ExmorR just Exmor. But it is really big. :-)

Morpheus
 
The thing I wonder when I read " 0.005 lux @ (F1.2, AGC ON)" is that why do they specify they can reach F1.2 only if AGC is ON ? meaning this AGC feature produce "arificial" computed gain that is not reached thru physicial capabilities of the cmos/lens couple and so result will be not as good as "real" F1.2 ?
 
none of those cameras came with F1.2 optics IIRC so dont really matter much since they giving you specs for a lens they not providing
 
  • Like
Reactions: hmjgriffon
I guess it's industry standard to quote sensor performance independently of the actual optics. All cameras tend to be advertised this way.
AGC ON is pretty standard too (and that's how you'll be using the camera in most cases so it's reasonable). However it makes the numbers even less objective.
All in all, these numbers are useful to some extent to compare models of cameras of the same manufacturer, but not much else.

For reference, while the sensor "min illumination" is usually quoted for f/1.2, what you actual get is (the Hik Europe website has all the specs unlike the US one):
* f/2.0 for most fixed focal cameras including the turrets
* f/1.4 for some vari-focal cameras
* f/1.4-f/1.6 for most lenses for C-mount box cameras
* f/0.95 (!) for the "DarkEye" lens that you can buy in combination with the Darkfighter xx26 cameras - $600 for the lens and it's not very wide angle

For reference on the new Exmor-based cameras:
* The 2MP Darkfighter (xx26) series is quoted as 0.001 @ f/1.2 color with a big sensor (1/2.0 or 1/1.8)
* The new xx25 Exmor-based series is quoted as 0.005 @ f/1.2 with a regular 1/2.8 sensor
 
  • Like
Reactions: spencnor
Dahua gives you lux ratings based off the actual packaged optics, its not industry standard.. just Hikvision Standard
 
  • Like
Reactions: morpheus
For reference on the new Exmor-based cameras:
* The 2MP Darkfighter (xx26) series is quoted as 0.001 @ f/1.2 color with a big sensor (1/2.0 or 1/1.8)
* The new xx25 Exmor-based series is quoted as 0.005 @ f/1.2 with a regular 1/2.8 sensor

xx26 Darkfighter uses the IMX185. On hikvisioneurope it is 0.0027lx @F1.4
xx25 Lightfighter uses the IMX290. On hikvisioneurope it is 0.007x @F1.4

IMX185 is 1/1,8", the IMX290 is 1/2,8". But it is not as simple as that. The IMX185 is big but uses old technology. It is a frontside illuminated Sony Exmor. The IMX290 is smaller but uses newer, more light sensitive technology. It is a backside illuminated Sony Exmor R starvis.

An IMX290 with the size of the IMX185 would be great. ;-)

Morpheus
 
  • Like
Reactions: kwijibo and nayr
How do they determine the acceptable level of video quality to get those published specification? I assume (from the above) that Hikvision uses the sensor manufacturer's supplied specs? Does Dahua do some sort of in house testing?

It seems like the only way to know if a particular camera would be acceptable is to test it for yourself.

Edit: on the spec sheet for the 3MP it gives the 0.005 lux rating at f1.2 with AGC on, then further down the page it says Iris f1.6. The 2MP version is the same. So I assume the maximum aperture is f1.6?
 
Last edited:
The exposure time used with any colour shots would be good info to include. For example, I can get quite good colour with my 4MP Hikvisionwith some ambient lighting, but that's with the shutter speed so slow anything moving is blurred. I've found 1/25th second a good compromise, cars are still a bit blurred but I can identify the local neighbour's vehicles from the colour and shape. (I live near the end of a dead end street so they're usually be travelling under 40km/h.)

Edit: my DS-2CD2342WD-I appears to be the turret equivalent of miapaglia's Hikvision bullet but with lower power IR (30m vs 80m.) Mine is also 4mm.
 
Last edited:
My DS-2CD2335FWD-I arrived in the mail yesterday. Currently testing it out :)

I'm very interested to hear how it performs compared to your I8 in colour at night.

I've been running my DS-2CD2342WD-I 4mm next to a Dahua 2MP Starlight, both locked in colour at night at 1/25th. There's a useful gain in the dark with the Dahua, but during the day the DS-2CD2342WD-I wins. I've been trying to export some example JPEGs from my NVR, but it seems to resize the 4MP down to 1080P (2MP) size.
 
My camera will be arriving on Friday according to DHL. I have to sign for it. I will be interested in the test between the 2mp and 3mp camera.