I was comparing specs of a couple of Dahua cameras and noticed many of their 2MP and 4MP cameras have the same sensor/size, same F-stop, but also have the same minimum illumination numbers. I don't understand how this is possible. If the optics are the same, the CMOS's type/physical size is the same, and you divide the individual pixel size down, it should have less sensitivity. Can anyone explain what's going on here?
Dahua Illumination Specs
- Thread starter fstop
- Start date
SouthernYankee
IPCT Contributor
Sure, if the sensor size is the same size. The light per pixel is 50% less on the 4MP then on the 2MP. People think more MP is better, so they buy more MP.
The current Dahua goto cameras are the 5442 which have a 1/1.8 sensor.
IPC-T5442TM-AS-LED . Review IPC-T5442TM-AS-LED (Full Color, Starlight+) - 4MP starlight
.................... Dahua IPC-T5442TM-AS-LED review
IPC-T5442TM-AS ..... Review-OEM 4mp AI Cam IPC-T5442TM-AS Starlight+ - 4MP starlight+
IPC-HDW5442t-ZE .... Dahua IPC-HDW5442T-ZE 4MP Varifocal Turret - Night Perfomance testing -- variable focus 2.7 mm-12mm 4 MP Starlight
IPC-B5442E-ZE ...... Review - OEM IPC-B5442E-ZE 4MP AI Varifocal Bullet Camera With Starlight+ -- variable 2.7mm-12mm bullet
IPC-B5442E-Z4E .... bullet 8mm-32mm variable focus zoom 4MP
IPC-HFW7442H-Z ..... Review - Dahua IPC-HFW7442H-Z 4MP Ultra AI Varifocal Bullet Camera -- 4 MP variable focus AI
The current Dahua goto cameras are the 5442 which have a 1/1.8 sensor.
IPC-T5442TM-AS-LED . Review IPC-T5442TM-AS-LED (Full Color, Starlight+) - 4MP starlight
.................... Dahua IPC-T5442TM-AS-LED review
IPC-T5442TM-AS ..... Review-OEM 4mp AI Cam IPC-T5442TM-AS Starlight+ - 4MP starlight+
IPC-HDW5442t-ZE .... Dahua IPC-HDW5442T-ZE 4MP Varifocal Turret - Night Perfomance testing -- variable focus 2.7 mm-12mm 4 MP Starlight
IPC-B5442E-ZE ...... Review - OEM IPC-B5442E-ZE 4MP AI Varifocal Bullet Camera With Starlight+ -- variable 2.7mm-12mm bullet
IPC-B5442E-Z4E .... bullet 8mm-32mm variable focus zoom 4MP
IPC-HFW7442H-Z ..... Review - Dahua IPC-HFW7442H-Z 4MP Ultra AI Varifocal Bullet Camera -- 4 MP variable focus AI
That's the issue. I see 4MP and 2MP sensors that are otherwise exactly the same yet they state the same minimum illumination specs, which doesn't seem possible.
I've noticed the same. I suspect a lack of testing on some models and just using the same specs from a model they have test results for.
sebastiantombs
Known around here
Keep in mind that a 4MP sensor only gets half the light per pixel that the same size 2MP sensor "sees". It's a simple matter of density/size. So if a 4MP camera has the same sensitivity as 2MP the 4MP is working quite well.
I think I'm not communicating my question well. The 4MP sensor is getting less light on each pixel than the 2MP, so the minimum illumination should NOT be the same, yet the spec sheet says it is.
sebastiantombs
Known around here
The reason the spec can be the same is because of the physical size difference. The 4MP is 1/1.8" versus the 2MP at 1/2.8". That doesn't sound or seem like a big difference, but it is.
No, the sensor sizes are the same. For example, the 59232XA-HNR and 59432XA-HNR. Both are 1/2.8" starvis.
| Image Sensor | 1/2.8" STARVIS™ CMOS |
| Max. Resolution | 2560 (H) x 1440 (V) |
| Effective Pixels | 4MP |
| ROM | 4GB |
| RAM | 1GB |
| Electronic Shutter Speed | 1/3–1/30000 s |
| Scanning System | Progressive |
| Min. Illumination | Color: 0.005Lux@F1.3 B/W: 0.0005Lux@F1.3 0Lux (IR light on) |
| Image Sensor | 1/2.8" STARVIS™ CMOS |
| Max. Resolution | 1920 (H) x 1080 (V) |
| Effective Pixels | 2MP |
| ROM | 4GB |
| RAM | 1GB |
| Electronic Shutter Speed | 1/3–1/30000 s |
| Scanning System | Progressive |
| Min. Illumination | Color: 0.005Lux@F1.3 B/W: 0.0005Lux@F1.3 0Lux (IR light on) |
alastairstevenson
Staff member
Another factor apart from sensor size and number of pixels is the sensor technology itself, which is showing steady improvement, and occasionally big leaps, such as back-side illumination (eg Starvis).
Given the model of sensor, the manufacturer's specs could be checked to see how the different sensors standardised performance compares.
For example, if improved semiconductor processing reduces the noise floor, higher-gain amplifiers become practical without as much image degradation, which then increases sensitivity.
alastairstevenson
Staff member
That does not necessarily mean the sensor design, or the process they are manufactured on is the same.
There is considerably more to it than simply the size.
There is a lot of technology in play also.
The 59232XA-HNR and 59432XA-HNR specs are posted above. Do you think they made improvements on the 4MP sensor to match the min illumination specs of 2MP sensor?
chaosengine
Getting the hang of it
no, 1/2.8"@4MP will not work well in low light conditions, consider 5A425XA-HNR. Data sheets not always show reality and sometimes they simply have errors.
Mike A.
Known around here
- May 6, 2017
- 4,199
- 7,066
I think that they're probably just using some base minimum as a number for the spec. I wouldn't over-think it too much.
There's no real objective standard for what's considered "minimum illumination." You'd think that within the same vendor they'd be consistent and you'd see some difference especially when they're giving it to 3 decimal places. But I think it's likely a case of showing precision without that necessarily indicating absolute accuracy.
You could hit up Andy here and see if he can shed some light (no pun intended) on how they do that aspect of their specs.
There's no real objective standard for what's considered "minimum illumination." You'd think that within the same vendor they'd be consistent and you'd see some difference especially when they're giving it to 3 decimal places. But I think it's likely a case of showing precision without that necessarily indicating absolute accuracy.
You could hit up Andy here and see if he can shed some light (no pun intended) on how they do that aspect of their specs.
I suspect error is the case. The min illum numbers are useful to me as I have a number of Dahua cameras so it gives me a rough sense of sensitivity comparison. The spec is also useful when trying to compare Dahua models. I'm guessing the correct min illum specs are for the 2MP sensors as those are usually the first batches on new chips. It seems lazy not to re-test on higher MP releases, and deceiving at worse.
Mike A.
Known around here
- May 6, 2017
- 4,199
- 7,066
Yes, could be that too. Not at all uncommon with spec sheets like that to have cases where they've just copied over from another earlier version.
Even if you have two different numbers I'm not sure that it means a lot. Unless you have you have more details as far as what "minimum illumination" actually means and you know that's applied consistently across models. And you know that's done in a consistent way. You'd hope that's the case but not necessarily so. Then you get to the next level of whether a shown spec of .0003 vs .0005 really makes any practical difference in actual use, When you see order of magnitude differences then, yes, that probably tells you something.
Even if you have two different numbers I'm not sure that it means a lot. Unless you have you have more details as far as what "minimum illumination" actually means and you know that's applied consistently across models. And you know that's done in a consistent way. You'd hope that's the case but not necessarily so. Then you get to the next level of whether a shown spec of .0003 vs .0005 really makes any practical difference in actual use, When you see order of magnitude differences then, yes, that probably tells you something.
You need the spec sheets for the actual sensors to do a comparison. While the camera vendor might say a sensor size is 1/2.8, Sony calls it "Type 1/2.8". For example, the IMX290/291 is a Type 2.8. The spec sheet says the sensor diagonal size is 6.46mm or 0.2543 inches. If my work experience is any guide, Engineering called it 1/4 and Marketing changed it to Type 1/2.8. The IMX178 sensor is a Type 1/1.8. Spec sheet says diagonal size is 8.92mm, or 0.3512". Sounds pretty close to 1/3 to me. The IMX347, also a Type 1/1.8 sensor is 9.04mm diagonal, which is about 1/2.81 .
I don't know how we could get that info except directly from Dahua. That may be true but doesn't explain how min illum numbers could be identical when Dahua calls both image sensors "1/2.8" STARVIS™ CMOS". I highly doubt they would be swapping complete sensor types when the other indicators suggest they are identical other than the MP. Either the sensors themselves are completely different and they happen to have the same min illum numbers, or the min illum numbers are wrong for one of them.