I would just be grateful at this point if the 54Pro could get level with the S3 at night, and during the day for that matter. Surely a lot of this has to be linked to the automation/processing, if so then it really is madness and at times like this frustrating. I can capture a great still image of moving vehicle on the S3 with less IR light, the 54Pro struggles with both freeze frame and general noise/quality as we know already. The techs at Dahua need to do some work on this firmware/camera double time.
Dahua WizColor 5x59-PRO and SmartLight 5x59-IL new series
- Thread starter steve1225
- Start date
There are added features in the PRO, like more LEDs, dual mics, and 12v output to power a camera. Not an exhaustive list, and not of any value to some users. I for instance use external IR and don't care about the white lights, so the extra LEDs are of no value to me.
My conclusion is not just wild guess based on public camera specs, I extracted the firmware and found that they all share the same sensor ID.
For all 8MP models, I know for sure it's SC850SL cause I have one teared down, sensor id is the same with the one in 5x59 series.
For all 4MP models, all models share the same senor ID too, I'm not sure about the specific model, but I do know for sure they use the same sensor, maybe it's SmartSens, maybe OmniVision. I said SC465SL because I saw config for this sensor in the firmware of my Chinese model
There are 10 2MP HWIDs in the firmware, maybe Dahua haven't released the hardware yet, but the firmware is ready. All 2MP and 5MP models share the same sensor ID, meaning Dahua just downscale a 1/2.8" 5MP sensor to 2MP on the 2MP models, pixel size may be 2um only. If released, we should sure avoid this model
That make sense, Pro models just have pro peripherals, others are the same with regular models
THIS is something I've wondered about.
I.e. what matters it the SIZE of the pixels. So if one 1/2.8" sensor has 5MP and another (same size) 1/2.8" sensor has 2MP, then the pixel size of the 2MP sensor is BIGGER and (assuming similar technology/etc.), you'd expect the 2MP to do better (at least per-pixel) than the 5MP in low-light conditions.
HOWEVER, if the SAME 1/2.8" sensor is used in both the 5MP and 2MP (basically just cropping) cameras, then the pixel size is obviously the same ... so no improvement in low-light.
CCTVCam
Known around here
- Sep 25, 2017
- 3,046
- 3,944
I think what you're seeing here is the limitations of the AI. It's probably enhancing the lighting but at the expense of other areas, If you think about when you do manual video adjustment in an editor, there are adjustments you can do eg brightness, contrast etc. However, one you go past a certain point, you introduce other artefacts such as noise into the picture. My guess (and it is) would be that the AI is making lighting adjustments that go beyond this point in order to give better low light vision, then probably filtering some of the induced noise back out. However, with motion it's probably more harder for it to discern what is noise and what isn't because unlike a motionless image, every frame is changing making referencing the norm much harder.
Altrhough quading chips can give gains, the only real answer is size. The larger the chip, the more surface area there is to capture photons. Quading achieves this, but when you look at "proper" low light cameras, they don't use 1/1.8 or even 1/1.2" chips.
Last edited:
I did some measurements of the T54PRO-ZE Field-of-View (and change in aperture/exposure) as focal length changes as seen below - more details in this thread.
