Overview / Testing Video of the New Hikvision ColorVu G2 with Acusense Hikvision DS-2CD2347G2-LU

Good morning,
I am planning to replace my selfmade RaspiCams with the one in this topic (DS-2CD2347G2-LU), until now I am not quite sure with model to choose (2.8,4,6mm) but thats a different topic to solve.
Maybe some of you can give me some hints where to buy the cam, I live in germany and it looks like that this cam is not available somewhere. The other question is regarding the deep of focus, in the datasheet I find a statement that the 2.8mm version has 2.5 m to ∞. Does this mean that if I under 2.5 m away from the cam the picture is not sharp or is it still good also in a shorter distance?

rgs
Jan
 
jasimo said:
Maybe some of you can give me some hints where to buy the cam
Click to expand...
See the start of this post :

Initial review of the DS-2CD2347G2-L(U) ColorVu 2.0 IP camera.

This is an initial review of the interesting Hikvision DS-2CD2347G2-L(U) 4MP IP turret camera, linked to the marketing terms ColorVu 2.0 and Accusense 2.0 The intention is to not overlap or duplicate the various other useful posts on the same device, but rather to try to pick out some different...
ipcamtalk.com ipcamtalk.com
 
  • Like
Reactions: jasimo
coney27 said:
@alastairstevenson Night setting with low illumination scene with no ambient light and no LEDs. Shutter speed is at 1/30th

Click to expand...



TBH, this video is exactly why I wouldn't be running out and ordering one...

In my opinion, it shows the light performance isn't that good despite the sensor size and I believe therefore a lot of the performance claimed in the specs could be down to the addition of the 2 led lights.

Just looking at the video above, the person is ghosted. Not surprising at 1/30 sec. However, for a low light as opposed to no light video, the picture isn't that bright or clear at 1/30. The most worrying aspect is looking at the ghosted person, you realise the amount of noise of noise is horrific which explains the fuzzy look to the rest of the picture where the low contrast makes it less obvious. My point is, if it's got so much noise and electronic gain and is so dim and low brightness at 1/30, what is it going to be like at 1/60 or 1/120 where 1/2th or 1/4 the light is available?

As above, what this is saying to me is the low light performance of this camera isn't all it's made out to be. I'd go so far as to say, looking at what we've seen so far, the Dahua 5442 looks far superior despite the 1/1.8 sensor.

This is one of those reasons why I'm not so keen to see the addition of white light leds to cameras to assist in the performance - in my opinion they distort and hide the true performance of the underlying chipset by making the minimum light levels artificially higher. Fine if you want to run another 8 watts of power per camera, are prepared to accept camera failure due to led failure down the line, and want led's blazing out all around you property. However, if you just want good low light performance, then in my opinion from what I've seen, you're probably better off with a camera that has good raw low light performance and if necessary adding your own sensor light, than having a built in white LED distorting the performance comparison figures and other aspects of camera ownership as listed before.

I'll be pleased to be proven wrong, but I'm just not convinced atm that this offers any performance increase. In fact, it may be a step back.
 
The LEDs were off in that video.
There was no ambient lighting, it was a fully dark scene.

edit
coney27 said:
Yea I couldn't see 5 feet in front of me. Shutter was set at 1/30th. Ill try and see if the low illumination scene improves that at all
Click to expand...
I'm wondering if "low illumination scene" has confused you.
It's a scene choice in the camera configuration, as opposed to a description of the scene itself.
 
Last edited:
alastairstevenson said:
The LEDs were off in that video.
There was no ambient lighting, it was a fully dark scene.

edit

I'm wondering if "low illumination scene" has confused you.
It's a scene choice in the camera configuration, as opposed to a description of the scene itself.
Click to expand...
No confusion on my part. I was referring to the camera setting not the environment setting that the camera was capturing in.
 
  • Like
Reactions: alastairstevenson
CCTVCam said:
TBH, this video is exactly why I wouldn't be running out and ordering one...

In my opinion, it shows the light performance isn't that good despite the sensor size and I believe therefore a lot of the performance claimed in the specs could be down to the addition of the 2 led lights.

Just looking at the video above, the person is ghosted. Not surprising at 1/30 sec. However, for a low light as opposed to no light video, the picture isn't that bright or clear at 1/30. The most worrying aspect is looking at the ghosted person, you realise the amount of noise of noise is horrific which explains the fuzzy look to the rest of the picture where the low contrast makes it less obvious. My point is, if it's got so much noise and electronic gain and is so dim and low brightness at 1/30, what is it going to be like at 1/60 or 1/120 where 1/2th or 1/4 the light is available?

As above, what this is saying to me is the low light performance of this camera isn't all it's made out to be. I'd go so far as to say, looking at what we've seen so far, the Dahua 5442 looks far superior despite the 1/1.8 sensor.

This is one of those reasons why I'm not so keen to see the addition of white light leds to cameras to assist in the performance - in my opinion they distort and hide the true performance of the underlying chipset by making the minimum light levels artificially higher. Fine if you want to run another 8 watts of power per camera, are prepared to accept camera failure due to led failure down the line, and want led's blazing out all around you property. However, if you just want good low light performance, then in my opinion from what I've seen, you're probably better off with a camera that has good raw low light performance and if necessary adding your own sensor light, than having a built in white LED distorting the performance comparison figures and other aspects of camera ownership as listed before.

I'll be pleased to be proven wrong, but I'm just not convinced atm that this offers any performance increase. In fact, it may be a step back.
Click to expand...
For the brief video that you are referring to that scene was totally dark and the only available light was coming from a doorbell which can be seen in the bottom right corner. As for the preformance of the camera with no IRs and no real light source IMO it preforms well. Obviously it could be alot better especially with some built-in IRs or a IR filter but in the end this camera is only taking in the visible light ( which is very little at best) and producing a halfway decent capture which is still useful. Even for the price of the camera im still impressed with its capabilities. Most scenarios where this camera will be used will have some sort of light source either a street light, flood light or parking lot light which will greatly improve the cameras capture. Ideally is this camera the best for a "no light" scenario? No. But thats why there are other options with IR capabilities. This is also why I produced a 20 minute overview video to really show the cameras capabilities in many lighting scenarios.

I am still using this camera to this day in its testing position and I am still very pleased with it. I have 10 more on order which I should hopefully get by the end of this month and will be selling them to some of my friends and clients who have the perfect environment for their use.
 
  • Like
Reactions: CCTVCam and alastairstevenson
coney27 said:
For the brief video that you are referring to that scene was totally dark and the only available light was coming from a doorbell which can be seen in the bottom right corner. As for the preformance of the camera with no IRs and no real light source IMO it preforms well. Obviously it could be alot better especially with some built-in IRs or a IR filter but in the end this camera is only taking in the visible light ( which is very little at best) and producing a halfway decent capture which is still useful. Even for the price of the camera im still impressed with its capabilities. Most scenarios where this camera will be used will have some sort of light source either a street light, flood light or parking lot light which will greatly improve the cameras capture. Ideally is this camera the best for a "no light" scenario? No. But thats why there are other options with IR capabilities. This is also why I produced a 20 minute overview video to really show the cameras capabilities in many lighting scenarios.

I am still using this camera to this day in its testing position and I am still very pleased with it. I have 10 more on order which I should hopefully get by the end of this month and will be selling them to some of my friends and clients who have the perfect environment for their use.
Click to expand...
Whats been your experience with Accusense on this one? I do have a "G1" Colorvue Bulletcam I recently installed but this was an OEM non-labled one from Andy. I can certainly say that while I think the Hardware of the camera itself is OK, the firmware on it leaves a whole lot to be desired. Is the labeled Hikvision firmware better? The OEM vers. seems to be of the "quick and dirty" sort.
 
alastairstevenson said:
The LEDs were off in that video.
There was no ambient lighting, it was a fully dark scene.

edit

I'm wondering if "low illumination scene" has confused you.
It's a scene choice in the camera configuration, as opposed to a description of the scene itself.
Click to expand...


Yes I read the low illumination part and presumed it was dusk. If it's totally black, then that's an impressive performance albeit 1/30 isn't realistic as shown and a higher shutter would result in a darker image and more noise. That said, very few people try and use cameras without at least some supplemental lighting.

Certainly going to be interesting if they can sort the IR washing the face out at close distance issue such as what Dahua have achieved with their 5442. Also, the price is impressive if it is delivering.
 
Last edited:
  • Like
Reactions: alastairstevenson
CCTVCam said:
Yes I read the low illumination part and presumed it was dusk. If it's totally black, then that's an impressive performance albeit 1/30 isn't realistic as shown and a higher shutter would result in a darker image an more noise. That said, very people try and use cameras without at least some supplemental lighting.

Certainly going to be interesting if they can sort the IR washing the face out at close distance issue such as what Dahua have achieved with their 5442. Also, the price is impressive if it is delivering.
Click to expand...
The video was taken just after midnight with no ambient light besides my doorbell. Obviously the lux isn't at 0 but I could not see more than 5ft in front of myself. Ill have to invest in a lux meter for my next camera test
 
  • Like
Reactions: alastairstevenson
Just been looking at Hikvisions Web Site. It seems Darkfighter X is using 2 x 1/1.2" sensors to see at night, one in ir and the other in colour:

Brochure/DarkfighterX.pdf

Looks impressive although would be preferable to see a test on here. Even some of Ring's marketing videos look impressive but I personally wouldn't be rushing out anytime soon!

I also have to wonder about the price of the X variant.....

Also beware, if trying to buy one. Hikvisions website has a video from a 4mp version. Went a UK dealers site on a price search and it seems they are are advertising a wide range of DFX (my abbreviation) models, yet the pdf only lists 2 - a turret and ptz in 2mp. Hmm.. Future launch? Out of date pdf? Mistakes by dealers - one I went to had 45 cameras listed under DFX but the low light specs didn't seem to match the Hivision website!
 
Last edited:
Hikvision has introduced several variants of the Darkfighter X within the last year to two years. Before that there was only one Darkfighter X. They normally employ 2 1/1.8 sensors and is Hikvision's best "night vision" product but that also comes along with a hefty price tag anywhere between $1500 to $2500 US. I'd love to be able to test one out but scoff at the idea of spending that much on a camera that I don't personally need. I believe one user on here actually has one and is mounted on an antenna which is mounted to the roof of their home. Hes posted pics and videos in the past but I forget their username.

Obviously it would be nice to be able to have a Darkfighter X to play with but any Darkfighter, not powered by darkfighter, camera would do a great job for a no light scene.
 
At that price it explains everything!

The stupid aspect is, considering the Colourvu uses a single 1/1.2" sensor and costs $170, it's difficult to see how they justify the cost of $1500+ for essentially the same camera with an extra sensor module added. I'm guessing a 2nd sensor adds between $70-100 to the production cost making the DarkvisionX probably possible for around £220-250. At that price they'd sell tens of thousands each month whereas at $1500+ I bet it's nearer ten or so. I'm guessing they'd make far more profit given the appetite for a large sensor, low light solution. OK lots of guesses, but I can't see the price difference justification in the specs. I could be wrong and something in manufacture makes it expensive, but I don't see it.

BTW has anyone reviewed the 8mp version of the ColourVu as I see in another thread, someone recommending this over the 4mp versions. It only seems to be available as a bullet but pricing seems reasonable at £198 in the UK.
 
Last edited:
You must log in or register to reply here.
Top Bottom