Dahua IPC-HDW5442T-ZE 4MP Varifocal Turret - Night Perfomance testing

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
@kbgator I would not recommend this for LPR unless you have the cam within 10' of a target zone. For LPR you need a cam with a longer lens, Z4 or Z12 variants depending on your install location. HTH

Good points!
BTW, could you, if you tried, use this weakly for LPR a little?
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
@ArnonZ
If you take the time to setup for FR with the right cam(s) you can get some amazing results. This does rely on the quality of your database (imported or created over time etc) however. Take a look at the 4K PTZ review I did where I show this in cam and NVR.

thanks m00st.
I am aware of those: facial recognition is good enough => good identification, minimal false positive.
the false negatives are due to a lack of identification that could be a result of: saturation/low light/wrong angle/cover/shade and a lot more. cannot identify a know person good enough => false positive.
however that, most systems can give you info about the % of fit to the known picture and you might even get the tools to play with the threshold. but that won't do much good cause it all starts with an image that is not good enough for identifying.

also, I did learn that the NVR I'm talking about can have 4 face identification or 16 with cam that supports FD. that's just why I was choosing that set.
I do expect relatively high lever of false negatives but if I'd use it for the night time, when I go to sleep and I not not expect any guests, I expect it to work as an alarm (maybe I aim to high, I know.... it's a test).

one thing I expect to work good in here is the human detection. while identifying a person is harder and requires shorter distance, I do expect the SMD to be reliable at night time, from a system that is capable of face recognition.
also, I expect the setup to act as tripwire SMD if the known person is not clear enough - set off for any human rather than skipping the identified ones.
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
@samplenhold thanks for the kind words. Think you meant my PTZ reviews and if thats the case then yes those cams have come on a long way including 2 of my faves, 49225 and 5A425 that have the advantage of being smaller, quieter and therefore less obtrusive than most. With that said they (PTZs) are certainly not for all installs or locations as you will notice I've stated in many posts and have to be configured correctly. The B5442 that you mentioned, I also reviewed and really liked that camera. @Parley I believe has the Z4 variant and loved that one too. If you need any assistance just reach out.

Thanks for the reply. I currently have ten outdoor cams with two more to be installed and two LPR cams to be installed just as soon as I can figure out how to get the Cat5e from the attic to the front porch. I have a mix of crossing cams on the driveway, wide angle view cams covering the front and side yards and four cams at the front door. I agree with you on having a mix of cams and fields of view. I only have one cam that goes to IR at night outside.

The reason I am so interested in your review is that I am considering yet another cam that will augment the two LPR cams at night. Specifically zoomed in to the T-intersection for the sole purpose of getting color and make/model of vehicles at night that my other overview cams can't get a good enough capture for that purpose.

We had an incident a week ago where some guy in a pick up pulled into the driveway of a neighbor and stole his tailgate. It took about 5 minutes at about 3:00am. I have the whole thing on one of my overview cams. But since this was so far away, the best I could do was to get a color and a few observations on the guy's pickup, but no make/model. And those observations (like no front plate) were mostly from other cams that picked him up as he drove by my house. I don't have the LPR cams installed yet. I felt really bad that I could not give him more info and the police asked me for footage since I have given them good footage in the past, one that allowed them to ID a perp and charge him.

I recently swapped out a 5231 varifocal turret that I had in my driveway for the T5442 in 6mm. I did a test using that varifocal the other night on the T-intersection and was not happy with the result. I also tested a
T5442 in 2.8mm that I have yet to install in it's final position. While the 2.8mm FOV was obviously not the right FOV for the task, the color recording was much better than the 5231 varifocal. So now I am leaning towards getting the B5442 Z4E for that position. That does mean I now have to run five wires not four from the attic to the porch!

I like your reviews, especially the POE ones. I almost pulled the trigger on one of your POEs for the above job, but decided that maybe the B5442 Z4E would be less creepy looking from folks walking by the house on the corner.
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
@samplenhold / All

As promised, here is the 2nd set of night testing.

This was taken with the newly released FW (6-20) which as I mentioned above would not feature any type of IR fix as it was just reported. I was going to lower the camera to the more optimal position that I would normally install at BUT for consistency sake (anyone comparing videos 1 & 2) thought I would leave in same place for now and can adjust later. I did however use a step to bring me closer to cam to show you the degree of washout this time on a face. Unfortunately I kept kicking the step in the near darkness so position changed a bit but IR washout remained the same (as expected). Therefore I also took the opportunity to shoot scenes at the end using the tips for dialing in that I mentioned in video 1 (same settings) so you can continue to see the difference that these can make while we await a FW update.

Captured 1/60 and 1/100 this time with IR only and IR + bulb. No flood lights or forced color this time as wanted to show any differences in FW relating to IR and the IR effect when at wide vs full zoom.

One thing I did not (and report back) about this latest FW is the presence of slightly more macroblocking than before. For those that aren't aware, macroblocking is a by-product of compression. Seems just slightly more pronounced in this FW on edges etc, not a huge deal but reported it anyway.

Shot again in native and kept in ProRes but unfortunately YouTube likes to do its thing with re-encoding. Make sure to select 1440 (when YouTube has finished processing) and can select 1080p as of typing this.

HTH, let me know with any questions

 
Joined
Aug 8, 2018
Messages
7,386
Reaction score
25,889
Location
Spring, Texas
OK, so zooming the lens and/or increasing shutter speed lessens the IR face washout.

So now I am little confused about IR. I have not used it much. But from reading this forum, I always thought if the cam was in IR mode, visible light did not affect it. But when you add the 60W bulb, one can now see the #30 on your shirt?
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
Increasing shutter speed absolutely will impact overall exposure (look at the 1/120 vs 1/60 in original video) and therefore yes will also impact washout as a by product of that. However the big difference comes when you adjust as I show at the end and balance the scene using HLC in conjunction with reducing IR light to compensate from the very strong IR currently on this cam.

You mention zooming, remember as I mentioned above that turrets don’t feature Zoom Priority IR therefore don’t increase / decrease in an attempt to counter for focal length like bullet cams do. So in the case of the turret it attempts to detect objects in scene closer or further away and should adjust accordingly which as we know this turret under current FW has issues with. However again zooming to 12mm won’t in itself increase the IR by focal length alone if the cam doesn’t detect it needs it.

On the last point if you are talking about the caps taken with IR=Auto then yes I would expect what you see in the video as the cam adjusts its IR again based on need within the scene. With the 60w bulb the cam decided it had enough visible light to adjust down the IR light (albeit not with super close items as demonstrated). As it does that it doesn’t washout the white areas of the shirt as much therefore more detail and texture is seen the more visible light the camera can use. This is why it’s important to balance and dial in your scene as I mentioned so you get great end results from the caps.

Last thing to mention is that if you were in IR mode only without any available light then you would expect what you see in the start of the video albeit that if the Auto IR was working 100% as expected then it should dial down dramatically as an object gets closer to it therefore decreasing IR accordingly.

Hope that helps


OK, so zooming the lens and/or increasing shutter speed lessens the IR face washout.

So now I am little confused about IR. I have not used it much. But from reading this forum, I always thought if the cam was in IR mode, visible light did not affect it. But when you add the 60W bulb, one can now see the #30 on your shirt?
 

ArnonZ

Pulling my weight
Joined
Feb 11, 2020
Messages
212
Reaction score
128
Location
here
Last thing to mention is that if you were in IR mode only without any available light then you would expect what you see in the start of the video albeit that if the Auto IR was working 100% as expected then it should dial down dramatically as an object gets closer to it therefore decreasing IR accordingly.
At last I understand why you expect a new F/W to fix the IR problem.
Thank you.
 

Rhodesy

Getting the hang of it
Joined
Mar 8, 2018
Messages
175
Reaction score
36
Thanks for the kind words. I meant 6 inches as I said it should be 6” lower than I had it ;)

Agree with you on the mount and have a selection of PFA130’s and if you want ultimate flexibility then I also use the Dahua PFA121 junction boxes that work with the Dahua bullet cams but the cool part is that you can buy in the 4 screw mount version (bullet cams) and also a 3 screw version (PFA122) which works with the turrets. Keep a couple of each spare then as you want to move cams just mount your new cam to your spare faceplate and switch one cam/plate out for the other on the junction box. Works really well in switching cams without having to remount junction boxes :)
Could I please double check, would the PFA122 be suitable to attach to a pole mount PFA152 to then attach the IPC-T5442T-ZE so two mounts and the camera is mounted vertically?
Would that risk water ingress?
 

EMPIRETECANDY

IPCT Vendor
Joined
Nov 8, 2016
Messages
8,259
Reaction score
23,717
Location
HONGKONG
Just get this firmware from dahua.
Australia guy can ask me for this firmware to test the IR, because you guys at nighttime now, lol.
IPC-HX5XXX-Volt_Eng_NTSC_Stream3_V2.800.15OG000.0.T.200721
 

biggen

Known around here
Joined
May 6, 2018
Messages
2,539
Reaction score
2,765
Just get this firmware from dahua.
Australia guy can ask me for this firmware to test the IR, because you guys at nighttime now, lol.
IPC-HX5XXX-Volt_Eng_NTSC_Stream3_V2.800.15OG000.0.T.200721
This is supposed to fix IR? Will that work for the non-varifocal 5442 turrets?
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
All, I wanted to test the latest FW (7/21) which is the first one that is aimed at resolving the IR washout seen on this cam. The good news as I state in the video is that progress is being made and you can see this in the clips I took comparing both FW’s (original and test). You notice this more in the 1/60 but still, adjustments have clearly been applied to this FW.

With that said, the IR is still too strong in the center of the image. Near IR still needs to be lowered down further and the response of the AUTO IR (Smart IR) needs to be quicker. I’ve flagged this with @EMPIRETECANDY & Dahua that changes still need to be made to these specific areas and am sure these will be addressed soon.


** SPECIAL NOTE - On the videos, note that when you step up to 1/100 you expect the darkening of the scene so this is not indicative wholly of IR adjustments. Also, in full zoom, while IR strength is less than before in this FW please note that on the distanced captures you are also seeing the natural IR falloff as well **

Either way, still a great cam with great image and if you use the workaround settings I mentioned in my other videos you can easily counter the IR washout while we await the next FW to try.

Thanks to @EMPIRETECANDY & Dahua for their quick attention and continued focus on this. I will report back when I test the next version

Video Link
Uploaded in Native 2560 x 1440 then split screened to show both FW's. Make sure you select 1440 once YouTube finishes its processing dance, if not then choose 1080p




For reference, in this test FW there is a Smart IR radio button added in the exposure menu, while the AUTO option remains in the illuminator menu as shown below

Smart IR Radio Button In Exposure Menu.jpg Auto IR Still Present In Illuminator Menu.jpg
 
Last edited:

David L

IPCT Contributor
Joined
Aug 2, 2019
Messages
7,932
Reaction score
20,757
Location
USA
All, I wanted to test the latest FW (7/21) which is the first one that is aimed at resolving the IR washout seen on this cam. The good news as I state in the video is that progress is being made and you can see this in the clips I took comparing both FW’s (original and test). You notice this more in the 1/60 but still, adjustments have clearly been applied to this FW.

With that said, the IR is still too strong in the center of the image. Near IR still needs to be lowered down further and the response of the AUTO IR (Smart IR) needs to be quicker. I’ve flagged this with @EMPIRETECANDY & Dahua that changes still need to be made to these specific areas and am sure these will be addressed soon.


** SPECIAL NOTE - On the videos, note that when you step up to 1/100 you expect the darkening of the scene so this is not indicative wholly of IR adjustments. Also, in full zoom, while IR strength is less than before in this FW please note that on the distanced captures you are also seeing the natural IR falloff as well **

Either way, still a great cam with great image and if you use the workaround settings I mentioned in my other videos you can easily counter the IR washout while we await the next FW to try.

Thanks to @EMPIRETECANDY & Dahua for their quick attention and continued focus on this. I will report back when I test the next version

Video Link
Uploaded in Native 2560 x 1440 then split screened to show both FW's. Make sure you select 1440 once YouTube finishes its processing dance, if not then choose 1080p




For reference, in this test FW there is a Smart IR radio button added in the exposure menu, while the AUTO option remains in the illuminator menu as shown below

View attachment 66979 View attachment 66980
Great improvement in new FW. Seems like the 60w bulb adds to the washout. Probably best to have total darkness with IR, right?
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
It does add to the washout but with balanced IR you should see this even out the close object as it dials back IR due to ambient in Auto/Smart IR. Therefore the final iteration of this FW should accommodate for both. HTH

Great improvement in new FW. Seems like the 60w bulb adds to the washout. Probably best to have total darkness with IR, right?
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,046
Reaction score
5,861
Location
US
All

Quick update.

I concluded about 2 hours of capture testing tonight using the latest test FW (7/31). Unfortunately the IR on this is still too strong center of image on SmartIR/Auto IR even when using Auto Iris. I have passed the feedback and my suggested tweaks to IR to @EMPIRETECANDY and Dahua who continue to work on resolving this FW issue, THANK YOU !

I did want to share my results at 2 main exposures to help others if they want to dial in using Manual IR only (I show how to do this with IR, HLC and other combinations in my original night videos for reference but wanted to focus on IR only adjustments on this FW test here).

I did some testing with Manual IR so I could let you know what the IR strength needs to be at center of image with targets as close as 6ft to 8ft away. All testing was completed at full wide because at full zoom you are just benefiting from IR drop-off and distance of target to camera. Please bear in mind that as always your location may need little more tweaking but this should get you close. Here are those results:

  • When running exposure 1/60 and using 1 x 60w bulb to light the scene then the IR needs to be dialed down to strength 16 at 6-8ft
  • If you run this same exposure (1/60) with IR only then you can run between 26 and 40 maximum for IR strength

  • If you move to an exposure of 1/100 with IR and using 1 x 60w bulb to light the scene then you can run up to a max of 25 strength
  • If you run this same exposure (1/100) with IR only then you can increase to a maximum of 40 for IR strength. You could push this to 50 but then start to see a little too much washout for my liking at this distance
HTH. Will continue to test and report back
 
Top