Worlds First Review - Dahua DH-IPC-HDW5849H-ASE-LED / IPC-Color4K-T - 2.8mm Turret

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
1,871
Reaction score
5,082
Location
US
Happy Halloween everyone ! For Audio would recommend the settings I've shared before. Codec = AAC, 16K bitrate, Noise Filter = OFF, Mic / Speaker (depending on your unit) set to 90 / 90 accordingly. You can always lower to 80 or 75 on both but shouldn't need to go below that. I'll do some audio caps from 2.8T, 3.6T & 4K-X (Bullet Cam) 2.8. I've shared the 4K-X before but I'll re-record some samples in a room that I know will push echo. Then attach to YT and let you know when available.
 

CCTVCam

Known around here
Joined
Sep 25, 2017
Messages
1,905
Reaction score
2,380
Yes, I also test the hikvision 4k 1/1.2CMOS turret too DS-2CD2087G2-LU 2.8MM
Right now daytime here, but i also find the dot too when digital zoom it a little. Will post more pics later at night.




View attachment 144065
View attachment 144064

Having seen these now, I don't believe they're hot pixels. Look more like gain artefacts. The fact they disappear with more ambient light suggests very much to me they're gain related. Probably the price to pay for amplifying the light so much to enable this camera to see so well in the dark.
 
Last edited:

wittaj

IPCT Contributor
Joined
Apr 28, 2019
Messages
14,835
Reaction score
27,552
Location
USA
^while the mavic camera is great, have you taken apart a camera to see how small the actual camera is inside? And then the processing and other stuff for the Mavic is in the drone itself and not next to the actual camera and sensor. Plus the zoom is digital not optical.

The difference between a drone camera and a surveillance camera is that the surveillance camera has to operate in a lot more harsh environments and work at night and if you set them side by side 24/7 you will see the limitations.
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
1,871
Reaction score
5,082
Location
US
Having seen these now, I don't believe they're hot pixels. Look more like gain artefacts. The fact they disappear with more ambient light suggests very much to me they're gain related. Probably the price to pay for amplifying the light so much to enable this camera to see so well in the dark.
Yes, agreed. As I mentioned earlier, this is part of the algorithm related processing
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
1,871
Reaction score
5,082
Location
US
As mentioned above, I just ran a super quick audio test against all 3 cams (4K-T 2.8, 4K-T 3.6 & 4K-X 2.8mm). Put a very simple video together showing these. These tests were using the same audio sample played back from an iPhone 12 Pro Max internal speaker at both 20ft & 8ft away from the cameras. The volume was set at 50% then 75%. Interesting results for sure. All camera samples are straight from the cams SD cards, nothing was changed in post. The cut-out/repeat of audio on the 4K-X samples is due to the camera dividing the files (hard to keep movement when just recording audio) and I didn’t want to adjust the settings for this simple test.

All cams were set exactly the same. Namely, per above CODEC = AAC, BITRATE = 16K, NOISE FILTER = OFF, MICROPHONE = 90, SPEAKER (where present) = 90. Double checked the settings on each due to the very different pickup and noise processing across these 3.

My thoughts below

Volume Ratings In Order of Loudest (1 being loudest)

1 - 4K-T 2.8mm​
2 - 4K-T 3.6mm​
3 - 4K-X 2.8mm - Comparing in an audio processor (very quickly) the 4K-X is about approx 6 db quieter than the 4K-T's at the closest 8ft test​

Quality in Order Of Captured Audio (1 being best)

1 - 4K-T 2.8mm​
2 - 4K-X 2.8mm - Although quietest (which plays a role), the audio wasn’t affected by as much digital noise as the 3.6 even when I later bumped the audio in post by +12db (not part of this video)​
3 - 4K-T 3.6mm - Louder but more digital noise and processing so bumped to 3 behind 4K-X on quality​

Background Noise Reduction (Even Without Noise Filter On) By Cam Mic/On-Cam Processing (1 being best separation of background noise to foreground audio)

1 - 4K-T 2.8mm​
2 - 4K-T 3.6mm​
3 - 4K-X 2.8mm - The reason I drop this down here is that when you bump the 4K-X to +12db (which is still quieter than the 4K-T’s) you hear more of the background noise (network equipment rack) that you do on the 4K-T’s​

So in this very quick test, the Wildcat conclusion is that the 4K-T 2.8 offers the best of the bunch overall BUT lets all keep it really by stating, no on cam Mic will ever win any awards for the audio as they are definitely more utilitarian in use and in on-camera processing.

HTH



EDIT - Timestamps added for each camera
 

IAmWatchingYou!

Getting the hang of it
Joined
Sep 13, 2019
Messages
88
Reaction score
34
Location
Raleigh, NY
@garycrist & @Wildcat_1

I have the encode settings at AAC, 16000, and mic volume at 75% and have attached an audio file of the massive amount of noise I hear from the 4k-T I got from Empire Tech (hover over the blank spot below in this post to see the video dialog to play it). I do hear some of the background noise I hear in the video @Wildcat_1 posted, but this is at least 5x as loud.


audiosettings.jpg

View attachment badmic2.mp4
 

Steviedee

n3wb
Joined
Apr 16, 2018
Messages
13
Reaction score
11
Location
Australia
Hey guys, is there any word on a new firmware update to fix the profile schedule/day/night settings issues. Currently the cam is either useless in the night or useless in the day, I cant have it both ways, if i set it up for minimum shutter speed at night, then when daylight comes, the camera whites out/overexposed. And if i set it up for daytime exposure settings, when it gets darker, the image becomes black. So temporarily, i've left it on auto, where i have useable footage during the day, but night footage can get a bit blurry. I've tried multiple factory defaults, used both firefox and explorer, dleted caches from both, and still no luck with having the day/night settings work properly.
 

EMPIRETECANDY

IPCT Vendor
Joined
Nov 8, 2016
Messages
6,925
Reaction score
18,260
Location
HONGKONG
Replying to myself here with an update for all. I've completed thorough testing again across 3 different browsers Safari + Firefox (on Mac) + IE (on PC). I specifically went through each scenario people have reported, noted below for reference:
  • Schedule not changing in Live View mode
  • Settings not 'sticking' after being applied
  • IVS rules being impacted (creation / adjustment)
  • IVS rule resetting Exposure when clicking Apply
  • Play back not working in browser for recorded clips from SD card
In each of these, I can state (and also show in this video) that they are not happening in my testing. The cameras are working correctly and as expected in each of these scenarios in my tests again today. I went back into each camera (the 3x 4K-T's I have) and tested. For reference, the camera that I recorded the footage against for the test in the video is a production (not beta, not pre-production) 2.8mm unit that Andy has been shipping out to all.

With the exception of the Mac based bugs for Safari and Firefox that I filed with Dahua when I first started testing these cams (applies to all 5.0 GUI cams on Mac), no issues are seen. Those bugs for reference are that Image Adjustment (for viewing parameter adjustments on Live Mode such as adjust contrast of the image for Live mode screen only) + on screen graphic functions such as E/PTZ creation (A180 cam) do not work with the embedded plugin and prompt you to download a Windows plugin ! This bug for Mac will stop IVS and E-PTZ graphics even being saved (you won’t even have an option to apply them as it will error out) BUT again this is ONLY for Safari & Firefox on Mac. This is NOT an issue for Firefox on PC as well as NOT an issue for IE as I show. So with that stated for continued reference, as you'll see on the video (11 mins but wanted to be complete), all the above scenarios tested fine.

Video being uploaded now and will update as soon as available thanks to the usual YouTube re-encode dance.

Almost forgot, didn’t show in this video as it was already 11 minutes in length BUT I did also test on Chrome on PC (v106.0.5249.119) as well as Firefox on PC (v102.3.0) and both of these are also working fine too in each of the 4 areas above. If people do want to see this as well, I can always capture more footage.

HTH show what I’m seeing across 1 x pre-production and 2 x production line 4K-T’s and critically that these issues are not present in this testing.

UPDATE - YouTube finally is ready in 1080p HD (make sure you manually select), still awaiting the YouTube 4K re-transcode (rant: wouldn't mind but the 4K file I upload already meets their standard requirements for encoding !, rant over)

Here is the video of the testing and the results

Additional Notes

  • Timestamps added for each browser test + intro at start mentions versions used in the test
  • At 07:47 of the video I mention setting night to 12:18:01 but my note I add shows 12:12:01, that was a Wildcat slip-up in the note, the actual time set as you can see on video is 12:18:01. You’ll also see the Windows Not Genuine popup when I have a fully licensed machine, gotta love Microsoft and one of many reasons why I moved to Mac :)
  • Also while I show it on screen in the Safari on Mac test, I didn’t call out specifically (so doing it here) that exposure settings etc never changed when applying IVS rules. Just calling it out more clearly here so people saw that it wasn’t an issue in this test either.
@Steviedee here has some setting guide by @Wildcat_1 you make some time to watch it and try some testing and check if working on it.
 

EMPIRETECANDY

IPCT Vendor
Joined
Nov 8, 2016
Messages
6,925
Reaction score
18,260
Location
HONGKONG
Last edited:

IAmWatchingYou!

Getting the hang of it
Joined
Sep 13, 2019
Messages
88
Reaction score
34
Location
Raleigh, NY
After testing today, I can confirm that the mic noise varies from power source - the camera produces different volumes of digital noise with different PoE switches and power supplies. So, it seems like Dahua didn't properly isolate the microphone circuitry, or used bad components to do so, in at least my camera.
 
Last edited:

IAmWatchingYou!

Getting the hang of it
Joined
Sep 13, 2019
Messages
88
Reaction score
34
Location
Raleigh, NY
The noise that was in the audio captures I posted was with the camera connected to a new PoE+ switch. Surprisingly with the power supplies the noise was about half as loud.
 

wittaj

IPCT Contributor
Joined
Apr 28, 2019
Messages
14,835
Reaction score
27,552
Location
USA
I thought I would give an update after playing with this for several weeks.

While we still have some quirky firmware issues to resolve, we do have some work-around solutions in the meantime. But after we initially set it up and dial it in to our field of view, we are rarely ever in it again, so I was never overly concerned about the quirks.

I find the image to be one of the better ones out there. This and the 4K/X bullet are the first cameras that I feel comfortable going with a little bit of digital zoom. Now of course many will digital zoom other cameras and they feel the images are ok, but I prefer the optical zoom. You can't get carried away with it LOL, but I feel it does a decent job.

The 4K/X and 4K/T are the first cameras I have had that I don't have to be pushing the limits of the camera parameters to get color at night in order to get a good image. We don't have street lights and most neighbors do not turn their outside lights on, so I am dealing with some challenging light conditions.

I will explain below, but for many of my other cameras that I have in color, I am running high gain, gamma, etc. and in some cases at around 70 just to be able to get color, but that does introduce artifacts and other issues. With the 4K/X and 4K/T I can run faster shutters and keep the parameters under 50, which helps tremendously.

Now with this being a 4K full-time color camera, it draws a different buyer to this prosumer market that we play in. These people are not aware of focal lengths and the value of varifocal and believe that the wide angle view allows them to make people and cars near and far. These people are chasing MP and full color. In short, they think these cameras are magic and can be the be all, see all, and that just doesn't exist. We have many posts within this thread and in other posts trying to educate the NOOB on the nuances. A great camera put in the wrong field of view or trying to do too much with one camera will result in poor performance, and we have seen that already with posts about the 4K/X or 4K/T. Like every camera, you need to select the correct camera for your needs and field of view.

So to build upon the thread regarding chasing focal length over MP, I wanted to take a moment to show some comparisons of the camera at a distance compared to their 4MP and 2MP equivalents, so I re-positioned several cameras to capture the same target at various distances.

To help guide the eye to the target (white car), I have blacked out all but the car moving at 25MPH out at a distance.

First up is the 4231 2.8mm 2MP Boobie cam. The white car is 100 feet away from the camera. So well beyond IDENTIFY distance.

I use this as an overview to "see the whole neighborhood" and not for IDENTIFY. I have to run a 1/30 shutter and high gain and gamma to keep it in color, but since it is for OBSERVATION and not IDENTIFY, it works and surprisingly with not a lot of blur/ghost. We can tell it is a white car and someone that knows cars may be able to identify it, but one shouldn't count on it. A lot of headlight bloom due to the high gain and brightness.

white car boobie cam black.jpg


Next up is the 5421-Z12E varifocal camera set up at around 50mm. The white car is at 60 feet away at this location. We start to get to IDENTIFY at that combination of distance and focal length if the camera parameters are set-up correctly (shutter and other parameters).

Similar to above, I have to run the camera at 1/50 shutter and at much higher gain and gamma and brightness in order to make anything out in color. As such there will be some blur for a moving car.

white car 5241 black.jpg


Next up is the 5442-ZE varifocal camera set up at 3.6mm focal length. The white car is 55 feet away at this location, so about double the IDENTIFY distance of 26 feet.

Similar to the other cameras so far, I have to run the camera at 1/60 shutter and at much higher gain and gamma and brightness in order to make anything out in color, so it does introduce some blur with a car in motion.


white car 5442 36 black.jpg


Next up is the 4K/T camera with a 3.6mm focal length. The white car is at 110 feet away at this location, so well beyond the IDENTIFY distance of 24 feet.

I am able to run this at a much faster 1/120 shutter speed and with all the parameters below 50. And the image is much better than the 2MP above at comparable distance.

white car4k black.jpg


Now unfortunately, this camera is double the distance away from the other camera samples, so obviously the car is going to be smaller. Now you may not be able to tell here, but on my system, you can tell that even though this is more in the OBSERVE range, it captures a lot of detail. It has the least amount of blur (due to the larger sensor and faster shutter) and less headlight bloom due to not having to force the camera in color and running parameters like gain and gamma too hot. I am sure if the camera was closer at about the 40-50 foot range then the results would be even better.


Based on my experiences, for anyone with any light available (street lights or outside lights on) and is looking for a great OVERVIEW color camera that could be placed on a 2nd story to OBSERVE what is going on, I think the additional $70 to go up from the 5442 to the 4K/T is worth it.

I also think it is worth it if you have a location that is around the 6-7 feet high and you want to IDENTIFY at 15-20 feet out and have some light, then the 4K-T is a great camera that the increased resolution can provide more detail for a digital zoom in certain situations. I know some have expressed a concern about near and far distance focus, but I do not see that with my camera, especially once I dialed in the settings.

And then of course, if you do not have ambient light and do not want to use the built-in white LED, then you are better off with a camera with infrared capabilities.


Now I would be remiss LOL if I don't show what one should use if they wanted to IDENTIFY a car at the street from a distance (hint it isn't a 2.8 or 3.6mm fixed cam beyond 25 feet).

So this is the 5442-Z4E varifocal camera set to around 28mm to capture the car at 75 feet. Similar to the other cams, I have to run a 1/90 shutter and much higher gain and gamma and brightness, etc. than we typically like to see, but I spent a lot of time tweaking it to get the best picture possible. Like literally if I up the gain one number, it will be a ghost mess. The slower shutter results in some blur, but I am willing to accept it to get color.

white car 5442z4E black.jpg
 

samplenhold

Known around here
Joined
Aug 8, 2018
Messages
5,399
Reaction score
17,527
Location
Spring, Texas
^ Damit @wittaj now you got me thinking of using this cam as a replacement for my SD12203T-GN that are used as overview cams in second story windows. I was all set to really consider the new mini-PTZ, but was not happy with 4MP in a 1/2.8" since the SD12203T-GN is a 2MP on 1/2.8" sensor. Was thinking of the mini-PTZ since it has way more zoom. But how often do I REALLY use the zoom? Both these high-mounted overview cams have been important in their normal 2.7mm settings.

Ahhhhrrrrrggg! Now I really need to think about this and define/redefine the use case for that position.

On the other hand, I COULD get both and mount them both in the same window. That way I get the great low light color performance of the 1/1.2" sensor AND the really long reach of the mini-PTZ! I mean I have room on the board in each window!

Rule #1 - Cams multiply like rabbits.

But that means I need another connection in those rooms. Or buy two more POE switches and two more UPS for them. Maybe get the solar panels on the roof then I don't need the UPS for each cam. Lets see, UPS for each room, that's about $200. Whole house solar with two Tesla battery banks, about $30k? I have to think about this. Hmmmmmmmmmmm.
 
Top