Worlds First Review - Dahua DH-IPC-HDW5842T-ZE-S2 / IPC-T5842T-ZE - 8MP / 4K Vari-Focal Turret

amrogers3

Pulling my weight
Joined
Sep 1, 2021
Messages
176
Reaction score
109
Location
Houston
Thank you for the clarification. I don't the "s" or "2" defined in the Dahua wiki. What does the S2 mean?
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,047
Reaction score
5,864
Location
US
Thank you for the clarification. I don't the "s" or "2" defined in the Dahua wiki. What does the S2 mean?
In terms of Dahua and how it uses S2. Usually an iterative release of a model and often times comes where internals have been updated no matter how small, such as changes to memory provider, SOC tweaks and other internals. Sometime can lead to larger user facing changes but generally not as that is handled by newer models etc.

In some cases you’ll find an S2 model number for a cam without a previous (non S2) model release due to changes between test and production units before final, readily available release. You can also see this if part availability slips for initial builds and therefore S2 version is released with more readily available parts and amended internals over initial planned units.

HTH
 

mecham56

n3wb
Joined
Feb 3, 2021
Messages
19
Reaction score
15
Location
USA
So, would this be a good option instead of getting various 5442 focal lengths? Just out of pure simplicity, I'd rather get a bunch of the same camera than getting a handful of different ones. I'm anxious to see how well this handles at night.
 

ljw2k

Known around here
Joined
Jun 9, 2014
Messages
1,486
Reaction score
2,260
Location
United Kingdom
I will stick to what Wildcat always states in many threads that a 8mp with a 1/1.8” is not going to work well at night and the Max he recommends is 4mp with the 1/1.8" Sensor like the 5442. Daytime any cheap 8mp look good even Reolink :)

5842 Wins hands down in Daytime.
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,047
Reaction score
5,864
Location
US
As ljw2k mentions, I'm not an advocate for just increasing pixel density on a sensor to increase resolution overall, that approach generally leads to failed expectations and poor implementation of a cam product noticed mainly at night or darker FOV's. Daytime is easy to get right (within reason) ;)

Now just one thing I want to re-state though is I've mentioned, that the max I recommend ('pushing it' that is) is 4K on a 1/1.8", never want to see MP's higher than that AND you must be ready for potential impact to noise etc of doing that IF not implemented well by the manufacturer. This means, don't assume all 4K/8MP 1/1.8" cams (or any cam you are comparing on MP & sensor) are the same, they're not and can perform very differently. This depends a lot of course on the SOC/Sensor/Optic pairing along with improvements to underlying code etc too which thankfully are continually improving.

I did a couple of posts on this, 1 is HERE about why you shouldn't chase megapixels and in the case of the 5A (PTZ review in the link I share) mentioned that with that implementation, 4MP was the limit. However as I also mentioned in that thread (quote myself) and emphasized appropriately here '....the lower the ratio of MP’s to appropriate sensor size will result in a better overall image and certainly result (for the most part) in higher quality, more usable, less noise in night time work. An example of a bad pairing would be the 4K cams we saw early on with 8MP crammed onto a 1/2.8” sensor. The 1/1.8” versions improved that but I can tell you from experience that for 8MP you need a larger sensor still to truly drive great quality across conditions. At a push though 4K/8MP on a 1/1.8” if you are looking at current availability but no smaller.'

The above was in May 2020 for context and I'm pleased to state that we are now starting to see improvements albeit there will always be a theoretical max or ceiling that you will never break through for performance :) Now in the case of this camera, the 8MP version definitely shows some improvements in its implementation for sure. To be honest I'm actually pretty impressed with the results but will save further comments and overall thoughts for my wrap up.

Either way, some key pointers:
  • Don't chase megapixels
  • Be Cautious of pixel density
  • Don't assume that every camera listed with same MP & sensor size is the same or implemented the same way whether across same or competing manufacturers
  • Beyond 4K definitely need to see manufacturers move to sensors larger than 1/1.8" and to be fair for improved 4K you want to start keeping an eye on that as you can only pull so much out or mitigate any issues generated from the overall limitations of Sensor/Optic/Soc pairings

HTH
 

ljw2k

Known around here
Joined
Jun 9, 2014
Messages
1,486
Reaction score
2,260
Location
United Kingdom
1/1.8" 8mp with a F1.4 lens, just a mix/match of old tech in different combinations on a so called new camera. They should have done the 1/1.2” STARVIS™ CMOS with the F1:0 lens with Vari -Focal as that combination would have been a much better camera that people would acually want.
 

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,047
Reaction score
5,864
Location
US
Here is the night time video as promised along with comparisons to 5442. A longer 11 minute video but night is a key area to test any cam and as always I include timestamps for your viewing pleasure :) Couple of caveats:

  1. As I mention in the video, this FW does NOT include SmartIR adjustments I’ve had added by Dahua to 5442’s (for those that have followed the journey in getting that in place :)) therefore would wash out the image. I’ve reported this back to Dahua for inclusion in next test FW (and will note on bug list). Due to this I manually dial down Exp. Comp but this can crush details in the blacks so just be aware there is more opportunity to adjust this and bring details back manually for now BUT in time will be addressed by the FW
  2. YouTube HATES side by side video sources from a re-encode perspective and truly screws what is ultimately good video being uploaded. Therefore make sure you take time to look at the full screen 5842 to understand (and reach your own conclusion) on quality of image. The side by sides will appear softer than full screen due to this PITA YouTube re-encode. This is even more noticeable on the color at night section I include where in some places you see smear on the side by side
  3. There is a rolling shutter issue on higher shutter speeds (1/250 its evident more) at night (in IR mode) which also needs to be corrected in next FW (reported and will add to bug list too)
  4. I include color at night in the test as I know people are interested in this. I don't advocate for it without appropriate additional lighting but again wanted to make sure this section was included for you

Couple of quick observations from me on this cam after testing at night:
  • I include distance markers and you can see that in IR only that around 22 feet is where the 5842 starts to clean up nicely from a detail perspective
  • In comparison to other 8MP/4K implementations on a 1/1.8" sensor, there are improvements in keeping noise down on this implementation for sure. Even when running a NR of only 36 (relatively low for night) it handles it well
  • There will always be a sweet spot for these cams and you should/will expect fall off outside of that in detail and especially over distance. This will always be more evident at night and especially if IR is your only source. This 5842 is no different from any other cam in that respect and as mentioned above, sweet spot for IR only FOV’s is around 22 feet (in this implementation with current FW) BUT add light to the FOV (as you should and I advocate for constantly) and you’ll adjust the sweet spot and improve the overall resulting image. If you look at timestamp 4:56 in IR with just 62.5w equivalent light added you’ll hopefully see how this cam improves further. In real terms it added approx 10 feet to the distance (therefore 30-32 feet total reach) at which to expect detail and allows 1/100 to become the sweet spot for this amount of light on the FOV. Increase the light further (check out the 100% 250w test) and you can push (while keeping in IR mode) to approx 40 feet but that should be your furthest expected limit for this cam in my testing (with decent config rather than using Auto which I never recommend) at night. Not to say you cannot go further out (or wider coverage) by adding in additional lighting in front and off center but 40 feet is where to gear your expectations on this one.

Nighttime Test Link - Remember to Watch on YouTube in Full 4K Quality

HTH, enjoy !
 

EMPIRETECANDY

IPCT Vendor
Joined
Nov 8, 2016
Messages
8,259
Reaction score
23,719
Location
HONGKONG
Everyone want to get a 4K with crisp pics day and night , but the price on the 1/1.2cmos will be extremely high on a motorized lens. And also need a bit good requirement on you computer, network and NVR.
This 4K will be a good solution as a traditional IR cam. The Color4k-x will be a good solution for low light full color. 4mp 5442 series will be our affordable series.
 
Last edited:

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,047
Reaction score
5,864
Location
US
As I was writing this I thought, Wildcat, why not use this as part of your wrap up (may still do) but thought it was a good point in the conversation here to layout some thoughts on the state of this cam and its place in the Dahua portfolio along with other recent reviews.
The benefit that Dahua is bringing to the portfolio is that there are now good choices of camera for each environment that complement your overall setup and most importantly each other and come in at a range of budget options.

For example:
  • Dual Lens (Fixed) - Great for traditionally dark areas, benefits from a small amount of light, allows for color capture where possible and fuzed image ensures you don’t have to choose Color vs B&W up front. Great for music videos too ;)
  • Color4K-X (Fixed) - Extremely good at what it does, 4K in full color all the time with a larger sensor that benefits from extra light which also does not attract bugs !
  • 5842 (this cam) (Vari) - Benefits those with little to no light that want 4K quality but have to rely on IR or small additional light (will awakes benefit from a bulb or 2), don’t want a ‘fuzed image’ as above but want low noise, usable caps
  • 5442 (Vari) - A leader in its own right, well tuned at the 4MP/1.1/8” pairing and offering a cheaper option than the cams above with similar great quality and traditional IR

Therefore, with the above cams + any speciality application requirements (LPR/ANPR, PTZ, Face Recognition, Access Control etc) you have some great options available to you that suit most deployments. This also gives you the ability that I call out many times and that is to have a staggered setup of B&W vs Color cams covering a given area and gaining you both color (lost in areas only covered by IR) + details (lost in dark areas when capturing in color) which increases your chances of critical ID. It also increases your ability for good PTZ activation (if you’ve seen my posts on PTZ call to preset in single, triangulated or quad rollouts) as you’re improving quality of target acquisition by your non PTZ cam. All of this with a choice or blend of Turret vs Bullet and Fixed vs Vari cams to fit your needs.

Could you have done this before with older cams, absolutely but with a lot of compromises (resolution, noise, image quality). Now you have the ability to do it with a base of 4MP and up to 4K (now with very usable quality based on install, FOV, lighting etc) and that takes this up a level IMO. This is great to see that we are moving towards a more consistent (HW wise but FW needs work ;) ) approach to cams in series and with less compromise as an end user or installer.

Just my 2c
 

enilm

n3wb
Joined
Oct 31, 2016
Messages
25
Reaction score
9
I ordered two of these bad boys to replace my 5231 that were I think 6 years old. So far I'm liking them.

I've been playing around with the settings for the past few days. I think I've got the Daytime dialed in with some great results. Still working on the Nighttime settings.

@Wildcat_1 I've noticed that the cameras are having a little hard time staying between day/night mode in the transition times. I noticed this morning I had a bunch of alerts at sunrise. I've just noticed now that with the sunset also It's going back and forth between day & night mode. I guess it's having a hard time deciding whether to stay in color or B/W LOL

I've only noticed it after messing around with the settings. It wasn't doing this while on Default settings.
 
Last edited:

Wildcat_1

Known around here
Joined
Dec 9, 2018
Messages
2,047
Reaction score
5,864
Location
US
I ordered two of these bad boys to replace my 5231 that were I think 6 years old. So far I'm liking them.

I've been playing around with the settings for the past few days. I think I've got the Daytime dialed in with some great results. Still working on the Nighttime settings.

@Wildcat_1 I've noticed that the cameras are having a little hard time staying between day/night mode in the transition times. I noticed this morning I had a bunch of alerts at sunrise. I've just noticed now that with the sunset also It's going back and forth between day & night mode. I guess it's having a hard time deciding whether to stay in color or B/W LOL

I've only noticed it after messing around with the settings. It wasn't doing this while on Default settings.
I generally recommend people either run the sunset utility often mentioned here OR personally I setup the individual day/night profiles schedules and do it that way rather than letting the cam decide. If you need any assistance dialing in just DM me, happy to help.
 
Top