Dahua WizColor 5x59-PRO and SmartLight 5x59-IL new series

wittaj said:
It's posts like this we need to remind the NOOB of chasing MP - most of us with 4K cameras have ended up swapping out for 4MP as that is the current sweet spot!
Click to expand...

Yeah but I don't think it's down to sensor size or should that be pixel size on the sensor as the larger sensor on the 4K colour series largely maintains the pixel size given it's 8mp on 1.2 vs 4mp on 1.8. I think the issue is rather more with the relationship between sensor physical size and dof. There appears to be complex arguments between photography experts as to what causes shallow dof with larger sensors with many saying it's not the sensor size but other factors. However, whatever those factors may be, many people seem to observe shallower dof on larger sensors which suggests the ideal move is probably to stop down a little with a larger sensor (or correct whatever complex issue there is that causes the shallower dof effect). It appears on the 4k colour series this didn't happen and the dof was quite shallow.

I actually rather like Steve's suggestion above:

"For me they should take new released in 2024 1/1.2" image sensor from OmniVision, which supports both white and IR and do only 8Mpx line of 7 series with various varifocals (ZE and Z4E minimum).
Latest Ubiquiti G6 PRO bullet showed that this is possible (put 1/1.2" and varifocal in small bullet chassis).
Also HIK is using this sensor in new ultra PTZs...

this way we could have 8mpx cam with super night performance using 1/1.2" sensor and AI-ISP / WizColor video processing.
comparable to performance 4Mpx on 1/1.8" (the same pixel pitch).."

I can't help but feel this may produce a better result, but we'll have to wait and see if we'll all be going Hik if that's what they're planning...could be some interesting testing ahead - Wizcolour Pro vs Hik Ultra.

I should add, I also was never impressed with the lens quality on the 4k series. Pictures never seemed very sharp even on the sweet spot, not compared to the 541R. Maybe lens cost / complexity is one reason why Dahua have stepped down to the 1.8 again, although with so many phones and action cams using 1inch sensors, it's hard to see what the issue is and not question why 1inch and not just 1/1.2 isn't actually available for CCTV at a reasonable price.

I also wish a Dahua factory rep visited these forums and engaged and took on board feedback as I think it would be beneficial to both parties. It's alright thinking many of us are amateurs, which many of us are, although there are pros on here as well. However, you have to ask yourself, how much actual feedback does Dahua get from companies buying CCTV? I think most just have cameras installed and never give feedback as the end users know little of CCTV. They just use the pictures and accept the spiel from the installer that it's the best or best suited. Of those that do give feedback, it probably goes to installers and they never pass it on which potentially leaves Dahua insulated from feedback from end users.
 
  • Like
Reactions: alekk
CCTVCam said:
I should add, I also was never impressed with the lens quality on the 4k series. Pictures never seemed very sharp even on the sweet spot, not compared to the 541R. Maybe lens cost / complexity is one reason why Dahua have stepped down to the 1.8 again, although with so many phones and action cams using 1inch sensors, it's hard to see what the issue is and not question why 1inch and not just 1/1.2 isn't actually available for CCTV at a reasonable price.
Click to expand...

this is why I think we never see Color4K replacement (in fixed lens version) from Dahua.

those 8Mpx 1/1.2" F1.0 cams were one-time exercise from both Dahua & HIK, using available that time old Sony Starvis sensor which was designed for vehicle recorders (lack of IR support).
HIK replaced those cams 1-2 years ago with 1/1.8" and never come back. Dahua stil produce old model without any upgrades.
Those constructions were to problematic and limiting...

But I hope (I have finger crossed) that new Dahua 7-line PRO cams should be based on 8Mpx 1/1.2" sensor.. the new one, released in 2024 and produced by OmniVision with full support for IR.
Combined with good big max F1.2 aperture varifocals (ZE & Z4E) and p-iris - where you can close aperture in day to have nice strong almost infinite depth of field / sharpness.

HIK is using this sensor in ultra PTZs and ultra dual sight cams.. Probably (not 100% sure) Ubiquiti is using it in G6 PRO...

varifocal with P-Iris and closing aperture + modern WizColor processing (with new very good sharpening algorithms which don't over do like Color4K or 5442-S3) should fix almost 99% problems which we had with Color4k..

Of course this will be camera from Ultra Line, done as Ultra line (big one with heavy optics) and priced as Ultra Line..
 
Last edited:
  • Like
Reactions: alekk, bigredfish and CCTVCam
All else equal, it's just optical physics that larger sensor size will result in less depth of field. Here's a DOF calculator to show that.

Ditto that @steve1225 is super knowledge (sign me up for "his" Ultra!) and has some great suggestions (ditto others) so yea, I agree that it would be good if a "Dahua factory rep visited these forums and engaged and took on board feedback as I think it would be beneficial to both parties." ... but it's really great that @EMPIRETECANDY does this.
 
Last edited:
  • Like
Reactions: CCTVCam
steve1225 said:
There is new Dahua Xinghan video presentation - on Youtube...

I selected part, where they show new perimeter protection and new animal detection using new 54PRO WizColor camera..
But you can skip to beginning or other parts of the presentation

Click to expand...


Features look great in theory although it looks as if you need a Dahua NVR to take advantage. Wonder if Ken can make them work on BI as surely the software is in the camera regardless even if it use off camera processing, otherwise if it was in the NVR, surely it would work with any camera, old or new, and be rolled out to all. My only other concern is the low light. Looking at the object distance detection picture again with the grass one side and fence the other and city in the background, you can see motion blur on the person and ghosting suggesting very slow shutter and very high gain were used to achieve it. We don't know how dark it was though and with only distant lights, it is possible it was pitch black in which case any detail retreival was remarkeable and performance should be much better in a more urban situation. Can't wait for the tests. At least it has IR if not and the ability to switch to colour on detection and the addition of light whether from it's own led's or external lighting.
 
  • Like
Reactions: flynreelow
CCTVCam said:
Features look great in theory although it looks as if you need a Dahua NVR to take advantage. Wonder if Ken can make them work on BI as surely the software is in the camera regardless even if it use off camera processing, otherwise if it was in the NVR, surely it would work with any camera, old or new, and be rolled out to all. My only other concern is the low light. Looking at the object distance detection picture again with the grass one side and fence the other and city in the background, you can see motion blur on the person and ghosting suggesting very slow shutter and very high gain were used to achieve it. We don't know how dark it was though and with only distant lights, it is possible it was pitch black in which case any detail retreival was remarkeable and performance should be much better in a more urban situation. Can't wait for the tests. At least it has IR if not and the ability to switch to colour on detection and the addition of light whether from it's own led's or external lighting.
Click to expand...

ONVIF is a dead end..

As I wrote many times on this forum, more & more new features will work only with Dahua NVRs, Dahua IVSS servers or Dahua VMS software (DSS Pro)...
And exported only by Dahua APIs, not ONVIF...

Also processing for more & more features are divided between cameras and NVR's - today AcuPick and Face Detection/Recognition works that way.. Next will be WizSeek.

the same is in Hik-vision world.. or many brands of high-end CCTV (Avigilon etc)...
 
  • Like
Reactions: CCTVCam, looney2ns, EMPIRETECANDY and 1 other person
steve1225 said:
ONVIF is a dead end..
Click to expand...
This sucks. I understand what you're saying and it makes perfect sense. After taking over a year to get up to speed with BI and finding I can do some viewing related things with it in seconds instead of minutes with the NVR, I have to face not being able to use new camera features with BI. If I eventually have to go with a closed system, it's sounding like unifi might be the best choice. I'm getting one PRO camera to check it out. If it's not much of an improvement over the 5442-S3 because of the ONVIF limitations, replacing everything with unifi would seem to become a viable option . (Opinion is based only on word-of-mouth and reading about unifi, with no first hand experience.)
 
  • Like
Reactions: CCTVCam
tigerwillow1 said:
This sucks. I understand what you're saying and it makes perfect sense. After taking over a year to get up to speed with BI and finding I can do some viewing related things with it in seconds instead of minutes with the NVR, I have to face not being able to use new camera features with BI. If I eventually have to go with a closed system, it's sounding like unifi might be the best choice. I'm getting one PRO camera to check it out. If it's not much of an improvement over the 5442-S3 because of the ONVIF limitations, replacing everything with unifi would seem to become a viable option . (Opinion is based only on word-of-mouth and reading about unifi, with no first hand experience.)
Click to expand...

there is one solution for that problem, which BIG VMS'es are using...

those have 'drivers' to cams.. so they have basic ONVIF support...
plus some extended 'Dahua driver', 'HIK driver', 'AXIS driver', 'Avigilon driver' etc, which are using Dahua/HIK/Axis/Avigilon private API to implement extended functionality (events / AI / configuration etc)...

But BI is too small for that.. And to much centered on universal support for cheap consumer brands...

For me BI has strange implementation problems somewhere in the core.
When I read on the forum that encoding must be h.264 not h.265, no SmartCodec, no AI Codec, only CBR, with small FPS - because BI supports well only that - It makes me laugh.

It's 2025. Such limits indicate that the architecture of this app is truly flawed.
 
  • Like
Reactions: Chokolinho
steve1225 said:
When I read on the forum that encoding must be h.264 not h.265, no SmartCodec, no AI Codec, only CBR, with small FPS - because BI supports well only that - It makes me laugh.
Click to expand...
I think you're being too harsh with BI on some of these things. Others that know more would be better with this, but I'll state what I know. h.265 can be used. I think there is some place I haven't run into where it's a disadvantage. I've used it, and the major debate is over if it's worth it because of the file size vs. image quality tradeoff. VBR works for me but I don't like what it does to the image. I'm not aware of any particular FPS limitation. My big beef with BI is it's complexity that often makes it a big pain in the rear, but others thrive on the same thing because of the capabilities it opens up. On the other hand and by comparison, viewing clips and events using the NVR or smartPSS is truly flawed. It is so much faster and easier with BI.
 
tigerwillow1 said:
I think you're being too harsh with BI on some of these things. Others that know more would be better with this, but I'll state what I know. h.265 can be used. I think there is some place I haven't run into where it's a disadvantage. I've used it, and the major debate is over if it's worth it because of the file size vs. image quality tradeoff. VBR works for me but I don't like what it does to the image. I'm not aware of any particular FPS limitation. My big beef with BI is it's complexity that often makes it a big pain in the rear, but others thrive on the same thing because of the capabilities it opens up. On the other hand and by comparison, viewing clips and events using the NVR or smartPSS is truly flawed. It is so much faster and easier with BI.
Click to expand...

Correct. A lot of these suggestions come from the time before substreams were the thing.

When it was just mainstream, it could struggle with many systems over 15FPS as an example.

Prior to cameras having AI, most would use BI motion detection and things like H265 and VBR, and smart codec would wreak havoc on BI motion detection with the way the algorithm was set up.

But if someone is using the AI of the camera, go to town with H265, smart codec, VBR, etc.

Just keep in mind that many NVR users also do not use H265 and smart codec and AI codec for what it does to the image. It was really only the FPS that was the major difference between BI users and NVR users.

But if you have a field of view and situation that H265 and smart codec and higher FPS makes sense and you using BI with substreams, then go have fun with it LOL.
 
  • Like
Reactions: CCTVCam and tigerwillow1
steve1225 said:
ONVIF is a dead end..

As I wrote many times on this forum, more & more new features will work only with Dahua NVRs, Dahua IVSS servers or Dahua VMS software (DSS Pro)...
And exported only by Dahua APIs, not ONVIF...

Also processing for more & more features are divided between cameras and NVR's - today AcuPick and Face Detection/Recognition works that way.. Next will be WizSeek.

the same is in Hik-vision world.. or many brands of high-end CCTV (Avigilon etc)...
Click to expand...

I can't help but feel Dahua are shooting themsleves in the foot here. For sure, their reason is probably they're wanting to bring everything in house and give people a reason to buy their systems. However, there must be hundreds of thosuands if not millions of users worldwide who don't want an NVR becuase of NVR's inherent limitations and cost for performance, and who can't afford $550 for the Dahua Pro software (that's the purchase price I've seen, unsure what the update price is and the purchase price is the base price with many add ons costing hundreds more from what I've seen).

By bringing their features to only their own eco system, they're giving those hundreds of thousands / millions of non Dahua NVR / Software users no reason to upgrade their cameras unless there is a significant non software locked improvement in picture quality. That could slow sales.

The answer is simple, either don't lock features into your own eco system or produce a cheap non commercial version of your software aimed more at the home user / small business somewhere along the lines of the BI model. ie. $60 purchase, $20 a year update fee. How could they do this without losing their commercial software sales? Lock the number of cameras supported to eg 10. That's more than enough for the home user / small business user and if they need more, they can always run a 2nd pc and another licence. How do you stop larger businesses buying multiple PC's and running dozens of cameras on multiple pc's for less than the commercial software? Limit the number of licences available to any one physical address / company / user to 2. That way you cap the home version off at 20.

However, a far lesss complicated way, and one that avoids software licence cheats, is simply make your features available to all over ONIV or some other open standard. To do otherwise, is to lock your market and potentially lose sales as for anyone not using your eco system, there's no longer any incentive to buy a new cameras at all this side of camera failure as the new features make little difference to buying incentive if the end user can't use them.
 
  • Like
Reactions: tigerwillow1
CCTVCam said:
I can't help but feel Dahua are shooting themsleves in the foot here. For sure, their reason is probably they're wanting to bring everything in house and give people a reason to buy their systems. However, there must be hundreds of thosuands if not millions of users worldwide who don't want an NVR becuase of NVR's inherent limitations and cost for performance, and who can't afford $550 for the Dahua Pro software (that's the purchase price I've seen, unsure what the update price is and the purchase price is the base price with many add ons costing hundreds more from what I've seen).

By bringing their features to only their own eco system, they're giving those hundreds of thousands / millions of non Dahua NVR / Software users no reason to upgrade their cameras unless there is a significant non software locked improvement in picture quality. That could slow sales.

The answer is simple, either don't lock features into your own eco system or produce a cheap non commercial version of your software aimed more at the home user / small business somewhere along the lines of the BI model. ie. $60 purchase, $20 a year update fee. How could they do this without losing their commercial software sales? Lock the number of cameras supported to eg 10. That's more than enough for the home user / small business user and if they need more, they can always run a 2nd pc and another licence. How do you stop larger businesses buying multiple PC's and running dozens of cameras on multiple pc's for less than the commercial software? Limit the number of licences available to any one physical address / company / user to 2. That way you cap the home version off at 20.

However, a far lesss complicated way, and one that avoids software licence cheats, is simply make your features available to all over ONIV or some other open standard. To do otherwise, is to lock your market and potentially lose sales as for anyone not using your eco system, there's no longer any incentive to buy a new cameras at all this side of camera failure as the new features make little difference to buying incentive if the end user can't use them.
Click to expand...

But this is not Dahua decision...

Simply ONVIF as a standard & organisation is not here.. And never have been. Don't support new features..

And no one is extending this standard fast enought... Or is extending with features a few years later after manufacturers implement them.

Big plus with Dahua is that we have public documentation of Dahua HTTP API. Last version which I have are 860 pages.

You can do hundreds things using this API, which are not available over ONVIF.

PS. Latest Dahua cams supports 4 ONVIF profiles: Profile S & Profile G & Profile T & Profile M - so in theory you have access to AI functionality defined by ONVIF M (metadata) like object detection & classification (humans, cars, faces, car plates). In theory full Video Meta Data functionality should be available to BI over ONVIF M (metadata). How much is implemented by BI?
 
Last edited:
  • Like
Reactions: tigerwillow1 and bigredfish
steve1225 said:
It's 2025. Such limits indicate that the architecture of this app is truly flawed.
Click to expand...

Still asking myself and the dahua marketing expert...

Why is the "AI" of dahua and HIK cams/nvr so bad and weak compared to a 50USD coral TPU paired with 10y old hardware and frigate ?

Why is are two billion dollar companies owned by the chinese government unable to program an UI for their hardware which works at least responsive ?

Why is a billion dollar company unable to clone software (search over multiple cams, create timeline) which others use 10y+ ?

Why is the firmware of "new cams" always a nightmare ?

Why does a billion dollar company show marketing material of blurry somehow overexposed people ?

Screenshot_20250912-165909.Drive.png
 
duplo said:
Still asking myself and the dahua marketing expert...

Why is the "AI" of dahua and HIK cams/nvr so bad and weak compared to a 50USD coral TPU paired with 10y old hardware and frigate ?
Click to expand...

I played coral TPU with Frigate 1-2 years ago - and for me Frigate AI detection (using coral) is a toy..
Big number of false positives or wrong classifications.
IVS on modern Dahua cams worked for me 10x better...

Of course - probably Frigate AI today is much better than was 1-2 years ago.
But Also Dahua have new implementation of IVS (Large scale AI models) - which works much better that it was a few years ago.

Everyone is progressing on the market. We have bigger AI processors in cameras/NVRs every new generation. And we have bigger/better AI models...

duplo said:
Why is are two billion dollar companies owned by the chinese government unable to program an UI for their hardware which works at least responsive ?
Click to expand...

I asked those questions every time I meet any Dahua representative. Never had good answer :)
they are not good in UI/UX at dahua.. never been.

plus they use almost 20 years old code base (yes that long they are on the market). No one wants to touch that code.

duplo said:
Why is a billion dollar company unable to clone software (search over multiple cams, create timeline) which others use 10y+ ?
Click to expand...

Who have those functionality 10 years ago?
Please specify - I'm very interested in that :)

You have search over multiple cams on Dahua NVRs - you can search any events from IVS, FaceDetect/Recognition, VideoMetaData or AcuPick over many cams at once..
They can be sorted over time and all shows thumbnails of people / objects..
if you find someone suspicious AcuPick allows you to search all footage showing specific person (by look) from all cams.

duplo said:
Why is the firmware of "new cams" always a nightmare ?
Click to expand...

Are you asking why first versions of firmware for each new line is not very well finished?

because software is hard. much harder that developing new hardware.
and require years to fine-tune & finish.

On every new hardware on the market (mobile, tv, smart gadgets, cameras, drones etc) first versions of software/firmware when released are bad. And requires some time to find bugs & finish missing features.

ps. please remember we here on the forum have access to new Dahua HW sometimes months before normal market start to use/install specific models.. So also we see early firmware version with all bugs & problems.


duplo said:
Why does a billion dollar company show marketing material of blurry somehow overexposed people ?

View attachment 228420
Click to expand...

ROTLF :)
because this is simply a real picture from camera footage done in real environment at full night conditions which shows what you will get from camera in that situation :)

Not professional done picture using DSLR which fake results (as often other companies do in presentations & video promotion materials)..

And I'm happy they use this not ideal picture to show real performance - not faked one...
 
Last edited:
  • Like
Reactions: looney2ns
You must log in or register to reply here.
Top Bottom