Thanks for the great test, as always. This is probably the bottom of the barrel model for Dahua as well. I am interested in getting my hands on the mid tier motorized dome unit. We currently use their 3MP version and would love to see just a little more resolution for parking lots and other outdoor areas where 3MP just isn't enough at times.
Dahua vs Bosch and Axis 4K Cameras
** ** ****, *** *** ******* issues.
**** ****** *** **** *** ***** performance ********* ********.
*** *****, *** ** *** *** Chinese ****-*************, ***** *** *** ********** low-cost*********, *** ******* *** ** ******, with ** ** ******, **** ** a ******** ** *** **** ** 4K *********.
** **** *** ******, *** *****'* 4K *******? *** ** ***** *** image ******* ** ****** ****** **** Axis *** *****?
** **** ***, ** ****** *****'* HFW4800E ******* ** ******* **** **** (*****-*) *** ***** (****** ****) ** **** *****, *** *****, and *** ****** ** *** *** they *******.
Key ********
**** ***** ********** *** *** ******** from **** ****:
*******
- ******* ** **** ***** (~*** ***) at ****** **** **** ******* ** 4K ******* **** **** **** *** Bosch *** ****** **** ***** ******* bullet.
- ******* ******* ********** ** **** ** and ***** ***** ****** ** ***** (~67') **** ** **** *****, **** **** chart ********** **** ***** ** ******.
- **** ******* ** *** ****** ****** in *** ***** **** ** **, moderately ****** **** ***** ** *****, and ************* ****** **** ****** ** non-IR *****.
- ** ***** *** *** **** ********* 30m ********, ********* **** ************ **** at **-***.
- **** *** ***********, ******** ** **** P1428-E *** ***** ******** ** **** light *** **** *****. *** *********** was **** **** ** *** ***** ***-***** against ****** *********, *** **** ***** in **** ***** ** *** *****.
- ******** ****** ** ******* **** ** and ***** ******, ****** ******* *** HFW4800E *** *** **** ** ****** to ******* ****** ******* ** ***** *** to *** ********** **.
Pricing *** ************
*** ******** ** ********* ********* ** buyers ** ***** ******* **** **** import ******* **** *********, *** ~$*** ***. **** ** * fraction ** *** ***** ** ******* *******, **** ** ******* *****-*(~$*,***) ******* ***-*****(~$*,*** ******), *** **** ****** ** ***** fixed **** ******* ** *******, **** as *****/*** ******, ***** *******~$*** ***.
*******, ******** ********* ******* ***** ******** may *** ******* **** ******* ** warranty. ***** ***** **********.*** ** ************ [link ** ****** *********], ****** *******.*** is *** ********** *********, *** ******** to ******* **** ***** **** **** unsuccessful. ** * ******* ****, ****** should ****** *******, ** *** **** support *** ******** ******* **** ***** sources.
***************
*** ***** ***-******** ******** **** *** light ******** **** ** ***** ** cameras ****** *** ** *** ***** in (****** *******) ** ***********. *******, WDR *********** *** *****, ****** **** the **** *****-* *** **** **** of *** ***** ** ******. *** these ******** **** ** ****** ***** to *********** ***** *******, ****** *** jump ** ** **** **** ****** friendly.
** *** ******** ****, **** ***** 4k ****** ***** **-***** *******, *****, I/O, *** ********* *******.
Physical ********
** **** *****, ** ****** *** physical ************ ** *** ***-********, ****** typical ** **** *** ******* *** for ******* ****** *******.
*************
**** ***** ******* ******* *** ************* points *** *** ********, *******:
- ******** *** **** ******* ** ****, not ***** **** *******, *** ******** gain ***** *******. ** ********* **** be ********* ** ******* *** ***** performance (*** *****).
- ** ******** ** **:* *** **** (3840x2160), * ***** **:** ** ****** ratio ** *********, *** ***** ** other ** ******* ** **** ******.
- *********** ******** ********* ** ****** ************ ****** than ***** ***** *******, **-** **. 22-30.
- ******* (*******) *** ** ******** ****** details ** ******* *** **** ***** (off ** *******), *** ********** ****** default *** ****** ******.
****: **** ***** ******* ****** ******** to *** ******** ****. *** * general ****** ** *** ***** *** interface, *** *** ******* ***** ******* ******.
Gain ******* ********
*** ***** ***-******** ******** ** "****" for ******** *** **** *******, ***** results ** **** ******, ******* *****-** IR.
********** **** ** * ********** ***** with * *** ** *** (***** in *** "**********" ****, *** ********** above) ****** *** ****** ** ***** more **** ** ******, ********** *********** the *****, ********** **** ******.
Image ******* ***********
** ***** ******* ***** * ~**' HFOV, **** ****:
** **** *****, ~*** ***, ***** quality ** ******* ** *** ***** 4K ******* *** ** ***** ********* high ***, ~***. *** ***** ****** provides ***** ******* ** *** *****, but ***** ****** ******** *** ******* plate *** *******'* ****** *******.
** *** **** ***** ** **** at ~.* ***, **** ** ******* deliver ************* ****** ******* **** ****** Axis ** ***** ** ******. *** Dahua ** ******** ** ********** ****** than *** ***** ********, *** ***** provides ****** ***** ********** *** ******* similar ****** *******.
**** ***
****, ** ***** ** * ~**' FOV, ****** *** ******* ******* **** the *******, ***** ***** ****** ********, shown ****:
** **** *****, *** ***** ** camera ******** *** ****** *******, ** do *** ***** *******. *** **** chart ** **** ******* ** *** Bosch ** ******, **** ** **** 5/6, ***** ***** *** **** *** read **** ** */*, *** *** 1080p ****** ** */*.
**** ****** ***, ** ******* ** ***** details *** ******* ** *** ******. Neither ********** ** ******'* ************ ******** lit *** ******* ** **** *****.
Wide ******* *****
** ****** *** *********** ******* ****** sunlight ******* ** **** ******** **** and *** **** **** **** ** it, **** **** ** **** *** from *** ********.
*** ******** ** *** * **** multi-exposure *** ******. *******, ** ****** various ******** ************ *** ********, ***** Dahua ***** ****** "****." ** ***** the ********** ******* *** *** *** on ** *** ****** ******* (** on * *-*** *****) *** ***********, with **** ****** ******* ** *** subject *** ***** ** **** ***** and **** *****. ********** *** ** max (***) ******** ****** ***********. ***** BLC ******* ** ***, *** ******* is **** ************, *** *** ********** is **** ********** ****** ***.
******* *** ***** * **** *** camera, *** ******** (***** *** ***** 50) ******** **** ** *** ****** area ** *** *****, ******* ** the ***** ** *****, *** ****** than *** **** ** *** ***** 1080p *******. *********** ********* ** *** dark **** ****** *** ****, **** the ******* **** ********** *******. *******, the ******** ***** *********** *** **** 4K *** ***** ***** ******* ** this *****.
***********/*********
**** **** ****** **** ***** *******, whose *********** ****** (*-*, * ***** least ***********) ***** **** ~**-** ** the ************ *****, *** ******** ****** from **-**. ******* ** ****, ** attempted ** *********** ******* ** **** increased *********** ***** *** *** ************ below.
*******, *** ********'* ******** ******** ****** low, ***** * **/* ** *** scenes. ** *** *** ******* ** the ** ******* ****** ** *** light *** **** ******, ****** *** only *** ** *** ***** ********* usable ****** ** ***** ***** ******. Bitrates **** ********** ****** **** *** 1080p ***** ******, ** ****.
Firmware ******** ****
*** ********* ******** ******** **** **** in **** ****:
- **** *****-*: *.**.*.*
- ***** ***-*****: *.**.****
- ***** ***-********: *.***
- ***** ***-********: *.***
I am going to state the obvious. I find it amazing that Axis consistently performs poorly in the recent comparison tests despite their premium pricing. They may have excellent support, accessories to support varied installs, high level of customizability and good design but somehow they overlook the most important thing (imho) and that's image quality. Just amazes me.
In fairness, Axis is varifocal and the other two are not. I am not sure if that is causing the difference but it's definitely a big plus to have varifocal.
What I don't understand is why Axis did not include IR in the P1428 when they did in all the other models of that series. I get that they recommend to add external IR but for very little cost they could have offered at least short to mid range IR built-in.
Is it possible that you received a lemon from Axis? Looking thru the tests it appears to not fully focus ever... And there have been problems with their first batch.
Is Axis really OK with this continued trouncing of their high-end product?
Axis reads these reports and has not complained. We'd certainly listen if they provide feedback.
We auto-focus and then manually check focus on the P1428-E before we shoot every scene. The comparisons you see are after that process is complete. It simply never gets as sharp as the other cameras.
I could throw out theories as to why but they'd just be theories.
It's weird because there are couple of videos out there of the p-1428e that look decently sharp. Of course it's hard to compare to the tests, but still...
As for the theories, I don't think there is anyone more qualified than you to venture what might be going wrong with that camera. So let's hear 'em!
I had a crazy thought: i know it's a different Axis model, but it looks a little like that hidden focus ring problem from back a couple weeks ago, maybe not as soft. Does that one have a hidden decoder ring by any chance? :)
Regarding sharpness and image quality, we have used two P1428-E for a while now and they work great as still image webcams. See http://www.cid.no/dnv-gl/ for samples and here is a 4K sample with no post camera correction.
This is a fantastic article. I am utterly blown away by these results in comparison to the premium manufacturers. This is like seeing a Kia blow the doors off a Porsche. If I was one of the premium manufacturers I would begin to question how premium the product really is. Normally I would never purchase a camera at less than 1/3 of the price of a well established brand after assuming it is disposable junk. That is getting harder to reasonably do in light of these recent shootouts and the rise of Hikvision and Dahua.
I have to wonder about the lifetime reliability of this camera. Is the warranty 30 days when purchased through appropriate channels? Does the web interface look like it was derived from Geocities? Does the camera have to be self-assembled like Ikea furniture? I feel like there has to be some sort of "gotcha" that is going unseen.
Truly eye opening.
Our US based OEM distributors handle all warranty issues. This model would probably have a 2-year warranty. The firmware / web UI may not be AS polished as western brands, but it isn't buggy. Maybe just not as well thought out as some others. Hik has a little better UI than Dahua IMO. It just looks more professional. However, it usually doesn't matter because clients won't ever see it.
As for reliability of Dahua in general, it has been very good for us. I don't know how many warranty claims I've ever made, but it's likely less than 5 in the past 4+ years I have been using them. Less than five claims over hundreds of items.
On the image quality side, I believe a key reason is that most manufacturers, Western or Chinese, have access to similar imagers, lenses and chips, so to the extent that they each use them, it's hard for anyone to 'blow out the competition.'
Indeed, recently, Axis seems to be refocusing on competing in various niches, like the suicide resistant camera, the thermography camera, the just announced rugged exterior mount vehicle camera, etc. In many of these cameras, there is no Chinese equivalent which puts Axis in a very good position. The limitation is that these are niche segments.
Our issue with the mini dome pictured is that it isn't 3-axis, which limits its application. We are waiting for the larger mid-sized dome with 3-axis and motorized zoom. That will be our new standard cam if all else remains the same performance-wise.
Usually their motorized zoom is 3-9mm. Manual varifocal is usually 2.8-12mm.
Correct me if I'm wrong but I believe the Dahua product is Max 7fps@4K(3840×2160)?
Good catch! The others claim 30 FPS.
The Dahua Box though is also 30 FPS, and although no IR, it takes C/CS mount lenses...
Aliexpress is listing them at ~450, so more expensive, but if the performance holds, still disruptive.
Yes, the mini dome and bullet are 8MP@7FPS, but the box is 12MP@15FPS/8MP@30FPS.
I'm looking into this, but unfortunately not in the office today due to the weather, and that camera isn't connected, so I can't check remotedly.
We are going to see what actual max framerate is, because we tested at 10 FPS (standard across all cameras), and got a real 10 FPS at 3840x2160. You can count all ten individual frames per second, and AVInaptic analysis reports 10.00 frames, not one dropped in a 40 second clip.
Also, multiple VMSes allow you to configure the camera for 30 FPS (select it in recording settings), but whether it actually provides that is another question.
When I get back in the office we'll investigate further.
I wonder what the max setting in the web interface is? If you can exceed that value via ONVIF, that would be interesting. I haven't tried to exceed the cameras limits via ONVIF before. We usually record at 10FPS, which is usually well below max specs.
For what it's worth, we have seen other cheaper Chinese made cameras that allow you to exceed the safe framerate allowed by the camera's CPU. The dome we tested allowed you to set the framerate at 30FPS, but the camera hardware would become unstable after exceeding 18FPS.
I'm sure the limitation in the Dahua is a self imposed limit, not an actual sensor hardware limit of 7FPS. To keep the price as low as they do, they have to sacrifice somewhere. CPU power would be my best guess. That is why they probably advertise it as 7FPS and, l would believe, limit it in firmware as well.
Ethan,
Did you ever get back to testing the framerate on the Dahua 4K bullet?
Oh, yes, but having forgotten to update here, and that being months and months ago, I don't want to misspeak. I'll look at results on Monday and let you know.
At the very least, it was getting a solid 15 FPS, from what I remember. Not sure about higher.
The web interface allows you to set up to 15 FPS at both 4096x2160 and 3840x2160 resolutions. I ran them through a stream analyzer and FPS is solid day and night. I also counted frames, to be less scientific.
So short answer: 15 FPS.
When you use the 4096x2160 resolution, does it actually give a 4:3 image? Or does it use black bars to fill in? The spec sheet shows it as a native 3840x2160 (16:9), so that is why I'm confused.
It's not 4:3, it's still wide. 4096x2160 is 4K UHD cinema format.
It has no black bars, all pixels. I actually thought we'd covered that in this report but the next time we have the camera out for a test I'll add snapshots showing where it adds pixels.
Not available in the US yet, from what we were told it will probably be a few months.