Avigilon H4 5MP and 4K Cameras Tested

Published Apr 25, 2016 18:09 PM

******** *** ******** ***** ** ** line, **** ******** ************ ********* ***** most ********* *** ***** ****** (*** 5.0L-), ********* ** *** ******, *** a ***, **** *********, **/*** *****.

** ****** *** ** *** ** H4 ** ***** ** *** *** they ******* ** ****** **** ****, Bosch, *********, *** ****, **** **** during **** *****:

******, ** ******* *********** ** *** light *** **** ***** *****, ********* comparisons, *** ****.

*******

********'* ** ** *** *** ** domes *** **** ***** ********** ** low ***** *** **** ***** ****** relative ** *********** ******, *** ****** competitively ** ******** ****** **** ******* models.

*******, ******* ****** *** ******** **** performance ** **/*** ******* ** **** tested, **** ****** **** ** *********'* 4K *** *** ***** *** ********* performing ****** ** **** ******. ************, the ** ** ******'* ******** **** among *** ******* ** ** ******* we **** ******, ****** **** **** Axis *** *********'* ** ******.

******** *** ** *** **********

*** ** *** *** *** ********** in ******** *** ** *** *******, however:

  • ***** ** ***** *********, ***** ********* only **** ******** ******* ******, **** been ***** ******* ** **** ********* *** *** ******** ** *** entire ** ****** ****. ****, ** have *** ****** *** ** *******' analytics ** **** ******, *** **** to ** * ****** ****.
  • ******** ******* ********* **** *** **** deeply **** ***** *************, **** ****** all ************* ******** *** *** ***, instead ** ******* ** *** *** camera's *** *********, ***** ******* ************ time.

*******

*** ******** ** ** *** ** estimated ****** ***** ** ~$*,***-*,*** ***. This ** ********** ****** **** ****, such *******************'*** ******* *****, ***** **** *** $800-900 ******. *******, ** ** ***** than ** ** ****** ********************, **** ** ***** **** *** $2,000+

*** *** ******** ** *** ** cameras ** *** **** ********** ****:

*** *** ** ***** *** ** estimated ****** ***** ** ***** $***-*,*** USD. **** ** ******* ** *********** 5MP ** ******, **** ** ********** ***-*****. *******, *** ***** ***** *** models **** ********** ** *** **** lenses **** ****** **** ******, **** as ************************** ***** (~$*** ******). ***** ****** provide ******** ****** *** *** ** their **:* *** ********** **. ******** and ******' *:* ***, **** **. 2592, **** **** ** *** **** features (********* ****/*****, ** ***** *********, etc.).

*** ***** ******* ** *** **** comparison ****:

Key ********

**** *** *** *** ******** **** our **** ** *** ******** ** lines:

******** ** ** ***** *******

*** ******** ** ** ****** ******** moderately ****** ** ****** **** **** some ***********, **** ** *** **** P1428-E (** **),********* **** *** ***** CM-6208 (** *****).

** ***** **** (**’+) *** ** longer ******, ******* *** ******* ** other ** ** ******, **** ** clear ****, ****** *** ********** *** non-IR **** *****-*.

** **** *****, *** ******** ** H4 ******** ***** ******* **** *** Panasonic **-******* ** *** ******, **** reduced ******** ** *** ******* *** test *****.

******** ** *** ***** *******

*** ******** ** *** ********* ****** than *** ******** ********** ** *** camera ** *** *****, **** ****** increases ** ****** *** ******** ******:

*** ** **** *****, ******* **** increased *** ** *** ****** *** compared ** *** *** *****.

*********

*** ******** ** ** *** ************* higher ******** ** **** ***** *** low ***** ******** ** **** *** Panasonic's ** ******.

*** *** **'* ******** **** ********** higher **** *** ***** *** ****** in **** ******, ****** ***** **** Avigilon's ******** ** *** *****.

**** ***** ****

********'* **** ***** ****, *** ** the ** ****, ****** ******* ****** in *** *****. **** ******* ******* framerate *** ********* *********** *** *-***** interval **** *** ***** ** ***** to ****** *********. **** **** **** is ********* **** **** ***** ******, which ******** *********, *** *** ** unusable ** ***** ***** * ******** framerate ** ********.

Physical ********

*** ** ** ***** *** ******* in **** *** ************ ** ***** outdoor **** ******, **** ******* ********, but *** ********* *** ******** *****:

  • *** **** *** ******** ******:*** ** ***** ******* * ***** USB **** ***** ****** ********** ** a **** ****** *** *****/******. **** is **** ** **** *********** ******, such ** ****'* ***-******, ****** ** not ******.
  • ********* ****** ******:***** *** ****** *** ****** ****** from *** **** ******* ** ********* threw ******. **** ****** *** ****** module ** ** ******** ** **** of *******, ** ******** ** *** future, ******* ****** ** ********* *** entire *******. **** ** **** ***** in **** *********** ******, **** ** Axis */* *****.

** ****** *** ******** ************ ** the ****** ** **** *****:

*************

*** *** **** ****, ************* ** the ******** ** *** ********* ** similar ** *** **, **** **** slight *******:

  • ***** ******* ******** **** **** ******* into *** *****, **** ***/*****, ********, WDR, *** ** ******** ** *** and **********/********/********* ** *** *****.
  • **** **** ******* *** ***** ** and ******* * **** ** ******** to **** *** ********.
  • **** ***** **** *** **** ***** to ***** ********.

** ****** ***** ******* ** **** video:

IR ************ ***********

** ***** ** *** ******** ** camera ** ********* ** * ********** short ***, **** ******* **** **** outdoor ** ** ***** *****, **** as *** **** **** (***), ********* DS-2CD4585-IZH (***), ** ********* **** (***). In *** *****, ** ***** **** spec ** ** ******* ********, **** greatly ******* ******* ****** ***, **** below ** *** ******* ****** *** parking *** **** *** ***** (~**') and ********** *** ******:

** *** *****, *** *** ******** H4 ******* *** ** ******** ******** its *** ** *****, **** *** subject ******** ******* ******* ***** ** passes *** ******* ****** (********* ** the *****), ****** ** ~***'.

******** ** ********

**** ******** ******* ******* ******** **, which **** ********* **** ** **** for ********* **** ******. **** ** shown ** *** ********* *****, **** the ** ****** ** **** *** camera ** *** **** ********* (*********) zoom *****. *** ****** ** **** zoomed *** ** *** ****** ***, with *** ****** *** ** **** switching ** *** *** ***** *** dimming *** ******* ******* ***.

Low ***** ***********

** ****** *** ** ******* *** competitors ** ** **** *****/******* *** to *** *** **** *********, ***** a **** ***** ** **** (~**°):

**

** **** ~***' ********/~** ***, **** of *** ******* ******* ****** ****** of *** ******* ** **** *****. Note **** **** ******** ** ****** the ** ***** ** *** ******* tested.

****** ** *** ******, **** *** higher, ******* *******, **** *********** ******* to ** ******** ****** **** *** Panasonic **-******

*******, ******* ** *** *******, ***** PPF ** ****, *** ******** ** provides *** **** ******* ** *** subject *** **** ***** **********, **** the ********* **** ******* *** *** chart ********.

***

*** ******** *** ******** ************ ********'* previous ********** ** *** ***** ** our *****, **** *********** ****** ** lower ********** ** ****** **** **** and *********, **** ***** ** ~** PPF/60' *****.

** ****** *****, ******* ********* *** strong, **** *** **** ***** ******* to **** * *** *** ******* roughly ***********.

*** ***** ***** ***********

********* ** ***** **** ** *****, with ** ***, *** ******** ** 5MP ******** *** **** ******* ** the **** *****:

** **** *****, ~**' **** *** camera, ***** *** ** ******, *** 5MP ***** ******* *** ********* ****** tested, ****** ** ** ************ ** other ****** *** ** **** ************ of *** ******* *** ********** *** test *****.

Full ***** ***********

** *********** **********

***** * **** **** ***** ** view (~***'), *** ******** ** ****** produces ******* ** *** ******* ******* to *** ***** ** ******* ******, with ******** **** ***** ********** **** Axis *** *****, **** ** ***** wash *** *** *****. *** **** chart ** **** *********** ** *** Panasonic ****, ***** ******* ***** *** first *** ****** *****.

**** *** *******, **** *** **** higher (~** **. ~**), ******* ********* by ********, ****, *** ***** *** again *******, *** ********* **** ***** produces **** ****** *******.

*** **. ***/*****

******* ** *** *** ******** ** full *****, ** ******** ********** ***** at ****** *****/***** *** **** *** Bosch ***, **** ******* ******** ** the ******* *** **** *****, ****** provides ******** ****** ******* ******* **** the ** ***.

*******, ** ~**' ********, ***** *** is ******* (~** ** *** ** 5MP), *** ******* ******** ** *** 5MP ** ******** ******* ******** ** the *******, ***** ******* ***** ******* than *** ***** *** ***** (***** also *** ******** ****** *** *** to *** **:* *** ********** **. 4:3).

Bandwidth **********

** ****** ********* ** *** ****** above ** *** *** *** ******** H4 ******* ******* ** ******.

** **** *****, *** ******** ** 4K's ******* ******* *** ****** **** the **** *** ********* *******, ********** in *** *****, ******* ** ~** Mb/s.

*** ** ***'* ********* *********** *** similar ** *** ***** *** ********* in ****-*** ******, ***** *.* **/*, and ***** **** *** ******** ********** H3 *** ***** (~*.* **/*). ** low *****, ******** **** ******** ***** than *** ** ***, ~*.* **/* vs. *.*.

**** ***** ****

******** *** ********** "**** ***** ****" in *** ** *******, ***** ****** users ** ****** *********, *******, *-***** interval, *** *********** **** *** ****** detects ** ******, *********** *********** ******** bitrates. ** *******, *** ****** ******** to * ***, **** * ******** interval ** *.

** *** *****, **** ***** **** worked ******* ****** **** ********'* *** VMS ** **** ** *****. ******* gains **** **** ******, ********* ** specific ******** ****, **** ****** **** framerates *** ********* ********** **** ********* than ****/**** *-***** ******** ********, *** example.

******** **** **** ***** **** **** been **** ** ***** *******/*********, *** are ****** ****.

Test **********

******* **** ****** ***** ******* ********, with *** ********* ***********:

  • ******** *** *** ** */*** ***
  • ************ *** *** ** ** ******* of ~**
  • ***** ****** **** **** ***** ***** above (***** ******* ****)

*** ********* ******** ******** **** ****:

  • ******** *.**-***-***: *.*.*.**
  • ******** *.**-***-***: *.**.*.**
  • ******** *.*-***-***: *.**.*.**
  • **** *****-*: *.**.*.*
  • **** *****-***: *.**.*.*
  • ***** ***-*****: *.**
  • ***** **-****: *************
  • ********* **-**********-***: *.*.*
  • ********* **-*******: *.**
Comments (21)
IK
Itamar Kerbel
Apr 25, 2016

Thank you for a very good review. Saved me the need to do it myself :)

Since Avigilon uses built in lenses what lenses were used for the rest of the cameras?

Do they come lenses or are they purchased separately?

Avatar
Ethan Ace
Apr 25, 2016

All the cameras tested included lenses. Most of them were motorized focus/zoom, except the Bosch 5MP starlight, which is varifocal.

IK
Itamar Kerbel
Apr 25, 2016

I'm sorry, but I don't understand something.

"All the cameras tested included lenses" - means that when you buy an Axis 4K (for example) you get a lens, and that you tested with the included.

"Most of them were motorized focus/zoom, except the Bosch 5MP starlight, which is varifocal" - if most of them were motorized zoom, doesn't it mean that they were varfocal?

Avatar
Ethan Ace
Apr 25, 2016

The Axis 4K (and nearly every other camera here) includes an integrated zoom lens, not an interchangeable lens, like C/CS mount. The Bosch is CS mount, but the camera ships with that lens. It's not motorized.

Yes, motorized zoom is varifocal. But varifocal most often means that you're manually focusing it, instead of doing it via the camera web interface.

IK
Itamar Kerbel
Apr 25, 2016

Thank you for the clarification.

Since it appears to me that the Bosch camera gave the best fight can you tell me the lens manufacturer for the 4K if it is known?

Avatar
Bryon Wheaton
Apr 25, 2016
Ava Security - A Motorola Solutions Company

I wonder on your bit rate comparison, was the H3 3MP camera measured with WDR turned on? The H4 5MP lacks WDR, so it would be interesting to know.

Avatar
Ethan Ace
Apr 25, 2016

That measurement actually was with WDR turned on. I'll make a note of that. All the cameras in the test which include WDR had WDR turned on, as it normally would be.

(1)
Avatar
Ethan Ace
Apr 25, 2016

I should also specify that WDR was turned on during the day, and off at night if it did not turn off automatically. Just to be clear.

(2)
UM
Undisclosed Manufacturer #1
Apr 25, 2016

A little off-topic, but how does IPVM acquire the cameras used for testing? Are they all purchased outright, do you get them on a demo/loaner program from manufacturers, or donations from manufacturers? What do you guys do with your obsolete models?

Avatar
Ethan Ace
Apr 25, 2016

We buy almost everything outright. We purchase online or through integrators, depending on the line and availability. A very very few we get on loan, and that's typically only if it's very expensive and we won't use it in the future. If it's very expensive and we plan to keep using it, we'll buy it.

As for what we do with older models...I have about 40 cameras staring me in the face from the shelves of my desk, another 5 sitting here waiting for testing, a stack of recorders about 3' tall to my left...and that's not even including John Scanlan's desk or the shelves or the closet. We have given away some.

We've been planning to get a lab setup together (see: Idea: Should IPVM Open An Online Lab?) but space planning and moving have been lower priority. Still aiming for the next couple of quarters.

UM
Undisclosed Manufacturer #1
Apr 25, 2016

It would be funny if you guys opened something like a library, but with cameras, recorders, switches, etc. One could "check out" an XYZ camera for 30 days for $x...Ill take a free "library card" for the idea.

JH
John Honovich
Apr 25, 2016
IPVM

1, that's the goal of the lab, at least virtually. Have lots of cameras and VMSes online so anyone can access and try out as needed.

AR
Austin Rich
Apr 26, 2016

This would be well worth the price of membership. If IPVM kept a backup image of each VMS server or VM and let everyone have admin rights (to the VMS) to fully test them out that would be incredibly useful. I would, however, count on someone breaking it every few days... hence the backups.

(1)
Avatar
John Scanlan
Apr 26, 2016
IPVM • IPVMU Certified

The idea would be to use imaging software to quickly restore each server before / after testing.

We have started this for our internal testing needs.

UM
Undisclosed Manufacturer #2
Apr 26, 2016

"Their on board analytics, which integrate only with Avigilon Control Center"

Given that it's built on the ONVIF Analytics Service Specification v 2.2.0, does IPVM know of any third party VMSs building support for these cameras? What are IPVM's thoughts on the current lack of support for what's proven to be a solid performer in the analytics camera space?

Avatar
Brian Karas
Apr 26, 2016
IPVM

There are many ONVIF specification areas that aren't well implemented yet. Try to find good support for Profile G for example.

It's good that Avigilon is using ONVIF as much as possible, but I wouldn't assume that widespread support is just going to come naturally.

What is Milestone's incentive for example to add support for this? Engineering is mostly a zero-sum game, if they work on this, some other feature most likely gets delayed. How many customers are going to use Avigilon's analytic camera with Milestone instead of ACC? Enough to make it worthwhile to Milestone? Unlikely.

Hopefully more analytics products will adopt ONVIF Analytics Service, which then means VMS vendors can pick up support for many products by adding support, instead of just one or two brands.

It also would appear the Milestone (or any VMS) would have to add support not just for receiving and handling notifications, but also configuring rules as well. How would a customer with an Avigilon camera handle adjusting rules and such with another VMS? It seems like that VMS needs to do a lot of UI work, which may not translate to all other products (eg: other products will use different kinds of rules, or different names. How do you normalize this?), or Avigilon would need to have some kind of free configuration/test only VMS license so that customers could be assured they can handle the basic camera config functions.

Overall, I don't think it's all that surprising that other VMS's aren't rushing to add support for Avigilon analytics devices.

DS
David Schulman
Apr 26, 2016

Great recap. One question, you do a bandwidth comparison which is very useful to me as I design wireless systems everyday and throughput is a main concern. Is there anywhere within your resources where you list maximum bandwidth usage per camera (realizing that spikes occur that would take the camera above said max)? For example you list 10 fps in your above comparison, would be good to know what the maximum throughput would be should the customer utilize full frame rate. Thanks.

UM
Undisclosed Manufacturer #3
Apr 27, 2016

@IPVM - is it possible to share what SoC Avigilon is using for 4K models? Amba S3?

(2)
UM
Undisclosed Manufacturer #4
Apr 29, 2016

The picture you posted of the setup shows the dome being left off of the cameras. Was the testing done without the glass? If so, doesn't that somewhat distort the results?

Avatar
Ethan Ace
Apr 29, 2016

That picture was taken while we were setting up. We hadn't done final aiming and focusing yet. Testing was done with the domes on.

JH
John Honovich
Apr 29, 2016
IPVM

Testing was done with all domes / bubbles / glass on. The photo was of setup.