The Super Low Light Shootout 2016

Avatar
Ethan Ace
•Published Mar 28, 2016 12:16 PM

*** **** *** ************ ******* ** low *****? **** *** ******** ** integrated **, ***** ********** *** ****** imagers **** *** **** *** *****, we ****** ** **** ***.

** ****** ** ******* ** *** *** scenes, * **** *****, ***** *.* lux, *** * ******* *** **** outside ********, *-* ***, **** *****:

******* ******

** **** * ***** ** ****** cameras ** **** ****, ************ * range ** ****, *******, *** ***********. 

** ******** ***, ******, *** **** end ** ******, ********* *** *** and ** ****** **** ********* *** Panasonic, ** **** ** ***** ****** super *** ***** ****** **** **** and *********. *******, */*" ***** *** light ****** **** ********, ***** **** a ********, ***-*** ***** ********* *****.

Key ********

**** *** *** *** ******** **** this ****:

Integrated ** ***** ****** ** ***** *** *****

** **** **** ****** (and ***** ** ****** ******, ********** IR ********* ******** ****** *******, **** more ******* ** ***** ***** *** detection ** ****** ******. *******, ***** generation ***** *** ***** ******, ********** 1/2" ***** ****** ******, **** ******** this *** ************* ** *** **** few *****. 

*** *******, *** ***** ***** ***** a ***** ********** ** ****** ****** large ****** */*" *** ******* */*" super *** ***** ******. *** */*" model ******** ********* ** *** ********** IR *****, ***** *** */*" ** only ******** *****.

** **** *****, ********** ** *** much ****** **** ***-** *** ***** models, ** **** ******* **** *** 2013 **** *****. *** ***-** ***** produced ****** ** *****, ***** *** IR ****** ******** ****** ******* ** the *******.

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

Super *** ***** ********** ** ******** ******

** ******** ******, **** ** *** lit ******* *** (******* **** *-* lux), ***** *** ***** ****** ******* image ******* ******* ** ********** **, though **** *** ***** ********* ** color ******. *** *******, ** *** 1/3" *** */*" *** ***** ****** below, *** *******'* ***** ****** *** blue ***** *** ******* ** ****, where **** ****** ****** ****/***** ** the ** *****. *** ***** ****** of *** *** ****** *** **** easily ** ****.

** ****** ******* **** ***** ******* in *** ***** *****:

Super *** ***** ***** ** ** **** *****

** **** ******, ***** ****** ***** low ***** ****** ******* ******** ********** over ** *******, **** ****** ********* in *** ***** *** ** ********* contrast ******* ****** ******* *** *** background, ********** **** ****** **** ***** mode, **** *****. **** ******* ***** the******** ** *-********** ****** ****/*" ********* *********** ****, **** *** ******* ****** ******* in *** ***** *****, *** ******** into *** ********** ** *** ** model:

** **** *****, ** ****** ***** findings ** **** ******:

Limited ******* ** *** **** *********** ** *** *****

* *******/** ************ ** ******* ******** *** ********* benefits **** ***** ** ******, ****** the **** ***** ******** **** *******, but ********* ** ********** ******* ** the *******:

Low **** **: **** ***** *****/**** **** *****

*** **** ** *******, **** ** the ********* **** ** **** ****, provided ***** ******* ******* ** ****** than **** **** ********* ********** ** and ***** ***** ***** ****** ** close *****.

*******, ** ****** ******, ****** ***** cameras' ******* ** *****, ***** ****** provide ****** ********* ** *** *******, who *** ****** ** **** ** the **** *****:

 

Average *** ***** ******

** * ******** **********, *** ***** below ***** "*******", ***-***** *** ***** model ******** ** ******. ****** *** notably ******, **** **** ******** **** the */*" *** */*" *** ***** models *** **** ********** ** *******.

Test **********

******* **** ****** ***** ******* ******** with *** ********* *************:

  • *.***, ** ***, ~**-** ************ *** used
  • */*** ******* ******* *****

** ******* **** ****** *** ** a **** ** ******* ************ ******* models.

*** ********* ********/******** ******** **** ****:

  • ******** *.**-**-***: *.*.*.*** 
  • ******** *.**-**-***: *.*.*.*** 
  • ******** *.**-**-**: *.*.*.***
  • **** *****: *.**.*.*
  • **** *****-***: *.**.*.*
  • **** *****: *.**.*.*
  • **** *****: *.**.*.*
  • ********* **-*********-*: *.*.* ***** ******
  • ********* **-**********-**: *.*.*
  • ********* **-***********-*: *.*.* ***** ******
  • ********* **-*******: *.**
  • ******* ***-****: *.*********
Comments (18)
JH
John Honovich
•Mar 28, 2016
IPVM

Just 5 years ago, people routinely warned about using megapixel in low light. And there was certainly a lot of truth in that as many cameras performed extremely poorly.

And now, there are megapixel cameras, without IR, in very dark scenes, producing images that rival IR. It's amazing progress.

If this continues, what does this mean for integrated IR and for thermal? Does integrated IR become more of a cheap low light option where super low light cameras are the high end, even for very dark conditions? Does thermal get relegated for very niche applications, even further than today?

(4)
(1)
TD
Terry Delmonaco
•Mar 28, 2016

We have had really good results with the Bosch Starlight series of cams. They come in PTZ, dome and box configurations. In Florida we have to deal with the sea turtle nesting season. All lighting along the coast must be minimized to the point our beach surveillance was compromised or non existent during the nesting season. We replaced the beach cams with Bosch 7000 series Starlight and the difference is night and day no pun intended. We were so happy with the Starlight cams we installed them at remote well sites with equally good results. You guys mightn't try testing the Starlight cams also.

(1)
(1)
JH
John Honovich
•Mar 28, 2016
IPVM
In reply to Terry Delmonaco

Terry, we have tested multiple Bosch Starlight cameras:

They are certainly good but the market keeps on advancing and there has been many strong new offerings from rivals in the past year.

(2)
(1)
TD
Terry Delmonaco
•Mar 28, 2016
In reply to John Honovich

I missed those tests. Thanks John.

(2)
UM
Undisclosed Manufacturer #1
•Mar 28, 2016

Is there any particular reason IPVM chose to test discontinued H3 cameras? http://avigilon.com/news/security/discontinuation-of-the-avigilon-h3-portfolio-of-cameras/

(3)
(1)
Avatar
Ethan Ace
•Mar 28, 2016
In reply to Undisclosed Manufacturer #1

Because the Avigilon H3 IR cameras were quite strong in our tests and were used as one of the examples of high performing cameras. The Avigilon H3 Non-IR cameras were used as an example of an 'average low light camera'.

Given the H4 HDs were just recently announced, we plan to test them in the near future. This test was actually already in the works before the H3 discontinuation and we're still waiting for our H4s to arrive.

(1)
UM
Undisclosed Manufacturer #3
•Apr 09, 2016
In reply to Ethan Ace

Yes, but 2 or 3.0C-H3A would have much better results also.

(1)
(1)
Avatar
Kestutis Nagys
•Mar 29, 2016

Good job, IPVM. So next expected step for manufacturers can be super low light cameras with integrated advanced IR, which can outperform thermal cameras on detection range, resolution and price?

U
Undisclosed #2
•Mar 29, 2016
IPVMU Certified
In reply to Kestutis Nagys

...super low light cameras with integrated advanced IR, which can outperform thermal cameras on detection range, resolution and price?

They are already clobbering them on resolution and price. As for detection, although super low-light and advanced integrated IR cameras will have improved detection range, it's fundementally different than the way thermal detects, and I'm not sure they are directly comparable.

Since thermal is working off a heat differential, it can detect people hiding in the grass, or hiding in a car etc. Of course at low resolution.

Hybrid thermal/visible cameras on the other hand may benefit by having the ability to illuminate in the visible that which is detected in the infrared, with enhanced resolution.

(1)
OE
Ola Edman
•Apr 11, 2016
In reply to Kestutis Nagys

As super low light already outperforms thermal cameras on resolution and price, two of your criteria are already true (and has been true since thermal cameras were introduced).

As for detection range, as long as we're talking perimeter protection, I am certain that thermal cameras will continue to reign in this area (with the exception of a select few scenarios) for the foreseeable future. Although thermal cameras are not immune to poor conditions (fog/rain/snow/pollution) they are much more resilient to such conditions and will provide much more consistent results for perimeter protection.

The false alarm rate is also lower overall for thermal cameras, something that has to be considered.

The gap between regular cameras and thermal cameras will however certainly be smaller, and keep shrinking, making it more cost efficient to use regular cameras for simple/low-risk scenarios.

JH
John Honovich
•Apr 11, 2016
IPVM
In reply to Ola Edman

Ola,

Good points. In an upcoming thermal vs super low light shootout, we plan to test both video quality / ability to capture details as well as video analytics performance. It will be interesting to see how close (or not) super low light cameras do given their noise and the risk of stray lighting impact their performance.

AG
Alexander Gutierrez
•Mar 30, 2016

Would you test Hikvision Low Light (Darkfighter) with integrated IR? i.e. DS-2CD4A26FWD-IZ

KW
Kris W
•Apr 22, 2016

In my opinion you guys should be testing with Avigilon H4 bullet cameras as they have light catcher and adaptive IR technology. The H3 is older and being phased out

(1)
JH
John Honovich
•Apr 22, 2016
IPVM
In reply to Kris W

Avigilon H4 test report is being released on Monday.

KW
Kris W
•Apr 22, 2016

Thanks John

Avatar
Ethan Ace
•Apr 25, 2016

Our Avigilon H4 test is published. We tested two cameras, the 4K (to see the state of 4K integrated IR models) and 5MP (which Avigilon specs as their most light sensitive model). Neither really moves the needle per se, or has huge impact on the results of this test, and general findings remain the same.

However, the 5MP Avigilon H4 (left) is clearly more light sensitive than the H3 3MP (right). This is most noticeable in non-IR scenes at night seen here.

The 4K Avigilon H4 is slightly/moderately better in some scenes than other 4K models, but ultimately, as we say above, 4K's benefits in low light are few, if any, versus 1080p/3MP.

See our full test here.

KW
Kris W
•Jul 25, 2016

FYI, street prices on the H3 cameras went way down

JH
John Honovich
•Jul 25, 2016
IPVM
In reply to Kris W

Kris, we covered that here: Avigilon Quietly Slashes H3 Camera Pricing 30-50%