Testing Low Light Vs Image Quality

Author: Ethan Ace, Published on Jan 07, 2013

While minimum illumination ratings imply a magical point where image quality goes from good to bad, the reality is that image quality gradually degrades as lighting levels lower. In this report, we share test results showing what this looks like and the rate of decline.

We took three well-known surveillance cameras (from Axis, Arecont and Sony) and put them side by side for a simultaneous shootout from 1 hour before to 1 hour after sunset, recording images and lux ratings at 10 minute intervals. We placed a subject carrying an eye chart at a 30 foot wide FoV. Here's an overview of the scene used:

lowering scene overview

These tests were performed at progressively lowering light levels of:

  • 7,000 lux
  • 4,000 lux
  • 3,500 lux
  • 2,300 lux
  • 1,800 lux
  • 900 lux
  • 330 lux
  • 40 lux
  • 6 lux
  • 3 lux
  • 2 lux
  • 1.5 lux
  • <1 lux

The minimum illumination ratings of each of these cameras also varied, as follows:

  • 3.1 MP Arecont: 0.3 lux color/ "Day/Night: 0 Lux, IR sensitive"
  • 1080p Sony: 0.4 lux color/0.25 lux black & white @ F1.2
  • 720p Axis: 0.4 lux color/0.06 lux black & white @ F1.2

Did these lux ratings hold true? Ha!

Overview / Key Points

Image quality fell modestly when light levels were over 500 lux, but then started to drop significantly from the 300 to 3 lux levels. All of the camera experienced significant visual quality loss far above their rated lux ratings.

This image shows a comparison of seven decreasing lux levels, for reference (click for full size):

Side by side lux levels

Get Video Surveillance News In Your Inbox
Get Video Surveillance News In Your Inbox

Lux Level Examples

Official sunset on the day of this test was at 4:51 PM. We set our scene at 3:50 PM and began recording starting at 4:00. At this time, the day was still very bright at over 7,000 lux. All cameras displayed clear video with little noise.


At 4:10, the results were approximately the same, with light levels dropping to about 4,000 lux.


At 4:20, dropping down to ~3,500 lux, the 3.1 MP camera began to display a notably dimmer image, while 1080p and 720p models remained fairly close to earlier ones.


At 4:30 PM, at ~2,300 lux, results were similar to the previous, with little change in any image quality.


Little change at 4:40 PM and ~1,800 lux:


Around sunset at 4:50 PM and ~900 lux, close inspection reveals visible noise beginning in 3.1 MP and 1080p cameras, though not a lot:


5:00 PM, at ~330 lux, continues this with noise more pronounced especially in the 1080p camera. At this point, light levels were reduced by about 95% from our starting point of 7,000 lux.


As the sun continues to set at 5:10, down to just ~40 lux from ~300 lux ten minutes prior, we see a jump in visible noise in the 1080p camera. The 3.1 MP and 720p models remain pretty steady, with the 720p camera only modestly dimmer than when the test began.


At 5:20 PM and ~6 lux, the 3.1 MP model switches to night mode, the first camera to do so, at around 6 lux. Noise continues to increase in the 1080p model, and sharply increases in the 720p camera, both still in color mode.


At 5:30 PM and ~6 lux, the increase in noise in the 720p model begins to negatively impact our ability to recognize the subject. The 1080p camera has cut over into night mode at this point, and noise is somewhat lower than in the previous color image.


Noise continues to increase in the 720p image at 5:40, with the camera still in color mode at ~2 lux. While dimmer, the 3.1 MP and 1080p models provide approximately the same image quality as at the prior light level.


The 720p camera finally switched to night mode at 5:50, though noise remains constant. At this point, the sun had fully set, with outdoor lighting providing a constant 1-2 lux scene.


Switching off the outdoor lights, the light level drops below 1 lux. The subject is still detectable in the 3.1 MP and 720p cameras, though details cannot be made out. The 1080p camera provides no usable footage of the subject at this light level.

No outdoor lights


Key steps for setup included:

  • All cameras were reset to factory default.
  • Shutter speed was standardized to 1/30s.
  • Auto back focus was run on the 720p and 1080p cameras, but was not supported on the 3.1 MP model, which was manually focused.
  • Day/night cutover settings were left defaulted. This resulted in all cameras cutting over at differing times, as can be seen in the images.

The specific models used were an Arecont AV3116DN, an Axis Q1604 and a Sony SNC-CH240.

4 reports cite this report:

Camera Coverage Areas on Feb 10, 2016
How wide and far on an area can a camera cover?  A fundamental metric in analyzing this is pixel density (aka PPF or PPM). Three key factors...
Surveillance Commissioning / Install Checklist on Feb 08, 2016
This 60+ point checklist helps end users, integrators and consultants verify that installation is complete. It covers the following...
Avigilon Real Time Exposure And Gain on Feb 24, 2015
Knowing how exposure and gain impact image quality, bandwidth, and camera performance can useful in installation and troubleshooting. Very few...
How Well do IP Cameras Work in Low Light? on Jun 06, 2009
[2016 Update: Camera technology has changed substantially since 2009. The same pattern occurs but cameras are generally much better in low light....
Comments : PRO Members only. Login. or Join.

Related Reports on Low Light

IR Camera Reflection Test on Oct 17, 2016
All that might stand between good video and bad in your IR dome is a ring of foam rubber, keeping infrared from reflecting into the lens....
Bosch First Starlight 1080p Camera Tested on Oct 13, 2016
Bosch has released their first 'Starlight' 1080p camera, the Flexidome IP starlight 7000 VR. We tested this new dome against competitive super low...
4 Biggest Low Light Problems on Oct 10, 2016
100 integrators told IPVM what their biggest problems were with low light images. The most commonly cited themes were: Limited IR Range Uneven...
Samsung 'Q' Low-Cost 1080p Dome Tested on Oct 10, 2016
The most competitive part of the market is for low-cost cameras. Hanwha Techwin / Samsung has released its new low-cost Q series to go after this...
Axis 4MP Camera Tested (M3046-V) on Sep 28, 2016
Axis has brought 4MP to its camera line in the new M3046-V, the highest resolution model in their revamped M30 series. We bought and tested this...
History of Video Surveillance on Sep 22, 2016
This is a concise history of video surveillance covering the past decade.  The goal is to help professionals newer to the industry understand...
Camera Course September 2016 on Sep 15, 2016
This is the only independent surveillance camera course, based on in-depth product and technology testing. Lots of manufacturer training exists...
Hikvision 4K Camera Tested on Sep 09, 2016
Hikvision is the most common choice for low price entry level products but they are also competing with low light models, smart CODECs, WDR...
Pelco Optera 270° Camera Tested on Sep 06, 2016
Multi-imager cameras are typically 180° or 360°. Pelco has released a fixed 270° versions of their Optera intended to cover exterior building...
Panoramic Camera (Fisheye / Multi-Imager) Guide on Aug 31, 2016
Panoramic cameras, including fisheyes and multi-imagers, have become increasingly widespread, with most manufacturers offering fisheyes and an...

Most Recent Industry Reports

China Hot For Intelligent Video on Oct 27, 2016
The clear top theme at the 2016 Security China show has been intelligent video. This is a big shift from past years, and China's historical focus...
Flying Security Guard Startup Aptonomy Examined on Oct 27, 2016
Backed by Silicon Valley's biggest incubator and led by two robotics PhDs, Aptonomy is set to launch, what they describe, as 'flying security...
Bidding Divisions (08, 26, 27, 28) For Security Systems Guide on Oct 27, 2016
Navigating the world of system specifications and bidding work can be complex and confusing, but a standard format exists, and understanding it...
Dahua Says They Are Botnet Attack 'Victims' on Oct 26, 2016
'Victim' or 'accomplice'? Dahua has issued a new press release, referring to their products as 'victims' of the massive botnet attacks hitting the...
Milestone Launches $50,000 Developer Contest on Oct 26, 2016
Milestone wants independent developers to enhance XProtect. In a hackathon-style contest Milestone is soliciting developers to submit ideas for a...
Smart H.265 Samsung Test on Oct 26, 2016
In our first test of H.265 cameras (including Samsung and Vivotek), H.265 benefits were essentially nil, both versus leading H.264 cameras and,...
The Xiongmai Botnet 'Recall' Will Not Work on Oct 25, 2016
The Xiongmai 'recall' has been the topic of global news, following the unprecedented bot net attacks that use their equipment, among...
Hikvision Partners With Intel Movidius For Artificial intelligence Cameras on Oct 25, 2016
The world's largest camera manufacturer is partnering with the worlds largest semiconductor company to create a series of intelligent...
Intel Movidius Targets Video Surveillance Market on Oct 25, 2016
The most commonly used chips in IP cameras come from Ambarella, Hisilicon or TI. Now, Movidius, who Intel announced acquiring in September, is...
Favorite Access Control 2016 on Oct 25, 2016
Integrators told us "What is your favorite access control management software/system? Why?", and the responses are interesting indeed. While no...

The world's leading video surveillance information source, IPVM provides the best reporting, testing and training for 10,000+ members globally. Dedicated to independent and objective information, we uniquely refuse any and all advertisements, sponsorship and consulting from manufacturers.

About | FAQ | Contact