WDR Camera Shootout 2011

Published Feb 05, 2011 05:00 AM

One of surveillance's biggest challenges is bright sunlight, especially when that sunlight shines into a building's entrance. Sunlight tends to wash out images, often making it impossible to determine the person entering. Unfortunately, capturing people as they enter your building is one of the best and most effective ways of conducting video surveillance.

Camera manufacturers offer a variety of technologies to rectify this problem. The general name given to them is Wide Dynamic Range or WDR. By design, WDR should provide accurate image capture of both dark areas (e.g., shadows covering a person's face) and adjacent bright areas.

However, there are no standards nor easy way to figure out if one WDR function is better than another. While there is a metric for measuring dynamic range (dBs, e.g., 65 dB or 110 dB WDR), each manufacturer can choose their own measuring methodology so comparisons are basically pointless. (This is similar to the problem with determining low light performance. Manufacturers provide minimum illumination measured in lux but using a variety of conflicting metrics.)

In this report, we conducted a series of tests to better understand WDR performance using 3 well known manufacturers - Pixim, a chip manufacturer specializing in WDR, Axis and Sony. For Pixim, we tested 2 cameras - a March dome using the new Seawolf Pixim chip and an ioimage camera using the traditional Orca Pixim chip. For Axis, we used the P1347. For Sony, we used the SNC-CH140.

[UPDATED: We have done additional WDR shootouts focusing on megapixel cameras.]

We wanted to understand how different WDR featuresets and varying resolutions impacted overall image usability.

Inside the PRO section, we break down comparative performance and provide recommendations on tradeoffs and options.

Key Findings

Let's start with our overall key findings from our testing:

  • All WDR cameras provided notable increases in WDR performance over a high quality control camera
  • Not only did megapixel not hinder WDR 'ability', the combination of WDR and megapixel provided a superior image
  • Increasing resolution from SD to Megapixel provided substantial image quality gains even in tough WDR conditions; However, a 1.3MP camera (Sony) outperformed a 5MP camera (Axis) showing resolution benefits have limits
  • Sony delivered the most substantial image quality benefits and overall highest quality images of test subjects
  • Axis and the Pixim powered cameras provided similarly moderate WDR benefits; Axis's increased resolution delivered a better overall picture
  • SD cameras exhibitied slightly better low-light performance than megapixel
  • While Pixim does not have an option to turn off WDR, Pixim camera's low-light performance was still better than Sony 1.3MP even with WDR off

Recommendations

We recommend cameras with WDR features to improve image quality in scenes with direct sunlight. Additionally, we recommend the use of megapixel WDR cameras to maximize the overall image quality benefits. Specifically, the Sony's performance stood out as providing the strongest WDR benefits. However, note that Axis and Sony only support WDR in their premium series of cameras so do not expect to achieve such benefits from all Axis nor Sony cameras. For Axis, WDR is found in the P series (not the M) and for Sony, the V series (but not the E nor X).

The Pixim cameras clearly showed WDR benefits and the best low light performance overall. However, Pixim's limitation to Standard Definition resolution clearly limited the overall image quality. While the WDR benefits looked to us to be similar to Axis's, the higher resolution of the megapixel camera provided a much better surveillance image. Finally, while Pixim has a reputation as being the WDR provider, the Sony's camera performance looked superior.

Cost Considerations

Cameras with WDR feature sets will generally cost a substantial premium over non-WDR cameras. All else being equal, expect to pay $100 - $200 more (online price) for IP cameras with WDR (the variance depends on the supplier).

For wider Fields of View, the modest HD price premium over SD warrants consideration of megapixel. The price difference between Pixim SD and Sony's/Axis' 720p WDR cameras is around $100 to $150. Representative SD cameras featuring Pixim's WDR average roughly $675 online, while 720p Sony CH140 and Axis P1344 cameras average $800. As such, the cost to go from SD to HD/MP (even w/ WDR) is fairly small.

Pixim Representatives:

It's interesting to point out that the Axis P1344 is $759 online, while the Sony CH140 is $845 online. The trade-off for Axis' lower price point, is that the Axis does not feature as strong a WDR nor an automatic method to turn it off during low-light conditions. Thus, 24 hour surveillance applications, without low-light conditions (e.g. indoor with controlled lighting, but daytime requires WDR), would be a good fit for the Axis P1344.

While narrower FoV applications will look at price/performance trade-offs involving SD vs. HD, wider FoV applications will generally look at the entire sprectrum of HD to higher MP. Sony and Axis premium (w/WDR) 3MP camera's, the CH240 and P1346 are both priced online around $1100. The 5MP Axis P1347 used in our testing is $1450 online. Sony does not currently have a 5MP camera - w/ WDR or otherwise.

Test Methodology

We selected four (4) cameras for WDR comparisons. Each camera is equipped with a different implemenation of WDR. For example, the Axis P1347, features Axis' own unique brand of WDR called 'Dynamic Contrast'. Both the March CamPX and Ioimage wdc-100dn feature a Pixim chip for WDR functionality. However, it is important to note that the CamPX is based on the latest Seawolf while the wdc100dn uses an older Orca series Pixim chip. The Sony CH140 features Sony's own 'View-DR' (offered in numerous other Sony models).

We used the following four (4) WDR cameras in our tests (w/ max resolutions):

  • Axis P1347 - 5MP
  • Sony SNC-CH140 - 720p
  • March CamPX Microdome - D1
  • Ioimage wdc100dn - 4CIF

For all cameras, max resolution and frame-rates were used. Other settings were left at their defaults. Additionally, each camera's lens is adjusted to the same approximate FoV. Testing also includes a non-WDR IQEye 4 series camera, as a control.

The cameras/WDR are compared throughout three test cases:

  1. Indoor WDR (Hallway)
  2. Outdoor WDR (Park)
  3. Low-Light WDR

The first scenario is of an indoor hallway. The hallway is dimly lit, providing a 'dark' area (~100 lux), and the propped exit door provides a 'bright' area (~1300 lux) to the scene. These elements produce a very challenging WDR scene under which to make comparisons. We use the IQEye 4 series camera to provide a non-WDR point of reference.

The second scenario is of a park. The scene is in general, very bright (~12,000 lux). However, there are patches of shade (~4,000 lux) provided by trees and a small building giving the scene some lighting variation. We are able to turn off WDR in both the Axis and Sony cameras to provide a non-WDR point of reference. Note that the CamPX and wdc100dn do not support disabling of WDR.

The third scenario is a simple indoor low-light scene. The Sony and Axis cameras are tested at the 1.0 lux and 0.3 lux level with WDR-on and WDR-off. The March and Ioimage cameras are tested at the same lux levels but only with WDR-on (WDR-off not supported).

WDR Comparative Results/Performance

In the next 3 sections, we show and explain the specific differences we found. You can see the exported video for yourself: IPVM WDR Comparison Video Samples [link no longer available] (~500 MB).

Indoor WDR (Hallway)

The four WDR cameras demonstrated key differences in performance compared to the non-WDR control. For example, in the control the subject's facial details were very difficult to examine, while positioned in the highly bright area (1,300 lux).

Of the four WDR cameras Sony's performance indicates best WDR. In contrast, Axis, March, and Ioimage produce evidence of fair to moderate WDR capability. However, despite fairly similar levels of WDR between Axis, March, and Ioimage, the Axis provided overall better image quality (yet not as good as Sony's). Here, it is apparent that higher resolution is having a large effect on the video.

The Sony (1.3MP) and Axis (5MP) comparison underscores the importance of both WDR and resolution in determining overall video quality - especially under difficult lighting. In this interesting comparison, Sony's image quality was better than Axis' despite Sony's much lower resolution. The superior WDR on the Sony appears to 'compensate' for its lack of resolution with respect to the Axis camera.

Outdoor WDR (Park)

Under the bright lighting of this test, the Sony and Axis cameras demonstrate better picture quality with WDR turned on versus off. Further, with WDR-on the Sony and Axis appear to illuminate 'dark' areas to a greater degree than the March and Ioimage cameras.

As this test shows, a wider field of view will benefit from higher resolutions. While both the Sony (1.3MP) and Axis (5MP) WDR perform well in the test, Axis shows slightly better image detail.

Low-light WDR

Both the Sony and Axis exhibited slight to moderate differences in picture quality with WDR-on vs. WDR-off. At low-light levels of 1.0 lux and below, WDR tends to detract from the overall image quality in the Sony and Axis. Interestingly, the Axis WDR tends to 'wash-out' the scene while the Sony WDR tends to 'darken' it.

The best overall low-light performance (w/ WDR-ON) is demonstrated by the March CamPx (Pixim Seawolf) followed by Ioimage's wdc100dn (Pixim Orca). The Axis P1347 and Sony CH140 tended to produce lesser quality images than the Pixim based cameras with their WDR-ON.

In the low-light scenes, megapixel cameras (Axis and Sony), with their WDR off, competed better against the standard resolution (March and Ioimage) cameras. (Note, that the ordinal positioning in terms of low-light performance remained the same, with March best, Ioimage second, and Axis/Sony trailing Ioimage by a slight margin). Such findings may be congruent with the belief that SD has advantages over MP in low-light imaging, but the magnitude of difference was less than dramatic. Indeed, when WDR is turned off in Axis and Sony, the difference between low-light performance among the four cameras narrows perceptibly.