Testing FLIR IR PTZ
FLIR's integrated IR PTZ, the DNZ30TL2R claims a whopping 150m (~500') IR range and HD resolution. Distances like these have historically been possible only with expensive high-end positioning systems which required expensive illuminators and motorized lenses.
We bought one of these new integrated IR models and tested it against an Axis Q6045-E, a common 1080p HD PTZ, to see how it matched up.
Inside we look at these questions:
- How far can this camera see, day and night?
- How easy or difficult is it to track subjects?
- How does it compare to standard non-IR HD PTZs?
- IR range exceeds 150m (~492') specified range, reaching objects ~700' from the camera.
- At 0.3-0.5 lux, IR PTZ was easily able to provide recognition details of subjects at ~500'. The Axis 1080p non-IR PTZ was barely able to detect our subject at ~50'.
- IR illuminators were responsive to zoom changes, turning on and off and adjusting power to fit the angle of view.
- IR overexposure reduced details of both human subject and test chart at short ranges (<100').
- Due to their greater reflectance, license plates were overexposed at all ranges.
- Daytime images were moderately clearer than non-IR Axis PTZ, with better subject details and more legible lines of the test chart at long distances.
- When tracking moving subjects at night with IR on, autofocus failed to work properly while the camera was in motion. After stopped for a few seconds, image focused properly.
- FLIR DNZ30TL2R: 2.400.GM00.0.T.03.01.04, build: 2014-11-18
- Axis Q6045-E: 126.96.36.199
Here are our key findings from this test:
The FLIR DNZ30TL2R sells for ~$1,400 USD online. This pricing is quite inexpensive compared to most HD PTZs (~$2,300 average) which typically do not include integrated IR.
If surveillance of subjects at long range is the main goal of the camera, the DNZ30TL2R offers some of the longest IR range we have seen, at a price similar to many SD PTZs. However, at closer ranges (<100'), IR overexposure is common, so if subject details are a key concern, we do not recommend it. Finally, if license plate capture is a requirement, the DNZ30TL2R is unsuitable, as the camera was unable to capture at any range due to overexposure.
This video reviews the physical construction of the camera:
Similar to the Dahua HDCVI IR PTZ from past tests, the FLIR DNZ30TL2R did not properly autofocus when tracking moving subjects. Only after stopping the camera for a few seconds did autofocus properly resolve. This is shown in the following video:
We tested the cameras outside in a parking lot with minimal outdoor lighting, varying between 0.3-0.5 lux at night. approximate distances are shown in this overhead view:
IR Range Performance
The DNZ30TL2R includes 12 IR LEDs: two wide angle, smaller LEDs used when zoomed out, and four larger, higher power LEDs used for long ranges. These LEDs turn on and off and adjust power depending on zoom level, seen here:
At ~500', our test subject is clearly visible in the FLIR IR PTZ, though without usable facial details. Additionally, 2-3 lines of the test chart are legible, while the Axis non IR model is unable to produce any usable images. Since it is extremely reflective, the license plate is extremely overexposed and impossible to read.
At 250', subject details improve in the FLIR PTZ with some usable details such as facial structure and hairstyle. The test is moderately overexposed, limiting legibility to line 5/6. Again, the license plate is overexposed and illegible.
At 100', facial details are moderately clearer, but overexposure worsens, with lines below 2 washed out.
Finally, at close range (~50'), our subject's face is overexposed, reducing visible details.
At ~450' range, due to its longer zoom lens, the FLIR IR PTZ provides moderately better details than the Axis non-IR model, with 1-2 more lines of the chart clearly visible.
The same was true at closer rangest as well, though far less pronounced, as in this image, at 250':
And finally, at close range details are similar, though smaller lines of the test chart are more difficult to read in the Axis camera, due to slight overexposure.
The following firmware versions were used in this test: