Testing: P-Iris Lens PerformanceBy: Benros Emata, Published on Jan 30, 2011
A new form of iris control, called P-iris, is emerging as a challenger to manual and auto iris lenses. P-iris can more 'precisely' control the aperture size than auto iris lenses. P-iris providers claim this can provide important quality benefits. For the P-iris 'pitch', see Axis's P-iris whitepaper.
While Axis introduced P-iris, many manufacturers are going to offer cameras with P-iris lenses. For instance, Vivotek and CBC have already announced P-iris offerings. From our conversations with other camera manufacturers, quite a number of other companies will release P-iris enabled cameras later this year.
We tested the Axis P1347 which includes a P-iris lens. In addition to testing that camera's general performance, we were interested in understanding how much of a benefit the P-iris lenses provided specifically.
We executed a series of test cases relevant to video surveillance applications comparing P-iris against DC Iris and Manual Iris performance.
For each test case, we used the same P1347 camera with the same settings, cycling through a P-iris, dc-iris and manual-iris lens. We then review the results, and provide a video analyzing the differences found (if any) for each of the test cases. We key in on what impact to video surveillance, if any, these differences might provide.
Megapixel lenses tested:
- P-iris: Kowa F1.6 / 3.5 - 10mm (LMVZ3510-IR)
- DC-Iris: Fujinon F1.4 / 2.8 - 12mm (FVL2812AI-MP)
- Manual-Iris: Tamron F1.6 / 4.5 - 10mm
Note: The P-iris lens was included with the Axis P1347. The Manual iris lens was set to fully open for all tests.
We performed tests across 6 scenes:
- Depth of Field
The series consists of three different daytime tests. The daytime/indoor shot is a basic indoor scene with a fairly narrow FoV. The daytime/outdoor is of a busy traffic scene during the day featuring a wide field of view and plenty of objects in motion. The daytime/outdoor/bright features a park scene large degree of direct sunlight into the camera lens.
Within the series is also a nighttime/outdoor and a low-light/indoor test. The nighttime/outdoor scene is identical to the daytime/outdoor scene, except that it is at night and the camera switches to b/w mode. The low-light scene is a 1.0 lux and 0.3 lux environment.
Finally, we performed a depth of field test. This scene is of a long hallway roughly 50ft in length. This test includes four (4) different scenarios. Two of them focused the lenses to a rear-field subject, and the other two focused on a near-field subject.
The combination of these tests provide a broad range of common scenarios to demonstrate the potential differences P-iris might provide.
- P-iris performance gains unlikely to significantly impact surveillance applications
- P-iris outperformed dc-iris and manual-iris lenses in majority of test cases
P-iris performance gains find greatest utility is use cases involving changing lighting conditions. For example a 24 hour day/night scenario will call for continuous iris adjustments. The additional precision provided by P-iris will slightly increase the likelihood of usable images for security investigations and live monitoring.
The overall benefits are modest and should not make a huge difference over other lens types. If a camera includes a P-iris, it a nice feature to have. However, it is unlikely that users should require or pay significant premiums for cameras that support P-iris.
P-Iris Technology Overview
While the P-iris approach has been pioneered in joint effort between Axis and lens manufacturer Kowa, it is not proprietary and has been brought to market by other camera and lens manufacturers.
Like a DC-iris lens the P-iris lens is an automatic iris opening and closing technique that essentially makes adjustments based on lighting conditions. The difference is that P-iris systems can use other methods, namely, AGC and exposure times, in concert with mechanical iris adjustments in order to maintain optimal iris settings (e.g. an iris not too open or not too closed).
The approach relies on two key elements:
- Precise knowledge of the position of the iris
- Software which enables balancing of iris position, gain (AGC), and exposure time
For example, if a scene is extremely bright, both a P-iris and a dc-iris will tend to close down to a narrow opening. However, the dc-iris cannot be 'told' to stop closing as the opening becomes so small that it begins to produce ill effects due to diffraction. The P-iris on the otherhand, can be 'told' to stop when it gets to that point. And, if the 'apparent' light is still too strong, it can invoke a reduction in gain or decrease exposure time.
Consider the opposite low-light scenario, where both P-iris and dc-iris tend to open up. Again, the P-iris is able to invoke increased gain and increased exposure time so that less opening of the iris is required than that of dc-iris systems. Theoretically, this helps the P-iris avoid optical abberations which can result in a loss of contrast, loss of depth of field, etc. The dc-iris, on the other hand is prone to open up too much causing a reduction in image quality.
Iris Related Image Quality Effects:
- Loss of contrast (too open)
- Loss of depth of field (too open)
- Purple fringing (too open)
- Vignetting (too open)
- Diffraction (too closed)
Claimed Benefits of P-iris:
- Greater depth of field
- Better contrast
- Improved image sharpness and clarity
- Optimizes megapixel camera applications
Many of the observed differences were small, yet when detected favored the P-iris. The most detectable differences were found in the Nighttime/Outdoor and Depth of Field comparisons.
Below we provide video screencasts of each of these scenarios.
- Daytime/Indoor: No differences
- Low-Light/Indoor: Minor differences
- Daytime/Outdoor: Minor differences
- Daytime/Outdoor/Bright: No differences
- Nighttime/Outdoor: Modest differences
- Depth of Field: Minor to modest differences
No material difference was found between the test lenses in this simple, well-lit, narrow field of view scenario.
In the 0.3 lux scene we discovered minor differences in picture clarity favoring the P-iris. Little to no difference was found in the 1.0 lux scene.
Little to no difference between iris types was found in this busy, well-lit, wide field of view scenario.
No material differences were found in this scene, which featured bright direct sunlight into the lens. Theoretically, we anticipated a possibility that non P-iris lenses might show effects of diffraction, displayed as streaking of light around very bright light sources, but did not see these effects.
During nighttime testing outdoors, slight to moderate benefits were seen. The P-iris controlled blooming from light sources better than DC and manual-iris lenses. Additionally, slightly better contrast was seen in the grayscale of the P-iris picture.
Effects On Depth Of Field
The results examined in the video are from a rear-focused shot, resulting in near field objects being in poorer focus. Our look into some very near-field objects reveals a better depth of field produced by the P-iris.