ONVIF Mega Test 2014

Published Apr 16, 2014 04:00 AM

In this groundbreaking report, we share findings of in-depth ONVIF testing.

ONVIF Real World Integration

We tested 14 camera manufacturers with 5 VMSes, performing 70 total integrations.

The camera manufacturers included: ACTi, Avigilon, Axis, Bosch, Dahua, DRS, Geovision, Hikvision, Honeywell, Pelco, Samsung, Sony, Wodsee (the super cheap Chinese offering).

The VMSes included: Exacq, Milestone, VideoInsight, Axxon and Wodsee. The first 2 are ONVIF Profile S conformant, the middle 2 are 1.x / archived products and the last is not officially conformant, just claims it in marketing. This provided a range of products / conformance to evaluate.

The testing checked (1) connectivity, (2) video streaming, (3) setting resolution, (4) setting frame rate and (5) enabling camera side motion detection.

ONVIF Test Tool

ONVIF has an official test tool that we accessed by paying $500 to become a member. This is the tool that each camera manufacturer uses to verify their conformance. We tested this with 15 cameras to determine how useful this tool is for verifying conformance and understanding problems.

ONVIF Device Manager

Finally, we tested the ONVIF Device Manager, an open source tool, to understand what value it offered and how it compared to the official ONVIF test tool.

Key Findings

Here are our key findings from this test:

  • Connecting a sampling of cameras from 14 different manufacturers to 5 recorders, the majority of cameras, 90%+, connected and streamed, regardless of ONVIF version and Profile conformance.
  • Profile S VMSs were far more likely to support motion detection, with Exacq and Milestone supporting 50% or more cameras. Non-Profile VMSs supported motion on nearly no cameras.
  • Support for changing framerate, resolution, and image settings (brightness, contrast, WDR) varied widely depending on VMS and ONVIF version used.
  • Compared to our 2011 ONVIF test, performance was radically better, with basic problems essentially eliminated.
  • In our tests using the ONVIF Device Test Tool, no cameras (out of 15+) passed a conformance test, even after following ONVIF recommendations and settings manufacturers used for their own conformance tests.
  • The tool itself is complex to operate, with little guidance on what settings to use (timeouts, event settings, etc.), and little clear feedback on what caused failures without intimate knowledge of low level camera and network performance information.
  • ONVIF Device Manager allows basic management, viewing, and configuration of ONVIF cameras, but does not provide conformance information and little in the way of diagnostics.
  • However, ODM may report features are available which are not, or are not properly implemented, limited its use to basic functions only.

Recommendations

ONVIF, especially the introduction of Profile S, has done much to increase camera/VMS compatibility, with most listed cameras and VMSs able to connect and stream. However, basic functions such as the ability to change resolution and framerate are not implemented across the board, and motion detection worked in few combinations, less than 25% of the time.

The ONVIF Device Test Tool is likely useful only to manufacturer technical staff, as the low level technical details required in order to properly run a test and interpret results are not available to outside parties.

ONVIF DM may be useful for basic camera configuration and feature checks, but does not perform any conformance testing, limiting its use to these tasks.

Unfortunately, the only way to truly test if a specific camera/recorder combination will work is to test it in the field. This is especially true of camera side motion detection, as neither tool discussed here can confirm that the VMS has integrated motion events specific to each manufacturer.

ONVIF Camera/VMS Field Test

We tested 14 cameras with 4 VMS systems and an NVR, varying in ONVIF conformance from Profile S to version 1.02 to claiming conformance without official listing. We tested the following parameters:

  • Does the camera connect?
  • Does the camera stream?
  • Can resolution and framerate be modified?
  • Can other image settings such as brightness, contrast, etc. be changed?
  • Does motion detection work?

The following charts detail each of these features for each VMS:

Connect/Stream

First and most basic, we tested whether the camera could be added to the recorder using proper credentials, and video streamed. In the most instances (90%+), this worked fine, with some glaring exceptions. The Wodsee NVR, which is not officially listed as conformant, crashed when the Avigilon H3 camera (Profile S) was connected. The Wodsee WIP130 would not connect to 3 out of 5 recorders.

Framerate/Resolution

Next, we tested whether each VMS had the ability to change framerate and resolution of each camera from the client. In Exacq and Milestone, the Profile S VMSs, most cameras could be changed, with few exceptions. Axxon Next, an ONVIF 1.02 VMS, was able to change resolution and framerate on all but Avigilon and Vivotek. Video Insight does not allow changing of framerate and resolution from the client, nor does the Wodsee DVR.

The non-conformant (but marketed as "ONVIF") Wodsee WIP130 could not be changed from any client whatsoever.

Motion Detection

Milestone integrates motion from most cameras in this test (9/14), while Exacq handles six (though Avigilon may be integrated by inserting a few lines in one file). In the version 1.02 VMSs, motion detection does not work at all.

Most surprisingly, the Wodsee NVR received events from Bosch, Dahua, and Honeywell cameras, though it is not officially listed as conformant.

Imaging Settings

Finally, we tested whether the VMS could adjust imaging settings such as brightness, contrast, etc., from the client. These features are purely optional in the VMS client, with Exacq, Milestone, and Axxon integrating the majority of cameras, while VI and Wodsee could not adjust any.

ONVIF Device Test Tool

This video provides a brief overview of setup and operation of the ONVIF test tool:

Full tests take about an hour or more in many cases, depending on how many features are being tested, timeout settings, and camera performance.

Test Tool Usage Issues

Though it appears straightforward, we found several issues impacted conformance testing and we could not get any tested cameras to pass a conformance test. We contact ONVIF with these issues and failed test reports for feedback, but even after following their recommendations, we found issues still remained.

These issues include the following:

Timeout and Event Settings

On the "Management" tab of the Device Test Tool are several timeout settings impacting testing. These settings include the time between the tool beginning tests, the time allowed for the device to reboot (occurs several times during testing), time the tool awaits for response to commands, etc.

Running a test with default settings resulted in large numbers of failed test steps. ONVIF pointed out that we refer to the "Feature List" XML document which is attached to each device in the conformance listing (see sample from Sony SNC-VB630 [link no longer available]) which shows what settings were used in the test. However, even though these settings decreased the number of failed steps, conformance still failed.

This screencast details how to find this XML document and adjust settings accordingly, and demonstrates how much manufacturers may differ.

Limited Feedback on Failures

Limited detailed pass/fail information is provided by the test tool, except in raw XML data for each step. This leaves users without intimate knowledge of the camera and the ONVIF spec unable to understand exactly why the test has failed.

In this screencast we show a failed test, which was run using recommended manufacturer specific settings and review the detailed feedback that is received on each step:

Test Environment

As per ONVIF, tests should be run on a closed network. Other traffic may cause unpredictable results, and increase latency, affecting the timeouts that were set in test configuration. Working DHCP and NTP servers must be available on the network, as the test tool uses these in the conformance test. Finally, ONVIF also recommends the test tool should be run as administrator.

ONVIF Device Manager

The ONVIF Device Manager (ODM) was created by a third party developer for basic viewing and configuration of ONVIF devices. This screencast shows basic use of ONVIF DM:

Unlike the official test tool, ODM does not perform any conformance tests. Instead, it is used for basic configuration and viewing of ONVIF devices. Some of this information may be quickly used to check if devices support ONVIF functions, but almost no diagnostic information is provided if specific functions do not work.

Methodology

We used the following VMS versions in this test:

  • Axxon Next: 3.5.0.819
  • Exacqvision: 6.0.8.58249
  • Milestone: 2013 (8.5e)
  • Video Insight: 5.5.26.4
  • Wodsee: V4.02.R11.63400083.12001

And camera firmware versions as follows:

  • ACTi D11: V6.06.16
  • Avigilon 2.0W-H3-BO1: 2.6.0.14
  • Avigilon 3.0W-H3-B: 2.6.0.18
  • Axis P3354: 5.40.17
  • Axis M3004: 5.40.5.4
  • Bosch NBN-733V: 5.90
  • Bosch NBN-932V: 5.90
  • Dahua IPC-HF3101N: 2.210.General 01.0R
  • DRS WatchMaster IP: v2.1.5770
  • Geovision GV-UBX1301: v2.11 2014-02-14
  • Hikvision DS-2CD2232-I5: V5.0.2
  • Honeywell HCW1F: 40.0.6
  • Pelco IME119: 2.0.0.11.8280-A0.0
  • Samsung SNB-5004: 1.13
  • Samsung SNB-6004: 2.22
  • Sony SNC-VB600: 1.12.0
  • Sony SNC-VB600B: 1.12.0
  • Sony SNC-VB630: 1.12.0
  • Sony SNC-CH140: 1.82
  • Vivotek IP8371E: IP8371E-VVTK-0100j
  • Wodsee WIP130-S20: 1.2.1.227383
Comments are shown for subscribers only. Login or Join