Robert, Do you have the specific model number for the Vivotek one? Vivotek has a lot of PTZs and the specs on them vary.
It could be that the new PTZ is simply slower than the old one (i.e., the internal mechanics of it).
If it's a latency issue, you might be able to test this by switching the camera stream to MJPEG to see if that's faster or to try to control it from the cameras' web interface (in case the VMS/recorder is contributing to latency). See this recent discussion on PTZ latency for more comments and ideas.
IPVMU Certified | 06/26/13 03:39pm
I would think latency would cause 'over-travel' and position errors, but not affect panning speed. Is it possible for you to connect a camera directly to a laptop and compare speeds vs. over the LAN?
What's the brand of fiber-optic to ethernet converters you're using on this? If there's any latency, that's the first place to look.
Is there a difference between going directly to the web page of the camera and the VMS? There is some inherent latency but it's usually not bad. It may be the Vivotek camera as well, we had issues with their cameras.
Jason, sometimes we have seen differences, with greater latency in the VMS, whether it's because the VMS application (server to client) introduces delay or because the VMS has an issue implementing the PTZ controls.
I like to try in the web page because it's the most direct way to connect and, presumably, the camera manufacturer knows how to control their own cameras.
Robert, thanks for the clarification. It looks like it's a latency issue then with manual controls. If it was a pure speed issue, the PTZ would be slow at top speed as well.
The next questions become:
- What's the CODEC? If it's H.264, try MJPEG to see if it reduces latency.
- What's the VMS being used? I'd check what settings they recommend to minimize latency.
I am presuming the latency is either coming from the camera side or the VMS. However, as Seth point out, it could come from the network, though I suspect, statistically, it's a camera or VMS issue more often.
I don't know about Vivotek but in the Dahua IO OTZs you can adjust the speed of PTZ actions from 1-8. For most of our full size PTZ models a speed of 8 can pan 180 degrees in 1 secondbut at a speed of 1 the it takes over 20 sefinds. In our case the speed can be adjusted on the fly on both the web interface and the NVR UI.
Sorry for all the spelling mistakes in the last post. The point I was trying to make is that there are most likely PTZ speed settings that can make the Vivotek move faster.
Bohan, I don't think speeding up the PTZ movements will solve this. From what Robert is describing, it's the latency that makes it hard to adjust in incremental steps (tilt over, then slightly pan, etc.). If all you wanted to do was make one move faster, speeding up the actions would help but if the manual commands are more complex, you still are dealing with the latency holding up / messing up actions. That said, it's worth trying the setting out but ultimately, they likely will need to find the source of the latency and correct.
Thanks all for your replies. John, our technician will be here at the Mall tomorrow at 11 am to address all of the questions. We understand the converter is the "Black Box, LP5500A-MM-CC. The tech is scheduled to change the CODEC to JPEG because the current setting is H-264. We'll also identify the VMS at that time and its settings.
Update: In response to all of the fine suggestions and questions above:
The camera ptz speed is significantly faster from the web page; from the vms, the camera ptz speed is about 25% slower; thevms manufacturer is Exacq Vision 5, Client Version 126.96.36.199302, image rate is 30 fps; the cameras are Vivotek SD83X2. Our techs changed the settings of CODEC from H264 to JPEG, no improvement. The converter info is above. The Tech company has discussed the issue with the manuafacturers of the cameras and the vms at least twice with no resolution. This system is under one year old with the latest technology. Unfortunatey, we were not able to chase car theives in a speeding car (like we used to with anolog) last week. Any more suggestions.
Is the speed from the web page good enough to control the camera to, for example, chase car thieves? Do you know how long the delay is from initiating the controls to when you see a response? 1 second? 2 second? longer? shorter?
John, our tech just checked our system with our top gun guy who catches all of these car thieves from time to time. PTZ Results:
- Web Page - Instant controls with 360 rotation in 4 seconds.
- Server - .5 Seconds to Start with 360 Rotation in 6 seconds.
- Kalatel (through the New Server) Instant controls with 360 rotation in 3.5 seconds.
In all of the three instances, only the Kalatel cameras are able to Pan, Tilt, Zoom AT THE SAME TIME. The joy stick is not effective at all and we rely only on the mouse controls for the web page or the server that allows us to either pan or tilt, or zoom but not at the same time.
Since the old Kalatel cameras (and its joy stick) operate only the inside cameras(through the new Server), we are unable to chase the car thieves with the current set up for the new exterior cameras with or without the server.
Hi Bob, Ok, that helps. We will contact Exacq and ask them for recommendations on what to try / optimize.
Btw, what are you using to control the PTZs currently. You mention a joysick. Is that a traditional CCTV controller or a USB joystick or?
Also, Exacq has a PTZ keyboard controller. Have you looked at that?
The Kalatel PTZs are controlled by the KTD 405 Controller and the new cameras (Vivotek) are controlled by CH Products IP Desktop USB.
I can recall some months ago, our Tech Co. did bring in the Exacq PTZ keyboard controller without making any improvements but this was before we recently changed settings from H264 to JPEG..would this make a difference if we tried the Exacq keyboard controller now?