Network Optix HD Witness

I have been trying Network Optix and think the UI is really next generation - like I cannot imagine a better way to do things - its so simple.

However I have only played with Milestone, DVRs, NVRs and smaller VMSs so can't really justify the above statement. However from the Exacq screenshots and review I do not see that much user friendliness.

Have any of you guys tried HD Witness - I really would like a general opinion of what other people think just to make sure I am not somehow just chemically aligned and naturally biased towards this particular VMS.

The UI is definitely different. The ability to move, stretch, manipulate and rotate video feeds is unlike any VMS I have ever seen. That's for sure.

Whether or not it is simpler or more user friendly, obviously depends on the user. We have heard similar praise from others about its UI. On the other hand, it does violate a number of VMS conventions (which can be argued is a good or bad thing). Exacq definitely does not have the video manipulation/control that HDWitness has, but it follows well established conventions and, for what it does, it does it with minimal steps/complexity.

If you are looking for a product that has a different and 'sexy' UI, HDWitness is it.

There are some other things you should keep in mind:

  • The admin side of the VMS / client is fairly minimal. It also lacks a true admin/configuration section typically found in VMSes.
  • It's one of the newest VMSes on the market, which is good for those looking for differentiation. On the other hand, you should carefully check that it has the low level features that you need as you may need some specific 'bells and whistles' that Milestone has done years ago, but NetworkOptix has not gotten to yet.

Looks like a hybrid of Ocularis and ISScctv.. I'm going to download the free version, but from the video examples, its really not super impressive or game changing.

Hi Sean

To be honest I do not really agree with that statement: In OnSSI and ISScctv, the Live, Browse/Playback, Export and Timeslice (thumbnail) modes are all rigidly separated just like other VMSes.

For me the advatnage of HD WItness is that all the above functions are on the one and only screen in the GUI - with no extra buttons cluttering the whole screen.

Plus the zooming is implemented in the best way I've ever seen, ditto for the timeline zooming and the playback speed adjuster. Everything is just natuarally done using the mouse wheel.

Thats fair Bohan, like I said, that was only from watching the sample videos, I do plan on working with the actual software to see what it is like to interact with.

I recently tool at look at HD Witness and I was impressed with the GUI. It defiantly has some features that other more mature VMS systems should be looking to implement. Where HD Witness lacks is in back end with the Administration and setup. I am also not sure if I like there camera auto discovery. I didn't see anyway to add a camera manually so I am not sure what happens if the system doesn't auto discovery which could happen on larger installs where cameras can be spread across different subnets. Also the user level and permissions is lacking and doesn't have enough refinement. They defiantly have a ways to go to be a contender for large commercial or enterprise level installs but for small commercial jobs and home installs it is a nice system.

IPVM Edit: Author is a Network Optix employee.

It is a fairly easy process to add a camera manually. Simply right click on the server, select "Add Camera(s)" then select the IP or IP range you have the camera on, add in the login credentials and click "Scan". Once the scan is complete you can select the cameras you would like to add by checking the box next to them in the window below the scan. The camera is now added to the server you ran the scan from, if you want to change the server to another simply drag and drop the camera to the desired server.

Edit of IPVM edit... Actually John Alemparte is a former employee of Network Optix and a current employee of Digital Watchdog. Digital Watchdog OEM's a version of HD Witness in North America called DW Spectrum which the sell pre-loaded on their new NVR Server line through major distribution as well as sell as a software-only solution through the same.

Disclosure... I am a Network Optix employee.

Just took a look at a few YouTube video's of the software and I did like a few things and thought it was a unique way for a VMS but really not a unique way for the way we search are basic day to day data today. I was intrigued.

John, the need to right click on different elements for different admin tasks makes it challenging and confusing as you have to experiment and remember where to right click for different actions.

By the way, what if the camera does not come up on the scan. Can a user manually add a camera even without a scan?

I love HD witness too. I love its simplicity, its very easy to use. One of the best I have seen.

But I agree with Duncan, it lacks in backend administration. The fact is, you cannot manually add a camera. Doing a scan only picks up cameras on the local network. So what do you do if you want to add a remote camera off the local network? It basically relies on discovery to add cameras which I dont like. I think it could be really huge if they addressed these issues.

Its really sleek and simple and that is what people are looking for. You can literally pick it up without reading any instructions. I would like to see them keep the simplicity but add advanced features such as email alerts, manual camera add, and many other advanced features that you see in the bigger VMS'es.

Sean N and Duncan...

As you can see in this video promo of nVision, an OEM of HD Witness in Asia, the manual addition of a camera is quite easy and doesn't require doing a full scan. If you know the IP, you can drive right to it. That portion of the video is at time mark 1:25 in this video:

We will be releasing some new videos in anticipation of the new features we're about to release in V1.5 including a full Events/Alerts/Notification system as well as some other new enterprise and client features.


Any plans on releasing an iPhone app? What are you plans for remote viewing? Will you be implementing multistreaming or transcoding?

From our experience in Asia - where we face some of the most challenging support environments on the planet due to a general lack of expertise in working with and supporting IP video - the automatic camera recognition is a huge advantage.

As Nathan pointed out as well, the manual camera addition is easy and simple to use and works very well. As most camera manufacturers have their own camera finder tools we typically use their camera finders to quickly locate new cameras which have default IP addresses that are not on our local subnet, then change their IP address to match our network settings - at which point it's easy to manually add them with their newly addresses.

We use nVision (nnodal's OEM version of HD Witness) to stream and record from a variety of local camera manufacturers in China (Dahua, Hikvision) and the built-in ONVIF support is great. Works *most of the time with cameras we've never seen or tested before as long as they are truly ONVIF compliant.

The Events/Alerts/Notification system which we are pushing out along with Network Optix in version 1.5 adds another key feature set that has been missing in the platform - and is executed in a manner which is simple to configure without any guidance.

Here's a sneak preview of nVision 1.5:

Finally - I think one thing to keep in mind is that this is a brand-new piece of software. When comparing it to Milestone, ONSSI, Exacq, or any other major brand keep in mind this software is 2 years old. Take into consideration how much progress has been made by the Network Optix core engineering team in such a short amount of time - and throw in the radical new user interface - which we call "The Grid" - and what you have is an IP Video Management platform that requires little to no training for users, consolidates monitoring and admin functionality into a single user interface, and follows the latest in IP Video standards with the ONVIF S Proflie.

It's not going to be perfect for every project right now - but at it's core it seeks to do one thing - simplify IP video management - and I think it's making great strides in that direction.

John, Sean

(I am a Network Optix employee as well ).

->Can a user manually add a camera even without a scan?

Yes, he can. Scan It's just for your convenience. With scan you can add not only single camera( you can do it without scan) but add 20 cameras from other network with one click. So manual camera addition is 100% here. You even do not need to specify which camera are you adding( manufacture, model etc ) it auto detects everything for you if you just provide ip or ip range or even dns name.

Hi Nathan

Nice to have you around

Ok sorry, sorry for the misinformation, i do not have the latest version. Was not aware that manual add wasnt there. Glad to see it is now. I will play with it tomorrow. Would love to offer this to my customers as it has the best interface I have seen.

Hi Duncan,

The iPhone app is under construction as I type. The media server needed to be upgraded in v1.5 to support the client the way we wanted. It will be uploaded to iTunes for review in the coming weeks and will certainly be on display at ISC.

As for multi-streaming and transcoding... We already do. We dual-stream from the camera by default. Works with almost all dual-streaming Onvif Profile S cameras and several others we've done direct integration with. Full resolution and a lower subset stream by default. Both are used extensively in our CPU and bandwidth optimization engine as well as with our server-side motion calculations and smart searches.

Transcoding is currently only utilized when streaming from our media server to our Android and our Web client. We utilize WebM Video technologies for those which allow dynamic transcoding between 120p and 720p on the fly. Both are core components of how we do our most advanced capabilities in conjunction with GPU utilization and some other cool things we do.

Our R&D capabilities far outweigh our marketing capabilities as a software startup. We're going to be making some more detailed and compelling video demonstrations over the next few days and weeks.

Hi Tony,

Does the on-board server based Network Optix motion detection processing work with Dahua cameras? When I first tried it several versions ago, it would not work, but of course I havent tried it in a while.

Hi Sean -

We have only tried it out with 4-5 models so far - but the motion does work well. I can send you some video tomorrow with motion search enabled on the Dahua cameras if you'd like to check it out.

Hmmm, I downloaded this last night, installed the whole thing on my desktop... scan wouldn't find any of my cameras, nor would it find an IQ511 when I entered the IP.

One thing I noticed: the "localhost" server showed up in the client's tree as - my desktop has multiple alias IPs for various reasons (,, and, which is my LAN's primary subnet) and the cameras are all on In theory, the server should still be able to find the cameras, unless it's bound ONLY to, and I couldn't find anywhere to change the server's binding.

I'll probably have more success if I try it on a more "normal" machine, but this isn't a very auspicious start.

Yeah that'd be great

Hi Sean -

Here you go - a quick screencast of a Dahua IPC-HD2100 camera with server-side motion enabled.

The Motion detection works with any video source as it is software based.

When we say 'software', we mean server or camera based? I assume Sean means camera based motion detection that HD Witness can recognize and integrate with. Yes/no?

Hi Bohan,

That's not entirely correct. For the server-side motion detection and advanced motion search features to work, the camera needs to be fully capable of dual-streaming. It is part of how we do the advaned motion calculations. Sometimes the cameras do dual streaming but with a non-traditional approach like Arecont which can only provide a second stream at 1/4th the resolution of their primary stream. In Arecont's case we take motion from the camera and don't do it server-side.

However you're correct that most dual-streaming cameras, especially Onvif Profile S cameras are fully enabled for dual-streaming and server-side motion calculations in our system. I just wanted to clear up that it's not all.

Hi Matt,

Server does not bind itself to a specific NIC on a machine. It works with every NIC and looks for a cameras in each subnet you have. However not all IQ cameras are supported.


Thanks Sergey... I have an IQ511 and IQ753 (both MJPEG), and a couple Dahua cams (H.264). I have some HIK and Axis cams I can test with as well, but those are the ones that are up and running right now.

I don't want to turn this into a support thread, though - I was just reporting my initial impressions. I'll deal with it more when I have the time.

Hi Nathan

I am a bit a confused by your statement regarding the need of dual-streaming for motion detection to work. The purpose of dual streaming (AFAIK) is to have 1 HD stream and 1 SD stream so that your can do motion detection processing on the SD stream to save CPU cycles while still recording all the available pixels captured by the camera right? (Its also useful for remote viewing via cellular/WAN) So isn't it a good thing that Arecont's sub stream is 1/4 resolution (75% saving in CPU cycles)? For example all our cameras have a main stream that matches the sensor's pixel count (1-5MP) and a sub stream that is either D1 or CIF. Currently we found that HD WItness 1.4.2 (1.5 beta) will configure a CIF stream at 7FPS but I presume that this will become a bit more configurable in the final 1.5 version since CIF/7FPS/128kbps can be rather noisy and blurry.

Hi Bohan,

We can discuss this a bit more offline as there is still a lot of stuff about our platform you haven't been exposed to yet. As far as Arecont cameras, you're right that it's a 75% savings but think of it this way... If you're using a 10MP Arecont camera recording primary stream at full resolution, would you still want to have to pay the storage price for having it's second stream at 2.5MP? That's higher resolution than a 1080p camera just from the second stream. Same principle goes for their 20mp panoramic camera. Now the second stream would be 5mp.

There are some other components we can talk about offline if you'd like. Hit me up on Skype. My ID is: MegapixelNews

Thanks for the detailed video Tony!! Now I do remember when I tested it a few months ago, i was able to get it to actually detect motion and do smart searching but I was never able to record only on Motion Detection. Basically it was doing like your video above. It would have a solid green bar with red specks of motion on the timeline. So basically it was still recording 24/7 but just showing when there was motion events on the timeline.

But what I could not get it to do is record only when there was motion detection. We wanted to do this to save on hard drive space. Where you ever able to get it to work like this?

No problem, Sean.

Motion-only recording works well with the camera we have in place.

Took another screen recording (using the nVision native screen-recording application) for you, below:

Thats great! I am going to play with it alot more this week. Would love to offer a nice software based solution to my customers.