MaxIcon

Depends on their definition of full bit rate, I suppose. If you want to use multiples of 1024, this would require 4096 kbps for 8 cams to be under their spec. Bumping up to the next step of 5120 kbps, that would take you to 41 Mbps I run my 1080p cams at 8192, and 8 of those would be 65 Mbps. The Hik 1080p bullets will let you set a bit rate all the way to 16,384 max(!), which would be 131 Mbps for 8 of them. I'm not saying you need those bit rates, but the 16 kbps is full bit rate on the Hik 1080p bullets. It may be he meant the NVR won't let you set the cameras for higher than 4096 kbps, so that would be full bit rate of the NVR, not the cameras.

I had a Hik 1080p bullet that I updated to 5.0.0, and ran for a week or so with limited changes to the config pages. After leaving it off for a few days and re-connecting, it prompted me to download a new webcomponents.exe, which gave me a handful of changes to the config pages that definitely weren't there before. The only thing I can think of is that I hadn't restarted both the camera and the PC at the same time, though I can't recall if I powered down the camera, or just rebooted it. Here's a post with the US download location: viewtopic.php?p=217811#p217811 One good test of internal IR reflection is to take the camera out in an open space at night and point it straight up into the empty sky. If there's internal IR reflection, you'll see it on the image. If you don't see any, but you see some when it's mounted, you're getting IR reflection from the local environment.

Heh! Makes you wonder what they do with all this old gear people return.

Does anyone know how to unpack the Hik firmware files? You can unpack the Dahua files by renaming them to zip and ignoring the header error, but this doesn't work with the Hik. If you could extract the files, it would be easier to tell the difference. That said, my guess is that the version - V4.0.9 - is the important part, and the date change indicates the customizations for the resellers (adding the Swann or Lorex name, whatever). Just a guess, though. At 1/12 or 1/6 sec, auto exposure was still taking place. I tested this two ways - setting it at 1/6 in the day and seeing that the exposure didn't change, and walking in front of it to make sure there was no motion blur. If it were fixed at that setting, it definitely would have been washed out and blurred at walking speed. I also set progressively shorter exposures to see what it was actually using, but didn't write down the speed it changed at. I think it was around 1/500 or so, though. I'd like to see all these cams report the parameters on the web page - shutter speed, bandwidth, frame rate, etc.

Sounds like the high power IR is a main difference here. While longer range is always good, there are a couple of potential downsides to high power on-camera IR. - If it's not smart IR, you're more likely to get washed-out images of things up close, at least until the exposure changes to compensate. Some of these cams are not very good at exposing for a small, bright area (like a face 5' away). If you're mostly looking at stuff further away, it's not as likely to be a problem. The only way to tell for sure is to test it out. - If the focus depth of field allows it to focus up close, you'll see an increase in dust, small insects, and other small stuff lit up close to the lens. The Dahua bullets are more prone to this than the smaller Hiks from what I've seen. When I walk past my Dahua 2100 at night, it's like a snowstorm for a minute or two. - More power means more heat, though that's not usually a big issue at night unless you're somewhere like Arizona. - More IR isn't generally good for reading license plates, as they're very reflective, and will wash out even with normal IR. You can see that in this post of BW's pics: viewtopic.php?p=222014#p222014

That first review claims the daytime image is terrible, which doesn't make sense. These cams have pretty nice daytime images. Most people seem to be complaining about the noise or networking problems. Gotta feel a bit sorry for the tech support of a company selling network-intensive products to the masses.

Here are the headers on the Hik/Swann. On the one with the wires attached, the red and black are to the DC power jack, and the rest are to the network cable. Yeah, it would be a pain to make a longer adapter for this. Here are a couple of boxes without conduit, because the conduit I had was too stiff. I since got some more flexible conduit, but haven't gotten around to taking it all apart. The Y-cam was my first IP cam, and started this whole transition from analog to network cams. Y-cam mount watching the front door, with both DC and network cables. A bird built a nest on it last year, so there's a bit of staining, but no problems from any of that. And a Vivotek IP8332: Conduit will definitely help the security. Most of my cables are run outside and unprotected, because my attic is really hard to pull stuff through, so I gave up after the first few cables. The problem with the short pigtails, like on the Dahua and Hik, is that you're really limited on where to position the box.

It looks like it could be connected to large spots of reflected glare off the cars, but you'd think the glare directly from the early sun would cause a similar problem. One thing worth trying with the BLC would be to a fuzzy area, like the trees to the right, and see if they come back into focus with the exposure control centered on that area. That could help figure out if it's a processing issue. I'm still pretty suspicious of the Dahua processing. I have one pic where it turned a feature in my back yard into a bunch of blocky pixels due to the bright/shade softness thing, and I haven't been able to figure out why it has so much trouble in certain circumstances.

No, it's got an inline header, like most cameras these days, especially the compact ones. You could make a new connection to replace it if someone figured out what type of connector it is. I'll post some pics up later. If I did this, I'd probably put a longer network cable on - maybe 6' or 10' - which would let you run it inside through a wall or whatever, and would still be short enough to be manageable if you needed to take the camera down. The Vivotek has a standard (though metric) cable gland, so you can actually connect conduit directly to the camera. It's also got a standard camera screw mount, which I also like more than these dedicated mounts, but they're the standard on inexpensive cams these days too.

On both my 2100 and 3300 bullets, the way to have a minimum exposure and variable exposure setting is to select either Low Noise or Low Motion Blur, and set the range from 0 to 17 ms (for 1/60 sec) or 33 ms (for 1/30 sec). Both low noise and low motion blur come up with the same default numbers, and neither remembers the settings, so I don't know why they have 2 entries. I think it's been like this on all the firmware versions, but my memory gets fuzzy after a while.

Yes, it would be bad at night - this would just be to test during the problem time. So, while it's soft, I'd adjust the fixed exposure faster and faster until it was too dark and noisy, to see if the focus problem changed. It may have no effect, but worth a try. BLC probably won't help, but WDR might if this camera has it.

I like to put mine in junction boxes, though my Dahua 2100 is under the eaves and the connectors are just hanging there. As soon as I get the time, I'll put a box up. I hate this practice of having short bundles of pigtails that can't be changed. My Vivotek IP8332s have an RJ45 jack inside the camera and came with a spare sealed cable gland, so you can pull the pigtail and run a network cable right in. I'd love to see everyone doing this.

I see one main functional problem with each camera at this point: - The Dahua has the softness in the low light areas, as Buellwinkle points out. This isn't awful, but it's annoying. With luck, they'll fix it in a future release. - The Hik/Swann gets noisier in low light, with the horizontal lines of noise you can see on some of the screen shots and clips. I didn't notice this at first because I was testing several cameras together, and the dual IR helped it with better exposure. The Hik's a good bit noisier in the same conditions as the Dahua. I limit all my cams to 1/30s minimum exposure to avoid motion blur, so if you don't mind a longer exposure, the Swann image gets better. Likewise, if there's a bit of ambient light or additional IR, it cleans up. Turning on the WDR helps a lot in the daytime shadows, but makes the noise worse at night, and there's no way to enable it on a schedule. The Hik's lens is hard to change because it's glued in, but people have removed the glue (I haven't tried it). I don't know if the Dahua's lens is or not. That only matters if you want to change it from what you ordered it with. As BW mentioned, he Dahua's IR can't be turned off and is only controlled by the photo sensor regardless of mode, but on the bigger bullets, it can be unplugged and it doesn't care. The Hik's IR and mode/filter are controlled by the photo sensor, so presumably it can't be turned off, but doesn't need to be. On the Hik, if you leave it in night mode during the daytime, it turns the IR LEDs off when the light is bright. Both have features in the firmware that sound nice and don't work as well as you'd like them to - this is typical in this price range. Hik's support is far better, but if you get the Dahua from someone who can provide the firmware, that helps. Overall, the Hik's software is more mature and stable, but that can change with any given release. If I were to buy one today, I'd probably go for the Dahua, even though I like the Hik better overall, because of the low light noise. If that were equal, I'd choose the Hik, hands down.

Interesting. I couldn't find it on the averusa.com site (where I normally look), except under Downloads/Discontinued Products, but it's listed as a product here - http://surveillance.aver.com/model/NXU8000. It looks like it's the same basic software that runs with the NV series boards, except it uses a USB dongle instead of an installed card to activate the software features, which would be nice - I don't run any analog on my Aver boards. It also looks like it's only offered in Europe and the UK. I can't find prices except in pounds and euros. That may be why I've been unable to find it before.

Does aver have a stand-alone software NVR? I look for one now and again but never seem to be able to find it.

One interesting test would be to manually set the maximum exposure time to a much higher number, like 1/2000 sec (or whatever still gives you a decent image) and see if it's better. That could tell you if it's a processing issue, vs a heat/warping problem. If it gets better at a fast exposure that still gives a good image, it may be that the Dahua's firmware is having trouble with certain image lighting situations. For instance, the 2 and 3MP bullets, when they have a well lit scene that also includes shade, will be sharp in the bright areas, but go very soft and fuzzy in the shade, and look out of focus. As soon as the shaded areas are well lit, they're sharp too, so it's a processing problem. In this case, it may be that it's confused by the bright spots that are reflecting off the parked cars and is considering everything else to be lower light, applying its soft processing to them.

After reading this, I connected the one I'm testing to my NV6480 with about the same results. I didn't set any parameters - just the IP address and login - and it's choppy, with occasional ghosting from dropped frames. What parameters did you set? Looking forward to hearing what you find on this. I don't have much more testing time on this camera.

What version firmware are you running on the bullet? I tested mine with a long maximum exposure in the daytime on V5.0.0, and it appeared to be the limit, not a fixed setting. That is, at 1/12 sec, I still had good exposure and no motion blur in the daytime, so it looked like it was changing to shorter exposures. My Dahua 3MP bullet lets you set either a fixed exposure rate or a low/high range of exposures.

This came from the US FTP site on 5/10. I just downloaded it again and compared it to the original, and there are no binary differences between the files, so it had something to do with the reboots. Another option would be to delete the existing web components folder, which is under Program Files on mine, then connect. It should install the latest version. Here's the version reported by the camera: Firmware Version - V5.0.0 130412 Encoding Version - V4.0 build 130411 The webcomponents.exe version is 3.0.4.6, while the old one from version 4.0.9 was 3.0.3.18. and the image setup screen, with both the new WB settings and the Intelligent IR setting. My Blue Iris had trouble with high bit rate images from both the Hik and the Dahua 3MP bullet, and a recent update that included a smoothness fix (don't remember which version, but after direct to disk was added) got rid of that. I still see the occasional frame rate drop, but before that, it couldn't go above 4096 on either cam with a well lit scene with lots of detail. I'd get hundreds of dropouts in a sunny yard between 6 am and 6 pm, and none after the sun went down.

The main drawbacks of high frame rates, which may or may not apply to your situation: - More HD usage/less recorded time - Higher bandwidth usage or lower image quality if fixed bandwidth is used - Higher CPU usage if a PC based system is used - Some dedicated NVRs won't support full frame rate on all channels Some of these may not matter, depending on your setup.

Well, that's one of the funny things. The first time I checked it, that's what I had - auto and manual, and manual enabled 2 color sliders (the 3rd color can't be changed). After re-installed webcomponents.exe, I now have a big selection of presets under white balance. Before, I don't think I rebooted both the PC and the camera at the same time. When I connected the other night, both had been off for a few days, and it asked me to re-install webcomponents.exe. Try that; if it doesn't work, I'll post the new webcomponents and you can see if that helps.

So, I posted about bugs in the 5.0.0 firmware that caused the 3MP setting to not work right, but turns out I didn't have the latest webcomponents.exe loaded. I'm not sure why it didn't prompt me to install the latest version after flashing, but when I re-connected the camera after a few days offline, it did. Now, the 3MP works fine (so far), and there are new features that weren't showing up before, like Intelligent IR, more white balance selections, more network option tabs, and so on. I can't tell what the intelligent IR does, though. When enabled, there's no obvious difference, and with another camera watching the IR footprint, I can see no changes in IR when I get right in front of the camera. It sounds like the cameras are shipping with 5.0.0 now, and that's a good sign. It only allows you to set 15 fps at 3MP, like the spec says.

The Hik/Swann bullet sensor is listed as 1/2.5" on their web page, 1/3" on all the PDF files. I'm assuming the web page is the current spec, but haven't measured it yet.

Heh! No, just an aspie with many decades of coping strategies under my belt. We make the best engineers as long as we care about whatever we're doing, and understanding the technology behind camera sensors and optics is a current hot button for me. This, too, shall pass.

For a given size sensor, more MP means smaller sensor elements, so less light hits each one. This is listed as a 1/3.2" (0.31") sensor, at 2592x1944 resolution (5.03MP, 4:3 aspect ratio). The optical size is generally about 1.5x the actual size, so this is about 0.21" diagonal. With a 4:3 aspect ratio at that resolution, each sensor element is about 1.6 um. A sensor with half the elements in each direction would have sensor elements twice as big on each side - 3.2 um - and have 1/4 the resolution - 1.25MP. Since the area of the element goes up with the square of the sides, the 5MP elements would be 2.7 square um, while the 1.25MP elements would be 10.7 square um, giving 4x the light gathering area per element, but only 1/4 the total resolution. The way around this is with larger sensors - 1/2", 2/3", or more. 1/2" is pretty readily available from some brands, but is more expensive. Going above 1/2" gets very expensive quickly, both in sensor/hardware costs and in optics costs.