videotiger

So far I have three Hikvision 3MP IP camera installed, and one specific camera does randomly lock up every few months. I'm using a Trendnet single-line PoE to feed it, which is what the other two are using as well. All three are in completely separate locations. No idea why this one locks up, but I added the capability to remotely power-cycle the most inaccessible one.

I've had limited success against IR bounceback even with cameras with the gasket sealing between the center camera lense and the LED IR ring. I suspect the culprit is that the IR output bounces off of the *outside* front window surface and back into the camera's lense. What *has* worked for me are cameras where the front windows are physically two concentric rings of glass, the center round disk for the camera lense, and the ring for the IR LED's. The other configuration that works is the Hikvision set up like on their DS-2CD2332 where the IR illuminator is completely separate from the camera window.

Thanks! I'll give that a try tonight. Being able to pull JPEG images will help ease the transition from my current set up to the next IP-based solution. I originally searched the forum several times with keywords like "hikvision jpg" and "hikvision jpeg" and either came up empty--or glazed right over the answer Thank you, buellwinkle, as well!

Is it possible to pull a JPEG image from the Hikvision DS-2CD2332 camera? I mean is there a URL I can use to fetch a full-resolution JPEG image. I have been able to do this in the past with other IP sources, such as StarDot IP cameras and video servers, as well as the Sony SNC-P5, SNC-RZ25, and the Axis M5014. For example, on the Axis, it's: http:///jpg/image.jpg If so, is the URL common across ONVIF-compatible cameras, or is it manufacturer-specific. Thanks in advance!

Nice all-in-one solution and very cost-effective too. The battery life would be my only concern, but hooking up external power would alleviate that issue. I have used three generations of GoPro's, which have a 170° horizontal FOV and a polarizer during the daylight hours. Granted, they are significantly more expensive, but HD Hero's were the only 1080p solution back then. My wife used to make snarky comments about my usage for this purpose (this is before the Russian dash camera usage became more widely known)--until one day my camera settled a dispute over who was at fault in a fender-bender. The other driver tried to claim my wife was at fault until the insurance company's adjuster received a copy of the video. Since then my wife hasn't complained a bit over my GoPro purchases Due to the declining ability of GoPro's to endure daily use (frequent dead batteries, insufficient battery life, freeze-up requiring battery removal to reset), I'm switching to a Sony HDR-AS30 which has the same 170°-wide FOV.

Even if the lens assembly is sealed against the *inside* of the front glass window, I find that IR bounce back is still an issue when the exterior of the window gets dirty. This shows up as translucent spots on the viewed image, and if left unchecked, can easily obscure half (objective opinion) or so of the viewable content at night. The only viable solution I have found--if you must have a single-piece camera+IR illuminator solution--are cameras where the lense window and the IR window are physically separated, usually by concentrically-arranged windows separated by an opaque material, like a plastic ring. This configuration can be difficult to find. Otherwise the recommendations from others of a separate IR illuminator is the ideal solution. Environment plays a big factor. I successfully use cameras without this split-window arrangement in some inland locations, but in more humid environments, like locations near the ocean with salt/mineral deposits, gunk will build up on the front window. In those situations a split window is a must-have for the all-in-one configuration.

FFmpeg gives the following assessment of the mp4 health: FFmpeg version SVN-r22666, Copyright © 2000-2010 the FFmpeg developers built on Mar 25 2010 06:05:35 with gcc 4.4.2 configuration: --enable-memalign-hack --cross-prefix=i686-mingw32- --cc=ccache-i686-mingw32-gcc --arch=i686 --target-os=mingw32 --en ble-runtime-cpudetect --enable-avisynth --enable-gpl --enable-version3 --enable-bzlib --enable-libgsm --enable-libfaad --enable-pthrea s --enable-libvorbis --enable-libtheora --enable-libspeex --enable-libmp3lame --enable-libopenjpeg --enable-libxvid --enable-libschroe inger --enable-libx264 --enable-libopencore_amrwb --enable-libopencore_amrnb libavutil 50.12. 0 / 50.12. 0 libavcodec 52.60. 0 / 52.60. 0 libavformat 52.57. 1 / 52.57. 1 libavdevice 52. 2. 0 / 52. 2. 0 libswscale 0.10. 0 / 0.10. 0 [mov,mp4,m4a,3gp,3g2,mj2 @ 0037bd40]max_analyze_duration reached Seems stream 0 codec frame rate differs from container frame rate: 2002.00 (2002/1) -> 1001.00 (2002/2) Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'test.mp4': Metadata: major_brand : isom minor_version : 0 compatible_brands: mp41 Duration: 00:00:22.72, start: 0.000000, bitrate: 5170 kb/s Stream #0.0(eng): Video: h264, yuv420p, 1280x1024, 2955 kb/s, 16.90 fps, 1001 tbr, 1001 tbn, 2002 tbc Stream #0.1(eng): Data: stts / 0x73747473 My guess is that the generator of that MPEG4 file is generating an invalid MPEG4 file. I've run into that with NVR and standalone DVR's.

I use VLC (http://www.videolan.org/) as my go-to player for Windows and MacOS. It also has a function in the menus to display the video and audio encoding information.

You're very welcome! While I'm not too surprised with the results, I would have been pleasantly surprised if the Gadspot gave the StarDot a run for its money. I was curious to do some controlled condition testing. Next time, though, I'll compare frames within a few seconds of each other. I weighed the option to return the camera but I suspect that the image won't improve even with a replacement camera, so I would be out the shipping cost to send it back to the disti. I'll just use it in a less critical application where the subjects will be a lot closer to the camera. I also have an ACTi TCM-1111 in hand for another installation. Preliminary testing shows that the user interface is a lot less ambiguous. I had a couple questions about the camera that I posted to their online Support. Their response was prompt and accurate, which matches the experience someone else noted in a different thread. If time and weather permits, I'll do another apples-to-apples-to-apples comparison/shootout between the ACTi, the Gadspot, and my StarDot.

I have never posted with photos, so bear with me if the references to attachments is mixed up. I just completed a side-by-side comparison of the Gadspot GS9405E/S with the StarDot NetcamSC (5MP true day/night, MJPEG). Both were set up to comparable resolutions, approximately 1200 pixels wide by about 800 pixels high on a single mount, aimed at the same general scene (see attached photo). The NetcamSC image were taken directly from the camera's web server. The Gadspot image was taken from the NVR software that comes bundled with the Gadspot IP camera, the FCW-SE064 package. I also got some video clips, but as these are IP cameras, the video clips are simply a sequence of images comparable to the stills I have attached. These images were taken outdoors in a mostly-overcast setting. As the images show, the Gadspot image quality was very disappointing: hazy and washed out. I did double-check the Gadspot's lense, and it was spotless. While setting up the Gadspot, I discovered something funky about the web server UI: I was trying to focus the camera by accessing the internal live feed, and the image would not update in real time. But: the time-of-day clock in the right lower corner would continue to increment! To prove my point, I stuck my hand in front of the camera to mostly obscure the camera's view, and I watched in disbelief on my laptop's screen as the time-of-day clock continued to increment, but my hand never appeared in the live image. So... I am disappointed with the Gadspot camera. I will probably end up getting an H.264 version of the NetcamSC for permanent use. It's about 4X more expensive, but as the sample image indicates, the quality is beautiful.

I have bought five Gadspots so far. Two were installed in a remote location several thousands of miles away; one of those died about two months after installation, which left me pretty miffed. No more Gadspots in that location; they only go where I can readily access them for replacement. I also use some KT&C N600's (discontinued) and Veilux vandal-resistant ball cameras. I like the Veilux' low-light performance and resistance to scum build-up on the lenses. I find that the Gadspots have a reddish/pink bloom appearance to their images, and the color space seems to be shifted towards red. It's more obvious in low light conditions where the camera is about to shift between day and night mode (I use true D/N cameras only): the overall image turns very pinkish. The contrast also seems somewhat harsh. My image quality/attribute is towards the KT&C's which seem to give a truer picture quality. I really should do some color chart calibration and some actual human-subject testing to validate my empirical observations. Based on someone's comment, I just picked up a Gadspot GS9405 3 MP IP camera. I'm still testing it, but preliminary testing shows that my StarDot hybrid box camera with CS-mount lenses has better low-light performance and truer image quality and contrast. Then again, a true C/CS-mount lense will run circles around the smaller lense sets in all-in-one cameras. It just seems that the Gadspot image quality is harsher. Maybe that's what's needed for security surveillance to get identifying information. My paid profession is to produce as life-like an image on a computer screen as possible.

I second that sentiment. I've had bad experiences with Swann, and would rather go without and avoid the headaches than try to implement a solution with their products.

I have the non-dome 5 MP hybrid non-H.264. Great image quality and great feature set in the built-in server. The best feature is that it supports Java so you don't need a browser that supports ActiveX to view it. The only caveat is that the auto-iris function seems to not be working well. It could be my lense, though.

I have a friend with a "starlight" camera. He determined that it functioned by doing multiple or extended exposures. Unfortunately that means that a moving object/person becomes a probable indistinguishable blur on the resulting video.

Other IP cameras may have this feature, but StarDot servers and IP cameras do support dial-up modems on the serial ports on their cameras and servers. I believe the IP controller is also capable of dialing out when triggered or on schedule (or both).