Analog HD, HD-CVI, HD-TVI and IP systems

[nightrider 0]

Hello there!

 

I am doing research on the subject because I need to buy video surveillance system. I have a few questions that I cannot find definitive answers about and I hope some of you may give some insight. For the moment I've chosen Dahua as budget solution.

 

1. What exactly is "Analog HD", is it digital and if it is, then why the analog bit in it?

 

2. To me, it looks like HD-CVI and HD-TVI are essentially the same technology but developed and integrated under different names by different manufacturers. According to the specs there is no digital encoding/compression, nor error correction, and the video output is given as "Analog HD". They all are using quadrature amplitude modulation (QAM), which is one of the few digital to analog modulations. So, my questions is, if there is no obvious digital processing of the video/audio signal, then the technology must be vulnerable to most of the analog signal issues /attenuation, signal distortion, degradation, cross-talk, noise, electromagnetic interference and so on/ and should be regarded as analog system, right? Or is this another marketing stunt, or perhaps I've missed the elephant in the room?

 

3. Why would you choose HD-CVI over IP for new installation? I know all the things about the few hundred milliseconds delay/real-time video, easy installation, no complex IP configuration hassles, reuse of the existing coaxial cabling.

 

As a side note, my observation is that most of the chinese manufacturers don't like to give full specs or to give in-depth answers.

 

Feel free to make any suggestions!

 

Thank you in advance!

[Don Stephens 0]

1. The only information you're going to find about AHD right now is what comes from Nextchip, and that isn't much.

 

2. AHD, HDCVI, and HDTVI are all really the same type of tech; as you said, they're just developed and manufactured by different people. They are all analog signal types and I'm going to leave it at that. Realistically, there is so much contradicting information going around right now that it's impossible to really know what's what unless you've tested the products yourself.

 

3. You shouldn't choose HDCVI over IP unless IP isn't in the budget. Otherwise, take your pick from your list as to why people do it: existing cable, knowledge base, bad information, etc...

 

 

Chinese manufacturers couldn't care less about giving you an in-depth answer. You should feel privileged if they even acknowledge you without having to buy anything from them.

 

I don't know what it might be that you're looking for in a spec sheet, but most of them contain everything you'd need or should be interested in. If you're looking for any additional details about a specific component they have listed, you're going to have to do the leg work yourself.

[nightrider 0]

1. The only information you're going to find about AHD right now is what comes from Nextchip, and that isn't much.

 

2. AHD, HDCVI, and HDTVI are all really the same type of tech; as you said, they're just developed and manufactured by different people. They are all analog signal types and I'm going to leave it at that. Realistically, there is so much contradicting information going around right now that it's impossible to really know what's what unless you've tested the products yourself.

 

3. You shouldn't choose HDCVI over IP unless IP isn't in the budget. Otherwise, take your pick from your list as to why people do it: existing cable, knowledge base, bad information, etc...

 

 

Chinese manufacturers couldn't care less about giving you an in-depth answer. You should feel privileged if they even acknowledge you without having to buy anything from them.

 

I don't know what it might be that you're looking for in a spec sheet, but most of them contain everything you'd need or should be interested in. If you're looking for any additional details about a specific component they have listed, you're going to have to do the leg work yourself.

 

 

Got confirmation that it is analog technology and is developed to upgrade the current analog systems seamlessly and at low cost to the HD era.

[SunnyKim 2]

nightrider, please allow me to add my 2 cents.

 

1. HD SDI is the method to send digital video data in digital number- bit serial wise. So called SerDes. Gennum supplies most of receiver chipsets. Originally, they were for studios of TV station. They won't down the price as good as Chinese CCTV manufacturers want to. Gennum does not show much interest in CCTV market. Techpoint and Intersil have introduced 4 Channel HD SDI receiver chipsets. But the performance seems not as good as Gennum's, on our limited testings in our lab. The good thing about HD-SDI is that the signal can go up to 100 meter over conventional cable. That means you do not need to install a new cable.

 

2. AHD, HD-CVI, HD-TVI.

They are all slightly different, so non compatible each other. But they seem to be cheaper than HD-SDI. You are all correct they are all based on analog QAM, so no room for error correction digitally. You may regard them as an expanded form of NTSC or PAL, the conventional analog TV signal. They claim that it goes over 500 meter. But it seems they must be using solid copper cable or Gold. In our lab, it barely goes through 50 meter on a commercial cheap cable, distorting colors. So you may need to upgrade the currently installed cable. And you have to limit your choices to the pairs of camera and its DVR.

 

Of course, HD-SDI, AHD, HD-CVI(Daihua), HD-TVI(Hikvision) have advantage over IP, in realtime preview.

[joseph.chen0312 0]

Brief all three solution is adopt EOC (ethernet over coaxialcable). Basic structure layout for camera is CMOS sensor (OV, Aptina and SONY) + ISP board( Integrate DSP, DAC and EOC) for AHD and CVI but for TVI one more Intersil Board(EOC board).

[Lee Tracey 0]

I have discovered an interesting problem with HDTVI but have so far failed to discover the cause so I need help. The problem is interference on a four channel HDTVI DVR and four HDTVI cameras.The interference is a vertical line of blocks of many colours that moves across the screen horizontally. I have eliminated ground loop and fluorescent. My only remaining guess is that the frequency of the DVR DAC is interfering with the frequency of the cameras and my reason for this guess is that at night when the cameras switch on their infra-red LED's the interference vanishes - possible change of camera frequency for infra-red operation.

[SunnyKim 2]

To me, chances of interfering with DVR DAC and Cameras could be very slim. If the DVR supports HDMI out, you can plug into a TV for verification.

Please also check whether color components of TVI signal can be suppressed as an option in the night, or low lighted condition.

Please play some options(control buttons) on their Low Pass Filterings on Luminance or Chroma. If not properly suppressed down(completely low pass filtered), the decoder may be mistaken as a Chroma, being placed in the high frequency band.

[Lee Tracey 0]

I have now discovered the source of the rainbow interference with HDTVI DVR's. It is the HiSilicon HI-3521 SOC specified from the BOM that HiSilicon sold along with the reference design. Luckily for Hikvision they spotted the fault early on and dropped the 3521 in favour of an alternative. Apart from this change and a few minor inconsequential mods the Hikvision circuit is identical to the original HiSilicon reference design. The mystery still exists of what actually triggers the 3521 to generate the rainbow and why it actually fades over about 30 days.

 

Lee Tracey - dvrdigital@gmail.com

[survtech 0]

But it seems they must be using solid copper cable or Gold. In our lab, it barely goes through 50 meter on a commercial cheap cable, distorting colors. So you may need to upgrade the currently installed cable. And you have to limit your choices to the pairs of camera and its DVR.

Gold? Please tell me where I can find solid gold CCTV coaxial cable installed. I just might have to make a night visit to the site

 

Other than that, what "cheap cable" did you test that introduced distortion at 50 meters? Clad? Any analog system that used copper-clad coax instead of copper-copper was installed by amateurs or fly-by-night trunk slammers.

[SunnyKim 2]

Gold could be too much, if accepted that seriously. To me, RG59 cable must have been approved for carrying the conventional TV signal (NTSC, PAL). Please correct me if I am wrong. The TV signal can carry at most 6 Mhz. But for HD sized video, it could be well over 20 Mhz. We know that as cable length increases, it increases its capacitance value. High capacitance could destroy high frequence components. That makes difficult to decode out the video signal correctly. You may observe as the distance gets longer, they (AHD/CVI/TVI, others) start losing colors, which are placed in a higher frequency band.

We just hoped and concluded that pure copper could be better. But we know most of cables are manufactured in China. For cheap cables, like CCTV usage, hard to control over its quality. Each roll of cable turns out to have different quality, in our lab testing, even though they are of the same brand or distributor. It is true some cables are better than others by Luck. But Down to Too Cheap and No Margin, nobody dares to talk about Quality or Quality Control. The ones we used for testing were cladded one. When installing these you may test your cables at your site, in advance.

[Lee Tracey 0]

I am not sure why this debate on cables arrived in this section of the forum relative to "RAINBOW" interference from HDTVI DVR's. The problem arises and all my experience with the problem has been on my workbench using RG59 cables of two metre lengths and all connected to a common single earth. Out of curiosity I did also make the set up with 20 metre and 100 metre RG59 cables and the interference was exactly the same. The problem has turned out to be the HiSilicon HI-3521 H.264 Codec Processor and the FIX is to use the HiSilicon HI-3535 H.264 Codec Processor. However there is no FIX ( that I have discovered ) for existing p.c.b.'s using the 3521. Hikvision must have discovered this problem early as their HDTVI DVR is immune from the "Rainbow" but uses the HiSilicon 3535.

 

With the Hikvision DVR this use of the 3535 is concealed as the SOC top plate carries the Hikvision name and the reference 2013 and the heatsink is glued to the SOC by a very powerful glue. Removing the heatsink destroyed the SOC and the further work to separate the top plate of the SOC from the heatsink took a lot of effort and technique. Once separated the legend on the SOC top plate could not be read by the human eye.

 

I leave others to figure out why such a powerful glue was used and not heat sink compound and a different fixing method. What triggers the "Rainbow" with the 3521 is still a mystery and why it gradually vanishes ( to the human eye ) over a few days is also a mystery though Hikvision obviously think the hard drive is involved. On their p.c.b. with the 3535 they have used a four pin DIN for the 12VDC supply thus providing the 12VDC supply to the hard drive totally isolated from the p.c.b. DC rail. Some users of the DVR's with the 3521 have reported to me that providing separate power supplies to each camera solves the problem. Unfortunately this is not a reliable FIX as investigation exposes the inclusion of TIME between first discovering the "Rainbow" and actually obtaining and fitting the separate power supplies - the fading effect has taken place and this is being mistaken for a separate power supply fix.

 

My personal opinion is that the "trigger" is somewhere in the 50 Hz power supply as I have not heard from any North American installer who has seen the "Rainbow". All the HDTVI DVR's I have examined are LOCKED to 60 Hz.

 

Lee Tracey - dvrdigital@gmail.com

[survtech 0]

It all started because certain posters can't participate in a discussion without turning it into a marketing spiel.

[SunnyKim 2]

I am not sure why this debate on cables arrived in this section of the forum relative to "RAINBOW" interference from HDTVI DVR's. The problem arises and all my experience with the problem has been on my workbench using RG59 cables of two metre lengths and all connected to a common single earth. Out of curiosity I did also make the set up with 20 metre and 100 metre RG59 cables and the interference was exactly the same. The problem has turned out to be the HiSilicon HI-3521 H.264 Codec Processor and the FIX is to use the HiSilicon HI-3535 H.264 Codec Processor. However there is no FIX ( that I have discovered ) for existing p.c.b.'s using the 3521. Hikvision must have discovered this problem early as their HDTVI DVR is immune from the "Rainbow" but uses the HiSilicon 3535.

 

With the Hikvision DVR this use of the 3535 is concealed as the SOC top plate carries the Hikvision name and the reference 2013 and the heatsink is glued to the SOC by a very powerful glue. Removing the heatsink destroyed the SOC and the further work to separate the top plate of the SOC from the heatsink took a lot of effort and technique. Once separated the legend on the SOC top plate could not be read by the human eye.

 

I leave others to figure out why such a powerful glue was used and not heat sink compound and a different fixing method. What triggers the "Rainbow" with the 3521 is still a mystery and why it gradually vanishes ( to the human eye ) over a few days is also a mystery though Hikvision obviously think the hard drive is involved. On their p.c.b. with the 3535 they have used a four pin DIN for the 12VDC supply thus providing the 12VDC supply to the hard drive totally isolated from the p.c.b. DC rail. Some users of the DVR's with the 3521 have reported to me that providing separate power supplies to each camera solves the problem. Unfortunately this is not a reliable FIX as investigation exposes the inclusion of TIME between first discovering the "Rainbow" and actually obtaining and fitting the separate power supplies - the fading effect has taken place and this is being mistaken for a separate power supply fix.

 

My personal opinion is that the "trigger" is somewhere in the 50 Hz power supply as I have not heard from any North American installer who has seen the "Rainbow". All the HDTVI DVR's I have examined are LOCKED to 60 Hz.

 

Lee Tracey - dvrdigital@gmail.com

Lee Tracey, I am not that much suspicious on HiSilion CoDec. But rather, there must be another chip set on the board, "HD-TVI Video Decoder'' which captures and reconstructs video signal from HD-TVI camera. This provides HD Video to a backend CoDec, say, Hi3521. I think this video decoder (HD-TVI) chip set, probably from a company, "Techpoint", whose guide line on PCB routing might be wrong or the register(parameter) values of the TVI decoder chip could be set wrong.

FYI, Hi3521 might not be powerful enough for HD-TVI based DVR. And Hi3535 that you commented seems to be better fit for NVR, not for DVR. Please check again.

[Lee Tracey 0]

Reference opinions from " Sunny Kim".

 

There is, so far as I have discovered, a single reference design from HiSilicon and or Techpoint. This reference design has been adopted by about 60 Chinese manufacturers. Apart from differences in the Gerber layout of the p.c.b. they are all the same BOM except for the Hikvision. The common setup is:

 

HiSilicon Hi-3521 H264 Codec Processor.

 

Techpoint TP-2802 4 Channel 1080p Receiver.

 

Intersil TW-2964 Video Decoder.

 

The Hikvision has a chip marked KY-2013 where the Hi-3521 would normally sit but the chip has the mechanical image of a HiSilicon SOC. I cannot find any reference anywhere of a chip called a KY-2013 so I am sure that it is a chip from some other source with a phoney Hikvision legend - almost certainly from HiSilicon. It could be a modified version of the 3521 or a modified 3535.

 

The only other obvious variation on the Hikvision is that while all the others use a common DC plug and socket and derive the power for the hard drive from the p.c.b. rail the Hikvision has a four pin DIN plug and socket and provides an isolated 12VDC supply to the hard drive.

 

The "RAINBOW" interference problem, as far as I have discovered to-date, only appears where the mains supply is 50 Hz but is NOT known to areas where the mains supply is 60 Hz and as the majority of sales has been to North America the majority of users do not know of the problem. That suggests that the "TRIGGER" is within the 50 Hz cycle - but Hikvision has solved the problem as their HDTVI DVR does not show it even with the cameras that are used with the DVR's that do show it - and all HDTVI DVR's that I have obtained are LOCKED to 60 Hz and cannot be menu changed to 50 Hz. The menu offers selection of PAL or NTSC but the 60Hz stays for both as does the RAINBOW. But the RAINBOW fades after a few days as far as the human eye is concerned but can be detected by test gear.

Finally to Sunny Kim: who do you work for?

[SunnyKim 2]

Thanks for detailed confirmation, Lee.

Hi3521 is not powerful enough to handle 4 Ch of 1080PX30FPS. My Guess the DVR(PCB) was orginally designed for 4 Ch 720P X30 FPS. May be used for recording 4Ch X1080P X15FPS.

I can agree that KY-2013 could be the same as Hi3521, a special marking service for HikVision.

But I can not agree to you with The Rainbow interferences. The problem must have been known/ detected reported, first, in China of 50HZ system (PAL & Elec Power Source), so the ground loop noise issues must have been removed. It could be improved by providing wider grounds for HDD. I am still carrying Strong doubt on possible mistakes on register parameter value setting for TP-2802, assuming this chip is designed OK.

Please let me go further. The parameter of 60Hz /50 Hz, Field Rate, is definitely meaningful for NTSC/PAL, selecting conventional analog cameras. There could be no meaning for 60Hz/50Hz for HD-TVI signal. If 30Hz/25Hz mode exists. I may be giving it some credits. But 30 Hz of Frame Rate seems to be the Default Mode. Whatever value you use, I do not think it would affect any results in case of HD-TVI. Finally I work for a chip design company for CCTV applications, doing softwares as well.