Waveform 13 Analog Video, part 2
Posted: Fri Sep 23, 2005 9:06 am
Originally published 11/19/2004, HDTV Magazine, editor Dale Cripps
Updated 09/21/05
By Richard Fisher
Now that we have covered connector types the cable is the next piece of the puzzle. It must be 75 ohm impedance and wide bandwidth. Length is critical and determines how much bandwidth the cable must have for the given application. In cases where the equipment is around the display you can use 6 foot lengths normally. Monster V3 is highly recommended for these applications as it passes the ISF requirements and is not overly expensive. While others could be used the problem is most cable manufacturers do not test or spec out their cables, which is why I have been recommending Monster V3. We would be happy to recommend other commercial cables provided they will give us info about their products. I have also recommended Bluejeans Cable or any other establishment that uses similar product for custom long distance runs that will require higher bandwidth cable. People wonder why these cables are so expensive and it is not the wire but the connectors and properly terminating the wire to the connectors that is expensive.
Coax comes in all kinds of varieties. Most applications are for either RF such as cable or an antenna or for video applications. RF coax is designed for 50Mhz applications or higher so the center conductor of RF coax is a copper skin on top of a steel core as higher frequencies literally ride on the skin and do not penetrate the core. To pass lower frequencies properly does require the core and therefore solid copper. Video is a complex signal from DC to 37Mhz for HD and that still requires low frequency response. The center conductor of video coax then is either solid copper or stranded. While there are some cables that could be feasibly used for both applications the rule is use RF coax for RF and video coax for video.
Next comes switching. Again with equal lengths of 6 foot video cable for your input and output you will not run into trouble typically with most passive switchers. Passive switchers use mechanical switches or relays to switch the signal. These can be used with long runs as well but that requires BNC connectors to prevent ringing. Ringing is seen as edge enhancement on graphics and the high contrast edges of objects creating a white ghost or halo next to the edge. Greg Rogers of Widescreen Review checked out a passive switcher last year and brought up all these points as well in his review. With active switchers you will have a buffer amp between the input and output which helps to reduce the possibility of ringing allowing you to get away with more but if you are going long distance then the RCA connectors can still cause ringing problems. How do you know which one you have? If the switcher is silent when changing inputs then you probably have active buffers. If you hear the clickety clack of relay chatter when switching inputs then you have passive relays.
Video switching bandwidth is another topic of angst. Logic would seem to dictate that if your signal is only 37mhz then all you need is 37mhz. Analog does not work that way. For analog you need the response out to the 3rd or 4th octave to prevent signal degradation. Audio has always been this way. So for audio we have 20hz-20khz. 20X2=40, 40X2=80, 80X2=160 and 160 Khz, the 3rd octave, is the typical -3db point for audio preamps and amplifiers with many going out to 200khz. For video we need 37X2=74, 74X2=148, 148X2= 296 the 3rd octave falls right within the range of 300-400mhz as indicated in testing done by the ISF on video switchers for the broadcast community. This response is typically measured at the -3db point. Why would you need all this bandwidth? To prevent high frequency loss and ringing. Video information is a complex waveform of analog sine waves and digital looking square waves, and it is the square waves that cause the ringing (this does not include the sync pulses).
Reality is most CRT displays top out at 960 lines or 18.5Mhz not 1920 or 37Mhz. All STBs are significantly rolled off so again the 37Mhz is being degraded. On the production side HDTV is still limited to about 1440 lines. One could easily argue that all one needs is half the response.
For all of the above reasons we recommend switchers capable of -3db at 30mhz for SD 480I and 480P scan rates and -3db at 100mhz for HD 720P and 1080I scan rates up to 10 feet of cable using 75 ohm RCA connectors. For long runs we really should go back to BNC basics and call a company like Bluejeans Cables or a professional home theater installer for structured wiring.
In the end which is better for the consumer? Digital. Digital cable and connector requirements are less stringent than analog for quality consumer applications and prevents snake oil marketing as it only deals with on and off square waves sending data and not the final analog signal. In fact digital would have made 3 pages of this topic unnecessary. Due to copy protection I am also speculating that only in this domain will we be provided the full bandwidth potential of HD.
Which is better for the real world? Depends on the display. Analog CRT displays are limited in most cases so analog video is very effective and in a properly calibrated system you will be very hard pressed to see any difference between DVI and component. While digital displays will have some what similar results they are a different story. To avoid A/D and D/A conversion processes digital is preferred hands down. On top of that the 720P displays will go all the way out to 1280 vertical lines. 1080P digital is coming the end of this year and will also go all the way out to 1920 vertical lines. For digital displays using analog connections half the response may not be good enough. Rather than invest in higher bandwidth switching and properly terminated cables I recommend sources and digital displays with DVI/HDMI connections instead.
Next week I will conclude this series with AC power and lightning/surge suppressor concerns.
LINKS
The following link explains what we are trying to achieve even though it discusses SD video
http://www.imagingscience.com/isfwire.h ... ifications:
Bluejeans Cable for more reading
http://www.bluejeanscable.com/index.htm
viewtopic.php?t=3486
viewtopic.php?t=3727
viewtopic.php?t=3375
REFERENCES
Wire, Cable, and Fiber Optics for Video and Audio Engineers
Stephen H. Lampen
McGraw-Hill Publishing
www.ee.mcgraw-hill.com
(note that this book is based on common industry practices and I do not support all the conclusions given for audio applications. That said it has been a great reference for video and digital applications)
If you have any feed back to this article please feel free to post your comments now by clicking on POST reply. All emails received concerning this article are open to being posted and published along with the article unless the sender requests otherwise.
Respect Our Copyright! DO NOT copy and paste this complete review or article to post or distribute. That is a violation of US copyright. Please feel free to quote small passages and please provide a link to the full review. Thank you!
Updated 09/21/05
By Richard Fisher
Now that we have covered connector types the cable is the next piece of the puzzle. It must be 75 ohm impedance and wide bandwidth. Length is critical and determines how much bandwidth the cable must have for the given application. In cases where the equipment is around the display you can use 6 foot lengths normally. Monster V3 is highly recommended for these applications as it passes the ISF requirements and is not overly expensive. While others could be used the problem is most cable manufacturers do not test or spec out their cables, which is why I have been recommending Monster V3. We would be happy to recommend other commercial cables provided they will give us info about their products. I have also recommended Bluejeans Cable or any other establishment that uses similar product for custom long distance runs that will require higher bandwidth cable. People wonder why these cables are so expensive and it is not the wire but the connectors and properly terminating the wire to the connectors that is expensive.
Coax comes in all kinds of varieties. Most applications are for either RF such as cable or an antenna or for video applications. RF coax is designed for 50Mhz applications or higher so the center conductor of RF coax is a copper skin on top of a steel core as higher frequencies literally ride on the skin and do not penetrate the core. To pass lower frequencies properly does require the core and therefore solid copper. Video is a complex signal from DC to 37Mhz for HD and that still requires low frequency response. The center conductor of video coax then is either solid copper or stranded. While there are some cables that could be feasibly used for both applications the rule is use RF coax for RF and video coax for video.
Next comes switching. Again with equal lengths of 6 foot video cable for your input and output you will not run into trouble typically with most passive switchers. Passive switchers use mechanical switches or relays to switch the signal. These can be used with long runs as well but that requires BNC connectors to prevent ringing. Ringing is seen as edge enhancement on graphics and the high contrast edges of objects creating a white ghost or halo next to the edge. Greg Rogers of Widescreen Review checked out a passive switcher last year and brought up all these points as well in his review. With active switchers you will have a buffer amp between the input and output which helps to reduce the possibility of ringing allowing you to get away with more but if you are going long distance then the RCA connectors can still cause ringing problems. How do you know which one you have? If the switcher is silent when changing inputs then you probably have active buffers. If you hear the clickety clack of relay chatter when switching inputs then you have passive relays.
Video switching bandwidth is another topic of angst. Logic would seem to dictate that if your signal is only 37mhz then all you need is 37mhz. Analog does not work that way. For analog you need the response out to the 3rd or 4th octave to prevent signal degradation. Audio has always been this way. So for audio we have 20hz-20khz. 20X2=40, 40X2=80, 80X2=160 and 160 Khz, the 3rd octave, is the typical -3db point for audio preamps and amplifiers with many going out to 200khz. For video we need 37X2=74, 74X2=148, 148X2= 296 the 3rd octave falls right within the range of 300-400mhz as indicated in testing done by the ISF on video switchers for the broadcast community. This response is typically measured at the -3db point. Why would you need all this bandwidth? To prevent high frequency loss and ringing. Video information is a complex waveform of analog sine waves and digital looking square waves, and it is the square waves that cause the ringing (this does not include the sync pulses).
Reality is most CRT displays top out at 960 lines or 18.5Mhz not 1920 or 37Mhz. All STBs are significantly rolled off so again the 37Mhz is being degraded. On the production side HDTV is still limited to about 1440 lines. One could easily argue that all one needs is half the response.
For all of the above reasons we recommend switchers capable of -3db at 30mhz for SD 480I and 480P scan rates and -3db at 100mhz for HD 720P and 1080I scan rates up to 10 feet of cable using 75 ohm RCA connectors. For long runs we really should go back to BNC basics and call a company like Bluejeans Cables or a professional home theater installer for structured wiring.
In the end which is better for the consumer? Digital. Digital cable and connector requirements are less stringent than analog for quality consumer applications and prevents snake oil marketing as it only deals with on and off square waves sending data and not the final analog signal. In fact digital would have made 3 pages of this topic unnecessary. Due to copy protection I am also speculating that only in this domain will we be provided the full bandwidth potential of HD.
Which is better for the real world? Depends on the display. Analog CRT displays are limited in most cases so analog video is very effective and in a properly calibrated system you will be very hard pressed to see any difference between DVI and component. While digital displays will have some what similar results they are a different story. To avoid A/D and D/A conversion processes digital is preferred hands down. On top of that the 720P displays will go all the way out to 1280 vertical lines. 1080P digital is coming the end of this year and will also go all the way out to 1920 vertical lines. For digital displays using analog connections half the response may not be good enough. Rather than invest in higher bandwidth switching and properly terminated cables I recommend sources and digital displays with DVI/HDMI connections instead.
Next week I will conclude this series with AC power and lightning/surge suppressor concerns.
LINKS
The following link explains what we are trying to achieve even though it discusses SD video
http://www.imagingscience.com/isfwire.h ... ifications:
Bluejeans Cable for more reading
http://www.bluejeanscable.com/index.htm
viewtopic.php?t=3486
viewtopic.php?t=3727
viewtopic.php?t=3375
REFERENCES
Wire, Cable, and Fiber Optics for Video and Audio Engineers
Stephen H. Lampen
McGraw-Hill Publishing
www.ee.mcgraw-hill.com
(note that this book is based on common industry practices and I do not support all the conclusions given for audio applications. That said it has been a great reference for video and digital applications)
If you have any feed back to this article please feel free to post your comments now by clicking on POST reply. All emails received concerning this article are open to being posted and published along with the article unless the sender requests otherwise.
Respect Our Copyright! DO NOT copy and paste this complete review or article to post or distribute. That is a violation of US copyright. Please feel free to quote small passages and please provide a link to the full review. Thank you!