Probably not you...much of what said is a mixture of theoretical and maybe a small amount of marketing - but Sean's info and other voices of real world experience is the most valuable in this thread and is excellent, helping to sort out some of the info as well. You may not see the difference as you described between the 2 lengths of 's' cable, and in some cases, like I mentioned, a signal analyzer might be what does it - however, the problem is really that with the longer length you are more susceptible to interference that would degrade the signal (as well as signal loss of long distance that might not be noticable depending on the monitor - it may be hardly noticable on ntsc tv.) If no interference is present, than the longer length s-video is doing okay, I guess, but I like Sean's suggestion. But in your situation, I would still avoid having to go through the rca adaptor.
-Jon

Connectors is connectors and signals is signals. The BNC connector is a professional connector used in a wide variety of video, communications and laboratory equipment up to frequencies of over 100 MHz. It is designed for quick connection and disconnection over thousands of cycles while still maintaining good impedance matching and low insertion and return losses. The RCA "phono" plug was designed for guess what. Connecting phono cartridges to amplifiers in consumer equipment. It is designed to make a connection cheaply. It too can be used at video and r.f. frequencies but is not nearly as robust as the BNC either mechanically or electrically. Any type of cable of reasonable size can be used with either connector but as you might expect the best quality cables (low loss, good sheilding) will likely be found with BNC's on them and not RCAs. If you have a good cable with BNC's on both ends you should have no problem using an RCA adapter to get the signal from a video device that does not have a BNC connector.

On to signals. The NTSC signal consists of luma (base band) and chroma (modulated on a subcarrier) signals. In composite video they are simply added and put out over a single coax. In S-Video they are kept separate and each sent over its own coax. One can reconstruct construct a legitimate composite signal simply by adding the two S-Video signals (with one twist: the industry adds a DC offset to the chroma to indicate whether the picture is 4:3, 16:9 letterbox or 16:9 and this DC must be blocked). The reason all this is done is becuase the luma and chroma signals overlap in frequency. In cheap sets/monitors they are separated by lowpass and high pass filtering which reduces both the sharpness and color fidelity of the signals. In better sets they are separated by "comb" filters which permit the full bandwidths of both signals to be recovered. Keeping them separate will always work best but a composite signal to a set with a comb filter should look almost as good as S-video.

Awesome...I am learning alot on this thread. Very good to know. Thank you.
-Jon

the problem there is the phase thing that sean was referring to... prosumer y/c equipment doesn't typically include a way to manually adjust for an out-of-phase situation... so you'd never want to use that 50 ft. y/c cable anywhere other than a monitoring situation, and even then, with the right equipment, you might be able to see the phase delay in the picture itself.

a component video signal going thru the 100-yard cable run that i referred to earlier will most likely be out of phase with itself by the time that the signal gets all the way to the other end, but broadcast equipment has the capability to compensate for that.

earlier i stated that y/c cables don't have shielding, which isn't really true... it's just that the shielding on bnc cabling is so much better.

when i used to edit hi-8 with the fast video machine, y/c was the only way to go... the vm card and it's interface box both had y/c connections, and so did the $5k sony evo-9850 deck, as did the $6k panasonic ag7750 svhs deck... but you always used the shortest y/c cable that you could to hook 'em up together.

Well, yeah. I wouldn't have a reason to use a long cable for anything other than a field production situation anyway, that I can see.

I wouldn't worry too much about differential delay. The main problem would be with the relative phases of the color burst in the luma part and the subcarrier on which the chroma signal rides. As these are at 3.58 MHz they have a period of 279 nS. Nominal cable delay is 2 nS/ft so that if you cut one cable a whole foot longer than the other the phase error would be about 2.6 degrees and all your colors would be shifted by that amount. If you mismatched them by a tenth of a foot (1.2" and if you can't cut them that close you should turn in your cable crimper) the phase error would be undetectable at 0.26 degree.