[MattH]

Hi to all, have been reading these forums for waht seems like ages, have bought a plasma but after seeing projectors at the Bristol show decided that was the way to go (wife not happy about the cost, but....), I'm sure I will have many more questions but this one I cant find any info about.

I am intending to mount a projector from the ceiling, the question I have is, is it better to have a longer cable to the amp for sound from the dvd player or a longer component cable going to the projector from the dvd player. If I put the dvd player nex to the amp the coax cable will only be about 0.5m long but the component cable will be about 30 feet to the proj.

Is this going to make any detrimental effect to the quality of the signal received by the proj, or does this not really make any difference?

Anyone had any experience of this situation??

Any help greatly appreciated.

[severnsource]

My projector is fed with 10M svga and s-video cables and they do not introduce any visible degradation to the image. As long as you use good quality cables there should be no problem.

There shouldn't be any problem the other way either (long audio cables).

Bill

You will be better off having a long single audio cable than a long video cable.

This is common sense really and is driven by:

a) The video signal is higher bandwidth than the audio signal. This means that the video signal will be more sensitive to cable quality and to noise than the audio signal will be.

b) The video signal is analogue and therefore more sensitive to changes in signal/noise ratio than the audio signal which is a digital signal (I'm assuming you are using a digital line from DVD player to output). Again - this is a consequence of the coding scheme used and is a big point in itself.

c) For best quality, you need a component video cable from DVD player to projector. This will be three 75ohm coax cables that need to be electrically the same length for video. This is harder to acheive over a long run and needs more expensive cables. However, your audio signal is only one single coax and therefore you can spend more per unit length on the cable.

Bottom line:

Put the source as close as possible to the projector. Use a short video cable. Use a component video cable if you can.
The length of the audio cable is not as important.
Keene electronics will sell you a 10M digital audio cable at a reasonable price.

John

[severnsource]

If your audio is fed to your amp digitally I would agree with John, a digital signal can tolerate a great deal more interference than an analogue one before problems become apparent.

If your audio is analogue then it's not so clear cut. Audio will have a potential dynamic range of at least 80db. If your maximim audio signal is 2V, than an interfering signal of 200microvolts will be perceptible. Unbalanced audio cable in long lengths can be quite prone to hum pickup, particularly if run near to and parallel with power cables. Unbalanced audio cables are also very effective R.F. aerials, and in combination with a poor amplifier can produce various sorts of low level distortion and other problems.

Edit: For analogue audio, the cable capacitance can interact with the output impedance of the source equipment to give HF roll off, if the output impedance is high enough. Not usually a problem with modern equipment, but a reason to keep analogue audio cables short. (Video cables don't suffer from this problem as they are matched transmission lines.)
(End edit)

By contrast video has a very limited dynamic range, it is unlikely to approach 50db, let alone exceed it in domestic equipment, so interfering signals need to be much higher in level to become problematic. Video signals are clamped, which means that rejection of low frequency interference is very good. The amplifiers are designed for handling RF spectrum signals, so are much less likely to produce spurious signals due to distortion products produced from high levels of RF interference than audio systems.

Matching cable lengths for component signals; now there's an interesting subject.
The only effect that different length cables will have on a video signal, assuming the cables are identical, is that one signal will be delayed with respect to the others. If this were significant it would produce a horizontal convergence error; the delayed colour would be displaced to the right of the undelayed colours.

The speed of propogation in cables is approximately 2/3 of the speed of light, say 2x10^8 m/s. My projector has 960 horizontal pixels, this represents a time of (52 microseconds/960); about 54 nanoseconds. If we assume that any registration error of less than 1/4 of a pixel will be unobjectionable, that gives us a maximum time delay of 54/4ns. That is about 14ns. In order to delay the signal by 14ns we need (2x10^8) x (14x10^-9) meters of cable. That is about 2.8 metres. So if your cable lengths are within 2 metres of each other you will see no visible problems.

As long as you use half decent 75 ohm cable there is not the slightest problem in using any length that you like, and you don't even need to bother to make the lengths the same.

Bottom Line. If your audio feed is digital, make the audio cable the longest, as John recommended, if your audio is analogue make your video signal longer.

Personally, I would put the DVD player where it is most convenient for operation and aesthetics and make the cables reach regardless of the length, 'cos the actual difference will be non-existent or negligible, as long as you use reasonable quality cables and are not in an area prone to unusually high levels of interference.

Bill

Yes I agree in principle to what Bill says - I'm assuming that your audio feed is digital since this is very common nowadays.

If trying to set up your own component cable, I would still make sure that all the cables were the same length and were from the same manufacturer. This is because different "75 ohm" cables may not indeed be 75ohm and this would cause problems with reflections at the receiver end. As Bill mentioned, a 75 ohm cable carrying a signal can be called a "transmission line". What this means is that high frequency electrical signals can actually "bounce" off the end of the cable if the cable is not terminated properly. This can cause all kinds of nasty effects.
In addition although what Bill says is OK in theory, we are making the assumption that all the signals on the cable are perfectly in sync to start with with no jitter or timing errors in the clock. This means that the theoretical length difference could actually be much shorter. I wouldn't have a component video cable with different lengths just to be on the safe side.

You can get decent 75ohm cables from somewhere like Farnell and it's a lot cheaper than being ripped off to the tune of £120 for some "high end" video cable that will basically do the same job except in a fancy purple colour. Bear that in mind!

Tuning cable lengths is actually still used to fine tune delays on signals in the video industry I believe - interestingly enough.

Cheers
J

[severnsource]

Yes, if you're into DIY make your own cables up. You willl know that you are getting decent cable. On the other hand, if your projector is like mine and has its RGB inputs are on a high density D type its worth paying somebody 30 quid to put the end on!

If I were using separate feeds I would make them the same length, a 2M loop of cable looks naff. You just don't need to worry about getting them the same length for any technical reason.

Indeed cables are cut to lengths for timing in the TV production industry in PAL systems, but for an entirely different reason. PAL sources at a vision mixing point need to have their subcarrier phase very accurately timed to avoid colour phase errors. The subcarrier ideally should be timed to within 1 degree; you are now talking about differences in cable length of around 2 cms. Phase errors in analogue systems are much more visible than the sort of very minor colour delay error we are talking about here.

Modern digital systems timings are much less critical; most digital mixers will tolerate half a line of mis-timing on their inputs, and some have frame stores, so timing becomes irrelevant - apart from the lip-sync errors of course!

Bill