at which stage the signal cable should be shortest and where its length matters less...?

Providing you use quality cables it will not make any difference, however the shorter you can have them the easier they are to run and hide.

Don’t pay any more than £10 for a 1m signal cable and £2-3 per meter on speaker cable, as you will just be wasting your money.

Hope this helps

Bill

I've already spent £500 on Tellurium XLR..

Ah well, I hope you auditioned the cable before you bought it.

Costwise I would not consider spending a quarter of the price of your ancilliary equipment on a cable, but some people would and do.

Lengthwise you use the shortest you can get away with (and always remember you may re-arrange items in future so do not buy too short) to keep costs down. Over the 'normal' lengths of cable we use there is no difference in sound quality.

This may not be the case if you happen to live in Blenheim Castle and want to run speaker cables full width though.

The most important leads to keep short are amp to speaker.  This is where cable resistance has the bigest effect.

As a general rule, a short lead is a good lead, but in the case of a coax cable used to carry S/PDIF from CDP to DAC, there is a school of thought that suggests the lead should be longer than 1.5M.

For those with an engineering bent who remember their transmission line theory, the argument revolves around the impedance mistmatch of the 75ohm coax with the (usual) RCA connectors and the resulting signal reflections.

Finally someoe has related to my post.. Thank you

My xlr's are 3m each from preamp to mono amps.. i just thought maybe it's too long distance for a signal to travel and i should change it for 2m?

But in traditional hi-fi setup the cables are short between all boxes and long aaaall the way to passive speakers

Baloney!  The most vulnerable part of any system is the front end.  Any distortion/interference introduced at the front will be amplified (that's what an amplifier does to the input signal lol) and will affect what you hear the most.  Speaker cables are largely irrelevant.  The previous poster who I have quoted knows as much about physics as I do about Rumanian folk dancing!  (I do have an engineering bent as my first degree wa in Electronics. lol)

Chris

Hmmn.  Just back from my folk dancing evening.

Taking my points one at a time. (I can only assume you were asleep during the lectures on transmission lines).

RCA connectors have a characteristic impedance above 100 ohms, coax used for S/PDIF is generally 75 ohm. In a mis-terminated transmission line, a portion of the signal is reflected back toward the source.  Assuming the source is equally mis-terminated, this in turn is reflected back to the sink. In a short cable, the reflections interfere with the edges of the pulses, smearing the shape.  Longer cables allow the reflections to occur away from the pulse edges. Doesn't matter a jot if your DAC uses jitter supression techniques, it MAY matter if your DAC is sensitive to input jitter. For those interested:

http://ece.wpi.edu/analog/resources/transmissionline.pdf

As for speaker cables, it is clearly a trade off. Reducing speaker cable length implies longer signal level leads. In an electrically noisy environment, it may be better to keep signal levels short. My system uses balanced interconnects which are less sensitive to this. As a general statement, short speaker cables are better.  While manufacturers are not infallible, I quote the instruction manual that came with my Krell power amplifier:

"Place the amplifier as close to the loudspeakers as possible and keep speaker cable length to a minimum. Krell amplifiers drive the lowest impedances with ease, but long loudspeaker cables reduce the power that is delivered to the loudspeakers"  

Low gain: as short as possible

high current as short as possible!

 RCA connectors are junk & connect the "live" before the screen for starters. If one had a source with XLR outputs feeding a pair of mono blocks, you would have the option of long screened balanced signal leads over long speaker cables where theory would favour the former. I'd speculate that mismatch losses at audio fs are fairly insignificant & even digitised audio over short distances only slightly less so. I don't know how consistent the characteristic impedance of RCA connections are nor how much they effect signal integrity if poor.

balanced (XLRs) for audio & coax (BNCs) for RF/digital would be ideal. My own experience tells me that cheap XLRs are better than expensive RCAs, that I can hear differences between speaker cord but so little difference between RCA ICs as to not worry too much about them.

XLR connections are ideal for the longer cable runs associated with the music industry, and I mean long.

3m is not long. You will hear no difference if you replace it with the same at 2m length.

I am not sure all this talk of RCA connectors is at all relevant.

+1 for that Alears.

In addition, you won't find Telurium XLR cables in studios.

To the OP, you seem to have a very nice system that does not need faffing with in any way, shape or form. Use good quality interconnects, maybe try proaudio stuff.

To andyjm,

5m of copper speaker wire of 2.5mm with have a resistance of approximately 0.017092 ohms

10m of copper speaker wire of 2.5mm with have a resistance of approximately 0.03418 ohms

I don't personally consider this amount of resistance as particulary significant in any part of the audio chain downstream of the sources.

Exactly my thoughts. For a home setup, cable lengths shouldn't matter at all. Certainly not 3m vs 2m.

............ and that is why they make 110ohm coax SPDIF cables

"The resistance of 16-gauge or heavier speaker connection cable has no detectable effect in runs of 50 feet (15 meters) or less in standard domestic loudspeaker connections for a typical 8 ohm speaker."

http://www.roger-russell.com/wire/wire.htm