I conclude that bi-wiring does have a significant effect in my system.
Swapping cable lengths and adding more copper has no or very little effect.
Swapping cable lengths and adding more copper has no or very little effect.
I’m not sure what I’m looking at here in your measurement setup photo……..I see only the TMM section of the speaker?
Follow by the model Bruno proposed, one with observable difference with bi-wiring might want to check if the current distortion is high in their speaker drivers.
If tweeter is low inductance ribbon tweeter and you use some special low inductance braided cables for tweeter and mid, then you can see some difference in measurements.
That could be a pretty significant problem. There are various systems for determining where speakers should be placed in a room. Also, there may be a need to treat the room in places with damping or diffusing materials/structures. Such things don't have to look bad. A bookshelf with different sized books can help. Things can be hung on the wall that look sort of like art, but that have some acoustic control effect.
Regarding sound stage, it refers to a perceived area between, around, and behind the speakers in which various instrument and or vocal sources are perceived to be located. To make that work well, among other things the speakers need to be positioned to do their best in helping to produce the perceptual illusion.
If there were to be further interest, I could probably share a collection of various speaker positioning methods in my dropbox.
Good star-quad often works better than braid. I already gave away one secret to making star-quad better, which seems to be all around ignored. This reminds me that my friend told me not to give away valuable secrets in the forum. I said, yeah, I understand. But, they will ignore anything useful anyway, so there is no risk from most of them. The only real risk is that a few professionals might take note.
I would love to see some measurements, star quad vs. normal.
Would you do that, Mark? (Assuming you use star quad 🙂 )
Would you do that, Mark? (Assuming you use star quad 🙂 )
Here is a view of the magnetic fields around 2-conductor and star-quad cable as viewed from the end of the cable:
Star-quad has four conductors in a controlled-geometry configuration such that if diagonally opposing conductors are connected together at the ends where the connectors are, most of the magnetic fields cancel out. That lowers cable impedance (not just resistance), reduces radiated EMI, and reduces susceptibility to pickup of EMI from external sources.
However, that's just the basic idea. Cable construction can get more complicated and involve more factors than just one important technical feature.
Star-quad has four conductors in a controlled-geometry configuration such that if diagonally opposing conductors are connected together at the ends where the connectors are, most of the magnetic fields cancel out. That lowers cable impedance (not just resistance), reduces radiated EMI, and reduces susceptibility to pickup of EMI from external sources.
However, that's just the basic idea. Cable construction can get more complicated and involve more factors than just one important technical feature.
Last edited:
Regarding measurements, I am more or less on a crusade to get people to understand the limitations of steady-state (PSS) FFT measurements. Its hard to get people to understand more than they do now for various reasons. One big reason is that most people don't have a very good understanding of how FFTs work in the first place. They also don't understand that world has changed from when we were just trying to measure nominally linear amplifiers, and measure them in a mostly EMI-free environment.
To cut to the chase, we do measure here where I am but rarely with FFT spectral analysis. Why? Its good for some things but not for clearly showing everything that people can hear. Of course everything is measurable in principle, and even today new measurements are being developed to meet modern needs. However, if people don't really understand FFTs then how should we expect them to understand some of the new stuff? Also, Hilbert transforms have been around for a long time, and could in principle be applied to audio reproduction, but there are no studies relating things like measured envelope accuracy with audibility. Thus, such measurements would be largely ignored as meaningless anyway.
Bottom line for me, its cheap and easy to try star-quad, cheap and easy to strip off the rubber jacket and try it again. If you can't hear a difference with 3 or 5 meters of cable then your system may have other problems.
Whether the difference shows up in an easy to see way in FFT spectral analysis or not, it doesn't mean there is proof of audibility one way or the other (except maybe in pretty simple cases, such as for obvious changes in FR or HD). Why is that? The reasons are involved. Probably this is not place go deep into that.
To cut to the chase, we do measure here where I am but rarely with FFT spectral analysis. Why? Its good for some things but not for clearly showing everything that people can hear. Of course everything is measurable in principle, and even today new measurements are being developed to meet modern needs. However, if people don't really understand FFTs then how should we expect them to understand some of the new stuff? Also, Hilbert transforms have been around for a long time, and could in principle be applied to audio reproduction, but there are no studies relating things like measured envelope accuracy with audibility. Thus, such measurements would be largely ignored as meaningless anyway.
Bottom line for me, its cheap and easy to try star-quad, cheap and easy to strip off the rubber jacket and try it again. If you can't hear a difference with 3 or 5 meters of cable then your system may have other problems.
Whether the difference shows up in an easy to see way in FFT spectral analysis or not, it doesn't mean there is proof of audibility one way or the other (except maybe in pretty simple cases, such as for obvious changes in FR or HD). Why is that? The reasons are involved. Probably this is not place go deep into that.
Last edited:
Oh, like microphone cable. That should be good for your 50' speaker cable runs. It's a really easy way to remove, or for speakers, block noise. I think it's much better for loudspeaker wires than complicated braiding. I tried a good number of braids and heard no difference from two good conductors. With the quads you might lose some high frequency information but I doubt anyone can hear it.
You underestimate the knowledge of many engineers here.
I presume that you don't use star-quad, otherwise you would be more than happy to present us a graph like the one I did. No need for an FFT, a simple frequency plot would suffice. It would equally boost our confidence in your explanations.
I find my old ribbon tweeter measurements with different cables. Lines are normalized to one measurement, CAT3 named is 3 CAT6 cables braided together, rest are 2,5 mm2 standard cables, all are about 3 m long.
Thanks for that chart.
It seems like CAT3 isn't really suitable as a (ribbon) tweeter cable.
I assume it has a high capacitance?
It seems like CAT3 isn't really suitable as a (ribbon) tweeter cable.
I assume it has a high capacitance?
Yes, relatively high capacitance but very low series inductance. I had no problems with this cable and ribbon tweeter as my amp have low output impedance also in high frequency range.
Note that lines are the result of SPL measurements divisions, lower level means higher response compared to base.
Last edited:
You presume too much.
Here is a pic of my planar headphones with a stripped star-quad cable. Also shown is the braided cable that came with the headphones:
The two cables differ in various ways, including length, and total resistance. It wouldn't just be a comparison of star-quad geometry verses not.
For my big speakers, the cable is custom manufactured, then it is cut to length and the connectors put on here. They are a proprietary design, and are not sold. I have agreed not to talk about their construction, but I will say it does involve a symmetrical geometry.
Last edited:
- Status
- Not open for further replies.