Yamaha's "Physics of Bi-amping"

[kenshone]

Reading around the net, it looks like the topic of bi-amping is more polarizing than American politics.

I came across this article by Yamaha about the theoretical benefits of bi-amping: https://hub.yamaha.com/audio/a-how-to/how-and-why-to-bi-amp-your-speakers/

They have a section in the article which specifically discusses the benefits of separating the transmission of low and high frequencies, in the subsection "The Physics Behind It."

A quote of interest:

The problem is this: If that thud of a bass drum and the ting of a triangle are sent down the same wire at the same time, there is potential for audio degradation of the sound of the triangle. That’s because high frequency signals (with their relatively low currents of electricity) are susceptible to being unduly influenced by the higher-current low frequency signals (and their associated magnetic fields) simultaneously being sent through the same wire.

They are implying that bi-amping reduces the distortion that is generated within the speaker by full spectrum audio signals. Is my interpretation correct?

I assume Yamaha isn't lying, but what variables affect the magnitude of the effect that I could expect to see when 1) vertically bi-amping (which I'm strongly considering right now) and 2) horizontally bi-amping passively crossed speakers.

P.S. I am fully aware that horizontally bi-amping with an active crossover produces great results, but I want to limit the discussion to potential improvements in passively crossed speakers.

 

[Beave]

I guess Yamaha thinks that wires are now nonlinear devices.

I wonder what other crazy beliefs Yamaha would like to share with us.

 

[ryanosaur]

But... The audio signal isn't separated until it hits the Passive XO in your Speaker. From the Source, to the Pre-Amp, to the Amplifier, and ultimately to the Speaker, the Signal is the Signal. All content is included.
In a Passive Speaker, it is the XO Network that filters that Signal to each Driver (or set of Drivers).

The only time I would consider Passive Bi-Amping worthwhile is if I had a Bass section that was extraordinarily difficult to drive. Most Speakers may have some low Impedance Minima where more power may be required (say, below 80Hz) but in many situations, a single strong Amp Channel can handle the work. If there is a challenging Phase Angle along with that Impedance Minima, and you are not crossing to a Sub... what does it serve you to buy twice as many Amp Channels?

Active Speakers (whether Internal Amps and DSP or External) are a completely different story.

This all just comes across as so much marketing to make you question the efficacy of it and hopefully buy more Yamaha product.

 

[kenshone]

With the caveat that I've never formally studied electricity, in a vertical bi-amping passively crossed scenario, wouldn't each channel of the amp see a different demand for signal?

In general, I understand that all things equal, lower frequencies demand more power.

So the tweeter's channel on the amplifier (whether L or R) will generally demand less "power," even though it will be receiving the same audio frequencies. Will the fact that the frequencies are being transmitted to the tweeter with less power improve the tweeter's sound quality ceiling, irrespective of the quality of the passive crossover?

Same question for the woofer, although I understand that it's harder to perceive distortion as the frequency decreases.

 

[kenshone]

I've been reading through this post from someone who looks like he knows his EE: https://www.audiosciencereview.com/forum/index.php?threads/bi-amping-101.22817/.

According to points 6 and 8, he touches on the same theoretical benefits of biamping that I touch upon in my previous post. His analysis is amp-centric, which is great. Does anyone have a similar analysis focused on the effect on the speaker?

The poster of the linked thread says he's never tried passive biamping, so he can only give real-world performance predictions based on theoretical analysis.

I would also appreciate commentary based on first hand experience with passive biamping.

 

[sergeauckland]

I guess Yamaha thinks that wires are now nonlinear devices.

I wonder what other crazy beliefs Yamaha would like to share with us.

Exactly this. They should brush up on the Superposition Principle as to why they're wrong.

S

 

[kenshone]

Another angle on one of my lines of questioning: Will each driver in a 2-way passively crossed speaker see a difference in current between the vertically bi-amped vs the non bi-amped configuration? If so, does this altered allocation of current cause the speaker to behave differently (ideally better)?

 

[Beave]

Another angle on one of my lines of questioning: Will each driver in a 2-way passively crossed speaker see a difference in current between the vertically bi-amped vs the non bi-amped configuration? If so, does this altered allocation of current cause the speaker to behave differently (ideally better)?

No.

 

[kenshone]

So the speakers wouldn't see a difference in current allocation, provided each of the amp's channels are capable of supplying the asymmetric power that is demanded between them. Perhaps there is different behavior when only the LF channel exceeds the amp's power limitations, but the HF doesn't (assuming they have independent limits -- I guess this would depend on the amp's design?).

In an attempt to understand how some people can claim to hear an audible improvement in sound quality (SQ) when vertically bi-amping, based on my understanding so far, here is a hypothesis. I'm assuming that some audiophile is AB-testing between a conventional single-amp setup and a vertical bi-amp setup with the same kind of amp.

Assume:
- for some fixed sound (volume, frequencies, etc.) that a 2-way speaker is asked to produce, the tweeter demands some power level A and the woofer power level B. There is some power level C at which the amp could meet the speaker's demands, for producing this fixed sound, under a conventional non-bi-amped setup (not going to claim C = A + B. IDK enough about electricity)
- A < B < C
- the amp's performance varies significantly with power, such that noise and distortion differences are substantial at power levels A vs. B vs. C

And per @Beave above, under vertical bi-amping, one amp channel would be outputting power level A and the other power level B.

Under these circumstances, is it not plausible that there would be an audible difference?

---

One simpler potential factor for an improvement in SQ could be that the amp that's being AB tested vs. its bi-amped configuration has high cross-talk. As I understand it, vertical bi-amping eliminates cross-talk just like mono-bridging would.

And of course, it could be a combination of the above.

So to summarize, generally speaking, the hypothesis is: the worse the amp's SQ, the more likely that performance will be improved by vertically bi-amping.

 

[Beave]

The most common reason that people (claim to) hear a difference when vertical biamping is not due to any electrical phenomena.

It is due simply to sighted listening comparisons, ie, a flawed comparison methodology.

 

[ryanosaur]

So the speakers wouldn't see a difference in current allocation, provided each of the amp's channels are capable of supplying the asymmetric power that is demanded between them. Perhaps there is different behavior when only the LF channel exceeds the amp's power limitations, but the HF doesn't (assuming they have independent limits -- I guess this would depend on the amp's design?).

In an attempt to understand how some people can claim to hear an audible improvement in sound quality (SQ) when vertically bi-amping, based on my understanding so far, here is a hypothesis. I'm assuming that some audiophile is AB-testing between a conventional single-amp setup and a vertical bi-amp setup with the same kind of amp.

Assume:
- for some fixed sound (volume, frequencies, etc.) that a 2-way speaker is asked to produce, the tweeter demands some power level A and the woofer power level B. There is some power level C at which the amp could meet the speaker's demands, for producing this fixed sound, under a conventional non-bi-amped setup (not going to claim C = A + B. IDK enough about electricity)
- A < B < C
- the amp's performance varies significantly with power, such that noise and distortion differences are substantial at power levels A vs. B vs. C

And per @Beave above, under vertical bi-amping, one amp channel would be outputting power level A and the other power level B.

Under these circumstances, is it not plausible that there would be an audible difference?

---

One simpler potential factor for an improvement in SQ could be that the amp that's being AB tested vs. its bi-amped configuration has high cross-talk. As I understand it, vertical bi-amping eliminates cross-talk just like mono-bridging would.

And of course, it could be a combination of the above.

So to summarize, generally speaking, the hypothesis is: the worse the amp's SQ, the more likely that performance will be improved by vertically bi-amping.

I’m not an EE, but the Amp does not send a signal out based on the power demand of the speaker, rather it sends a signal out based on the power demand as selected by the user at the volume knob. With any Passive Speaker, it is the XO Network that is the gate for the signal coming to the Speaker.
Also consider that if the Amp attached to the Speaker was in fact sensitive to the Speaker, you would never blow Drivers. (This is not the case!)
There may be something that changes with the load each Amp Channel “sees…” but nothing that changes how many watts or volts are sent through the cable.

If there were truly some benefit, why is there no evidence to support it?

It is the marketing departments that seem to put these notions out. Never have I talked to a Speaker Designer who thinks Passive Bi-Amping does anything that you can’t achieve through a single Amp Channel of good clean amplification.

As before, in an active design, you see people running all kinds of unique setups attempting to tune the sound beyond even DSP (SS to the woofers, tubes to the tweeters, for example). Or you see some hybrid designs with powered bass sections and passive up top. Even there, the smart designer is tuning the DSP to work with passive network rather than leaving it in the hands of the user.

 

[Plcamp]

I would expect at best a small measurable improvement in the amps IM distortion simply because it has limited current bandwidth to deal with. I would not expect an audible difference.

It does allow user to select a lower power higher SINAD amp for the hf range, and many open baffle enthusiasts swear that’s an improvement (but I doubt it).

 

[kenshone]

I’m not an EE, but the Amp does not send a signal out based on the power demand of the speaker, rather it sends a signal out based on the power demand as selected by the user at the volume knob. With any Passive Speaker, it is the XO Network that is the gate for the signal coming to the Speaker.
Also consider that if the Amp attached to the Speaker was in fact sensitive to the Speaker, you would never blow Drivers. (This is not the case!)
There may be something that changes with the load each Amp Channel “sees…” but nothing that changes how many watts or volts are sent through the cable.

If there were truly some benefit, why is there no evidence to support it?

It is the marketing departments that seem to put these notions out. Never have I talked to a Speaker Designer who thinks Passive Bi-Amping does anything that you can’t achieve through a single Amp Channel of good clean amplification.

As before, in an active design, you see people running all kinds of unique setups attempting to tune the sound beyond even DSP (SS to the woofers, tubes to the tweeters, for example). Or you see some hybrid designs with powered bass sections and passive up top. Even there, the smart designer is tuning the DSP to work with passive network rather than leaving it in the hands of the user.

Right, I've just read that the flow of electricity has many similarities to the flow of water. I don't have the time right now to delve into it, but the amplifier dictates the total amount of current flowing into the system, and the characters of each driver and the crossover would dictate how that current is allocated between the two drivers.

I'm sure my hypothesis was way too handwavy and full of EE errors. TLDR: The crux of my hypothesis is that the difference in current flow through each channel under v. bi-amping would make a difference if the rate of flow has a significant effect on the amp's signal quality. If you look at the graphs of the good amps, their THD+n plotted against power has a gentle and predictable slope within their rated power boundaries. I actually haven't bothered looking at graphs of bad amps, but I'm sure it would be instructive...

 

[Soniclife]

there is potential for audio degradation of the sound

It's always a red flag when you see this type of language, it's just FUD, you can write this about every aspect of audio. If there is degradation detail it.

 

[techsamurai]

What happens when you bi-amp using an AVR? Does the channel send the full signal over both wires or does it separate frequencies?

 

[twsecrest]

What happens when you bi-amp using an AVR? Does the channel send the full signal over both wires or does it separate frequencies?

I'm assuming you mean bi-wire, where you use the same A/V receiver, but use two channels (one for woofer and one for the mid-range/tweeter) to drive the front speakers?

 

[techsamurai]

I'm assuming you mean bi-wire, where you use the same A/V receiver, but use two channels (one for woofer and one for the mid-range/tweeter) to drive the front speakers?

No, I meant when you use the bi-amp feature on an AVR that uses the extra 2 channel amps. Does the AVR split the frequencies across the cables sending high frequencies on 1 cable and Low Frequencies on another or does it send the full signal on each cable?

 

[HarmonicTHD]

No, I meant when you use the bi-amp feature on an AVR that uses the extra 2 channel amps. Does the AVR split the frequencies across the cables sending high frequencies on 1 cable and Low Frequencies on another or does it send the full signal on each cable?

No.

 

[theREALdotnet]

No, I meant when you use the bi-amp feature on an AVR that uses the extra 2 channel amps. Does the AVR split the frequencies across the cables sending high frequencies on 1 cable and Low Frequencies on another or does it send the full signal on each cable?

It sends the same signal via both sets of amp channels. The speaker’s internal cross-overs are still in place if you remove the bi-wire bridges, after all.

 

[gene_stl]

There are situations where this kind of setup can be useful. It at the very least should be somewhat better than using a regular simple full range set up.
I once tried to help someone do similar "bi-amping" on a pair of Infinity speakers. They had a rear firing tweeter (which I recommended disconnecting) and were a five way system (six way with the rear firing tweeter). Accordingly these speakers were famous for their gyrating impedance which required an amplifier that was regularly comfortable with a two ohm load, something which was NOT common when the speakers were new. Thus removing the link and operating as a two way , even with both amps running full range would make the load seen by each amplifier less complex. Of course it would be better if there was an active crossover but although the full range-ness of it could eliminate some of the headroom that could be available it is the impedance of the speaker that draws the power out of the amp and probably generates some of the intermodulation the elimination of which makes bi and tri amped systems sound better. It also would throw away the better voice coil direct damping of a real multi amp system.

Many AVRs (and processors) can do bi-amping using their multi channel outputs.

The notion of "bi-amping" without an active crossover to band limit the amplifier feed signals is a recent mostly stupid development but not as stupid as Bi Wiring which is only to sell double sets of speaker cables. But it probably is very slightly better than full range. For each amp out put the impedance rises sharply outside of that speakers range which in most cases would make the amplifiers job easier. But for a two way system you can build a passive RC network for almost no money so why not do that. I did that once in the early seventies on a two way JBL system and it improved the sound a lot.