There are great sounding and not super expensive 2-way designs. Why go for 3-way, then?

[Alexium]

Hi! I'm a noob in speaker design, but I'm trying to learn. For a while I used to think 3-way is better than 2-way, because duh, of course a specialty driver is better. Then I started noticing that prominent speaker builders create both 2-way and 3-way designs in seemingly equal numbers, and there does not seem to be a strong preference either way, even when cost is no object.
In my eyes, 3-way designs have some objective drawbacks:
1. More drivers = larger baffle area required, more difficult construction, higher cost and build effort.
2. More complex crossovers = harder to design right, higher cost.
3. Possible issues with the sound field summing up properly, more distance required from the speaker to the listener? I just made this up, please tell me if it can be an issue or not.

What are the benefits, then?
A use case that I'm most interested in: if we state that the speaker must only go down to 80 Hz, or even 100 Hz, what are the tradeoffs and considerations in the 2-way vs. 3-way debate? And no concert-level output required, let's say normal listening level is 90 dB at most, short peaks notwithstanding.

Obviously, a 3-way is great for getting deep base, but in my experience a free-standing dedicated subwoofer is better than one integrated into the speakers.

 

[AllenB]

The first thing when you decide to use a dome tweeter, you set a crossover frequency and find you can't use a woofer larger than 6.5", preferrably smaller.

 

[indianajo]

The first thing when you decide to use a dome tweeter, you set a crossover frequency and find you can't use a woofer larger than 6.5", preferrably smaller.

That's why great sounding 2 ways like Altec Lansing Voice of the Theater, Peavey SP2 (2004), Pispeakers 4pi use a compression driver. I've heard the first two and bought the second: the last one has great charts. I haven't heard a good 3 way since 1975. AR3, KLH5. I couldn't afford those then and 3 ways are unnecessarily expensive now. I live in a flyover state, and the hifi industry has collapsed to Circuit City & Best Buy, which don't demo anything but car buzzers & have huge warehouse rooms anyway. Hifi meets are in hotel meeting rooms shaped nothing like a home. I heard a decent 3 way in a friend's home, but those were gutless wonders, ie no bass. Build your room for listening, you don't need the measurements of a speaker set out on the front of an open stage. Much more bass is required for the open stage placement, that's why bands always use thousand watt sub.
The first tradeoff is top end. If you can hear 15-20 khz you need a 1" CD or a 3rd driver for that range. Those crossover pretty high. If you can't hear that high (I can't) a 1.4" CD covers 1.2 to 15 khz with great dispersion. Good dispersion means you can walk around the room and the sound doesn't change. Despite all the fear of beaming posted here a 15" can be engineered to cover 54-1200 hz with minus 3 db point of +-22.2 deg (Peavey SP2-XT) or minus 6 db point of +-45 deg (SP2(2004). In a narrow long room (mine is 14' w 11 ' h 30 ' long) that means sound is even from 9 ' away all over the room. The dispersion tweak is using a low excursion 15" and with the SP2(2004) pinching the back of the cabinet to make the reflections spread wide. The crossover of that is 2 khz. (Modern SP2 pushes the crossover higher to achieve 1000 w rating, and dispersion & harmonic distortion charts have disappeared).
Plus by backing the speakers into the hard front wall, my SP2(2004) cover 25-54 hz only 10 db down instead of 30 db loss the middle of the room chart shows. So I don't need a sub.

 

[newvirus2008]

The 3-ways attempt to separate the most important mid-range band from the lows and highs to reduce inter-modulation distortion. Also, the lows are usually not the subs, 3-ways with subs are actually 4-ways.

 

[6L6]

A 2-way with a designed roll-off of 80-100hz and intended to be used with a subwoofer is, effectively, a 3-way design. Yes, the middle driver doesn’t have an electrical bandpass filter, but relieving the middle driver of the low bass is really what it’s all about.

I have heard great 2-way designs. I have heard great 2.5-way designs. I have heard great 3-way designs. I’ve also heard awful examples of all. They all have advantages and disadvantages. My preferences fall into the category of middle-sized satellites + sub or 3-way with big woofers. There is a LOT of musical information in the 100’s of Hz and small speakers rarely get the job done right.

 

[andy19191]

A use case that I'm most interested in: if we state that the speaker must only go down to 80 Hz, or even 100 Hz, what are the tradeoffs and considerations in the 2-way vs. 3-way debate? And no concert-level output required, let's say normal listening level is 90 dB at most, short peaks notwithstanding.

90 dB average is pretty loud, add 10 dB for listening at 3 m. or so, add 15-20 dB for peaks and knock off one or two dB for room reflections means you are looking for a driver to deliver 110-115 dB cleanly at 80-100 Hz. That is a 10-12" driver (or a 2 x 6.5-8" driver). In addition asking for a passband of more than a decade tends to lead to a compromised passband in terms of resonances, radiation pattern, insufficient clean SPL,... Requiring high quality therefore tends to dictate a largish 3 way whereas settling for slightly compromised SPL, resonances, radiation pattern,... means a 2 way of lower cost and complexity. It is of course fun trying to minimise those compromises with design.

 

[mordikai]

Why have speakers when you can headphones and listen anywhere?

 

[youknowyou]

A 2-way with a designed roll-off of 80-100hz and intended to be used with a subwoofer is, effectively, a 3-way design. Yes, the middle driver doesn’t have an electrical bandpass filter, but relieving the middle driver of the low bass is really what it’s all about.

I have heard great 2-way designs. I have heard great 2.5-way designs. I have heard great 3-way designs. I’ve also heard awful examples of all. They all have advantages and disadvantages. My preferences fall into the category of middle-sized satellites + sub or 3-way with big woofers. There is a LOT of musical information in the 100’s of Hz and small speakers rarely get the job done right.

This is the truth

Obviously, a 3-way is great for getting deep base, but in my experience a free-standing dedicated subwoofer is better than one integrated into the speakers.

can you cite example of speakers you liked?
2 way and 3 way?

 

[Alexium]

90 dB average is pretty loud

Thank you for the calculations! Indeed, 90 dB is pretty loud. I have not used an SPL meter, but from what I understand, I'm hardly ever listening at more than 75-78 dB, with an occasional explosion or gunshot in a movie spilling over 80 dB. This is at 2.5-3 meters from the speakers. I guess I have picked the 90 dB figure to be on the same page with more of you guys, but I doubt I will ever need more than 86 dB at which THD is typically measured.

P. S. Have you forgotten to subtract 3 dB when calculating headroom because there are two speakers?

 

[Alexium]

A 2-way with a designed roll-off of 80-100hz and intended to be used with a subwoofer is, effectively, a 3-way design

Absolutely, I agree. But this not what we generally mean when we're talking about 3-way designs. Most 3-ways still need a sub in order to make me as happy with the 20-50 Hz region as 2-way bookshelves with a sub.

can you cite example of speakers you liked?
2 way and 3 way?

That's an interesting question, and I wish I had more experience (read: I wish I had more friends in the audio hobby, and at least one demo room in the area).
I like (but don't own) KEF LS50 Meta. I like the cheap 6.5" + soft dome tweeter DIY speakers I have now. They are made by another person and I don't know the specs, don't even know which drivers are used, but they're quite OK.
I liked ProAc 100, but that was a long time ago, today I might have a worse opinion if I got to listen to them.

I have almost no experience with 3-ways. Don't think I've heard any modern 3-way system. I have a pair of vintage Braun L830 which I liked at first, but now I think I overpayed and they need EQ in the mid-range.
I have looked into Philharmonic BMR and Wharfedale Evo 4.2, but discarded both based on measurements. BMRs have a bit too much distortion, and Wharfedale's frequency response is all over the place.

 

[Alexium]

The 3-ways attempt to separate the most important mid-range band from the lows and highs to reduce inter-modulation distortion.

That makes sense to me, but if that was the case, we should see in general better THD in 3-way designs than 2-way? I am not observing this when reading ASR and Erin's speaker reviews. 3-ways do have lower THD below 150-200 Hz, true, but above that I see no correlation between THD and the number of bands.

 

[andy19191]

Thank you for the calculations! Indeed, 90 dB is pretty loud. I have not used an SPL meter, but from what I understand, I'm hardly ever listening at more than 75-78 dB, with an occasional explosion or gunshot in a movie spilling over 80 dB. This is at 2.5-3 meters from the speakers. I guess I have picked the 90 dB figure to be on the same page with more of you guys, but I doubt I will ever need more than 86 dB at which THD is typically measured.

P. S. Have you forgotten to subtract 3 dB when calculating headroom because there are two speakers?

There is only 1 speaker if the sound source is on the left or right. One could perhaps justify adding a dB or two in the way one tends to add a dB or two for reflections bringing up the perceived average level (but not peak level). Listening above 80-85 dB average for several hours is not wise because it is likely to harm ones hearing. Laws requiring employees to wear hearing protectors kick in around this average level. Peaks and short term can be louder without problems.

 

[Alexium]

Listening above 80-85 dB average for several hours is not wise because it is likely to harm ones hearing

This means I overestimated the average comfortable loudness by 10+ dB.
My Denon AVR does 125 W into 2 channels at 0.1% THD, the speakers are rated at 6 Ohm, and I'm listening at the AVR volume setting of 60-62 (out of 100). I wonder if this information can be used to estimate the decibels. The speakers seem to be unremarkably average in sensitivity, 85-87 dB, I guess.

 

[andy19191]

That makes sense to me, but if that was the case, we should see in general better THD in 3-way designs than 2-way? I am not observing this when reading ASR and Erin's speaker reviews. 3-ways do have lower THD below 150-200 Hz, true, but above that I see no correlation between THD and the number of bands.

Harmonic and intermodulation distortion are different. To read about the differences and see some examples for 2-way, 3-way and subs click on the "Measurement description" here.

 

[Alexium]

Harmonic and intermodulation distortion are different. To read about the differences and see some examples for 2-way, 3-way and subs click on the "Measurement description" here.

Thanks. I thought that IMD is caused by non-linearity, and the level of IMD correlates with the level of THD. Apparently, I'm wrong.

That article is great, it directly demonstrates the benefit of adding a midrange driver.

Still, why is it that not all great speakers are 3-way, and not all 3-way speakers are great?

 

[AllenB]

Non-linearity is not a significant concern with speakers, as long as you make reasonable choices. You should focus on more important things such as why a larger woofer is not capable of bringing higher frequencies to a dome tweeter.

 

[Mario Pankov]

Most 3-ways still need a sub in order to make me as happy with the 20-50 Hz region as 2-way bookshelves with a sub.

Here lies a part of the answer if by a good 3-way you mean a 8-x-1 or even worse - a 6.5-x-1 well, it is not. It would be a step up from most 2-ways if price of drivers does not matter but not a big one.

Now, go on hear a 3-way using a large woofer, such as a 12" or even better - a 15", and you will never want to go back to a 2-way no matter of drivers in it. The dynamics of most live events are not achievable by small woofers, size does matter after all Add to this distortion reduction and port noise/shout (2-way will essentially be a ported one to achieve any bass output) and you`d have a winner with a big gap.

 

[Alexium]

Now, go on hear a 3-way using a large woofer, such as a 12" or even better - a 15"

What if I high-pass it at 80 Hz, do you still think such a large driver is needed (12"+) and will be beneficial?

 

[andy19191]

Still, why is it that not all great speakers are 3-way, and not all 3-way speakers are great?

Not sure what you mean by great but most room installations that take sound quality seriously in a technical rather than audiophile manner (i.e. studios, labs, some home installations) will use 3-4 way monitors/floorstanders with distributed subs. Subs plus 2 ways with large midwoofers and large deep waveguides with compression drivers can approach a fairly high standard of sound quality but normally at the price of more complexity and expense than equivalent 3 ways. They don't make much sense if the objective is high sound quality in the most straightforward cost effective manner. Look to midfield studio monitors from the large established manufacturers to see examples.

 

[Alexium]

Look to midfield studio monitors from the large established manufacturers to see examples

That's fair enough. But all the currently popular 3-way home Hi-Fi speakers in the $1000-2000 range that I've looked at are not great. The only exception is Kef R3, but it's not a conventional 3-way either, its woofer is small. KLH model 5 is more conventional in its layout, and it's not a great speaker.
At the same time, Genelec 8030C is almost flawless (other than bass extension below 60 Hz), and it's a small 2-way.
Genelec 1032 CPM mid-field monitor, EUR 2000 per one unit, is a 2-way. But you're right that it is rather an exception among mid-field monitors.

Why is it that most near-field monitors are 2-way, and mid-field ones are 3-way? Isn't the only difference that mid-fields monitors need to achieve higher SPL at longer distance? Or is it also because a 3-way won't sum up properly at the near-field listening position?