Hertz 101: How many is good and what is the big deal?
Hertz is the number of cycles or times a sound wave hits a peak every second. If you want technical knowledge go to The Physics Classroom. But wait! We don't need physics. In speakers, Hertz is the number of times the cone of the speaker goes in and out per second. So at 20Hz, the cone is going in and out 20 times every second. At 20k Hz the cone is going in and out 20,000 times per second. The faster the cone moves the higher the tone. The farther in and out it goes in each cycle, the louder it is. Very simple.
With a little logic we can surmise that a big heavy woofer cone is not going to get to 20,000 in-out "cycles" per second (at least not sounding very good), nor is a tiny cone tweeter going to push much air at 20 cycles per second. So we have woofers and tweeters to handle each part of the job they do best. Then there are all kinds of related problems like stopping the cone cleanly at each end of the in/out cycle and loudness which requires the cone to move farther in and out. Some cones have a ring that cause distortion at some frequencies, especially lighter cones (which help the clean stopping problem) like Kevlar, metals, fiberglass, and any harder material. The list of problems goes on and on. These are all things that either add to the sound wave or subtract from the clean wound wave. That is what speaker building is all about, getting the components to produce the correct sound with nothing added or subtracted. We have to use the right tool for the right job.
Why are we obsessed with 20-20k Hz? Because that is supposed to be the range of normal hearing. Most of us fall short of this range. I know I cannot hear the full range and most honest people will tell you they can't either. If you are over 35 you most likely have lost some of the upper range. If you have ever been to a rock concert you have lost some of your lower range as well. What is interesting is I know I can pick a favorite speaker out of a group and notice details I should not be able to hear according to my ear charts. So there is something more than just hearing the waves. It is clear that some people do have very good ears and can pick out very small changes in sound quality. It may be due to physical ability, genetics or training. But the ability is there. Can people really hear the difference between $100 and $1000 cables? Who knows? I hope I can never tell because I can't afford expensive cables. A reasonable guess would be that most of us can hear from about 30Hz to about 9000Hz.
Then there is recorded music. One would assume most CD and LPs (those larger black discs with a small hole in the middle) are mixed and designed to play the full range of music. This is mostly not true. The engineers who put these things together know 99% of the people who buy music can't tell it is full range or not, are going to play the CD on lousy equipment or in a car, and are not going to listen carefully to the mix quality anyway. The engineers have choices and technical limitations on what they can and cannot do. They mix CDs for a mass market and low cost. So without trolling for arguments with engineers, we can say that most CD are mixed to be less than full range and compressed so you can hear soft passages in a car. Home theater is just the opposite. They figure you have a sub turned way up and you want rumbles and explosions and little shakes when the bad guy comes on screen. They add low tones. This is why so many audiophiles cringe when people ask advice for a system that will play 70% music and 30% HT. Music and HT really have different needs and it is easier and cheaper to get a clean 40-20kHz for music than 20-20k for HT. And the last octave (20-40Hz) can get pretty sloppy with most HT stubs, killing the musical enjoyment.
One of my favorite systems right now is a vintage Marantz receiver and a pair of DIY monitors. It sounds great and I just love to listen to it. How often do I miss notes in the low end? Only sometimes. This system has an f3 of around 60Hz. That means the speakers have lost 3 decibels of volume at 60Hz and it will continue to lose volume as the notes go lower. An f3 of 60 is fairly high for most systems. But if I look at the Music Range Chart, you can see I am not really missing very much, the bottom few notes of the bass, tuba, harp and accordion. On the piano 60Hz is the 15th key from the left including the black keys. These notes are simply not played that often. I am probably missing some harmonics on some CDs but not much. Also, just because the F3 is 60, that does not mean all the notes below are gone, just that they are at lesser volume. When I do run into a CD heavy with low notes (like Holly Cole's Temptation), I go to the HT system with the subs or my system with 8" woofers. Many audiophiles will launch into discussions about the low pedals of the organ and the bottom bass notes. But I don't own much organ music. For most of us there just is not much in the low end on most CDs. You can test this by playing your favorite music CD on your HT and lightly placing your hand on the sub cone. You won't feel much unless you crank the volume. You will get more for your money if you abandon the bottom octave.
There is also a power issue. I don't know the exact percentage, but a lot of your amplifier power is for loud and low. If you are running a full range speaker and playing loud, probably 50% of the Watts are going to the bottom octave. If you cut the bottom octave or use a sub, you don't need nearly the number of Watts. This is why so many HT systems don't stress power, if the sub has a separate amp, the mains don't need that much. I quote my old professor (probably dead by now) when I say you really don't need more than a true 60 Watts for 99% of systems. Again, the message is to understand where you are putting your money and why.
So now I have convinced you to go for better sounding smaller monitors and you are thinking subwoofers are the answer. And if you want that bottom octave, subs *are* the answer. But I have two issues with subs: integration and integration. The first integration is in the general sound of subs. It is very hard to get them to blend with the monitors. Logic would dictate that the first octave would not be that tough since your cone is moving so slowly. But to me they just sound different. The second integration is with amplification rates and I think this problem also occurs in bi-amping. We may be able to match volumes of the sub and mains at a specific level but what happens with that level changes? Do all amps amplify at the same rate? If we match SPL at 80db, will there be a match at 70db? I don't think so. This is not such a big issue with HT but I think I really hear it with music. I will have to test this more thoroughly when the kids are in college and I have more time. So while there is a valid argument for monitors and a sub, now I suggest a full range will integrate better.
Buuuut, you say. My Super Duper Max Tech Boomer Kevlar Elite Stealth speakers have a published response of 20-20000Hz +-3db. Like published amplifier power specs, the response for most brand name speakers are misleading. In fact, the specs are probably less accurate for speakers than for amps because with amps you can more easily document distortion. If you run a 20Hz signal through a table top radio and the woofer moves, you are getting a "response". Does is sound good; no. How about the tweeter? I have yet to hear a tweeter that does not make noise at 20000Hz. I have one pair that will make dogs howl (special deals-today only). But that is not what I want in a speaker. So don't rely on specs. This goes for the woofers and tweeters we builders are buying as well. Get your own specs or rely on the experience of others for speaker parts and only your ears for speaker systems.
In conclusion, the goal is to evaluate what your goals in system building are and buy toward that goal. Getting the full range out of a system is cool but not really needed for the majority of music listening. You will get better sound for your money if pull back from a full range systems and build for specific goals rather than going for a mega system. But if you have the money and desire, go for it.
Now lets apply this to speaker building:
Look at the chart below. The green areas represent "Joe driver's" optimal operating range. The yellow areas represent the driver's extended range or less than optimal range (see disclaimer below). The lighter purple areas extend the instrument range to include rarely played notes, say the first and last 10 keys on the piano. Comparing the instrument range vs. the driver ranges, you should get an idea of the speaker building problem: no driver does everything well.
You can see why 6.5" two-ways are so popular. Almost any 6.5" driver can fill out the mid range, but the trick is getting a driver to do it well while allowing enough room for the next driver to take over. A 6.5" driver may be into its distortion range at the upper end, the lower end or both. Every driver is going to have a shortcoming at one end or the other. We are stuck between having a driver fall short in one area or adding drivers and crossover parts to try and blend them.
Take a look at the white papers from Linkwitz lab. Mr. Linkwitz was just looking for low distortion drivers, but his results support: 1) finding a driver to handle a large portion of the listening range is difficult, 2) expensive drivers measure better in a broader range, 3) smaller drivers measure better in the higher ranges while larger drivers measure better in the lower ranges. We all know this intuitively but it is nice to see someone use rocket science to get there.
Not only must have range extension to get to the next driver, the drivers must transition smoothly with like sound or you will hear the change-over, especially in the transition from mid to harmonics. In my mind, the order of difficulty is: Mid to Harmonics, Mid to Bass, Harmonics to Airiness. The larger the driver the harder it will be to blend upward, the smaller the driver the harder it will be to blend downward.
This whole line of thinking really supports the idea of building kits when starting this hobby. The kit designer will have already done the work in choosing drivers that match and designing a cross-over. Going it alone can be fun but it will be harder and good results will take longer.
Disclaimer: This chart is for conceptual use. I chose four octaves plus one octave for the distortion/roll-off range. This is to define the best operating range for any "Joe driver". Clearly this does not present all drivers in any size group. Many drivers have a broader operating range than the six octaves shown, particularly very expensive drivers and really cheap drivers ;) A driver's published range is usually the mechanical limit, not the distortion limit. Before you write me to tell me I am nuts, consider two questions: 1) how far below the f3 are most drivers at the bottom of the yellow range; 2) how bad is the off axis at the top of the yellow range? And please don't write me about your 4" driver that goes to 20Hz in a TL.
Copyright Peter Jay Smith 2005 Return to helarc.com