Sign in with
Sign up | Sign in

Objective Benchmarks: Frequency Response

Bringing Home The Bass: 2.1-Channel Speaker Roundup
By

We’re going to start with the objective benchmarks, and since this isn't an audiophile-centric site, we offer some background information to help our readers make sense of frequency-response charts.

Perhaps the best way to understand speaker frequency response is to learn how it is measured. A special audio signal is played over the speakers, which equally covers the entire audio spectrum at the same level. The speaker output is then measured across the part of the spectrum that humans can hear, from 20 Hz to 20 kHz, and the resulting data is the frequency response. A perfect speaker, in ideal conditions, would produce a completely flat response, the same as the test signal. Peaks and valleys in the response graph mean that the speaker isn’t playing back all portions of the audio spectrum equally.

Testing speaker frequency response is best performed in ideal acoustic conditions, such as a lab equipped with an anechoic chamber, and that is something rarely found anywhere except at a high-end studio facility. We’ve done the best we can, but in less-than-ideal conditions, we’re going to see a lot of bumps in the frequency response due to phase cancellation (sound waves from the satellites canceling each other out) or comb filtering (frequency response errors caused when sound interferes with the room it’s measured in or objects in that room). Readers need to keep in mind that our frequency-response measurements are most useful as a comparison tool between the speakers we’re testing, not as absolute representations of reference response.

Before we discuss the results, here’s a quick legend of the audio spectrum:

  • The sub-bass range is from 20 Hz to 60 Hz, and it isn’t heard so much as felt. Sound in this range provides a sense of power.
  • The bass range is from 60 Hz to 250 Hz. This is where you can hear the bass rhythm, especially in the 90 Hz to 200 Hz area.
  • The lower mid-range is from 250 Hz to 500 Hz. It’s also called the bass presence range.
  • The mid-range is from 500 Hz to 2 kHz. This is where you can hear instruments and vocals.
  • The upper mid-range is from 2 kHz to 4 kHz. Higher instruments and most vocals are heard here, and the human ear is very sensitive to changes in this range.
  • The presence range is from 4 kHz to 6 kHz. Too little sound in this range will make the audio seem like it’s coming from farther away.
  • The brilliance range is 6 kHz to 20 kHz. This is where you hear harmonics and clarity, and accented sound in this range will make it seem crisper and clearer with higher fidelity.


Here are the frequency responses of the products in our roundup. We’re not including the Rockus 3D|2.1 in this first graph and we’ll show you why in a little bit. We benchmarked all of these speakers using the maximum bass setting to show the maximum bass response they’re capable of. All five of these 2.1 speaker systems follow a similar line, but there are some notable differences:

There’s a lot of information here, so let’s examine the results line by line.

The green result is the Corsair SP2500. The large 8” subwoofer is capable of strong sub-bass performance. The frequency response curve generally follows the same characteristics of the other products, except for a valley at 500 Hz and a larger dip at 3.5 kHz. These concern us a little, but we acknowledge that it might be a symptom of limitations with our test setup, such as phase cancellation or comb filtering. Nevertheless, since the speakers were meticulously placed and aimed in the same manner as the other test subjects, we have to consider these dips and pay extra attention when we test the speakers subjectively.

Next, the purple line is the response of the Creative Gigaworks T3. This system puts out a little less sub-bass than its contemporaries, but the rest of the response is very neutral until we reach the brilliance range over 10 kHz, which is somewhat surprising when you consider the small 2” drivers. We’ll keep an eye out for a minimal brilliance response when we listen to games and music with this system.

The yellow line represents the Klipsch Promedia 2.1, a speaker system that does a good job of defining the average response of our test group. It rarely shows the highest or lowest response in our tests, and has no obvious weaknesses except a drop-off in the 10 kHz+ range, which is a little surprising when you consider that these speakers are equipped with dedicated tweeters.

The orange line is the result of the Logitech Z623 speakers, which perform similarly compared to the Promedia 2.1s, with slightly less audible bass, but stronger brilliance (although at over 5 kHz, the response does get a bit peak-ish).

Now let’s consider the Rockus 3D|2.1 on a separate graph:

The yellow line here is the same Klipsch Promedia 2.1 response from the last graph, which is something we’re using as a baseline for comparison purposes. The blue line represents the Rockus 3D|2.1 in music mode and the green line represents that system’s 3D mode. As mentioned previously, all tests are taken with bass settings at maximum.

When the Rockus 3D|2.1 is in music mode (the blue line), we can see a flat overall frequency response, but that’s tempered with some significant peaks and valleys and a relatively low bass frequency response combined with a lot of brilliance. In 3D mode (the green line), we see significantly boosted bass along with accentuated lower and upper mid-range output.

These results give us a little concern that the Rockus 3D|2.1 might have some uneven mid-range response and low bass. We should point out that excessive bass that drowns out the rest of the audio spectrum is not necessarily a good thing. And once again, we’re aware that phase cancellation and comb filtering may be to blame for these uneven results, so we’ll put more faith into our subjective testing.

Ask a Category Expert

Create a new thread in the Reviews comments forum about this subject

Example: Notebook, Android, SSD hard drive

Display all 97 comments.
This thread is closed for comments
Top Comments
  • 12 Hide
    damasvara , January 6, 2011 6:34 AM
    Audiophiles dissing common audio listeners = Hardcore PC gamers dissing console gamers

    Typical... :pfff: 
  • 11 Hide
    cleeve , January 6, 2011 6:03 AM
    clownbabyWow, those freq response graphs are pretty telling that computer speakers are basically all trash.


    If you know audio, you know that the environment has a *MASSIVE* effect on response recording. Even moving the mic a few inches in the same environment can change the result by a large amount.

    As we've stressed in the article, we don't have the luxury of a professional-grade anechoic chamber for testing. Our results are likely heavily degraded by comb filtering and phase cancellation, but they can be used to compare speaker output to each other because they were all tested in the same conditions.

    The bottom line is, don't write these products off based on a response taken in less than ideal conditions.

    Use these response graphs for what they are good for--not absolute, but relative measurements.
  • 10 Hide
    miaaron , January 6, 2011 3:03 PM
    A few comments:
    1) The Creative Gigaworks T3 does have 3 drivers in the sub enclsoure, but two of the drivers are not active/powered. They are more commonly referred to as passive radiators and they just tune an enclosure like a port would. If you want more info on how ports work, search for "helmholtz resonator".

    2) Don't knock the speakers for the freq response measurements, or any other speakers measured freq response when someone measures them in a regular room. What those freq measurements are showing is the freq response of the speakers in their room...not your room. That's why all the response curves look alike for the most part, because the acoustic patterns of the room are going to dominate the measurement. So don't look at the measurements themselves, look for deviations from the overall pattern to identify issues, like the dip in the response of the sp2500 between 2k-5k.

    I've done a LOT of room measurements and can say the author did a decent job here. If he had no other audio experience before diving into this project, it shows he really did his research...or got really lucky. lol The people with the nasty/mocking/snob comments should actually take the time to do some in-room measurements, they would be suprised.

    3) Don't let anyone tell you a PC speaker sucks because it is a PC speaker. I've built dozens of DIY speakers, and I bought a logitech Z-5500 for my bedroom. Why, cause I couldn't build them for the price I paid ($160 AR @ buy.com). I could have built something very similar if I wanted, as the Tang Band driver logitech used in the z-5500 was well known to the DIY community, but it wouldn't have been a nice without a lot of effort & extra money.

    4) Plastic enclosures aren't bad. The enclosure material doesn't matter as long as it blocks the rear wave of the cone output and doesn't resonate. Plastic actually has a huge benefit in small speakers like this, as a simple curve or some added thickness can add a lot to strength...something cheap and easy to do in a plastic mold.

    5) Don't be afraid to use the EQ built into many sound drivers.
Other Comments
  • 1 Hide
    hmp_goose , January 6, 2011 4:14 AM
    It's you: Where the hell are my 5.1 sets?
  • 4 Hide
    Mark Heath , January 6, 2011 4:20 AM
    Interesting.. thanks for the article. :) 
  • 1 Hide
    jazn1337 , January 6, 2011 4:40 AM
    Dang, I was hoping you guys would look at the Swan M10s.
  • 7 Hide
    clownbaby , January 6, 2011 4:50 AM
    Wow, those freq response graphs are pretty telling that computer speakers are basically all trash. The bass peaks and generally crappiness in the mid range seem to be a common theme. Almost no consideration seems to be given to music listening.

    2.1 is the ideal setup for a computer imo. 4.1 at most. A center channel just isn't needed for monitor sized screens.

    You can buy a cheap onkyo receiver, some low end bookshelf speakers and a small sub for a few hundred bucks and have sound that will destroy the best pc speakers.

    The fact is, pc speakers are toys. There is no high end option. What they market as high end would be laughed out the door by the regular audio comminuty.

    p.s. Plastic is not an acceptable cabinet material
  • 0 Hide
    clownbaby , January 6, 2011 4:52 AM
    this is what frequency response graphs of decent speakers should look like.

    http://www.speakerdesignworks.com/StatementCenterChannelResponsePlots.JPG
  • 3 Hide
    tigerwraith , January 6, 2011 4:53 AM
    My Logitech G51 speakers have a Headphone and mic passthrough on the remote.
  • -2 Hide
    clownbaby , January 6, 2011 4:54 AM
    this is a frequency response graph of the first diy speaker I built from a popular design. This is a super budget MTM speaker.

    http://www.speakerdesignworks.com/TritrixMTMfr.gif
  • -1 Hide
    d0gr0ck , January 6, 2011 5:03 AM
    clownbabyWow, those freq response graphs are pretty telling that computer speakers are basically all trash. The bass peaks and generally crappiness in the mid range seem to be a common theme. Almost no consideration seems to be given to music listening.2.1 is the ideal setup for a computer imo. 4.1 at most. A center channel just isn't needed for monitor sized screens.You can buy a cheap onkyo receiver, some low end bookshelf speakers and a small sub for a few hundred bucks and have sound that will destroy the best pc speakers. The fact is, pc speakers are toys. There is no high end option. What they market as high end would be laughed out the door by the regular audio comminuty.p.s. Plastic is not an acceptable cabinet material


    Pretty much this. I've been telling people for ages that their super-duper PC speakers aren't. Any brand that quotes max power over RMS values raises an instant red flag for me. Even 20yr old Radioshack shelf speakers can run circles on most modern PC speakers.

    I die a little bit every time I hear someone with a premium add-in sound card is running generic PC speakers.
  • 0 Hide
    tigsounds , January 6, 2011 5:25 AM
    This is all Go out and buy it junk. Build your own and end up with something that rattles the neighbors nerves if done right.
  • 3 Hide
    Mark Heath , January 6, 2011 5:46 AM
    For all those who trash all PC speakers, they're usually the best option on the lower end of the scale. There are people out there who have compared entry level (sub 400) active speakers to the Klipsch Promedia set (best active speakers ~150 for sound quality imo) and they say that they're not that different. If you do it right, then it's not as bad as you might think.
  • 11 Hide
    cleeve , January 6, 2011 6:03 AM
    clownbabyWow, those freq response graphs are pretty telling that computer speakers are basically all trash.


    If you know audio, you know that the environment has a *MASSIVE* effect on response recording. Even moving the mic a few inches in the same environment can change the result by a large amount.

    As we've stressed in the article, we don't have the luxury of a professional-grade anechoic chamber for testing. Our results are likely heavily degraded by comb filtering and phase cancellation, but they can be used to compare speaker output to each other because they were all tested in the same conditions.

    The bottom line is, don't write these products off based on a response taken in less than ideal conditions.

    Use these response graphs for what they are good for--not absolute, but relative measurements.
  • 2 Hide
    Anonymous , January 6, 2011 6:05 AM
    I would be interested to compare these with the Harman Kardon 2.1 soundsticks III, which i own and i am very happy with, even though they are (only) rated @ 20W RMS (Sub) + 2 x 10W RMS (satellites).
  • 5 Hide
    pandemonium_ctp , January 6, 2011 6:29 AM
    Quote:
    Is it just us, or are 5.1- and 7.1-channel speaker systems impractical for PCs?


    It's just you (guys). Gamer immersion? IMHO more channels is more important for gaming than for watching movies (which I also do with my 7.1 setup; TV speakers = 2 additional).

    Simple rules for buying good speakers (and anything peripheral):
    -Stay away from wireless
    -You can't really go wrong with Logitech, Bose or Altec Lansing (though not as good as they used to be)
    -High RMS/Watt output means next to nothing (unless you're deaf and need speakers at high volume all the time and don't care about quality of tone)

    TigsoundsThis is all Go out and buy it junk. Build your own and end up with something that rattles the neighbors nerves if done right.


    I'm all for building your own, except most people won't know the properties that are important for stereo systems and will wind up with sub-par performance compared to cheap store-bought crap. If you're just after rattling your neighbors nerves then you're just a douche. :/ 
  • 12 Hide
    damasvara , January 6, 2011 6:34 AM
    Audiophiles dissing common audio listeners = Hardcore PC gamers dissing console gamers

    Typical... :pfff: 
  • 0 Hide
    Anonymous , January 6, 2011 6:52 AM
    I upgraded my computer speakers this holiday season and let me just say they are AMAZING now. I can't believe how much better real speakers are than PC speakers. I also tossed in a high end sound card and my music experience is surreal.

    If anyone is curious, google for "WAF-1 Rosewood Pair" the price they are selling for right now is a STEAL right now. Unfortunately they ran out of the matching 2-channel amp, so you need to find an amp elsewhere. I also picked up a BIC V1020 subwoofer because I like dance music. I'll never go back to normal PC speakers again.

    Price wise, it's the equivalent of getting a top-tier video card, but for your sound system.
  • 2 Hide
    specter_jester , January 6, 2011 7:02 AM
    I've no regret till now of my 1.5 years old Logitech Z2300 ;-)
  • 1 Hide
    reasonablevoice , January 6, 2011 7:37 AM
    clownbabyWow, those freq response graphs are pretty telling that computer speakers are basically all trash. The bass peaks and generally crappiness in the mid range seem to be a common theme. Almost no consideration seems to be given to music listening.2.1 is the ideal setup for a computer imo. 4.1 at most. A center channel just isn't needed for monitor sized screens.You can buy a cheap onkyo receiver, some low end bookshelf speakers and a small sub for a few hundred bucks and have sound that will destroy the best pc speakers. The fact is, pc speakers are toys. There is no high end option. What they market as high end would be laughed out the door by the regular audio comminuty.p.s. Plastic is not an acceptable cabinet material

    I've been saying this for years. It is 100% true.
    As per damasvara's comment about audiophiles dissing casual listeners, nothing could be further form the truth. We are not criticizing you, we are criticizing these low end speakers! We are trying to tell you that there are much better products out there for the money. Come, join us, you'll never look back.
  • 0 Hide
    Anonymous , January 6, 2011 8:23 AM
    Much more expensive but please vbuild and review a system like this:

    B&W CM5 + Rotel RC-1550 + Rotel RB-1552 + Optical Out Sound Card
  • 1 Hide
    ZakTheEvil , January 6, 2011 10:01 AM
    I'll never fall for another overpriced computer-specific speaker system. SPDIF connected to Home Theater receiver and a set of decent 5.1 HT speakers works for me well and adds the flexibility of multiple inputs and a fully featured remote, also has EQ, compression for night time listening, etc. Movie surround modes work great for games too.
  • -1 Hide
    Hupiscratch , January 6, 2011 10:21 AM
    My Bose Companion 3 is doing great, and have the auxiliary jacks you´re asking. It should be on a next test.
Display more comments