barefoot
barefootsound.com
In my Monitor Mythology post I told an odd little tale to illustrate why frequency response isn't a good measure of monitor quality. Here is some actual data to prove my point.
I put three signals through a spectrum analyzer.
1. A pink noise test signal - this represents the frequency response of a very nice set of monitors with about +/-2dB response from about 20Hz to 20kHz.
2. A sort of 'frog' sound in the midrange about -6dB down.
3. The sum of both the pink noise and the frog - this would be the effect throwing a poor little frog into your monitors and measuring their response. (i.e. the distorted response of monitors with a frog in their throat).
http://www.barefeet.bigstep.com/generic.html?pid=5
The yellow curve is the 'frog', the pink curve is the pink noise, and the blue curve is the sum.
The most striking thing about these curves to me is how little difference there is between the strait pink noise response and the sum response. While there is an approximately 2dB step at 1639Hz, the overall variability in the sum response is no greater than the pink noise. This type of blip can, and often does, occur at crossover points. In fact, if I were going to try and improve the blip, I would not the touch the band where the frog sound was added at all. I would use a band pass filter to slightly boost (active) the frequencies between 1639Hz and 3139Hz, leaving the midband frog distortion completely intact. Even if I chose the distorted band, the most I could cut is 2dB. This would have virtually no effect on reducing the audibility of the distortion since it's only -6dB quieter than the base response.
I have attached a short mp3 of both the frog sound and the sum of the frog and pink noise. Hearing any signal inside broad band noise is always difficult, and it's even harder due to the degradation induced by converting to mp3. Still, I think you should be able to make out the frog sound within the noise. To give an even clearer example I'll add another mp3 to this thread of a piece of music with obviously audible distortion added, but the response looks virtually identical to the undistorted signal.
So, don't make the mistake of thinking that just because a monitor has a flat response it is necessarily accurate. Flat response is a relatively easy design parameter to obtain these days. Low distortion, on the other hand, is not.
barefoot
I put three signals through a spectrum analyzer.
1. A pink noise test signal - this represents the frequency response of a very nice set of monitors with about +/-2dB response from about 20Hz to 20kHz.
2. A sort of 'frog' sound in the midrange about -6dB down.
3. The sum of both the pink noise and the frog - this would be the effect throwing a poor little frog into your monitors and measuring their response. (i.e. the distorted response of monitors with a frog in their throat).
http://www.barefeet.bigstep.com/generic.html?pid=5
The yellow curve is the 'frog', the pink curve is the pink noise, and the blue curve is the sum.
The most striking thing about these curves to me is how little difference there is between the strait pink noise response and the sum response. While there is an approximately 2dB step at 1639Hz, the overall variability in the sum response is no greater than the pink noise. This type of blip can, and often does, occur at crossover points. In fact, if I were going to try and improve the blip, I would not the touch the band where the frog sound was added at all. I would use a band pass filter to slightly boost (active) the frequencies between 1639Hz and 3139Hz, leaving the midband frog distortion completely intact. Even if I chose the distorted band, the most I could cut is 2dB. This would have virtually no effect on reducing the audibility of the distortion since it's only -6dB quieter than the base response.
I have attached a short mp3 of both the frog sound and the sum of the frog and pink noise. Hearing any signal inside broad band noise is always difficult, and it's even harder due to the degradation induced by converting to mp3. Still, I think you should be able to make out the frog sound within the noise. To give an even clearer example I'll add another mp3 to this thread of a piece of music with obviously audible distortion added, but the response looks virtually identical to the undistorted signal.
So, don't make the mistake of thinking that just because a monitor has a flat response it is necessarily accurate. Flat response is a relatively easy design parameter to obtain these days. Low distortion, on the other hand, is not.
barefoot
