Does it matter what speakers are used with REW? Noobi question

[GSark]

Hi, this is a simple question and I can't find an answer. I've read a lot of documentation and watched a bunch of videos about REW, and am playing with it on my home system. I've not seen any advice on what speakers to use. The given directions are all the same: hook up your speakers. So does it make a difference which speakers you use or not? If so, what are the minimum requirements for these speakers? Can I use my computer speakers which are decent quality and include a sub-woofer? Does REW somehow account for speaker quality/linearity during the calibration phase?

Posted the same question on Gearslutz and got a couple of answers but no advice. The answer: yes it makes a difference. But no one states what difference it makes or what minimum speaker requirements are. Since I've not bought monitor speakers yet, would like to clear up the confusion on this point before I move forward on my studio design. Thanks for your help.

 

[Wayne A. Pflughaupt]

The purpose REW is to measure the performance of the speakers. You can use any speakers you like. The minimum requirements you saw for speakers are about what separates speakers that sound good from speakers that don’t. Obviously you want speakers that sound good.

Regards,
Wayne

 

[John Mulcahy]

Sounds like you want to do some testing to help with the studio setup before you get your monitors. To get an idea of the modal behaviour of the studio you need speakers that can cover the range of frequencies you are interested in. Computer speakers and sub might well do the job to see what the studio is doing and what treatment is helping, but you will need to make more measurements with your monitors when you get them.

 

[GSark]

John, exactly right. I'm in the process of understanding the acoustics of my room so I can do a proper job of acoustic treatment. Perhaps you can confirm my understanding of the process? In theory the result from REW is an impulse or transfer function of the room. These parameters define how the room affects sound delivered to it from a certain location. Ideally, the transfer function should be independent of the speakers that are used to generate the test signals.

But in the real world, the speakers used to generate the tone will affect the final results because the speaker has its own non-linear characteristics, especially in the highest and lowest frequencies. The more linear the speaker response, the more accurate the REW analysis, and therefore a monitor type speaker (that has flat response) will give better results than a typical hi-fi speaker system. So I'm guessing it doesn't matter much which monitor is used once a certain quality threshold is met.

 

[John Mulcahy]

To clarify some terminology, it isn't really the linearity of the speaker that matters but its bandwidth and its response flatness. The swept sine measurement technique is quite good at rejecting harmonic distortions (non-linear behaviour) but if the speaker can't generate a meaningful sound level below some frequency limit you won't learn anything about the room's behaviour below that limit. If the speaker's response is not flat then that will be reflected in the measurements, though typically the variations in a speaker's anechoic frequency response are far, far smaller than the effects of the room's modal distribution, so flatness isn't a big concern. Bear in mind that the room's transfer function depends on the position of both the source and the mic, a room doesn't have a single transfer function.