Sine sweep used for impulse response

[tony22]

I want to say first this post is in no way intended to be critical of REW. I’ve used it successfully for years on various speakers and consider it a great tool (I think at one time I was a member of this forum, but that was some years ago). However, my concern is driven by things I’ve been told recently by speaker manufacturers at shows, regarding their not recommending the use of REW for fear it could damage some of these newer driver technologies - specifically coated beryllium or ceramic tweeters of various sorts. I’m being told the sine sweep, especially at longer sample lengths, could damage these drivers. I have no way of knowing if this could be true.

I always carefully Check Levels and make sure I fall within the recommended operating parameters with REW, so I’m not sure if there’s anything that would hurt these tweeters. Of course my systems are not being inherently overdriven, for example, by poorly matched or underpowered amplification. So I don’t know if there’s any merit to these claims. I do occasionally use longer sample lengths, if that matters.

Guidance here is appreciated.

 

[John Mulcahy]

The sweeps are logarithmic, so spend equal time in each octave. For a 20 Hz to 20 kHz sweep for example, covering 10 octaves, that means 1/10th of the sweep time is spent in that last 10 kHz to 20 kHz octave. I have corresponded with quite a few speaker manufacturers who routinely use REW for their own measurements, so there is no inherent concern. You should, however, measure at levels you are comfortable listening to (mainly for the sake of your ears). For in-room response measurements there is no need for high levels or long sweeps, 128k or 256k sweeps are fine to get good results,

 

[tony22]

The sweeps are logarithmic, so spend equal time in each octave. For a 20 Hz to 20 kHz sweep for example, covering 10 octaves, that means 1/10th of the sweep time is spent in that last 10 kHz to 20 kHz octave. I have corresponded with quite a few speaker manufacturers who routinely use REW for their own measurements, so there is no inherent concern. You should, however, measure at levels you are comfortable listening to (mainly for the sake of your ears). For in-room response measurements there is no need for high levels or long sweeps, 128k or 256k sweeps are fine to get good results,

Thank you, John. It’s good to hear from you directly on this. Yes, I never go beyond the recommendations in Check Levels, and when I do the sweeps they typically result in 8-10 dB of remaining headroom.

 

[gatestick]

Sort of a related question. Speakers are designed to reproduce sound, so how/why would reproducing a perfect sign wave damage them? I would think that reproducing music with distortion (heavy metal) would be more likely to damage them.

 

[John Mulcahy]

Drive units have limits on the power they can handle. A small driver like a tweeter may be very limited. Long duration, high level, high frequency signals can exceed its ability to dissipate heat and damage or destroy it.

 

[AudiocRaver]

I have been using REW to measure speakers for 10 years with all kinds of tweeters, and with every measurement session, I always always ALWAYS ALWAYS start with a very low level signal and carefully increase it no more than 10 dB at a time to the 75 dB SPL range for measurements. I have never harmed a tweeter.

It is true that at the end HF end of a log sweep, all of the signal is going to the tweeter, while a pink noise signal is constantly and randomly changing in frequency. But pink noise, if averaged over time to get a nice smooth plot, can end up putting as much power into a tweeter as a sine sweep.

• Take your time
• be cautious
• use reasonable signal levels
• use shorter sweeps
• always use log sweeps

and you won't have any problems.

Used carefully, a short, moderate level log sine sweep is as safe for tweeters as pink noise.

BUT:
If you don't trust REW's sine sweeps, you can always use pink noise for short durations and put up with lower resolution results.

 

[AudiocRaver]

Metal music is no more likely to damage a tweeter than a moderate test signal, but an amp driven to clipping can sure fry one!

(I have never done that either).

 

[Igor Kurilovich]

Greetings everyone! My experience using REW is about 7-8 years. Of course, the first thing that attracted me was that the program is free! During the use of REW, I felt a significant professional growth! I began to understand the interactions of various physical factors and their influence on the formation of auditory images. This happened thanks to an incredibly intuitive interface, a large number of YouTube videos, and to say the least, this forum - a big thank you to everyone and John Mulcahy personally! My post does not question the achievements of the authors of other programs for acoustic measurements! But many times I was involved in disputes with colleagues - which software is better? The beginning of such discussions was almost always the statement that good cannot be free! But you can't fool your own ears! and most importantly, the laws of physics and mathematics do not work selectively. My conclusion is very simple - commercial software is usually either limited in functionality, or oriented to the user's qualifications, or to a limited scope of application, or all together. Approaching the initial question of this branch, I will express my personal opinion - manufacturers may not be satisfied with the existence of REW for one reason - it is a tool with which you can obtain objective parameters of loudspeakers, systems, rooms and compare the obtained data with the data of publications for free!!! P.S. I have great hope that my passion will turn into a business, then REW will turn into a commercial program for me!

 

[AudiocRaver]

Greetings everyone! My experience using REW is about 7-8 years. Of course, the first thing that attracted me was that the program is free! During the use of REW, I felt a significant professional growth! I began to understand the interactions of various physical factors and their influence on the formation of auditory images. This happened thanks to an incredibly intuitive interface, a large number of YouTube videos, and to say the least, this forum - a big thank you to everyone and John Mulcahy personally! My post does not question the achievements of the authors of other programs for acoustic measurements! But many times I was involved in disputes with colleagues - which software is better? The beginning of such discussions was almost always the statement that good cannot be free! But you can't fool your own ears! and most importantly, the laws of physics and mathematics do not work selectively. My conclusion is very simple - commercial software is usually either limited in functionality, or oriented to the user's qualifications, or to a limited scope of application, or all together. Approaching the initial question of this branch, I will express my personal opinion - manufacturers may not be satisfied with the existence of REW for one reason - it is a tool with which you can obtain objective parameters of loudspeakers, systems, rooms and compare the obtained data with the data of publications for free!!! P.S. I have great hope that my passion will turn into a business, then REW will turn into a commercial program for me!

Well said.