music frequency/energy content

[pjanda1]

I'm pondering some enclosure designs for fullrange drivers. Obviously most are fairly dynamically limited compared to what I think necessary. However, I'm just guessing at what I think is necessary.

I would like to know, for example, if a typical rock tune is reaching 105db peaks (A weighted), how much energy (in db would be great) is there between 80-100hz? (I know that can damage ones ears, this is an example.) What's the answer for a typical late romantic orchestral work?

I assume I looking for spectral analysis, but I'm not really sure. Anybody know where I might find this information? I've found tons of software via google, but of course, I'd prefer to find the work already done.

thanks,

Paul

[noodle_snacks]

I am not sure if i am answering your question but your real enemy here is cone excursion. Most full range drivers that i will know of will hit cone excursion well and truely before reaching 105db in the bass region. I suspect that a full range driver being driven under any large amount of power would probably also suffer from a good amount of power compression as well. In my opinion if you want to achieve 105db from 150hz down you would be better of going for a two way design, or you will have to consider the use of subwoofers.

If you want to look at where the peaks lie in the music you wish to listen to then i would suggest downloading the trial version of this software:

http://www.ymec.com/products/dssf3e/

Then setting the input to the output of your computer or running a cable from a cd player etc. Set the scale from 0db to -50db and note where the peaks are. It is simple to do and hte easiest way to see where the peaks lie in the type of music you like listening to.

[Pano]

This is an interesting subject and one which can hold some surprises.
I don't have any numbers for ready you, but there is some free spectrum analyzer software out there that will run on PC.

With the software you can have a good look at an audio file and study where the energy is. For rock and pop it tends to be around 500 and 1K hz, as well as 125-200 Hz.

It would be nice to have a software that would run a analysis of an entire file, or batch of files, then give you an integrated graph of the energy spectrum. Goldwave can do this for peak and average levels, but doesn't do it for spectrum (AFAIK).

That doesn't get you the answers you wanted, but if you have the time it is fun to play around with the software and see where the peaks and energy are.

[pjanda1]

Originally posted by noodle_snacks
I am not sure if i am answering your question but your real enemy here is cone excursion. Most full range drivers that i will know of will hit cone excursion well and truely before reaching 105db in the bass region.

Indeed, this is the problem. I'm using a sub, but I'm looking at enclosure options for drivers. Different vented alignments run out of excursion in slightly different bands. I'm currently favoring sealed overall, and crossing to a sub around 80.

Originally posted by panomaniac
This is an interesting subject and one which can hold some surprises.
I don't have any numbers for ready you, but there is some free spectrum analyzer software out there that will run on PC.

With the software you can have a good look at an audio file and study where the energy is. For rock and pop it tends to be around 500 and 1K hz, as well as 125-200 Hz.

My fullrangers don't look too impressive on paper. However, they are 8" with 2mm of excursion and many folks are happy with smaller drivers that have .5mm. They all have problems below 100-150hz. However, as you said most energy from rock is a little higher up. I know that most music resides in the midrange, yet I at least tend to think of needing similar dynamic capabilities everywhere.

I was hoping someone would have written a paper on this, as it seems like it should interest a speaker designer. This is an issue for smaller woofers in multi-way speakers. This knowledge would help people ensure that tweeters in two way systems have adaquate capabilities in the bottom of their passband.

I've got a Mac, so I'll try and find something that will work for it. It will be hard to amass enough data to give reputable results. I also don't necessarily trust the accuracy of all freeware. Were I an engineer, I'd write a paper with credible results.


Paul