Archive for the ‘audio’ Category

WinISD Accuracy and Subwoofers

with 2 comments

To those of you who plan to use WinISD for predicting subwoofer maximum SPL.

Be weary of using WinISD to predict maximum output. WinISD can be a great tool for predicting speakers at small amplitudes, but it is a little naive concerning the non-linearities that occur with high amplitudes.

WinISD makes all of its predictions based off of T/S parameters, which are small-signal specifications and they are not always scaleable to larger power levels. You hope they don’t change in order to use them as an indication of performance, but they are better used for on deciding on box size and box type.

It’s also important to remember two different speakers with the exact same T/S parameters can have extremely different parameters at a given X-max. With small drivers it isn’t that big of a deal since X-max isn’t a big factor, but high excursion drivers are a different story. It’s safe to say that frequency response changes with power output especially for lower frequencies and WinISD doesn’t model the changes since it models in an ‘ideal’ world.

By ideal I mean WinISD assumes that a driver won’t suffer from power compression and more generally assumes that a driver’s response won’t change as additional power is applied.

Take a look at AV Talks Tests and Ilkka’s Tests. You’ll notice that every one of the tested drivers suffer from power compression at lower frequencies and the lower you go the greater the problem becomes. As Keith Yates puts it, “Power compression is the audio equivalent of getting shortchanged.” WinISD does not give you any indication of when your design is going to get shortchanged, which makes it very hard to compare to subwoofers that are already tested.

Another important piece of WinISD that is missing is that it doesn’t indicate how THD levels relate to SPL and frequency. For example, WinISD may show a maximum ouput of 114dB at 40Hz, but it won’t tell you how much distortion exists at that SPL level.

Further reading on high amplitude scenarios: here

Written by syntatic

April 28, 2008 at 8:44 pm

Posted in audio