GratisVolver™ – Convolution for the Rest of US…

By Pat Brown

GratisVolver™ is a software convolution program. Finally it is possible to convolve the IR with dry program to listen to it.

So what good is an impulse response, once you have it? There are a number of “scores” that can be used to assess it. While these are useful, they don’t tell you the most important thing about it – what it sounds like.

For years, users of advanced acoustical measurement systems have looked at graphs of IRs and tried to figure out from the various metrics whether the response was good/bad, better/worse, etc. We didn’t have a simple way to convolve the IR with dry program to listen to it. Many sound practitioners shunned measurement altogether and relied solely on their ears to diagnose problems, as in many cases they found them more reliable than the measurements. It’s now possible to efficiently combine measurement and listening to extract all of the information contained in the IR.


GratisVolver is a software convolution program that is simple, fast, and best of all, it’s free. It was written by Bengt-Inge Dalenback, developer of the CATT-Acoustic™ acoustical prediction program. Bengt-Inge is a member of the Syn-Aud-Con listserv and generously cranked out the application to support the new Syn-Aud-Con IR Exchange (see article, this issue). GratisVolver allows anyone to perform fast, accurate convolutions on any of the IRs found on the exchange. Here is how it works (Figure 1).

GratisVolver™ Screenshot showing the convolution process

  • 1. An anechoic WAV file is specified in section 1.
  • 2. The IR WAV file is specified in section 2. The sample rate and other info is shown in the window. Note that the sample rate of the IR and the anechoic file must match.
  • 3. The convolution is performed by hitting the Convolve button. Automatic file naming and playing is supported. The file name will be created from combining the file names of the anechoic and IR files. Optionally, the user can overwrite a fixed WAV file, so that convolutions can be performed without accumulating a lot of files on your hard drive. I normally use this option, since the convolutions are nearly real-time.

That’s it! If it sounds easy it’s because it is. We use GratisVolver in our Sound Reinforcement for Designers course to evaluate the relationship between loudspeaker Q and listener distance in a number of auditoriums. Convolution allows the various acoustic metrics to “come to life” and reveal much more insight than a single-number score.

We naturally assess rooms by clapping our hands and listening. The first “look” at an IR should just be to play it back in a wave editor and listen to it. Many room problems, such as flutter echoes or “slap backs” can be easily detected with this simple test. The next step is to listen to speech or music that has been convolved with the IR. I can easily remember when a few seconds of convolved audio required a significant wait and some pretty high-powered software.

Since it is so small, GratisVolver can be run concurrently with your favorite measurement program (Fig. 2), allowing you to listen to the measured data as the various metrics are determined. I use some short sound clips of male voice, female voice, drum kit, and cello for the convolutions. Since measured IRs are likely to be either 44.1kHz or 48kHz, I keep a version of my dry files with each sample rate. This speeds up convolutions by removing the need to resample the IR to the needed rate.

GratisVolver™ and EASERA screenshots


GratisVolver can also be used to recover the IR from a recording made in the room. This process is called deconvolution, and it is accomplished by convolving a recorded sweep file with an inverse version of the dry sweep file that was played into the room (Figure 3).

Chart showing the convolution process

Here’s how it works:

  • 1. An inverse version of the “dry” sweep file is specified in Section 1. Note that the inverse file is an element-by- element division (hinv = ifft(1./fft(h)).
  • 2. The “wet” mono or stereo sweep that was recorded in the room is specified in Section 2. Note that the sample rates of the dry and wet sweep must match.
  • 3. The deconvolution is performed by hitting the Convolve button. The IR file name will be created from combining the file names of the anechoic and IR files. This file can be opened in any wave editor and trimmed to the desired length.

Almost any portable recorder can be used to record the sweep in the room. Just follow good recording practices (avoid overload, use an appropriate microphone and placement, etc.). A personal computer can also be used, but I prefer the freedom from the PC offered by simply making recordings on a portable device.

GratisVolver puts IR collection and convolution within the grasp of anyone. It makes an excellent supplement to any of the popular measurement platforms, since it processes WAV files as its native format. A user can use their favorite measurement application to capture IRs, and then use GratisVolver to convolve them, or they can use GratisVolver to gather the IRs, and their favorite measurement application to analyze them. You don’t need the latest-greatest PC to use GratisVolver. The program is about 50kB in size, and convolutions are nearly real-time.

We are pleased to add GratisVolver to the growing list of powerful measurement and prediction tools available to audio practitioners. It is rare for such a useful tool to be so small, fast, and available at no cost to the end user. We applaud the developer’s synergy in sharing this with the audio community. GratisVolver can be downloaded from

Supporting files and impulse responses can be downloaded from