Hi All,

Modeler Plus 6.5 has a unique feature that allows you to add background noise to your simulations. The software contains a library of pre-recorded noise files: the NCB curves, the RC curves and several custom noise files. We are now expanding the usability of background noise in Modeler® software by introducing two related applications, Modeler® NOISEGENERATOR and Modeler® NOISECONVERTER.

The first application – NOISEGENERATOR – is used when you have the spectrum of the background noise, but not a binaural recording of it. NOISEGENERATOR allows you to define the frequency response – either in Modeler or directly in the application – and the application will produce a loop of spatialized, binaural noise with the defined spectrum and a .noi file. Copy and paste the associated .noi and .wav files into your Modeler data folder, and then analyze background noise effects through Modeler and Auditioner simulations.

The second application – NOISECONVERTER – is used when you have a binaural recording of the background noise, and simply want to process it for playback in Auditioner. NoiseConverter reads in the raw binaural recording and outputs a loop of the background noise as a .wav file and a .noi file, that you copy and paste into your Modeler data folder, and then analyze the background noise effects through Modeler and Auditioner simulations.

For now the binaural recordings have to be made with the Earbuds and the ProRecorder, but we plan to extend this library based on requests from the field. If you need the Earbuds, please start with your local field engineer.
Enclosed you will find both applications along with manuals.

To run NoiseGenerator and/or NoiseConverter:
Install MATLAB Component Runtime, provided in the FTP folder. Upon successful installation, simply double-click on the NoiseGenerator.exe file to launch the application. An MS-DOS window will first appear, followed by the application itself. Please keep this window open in order to properly run the NoiseGenerator application. Clicking on the Help button in the lower right-hand corner of the application will bring up the User’s Guide.

The MATLABComponent Runtime is fairly large (150M) and can be found through the following ftp site: ftp://m6-files:qd34iw83@ftp.bose.com/

Here is the NoiseGenerator

NoiseGenerator is for now also located on the ftp site.

The NoiseConverter is attached to this note.

Have fun with the applications and please post any comments or issues you may have.


Original Post
Nice picture !

Welcome, Paul. Good to see you here.

I was recently asking myself whether you guys got ahead with your speech level measurements. Jan from TNO has now analyzed several 1000 speech files and we pretty much concluded what to recommend in terms of a simple measurement.

Hope to talk to you soon !


Hi Kevin,

there are currently no mics (Earbuds) and preamps (PR-1) available from Base ops. So you will have to look for somebody in your region who may have one left over from the good old days and can loan it to you. Note that you only need the PR-1 if you need to have level-calibrated recordings using the calibration tone.

If absolute level scaling is not important or if you have other ways of calibrating, you may get OK results using very small (1/8") mic capsules that fit into or onto your ears, but be aware that the frequency response may not be perfectly equalized as with the Earbuds and PR-1.

Morten may have further information when new Mics and PR-1 devices will be available from Base Ops.

As a workaround that does neither ensure proper calibration nor frequency response, there have always been Earbud-type mics from a German company called Soundman. Here's their website: OKM Mics
This mics sound pretty good and have been used for decades by bootlegger's around the globe to record concerts.

Hope this helps.

Hi Thomas

Thanks for the link to OKM. Looks like a useful product, but before I commit my pocketmoney, will we see the Bose calibrated unit sometime soon, and will it be a similar price?

If not, do I need the mic only or the one with the A3 adaptor, if I am going to use them with a MD recorder or digital recorder with 3v power supply? It’s not clear from their website. I realise you’re not a rep for OKM! But you usually know these things.

Best regards
Hi Thomas

Would love to pick this one up with you. Any chance of a chat soon?

Best regards

Originally posted by ThomasS-at-Bose:

I was recently asking myself whether you guys got ahead with your speech level measurements. Jan from TNO has now analyzed several 1000 speech files and we pretty much concluded what to recommend in terms of a simple measurement.

Hope to talk to you soon !


Hi Paul,

The OKM's run on standard "Plug-In power" mic inputs. Maybe SNR can be improved using a dedicated preamp. I would always recommend to record straight to .wav with an SD-Card type recorder since you need to post-process the recordings and conversion from MD to .wav is therefore needed.

I used the OKM's of a friend of mine to record his concert last Friday. Processed for Auditioner playback, it sounds pretty amazing ...

Here's a pic of what we did. The mics are mounted to a basketball that hangs from the (very low) ceiling. We used this Rastafarian, dreadlock-equipped cap to improve the rear localization (nah, just kidding...)

Regarding the Speech Level stuff, I'll send you a private mail.

Let me know in case you have any additional questions.

Best rgds

Hi Thomas,

I've recorded a childrens event in our church gymatorium (Sancta-noisium) using our ear bud mics.

I've recorded a 3 minute segment I would like to process. It looks like we're limited to a 60 second loop using Noise Converter.

When I listen to the Pre and Post NoiseConverter files, there does not seem to be much difference in the EQ or localization. It appears that all NoiseConverter doing is calibrating level, looping the wav, and creating the .noi file.

Is this correct?

Is there any need to process a longer file to be played through the PB if recorded with the Bose ear buds? (I do know I need to calibrate the level with the Calibration tone)

BTW - I did notice the overall pitch was slightly lower in the processed wav file. I'm assuming there is a little stretching going on to tie the ends of the loop together.

John R
Hi John.

Sorry for the late reply. Here are some answers to your questions:

Maximum recording length for Noise tools: Yes, the max length is around 1 minute.

The difference in terms of microphone EQ is relatively small but not insignificant: The swing in EQ is roughly +/- 2 dB with a relatively broad reduction in level between 500 Hz and 4 kHz. I haven’t tried myself but this should be clearly audible. There’s no other processing involved except the level scaling and the loop-crossfade. All processing to get from binaural to transaural (Auditioner) is performed during playback.

I’m not sure about your last question. For modeling work, I find that the 60 sec limit is absolutely sufficient. Nonetheless, one may want to listen to a longer binaural recording on Auditioner for other purposes (e.g. a recorded A/B comparison of two loudspeakers or 15 minutes of sheep or a bootleg concert recording, see above). For this purpose, you need to use the ‘WavProcessor’ application which you can find on the Modeler Plus CD. This app does the EQ (for both mic and PBU) and the binaural to transaural processing but not any absolute level scaling. Of course, this file must be played back via a standard .wav player, e.g. iTunes and not via Modeler.

I’m surprised to hear about the pitch change. There’s nothing like this going on in the crossfade-processing (why should it do that anyway ?).

Hope this helps.

Best rgds

Thanks for the reply Thomas.

I think the WavProcessor was the app I was looking for.

For Curiousity's sake, I've attached the pre and post processed noise files to let you hear the pitch shift.

The EQ difference you noted may have been masked by the pitch shift.

I've converted these to MP3 for file size. If you want the .wav files let me know.


Hi Tom,

we haven't updated any of these tools and as far as I know, don't plan to do so. I'm pretty sure you'd have to find appropriate Mathworks Runtime Components as well.

Sorry that I don't have any better news. Anything in particular you'd need to accomplish in this regard ? Can you use the WavProcessor ?

Best rgds,


Add Reply

Likes (0)
Having trouble signing in?

We recently updated our sign-in procedure and if you have old sign-in data cached, this can create a problem. Please:

  1. Clear your browser cache and cookies
  2. Then close the browser (not just the window)
  3. Open the browser and try again
Thank you

Please make sure that your profile is up to date