des sons dans l'air...

préludes à l'espace

écouter

l'Acousmonef

l'espace concret

l'espace du faire

galerie

l'acousmonde

voyages en acousmonef

télécharger

les acousmobiles

les acousmodules

les pages du faire

documents

calendrier

utopies et uchronies

de l'autre côté

l'acousmonhome

causeries illustrées

l'acousmathèque

bios

visions de l'acousmonde

le temps du faire

audio blog

des sons délyrants

liens

singularités

abécédaire

 
d o w n l o a d
 

 (french version)

You can download the files for the pieces in the Preludes to Space series here. These can be used for diffusions/concerts or listened to in an educational context (most are available in binaural reduction on the listen page). For longer pieces, please contact me.

I offer the following formats:

1. The 80-point nodal space files correspond directly to the allophonic composition format as I use it in the Acousmonef. They can be adapted for all "true 3D" systems, that is, those based on a multilayer mesh in the three dimensions of space. This mesh can be linear, like 4DSound, or irregular, as in some large acousmoniums, without any significant loss of spatiality. Obviously, the more diffusion points there are, the better... We can consider around thirty channels to be a minimum value for correctly representing layering in three dimensions, the absolute minimum being 18 in the form of two levels of 9 (8 around and 1 in the center).

2. For single-layer "immersive" surround sound systems, which are by far the most numerous today (domes, formats derived from cinema, etc.), I performed 56-channel spherical space transcriptions with distance simulation for the original surround sound, then encoded it in 7th-order (64 channels) or 10th-order (121 channels) ambisonics. Here too, around thirty points/speakers is a good value (starting at ground level), even though the process can handle greater reductions.

3. For acousmoniums that lack diffusion points with sufficient elevation differences, I propose reductions in the form of a horizontal grid of 25 points. These should be easily adaptable for systems with roughly the same number of broadband points, most evenly distributed across the entire surface.

4. I can also send the settings or optimized transcription files for any device (including, under certain conditions, for octophonic circles), the choice of compatible components depending directly on its characteristics (see examples).

All files are in Wavepack lossless compression format, 48 kHz, 24-bit. They are directly playable in Reaper.
Except for the ambisonic encoded files, their levels conform to the K-System in K-20. There is no dedicated low-frequency channel; the signals sent to the subwoofer(s) must be obtained using the Bass Management technique (see notes at the bottom of the page).
The plugins and software indicated for making the adjustments are free, but of course, everyone is free to use others ;-)

Download:
- Reaper for Linux, Mac, and Windows (unlimited trial version)
- the Acousmodules SimpleTest 64 plugin for Windows and Mac (it's not essential but very useful for making adjustments and calibration)
- K-20 test file

The following mini-tutorials assume that Reaper and the audio interface are already configured and set to send the required number of channels to the speaker device, and that the intensity levels have been calibrated.

1. the original files in nodal space on 80 points

The Preludes to Space series currently comprises around fifty pieces, all available in this format. I have selected here those for which the "multi-layered 3D" space is most essential, particularly due to the point density required both inside and at ground level. Contact me for others.

SpatGris software
Although its primary purpose is spatial mixing, SpatGris software also allows (beta version 4.0) real-time conversion of multiphonic formats. Since it supports 128 channels and more, has an efficient MBAP calculation method, and is open-source, it is poised to become the ideal application for adapting formats based on three-dimensional meshes.
Its operation as a standalone application is less convenient than a plugin, but the process is nonetheless quite simple:
- Place the 80-channel file on a track; the track is automatically configured with the required number of channels.
- Insert the ControlGris plugin onto the track and import the spatial configuration "Acousmonef-80-Cube-Compress" if the elevation of the diffusion space scheme is between 0 and 1, or the "Acousmonef-80.4-Cube-full" file if it uses a wider elevation (not yet supported by ControlGris 2).

- In the SpatGris application, define or import the spatial configuration corresponding to the speaker system for the diffusion.
- Note: In both cases, the spatial mode must be set to "Cube".
- On Windows, communication between the two applications can be done using ReaRoute, which is installed with Reaper and must be previously configured to support at least 80 channels (256 maximum). The advantage is that the communication can be bidirectional, meaning that the result of the conversion performed in SpatGris can be sent back to a Reaper track to benefit from further processing such as bass management and everything that can be done before the soundcard outputs.

- On Mac, you should use BlackHole (included with SpatGris) as the output sound card in Reaper and the input sound card in SpatGris. Bidirectional communication requires aggregation with the sound card used for playback.

Caution: Do not use this for peripheral systems, i.e., those without internal loudspeakers (hollow spaces)! For this purpose, use the following ambisonic encoded files.

Download:
- SpatGris 4.0  software for Mac and Windows (beta version!)
- Acousmonef-80-Cube-Compress and Acousmonef-80.4-Cube-full preset files

2. high-order ambisonic files for cubidomes

Ambisonic spatial representation is particularly well-suited to loudspeakers configurations  physically based on a sphere, such as domes, but it can also be used for other spatial arrangements of the periphery type (single layer), such as variations of cinema formats, by applying the necessary distance compensations (many "domes" are actually closer to cubes...).
Its drawback is that the double processing applied for encoding and decoding introduces a certain loss of precision and spatial depth, which is all the more significant as the order value decreases. The maximum generally available in our software is order 7 (= 64 encoding channels), but Reaper allows up to order 10 (121 channels), and thanks to the SPARTA AmbiDec decoding plugin, we can thus get a little closer to the original definition... But is it really worth it? While the improvement in sound depth is slightly audible, it doesn't seem to justify doubling the file size at this time. Therefore, these are the 7th-order encoded versions, and I can provide the 10th-order versions upon request.
In all cases, it is strongly recommended that the lowest elevation level be at ground level.
For educational purposes, these files can easily be decoded into immersive cinema formats like Atmos, with the resulting sound quality obviously proportional to their spatial reduction.

Sparta AmbiDec plugin
- Place the 64 or 121 channel file on a Reaper track; the track will automatically be configured with the required number of channels.
- Insert the Sparta AmbiDec plugin on the track and enter the angular values ​​of the speaker points corresponding to the sound system (note: for 7th order, the IEM AllRADecoder plugin is also perfect...).

- If the center distance of all speakers is not identical, insert the IEM Distance Compensator plugin after the decoding plugin and enter the center distance values for each speaker.

- Adjust the track gain according to the information provided in the footnote.

Download:
- Sparta and IEMplugins for Linux, Mac, and Windows
- The K-20 reference file encoded in HOA in the same way as the parts (see footnote)

3. 25-point mesh surface files, for acousmoniums

GRM-Tools SpaceMaster plugin
Although it's somewhat dated and no longer supported on macOS, it's perfectly suited for converting spaces with up to 64 channels. Its plugin functionality allows it to be used directly on audio objects, which can be very useful for playing multiple pieces created in different formats.

- Place the 25-channel file on a Reaper track; it will automatically configure the track to the required number of channels.
- Insert the SpaceMaster plugin onto the track (or object) and define the speaker layout corresponding to the sound system in the Layout Editor.
- Import the "Surface25" layout into the same window.

- Back in the plugin interface in Calibration mode, select the layout you defined for the broadcast system.
- Switch to Master mode and select the "Surface25" layout in "Init sources".

- Adjust the Spread parameter to around 25%
- Note: It is not necessary for the spatial representation to be exact, i.e., conforming to the actual distances between the speaker points; it is preferable that it be adjusted so as to cover the entire surface of the plugin and that the points that constitute it are as equidistant as possible.

Download:
- the GRM-Tools SpaceMaster plugin for Windows and older versions of macOS
- the "Surface25.lay" layout file

4. specific adaptations

For loudspeaker systems that don't fall into the two previous categories, or if you lack the resources to perform the adaptations yourself, you can provide me with the precise specifications of your sound system so that I can handle them myself. This imposes a certain rigidity on the output format— it can only be played on that specific system — but ensures that it is optimized, at least in terms of its artistic quality.
The result of this conversion can be either an audio file playable directly in Reaper or a preset file for the TransSpace 80 plugin. The latter solution can be more flexible because it doesn't require sending new audio files if the system is modified, as playback always uses the original 80-channel files.
In both cases, you need to send me the precise placement of the speaker points as well as the corresponding numbers for the audio interface output channels as they appear in Reaper (important!). I recommend using the VisualSpacer 80 plugin for this (see examples below).

TransSpace 80 plugin
Unlike spatialization software such as SpatGris, the conversion between the two formats here is not automatic. It requires some familiarity and experience, and, for more complex cases, necessitates making choices based on knowledge of the work and what you want to achieve... and that's precisely what makes it so interesting (see the videos https://youtu.be/-h-NLFMnY1g and https://youtu.be/tIScN4DMqHg).
If you choose this solution:
- You provide me with the setup: the speaker positions and channel numbers (if possible, using the VisualSpacer plugin)
- I will then send you the vst.preset file for TransSpace
- On the track where the 80-channel room file is located, you insert the TransSpace 80 plugin and import the preset file that was sent to you. That's all! ;-)

Download:
- VisualSpacer 80 VST3 plugin for easily communicating the speaker device's placement, for WIndows and Mac, and its mini-manual
- TransSpace 80 VST3 plugin for WIndows and Mac, and its mini-manual (note for Mac)

Note about level calibration

All pieces are set up with a listening level calibrated according to the K-System principle in K-20: a digital reference level (pink noise at -20 dB RMS) should theoretically correspond to the listening point(s) at 83 dB-SPL (C/Slow) for each channel.
In reality, this value obviously needs to be adjusted according to the number of channels, the room's dimensions and acoustics, not to mention personal preferences, but it is nevertheless important to try to maintain this equivalence as a starting point, even if it needs to be adjusted later.
The Loudness R-128 measurement and normalization can be used as a complement or replacement in the case of ambisonics, but it is not sufficient on its own because it does not reflect the sometimes significant differences in dynamics between rooms, and obviously does not take into account the equipment and its settings, nor the room itself, and therefore the actual listening level.

Using the K-20 test file:
- Place it on a track; both the track and the master should be set to "0dB".
- Configure a sound level meter to "C curve / Slow".
- Adjust the amplification level so that the sound level meter displays approximately 83 dB-SPL for each channel at the intended listening distance.
- Repeat for each output channel/speaker.
Or, more simply:
- Insert the SimpleTest plugin on the Master, set to 0dB (the plugin's gain setting should be -12dB Peak, which is its default value).
- Play each channel in turn and adjust the amplifier gains so that the sound level meter displays the same reading as before.

Bass Management

One or more dedicated low-frequency speakers not only complement the spectrum produced by the "normal" speakers but also contribute to the overall homogenization of the system. The Acousmodules SimpleBass plugins make it easy to:
- select the crossover frequency for each output channel
- choose the crossover filter curve
- adjust and send the low-frequency signal to one to four subwoofers independently
- apply a low-cut filter to all channels corresponding to the crossover frequency
- adjust the send gain to the four subwoofers
- choose the send channel for the four subwoofers

Download the SimpleBass 36 and SimpleBass 64 plugins for Windows and Mac, and place them at the end of your signal chain on the Master or MonitoringFX module in Reaper.

* Note regarding downloading Acousmodules plugins for Mac: Due to Apple's protection policy, files downloaded from macOS will not run on the latest versions of the OS because they are not notarized. One solution is to download them from Windows or Linux and then transfer them to macOS. Another solution is to temporarily disable Gatekeeper with Sentinel.