Information and resources for creating SpectraStrobe audio tracks.
SpectraStrobe and AudioStrobe are audio encoding/decoding technologies for use with Auditory & Visual Stimulation (AVS) devices, also referred to as Mind Machines, or Light & Sound Machines (L&S). These devices use specifically crafted auditory and visual stimulus to explore different states of consciousness, typically in an attempt to invoke the frequency follow response (FFR) of the brain through the use of techniques such as binaural beats, monaural beats, and isochronic tones.
A .zip file of the resources directory can be found in the releases section.
Although this repository is meant to primarily address the SpectraStrobe format, both AudioStrobe and SpectraStrobe work through similar principles, with SpectraStrobe being a bit more complex both in implementation and capabilities.
In AVS devices the auditory component is usually experienced by the user through a pair of headphones typically consisting of device-generated tones, recorded audio, or both. The visual component is usually experienced as pulses of light emitted from LEDs built into special eye glasses that the user wears in conjunction with the headphones. Some models offer a closed-eye experience while others allow for open-eye viewing. The goal is generally to synchronize the auditory and visual pulses to create the desired stimulus for the user.
AudioStrobe and SpectraStrobe are both focused on adding visual control signals directly to recorded audio material and do not focus on generating audio tones directly on an AVS device itself. This is achieved by encoding and decoding visual control signals into the high frequency content of recorded audio in ranges that are typically beyond the point of most humans' hearing, approaching 20 khz (kilohertz).
AudioStrobe and SpectraStrobe take a similar approach to encoding and decoding visual control signals from within audio recordings, however AudioStrobe only encodes a single stereo control signal, where as SpectraStrobe encodes three separate stereo control signals which typically correspond to individual red, green, and blue LEDs.
For encoding, high frequency sine wave tones are added into the audio content of the recording. Decoding involves using specialized filters during playback to isolate these high frequency tones from the rest of the audio content, which are then used to control the brightness levels of the AVS device's LEDs.
The brightness of the LEDs is based on the amplitude (volume) of the corresponding high frequency control tone(s).
The following table lists the various control tone frequencies:
Format | Frequency | Purpose |
---|---|---|
AudioStrobe | 19200 hz | controls the brighness of the LEDs; does not specify LED color |
SpectraStrobe | 18200 hz | reference tone used to indicate the SpectraStrobe format |
SpectraStrobe | 18700 hz | controls the brightness of the red LEDs |
SpectraStrobe | 19200 hz | controls the brightness of the green LEDs |
SpectraStrobe | 19700 hz | controls the brightness of the blue LEDs |
Note: The SpectraStrobe reference tone at 18200 hz also requires a specific stereo panning pattern to be applied in order for it to be correctly identified as SpectraStrobe audio.
Since both AudioStrobe and SpectraStrobe share a frequency space (19200 hz), in order for an AVS device to be able to detect which type of material it is playing back, a reference tone must be added at 18200 hz to indicate SpectraStrobe. This tone is also a sine wave but requires a special stereo panning pattern to initiate detection.
The SpectraStrobe 18200 hz reference tone must be panned hard left/right (100% left then 100% right) in a repeating pattern at a rate between 10-20 cycles per second (hz). Using audio mixing metaphors, and speaking with the terminology of a Low Frequency Oscillator (LFO), this would be interpreted as:
waveform: square, frequency: 20 hz, offset: 0%, amplitude: 100%
which would be applied to the stereo panning of the SpectraStrobe reference tone track within the mix.
This indicates the number of audio samples per second used in digital audio recordings.
Minimum: 44100 hz
Since the control tone frequencies are very high (nearly ultrasonic) they require enough bandwidth in the sample rate to be properly reproduced. The Nyquist theorem dictates that in order for a given frequency to be accurately reproduced, the bandwidth (sample rate) required is twice that of the desired frequency. The highest frequencies in AudioStrobe and SpectraStrobe are nearly 20000 hz (or 20 khz) so the minimum sample rate required would be 44100 hz (44.1 khz).
44100 hz has been the standard for digital consumer audio for decades starting in the late 70s and continuing as the standard for Compact Discs into the 80s and beyond.
Note: Higher sampling rates (48000 hz, 96000 hz, etc.) are completely acceptable during the audio mixing process, however most AVS devices will do just fine with 44100 hz and may only support playback at lower sample rates with 44100 hz and 48000 hz being safe targets. Bouncing the final stereo mix to one of these two lower sample rates is recommended for compatibility.
In digital audio recordings the bit depth indicates how many bits should be used to encode the audio waveform's amplitude (volume). The more bits available per sample, the more distinct/discreet levels of amplitude can be represented. In digital images the equivalent would be: the more bits used per pixel of an image, the more distinct colors available per pixel.
Target: 16 bits, unsigned integer
Besides having a sampling rate of at least 44100 hz, the final digital audio recording should also have a minimum bit depth of 16 bits per sample. During the audio mixing process it is recommended to use 32 bit floating point precision, although 16 bit integer will also suffice.
Note: For the final stereo mix/bounce 16 bit is recommended for compatibility and smaller file sizes.
Precision | File Format | Pro | Con |
---|---|---|---|
Lossless | .wav | accurate/uncompressed playback | large file size |
Lossy | .mp3 | smaller file size | possible compression artifacts |
There are two main types of digital audio file formats. Those are that are lossless and those that are lossy. Lossless formats will perfectly reproduce the original digital audio content sample-by-sample during playback whereas lossy formats use digital data compression to minimize the audio file size at the cost of introducing compression artifacts that cause inaccuracies during playback. Lossy inaccuracies can vary from minor/unnoticeable to severe and audible, which is generally related to the level of compression used: the more compression, the smaller the file, and the higher the inaccuracies similar to image compression artifacts in a file format like .jpg.
Note: All final mixes should be stereo.
Recommended Settings: 44100 hz or 48000 hz, 16 bit
Wave files (.wav) use uncompressed audio and offer perfect playback of the original digital recordings but have the largest file size.
Note: Although it is possible to have 32 bit wave files, 16 bit is recommended for compatibility.
Recommended Settings: 44100 hz or 48000 hz, 16 bit, 320 Kbps constant bit rate (CBR)
MP3 files (.mp3) use varying methods of audio compression to achieve smaller file sizes at the cost of introducing inaccuracies during audio playback. As long as a constant bit rate of 320 Kbps is used, the compression artifacts introduced should be completely neglibable both from a listening perspective and from an LED control signal perspective.
Note: Although it is possible to use both variable bit rate (VRB) and higher compression settings (like 128 Kbps, etc.) this is not recommended for use with AudioStrobe and SpectraStrobe as this will introduce too many sample inaccuracies during playback and likely lead to visual control errors or failure to detect visual control in general.
During playback of the AudioStrobe or SpectraStrobe content special filters are used to isolate the control tones in their respective frequency bands (18200 hz, 18700 hz, 19200 hz, 19700 hz). The amplitude of these tones is used to control the respective brightness of the LEDs on the AVS device's glasses. In the case of SpectraStrobe, the reference tone at 18200 hz does not directly control brightness, but indicates to the playback device that SpectraStrobe content is present.
For the color control tones (18700 hz, 19200 hz, 19700 hz) the amplitude of their signals in relation to the amplitude of the reference tone (18200 hz) determines the brightness of each LED channel. AudioStrobe does not include a reference tone, so it uses a fixed amplitude model.
In general -24 decibel (dB) indicates maximum brightness for the color channel. For reference 0 dB is the maximum allowed signal level for a digital audio recording.
Digital audio mixing usually consists of a piece of software or equipment known as a Digital Audio Workstation (DAW). Some audio software focuses more on simply editing single or multi track audio files as opposed to being a full-fledged audio workstation.
Although digital audio software may have different user interfaces, the mixing theory is typically the same:
A series of mono or stereo audio "tracks" (isolated channels of audio playback) that each have independent controls for volume (usually measured in dB), stereo panning (left/right speaker playback), and options to process audio independently in each audio track. Mathematically speaking, each independent audio track's waveform is added together (summed) to produce a final 2 channel, stereo mix, which is what listeners ultimately hear during file playback.
To add AudioStrobe or SpectraStrobe information to an audio mixing session, one should create separate stereo tracks for each tone required. With the exception of the SpectraStrobe reference tone, varying the volume of the track will vary the brightness of the corresponding LED channel on the AVS device, and varying the panning of the track will vary the left/right eye balance of the corresponding LED channel.
The following tables demonstrate the basic LED control theory for each tone track in the mix:
Track Volume Level (dB) | LED Brightness |
---|---|
-INF dB | LED color channel is completely off |
-24 dB | LED color channel is completely on; full brightness |
Track Panning Balance (%) | LED Balance |
---|---|
100% Left | LED color channel only displays on left eye; right eye off |
0% (centered pan) | LED color channel displays equally on left and right eye |
100% Right | LED color channel only displays on right eye; left eye off |
Note: These tables are only used to indicate the extremes of the range. As volume and panning are altered, the corresponding LED values (brightness or left/right eye balance) will gradually change in respect to the chosen level. Also, -24 dB is simply a standard value for the specification; with SpectraStrobe the overall brightness of each LED color channel is a relative relationship between the reference tone's dB level and the color tone's dB level. For example, you can set your reference tone volume below -24 dB to create more headroom in the mix, and you would drop the color tone tracks volume to match. If the inherent volume of the tones you are mixing already is -24 dB, your maximum volume level of the audio track on the mixer would actually be 0 dB / unity gain instead.
During audio mixing, all of the independent audio tracks will be summed into the final stereo mix. One must be aware of the available headroom in the overall mix. If the final stereo signal of digital audio reaches or exceeds 0 dB on either left or right channel it produces digital clipping, which is a form of digital distortion, and is generally considered to be undesirable.
Even those familiar with audio mixing and headroom will want to give special care when mixing an AudioStrobe or SpectraStrobe recording because even though the tones will not be heard (as they are basically ultrasonic) they still consume part of the available headroom in the overall stereo mix. You can always adjust the tones down below -24 dB to provide more headroom for the audible parts of your recording (allowing your mix to get louder), but as you reduce the volume level of the tones, it is possible that the audible content in the mix gets loud enough that it will interfere with the detection of the control tones during playback or even completely remove the ability for the special decoder filters to sense them at all.
One way to achieve proper signal integrity is to EQ or filter all audible tracks so that frequencies starting in the range of 13 khz and ending at 17 khz begin to taper off, with the steepest cutoff at 17khz. This leaves the frequency space for 18 khz and above free for AudioStrobe and SpectraStrobe control signals and improves track headroom since there won't be a stacking of signal levels from the actual sound/music part of the mix with the inaudible tone control signals.
Since AudioStrobe and SpectraStrobe insert their control signals into the audio recording itself, it is possible for unwanted frequency energy to "bleed" into the control and reference channels causing unwanted glitches and imprecise control of each color channel on the LEDs. This effect can be minimized using proper track equalization (EQ) and filtering. In general, the steeper the equalization or filters, the better, as it will minimize this frequency-based interference.
Ideally you would use the steepest (highest order/highest dB cut) bandpass filter available in your DAW on each track. You can also create your own bandpass filter by combining both a lowpass filter and a highpass filter together in series on each track, and setting their cutoff frequency to the frequency of the respective tones. Keeping Q/resonance at 0 is recommended as it may unintentionally amplify the energy around the filter cutoff point.
Audio Track | Suggested Filter |
---|---|
Reference Tone | bandpass (or combined lowpass & highpass) w/ cutoff @ 18.2 khz |
Red Tone | bandpass (or combined lowpass & highpass) w/ cutoff @ 18.7 khz |
Green Tone | bandpass (or combined lowpass & highpass) w/ cutoff @ 19.2 khz |
Blue Tone | bandpass (or combined lowpass & highpass) w/ cutoff @ 19.7 khz |
Note: If creating an AudioStrobe track it corresponds to the green color tone @ 19.2 khz.
As mentioned in headroom considerations, if you find that the audible portions of your mix (ie the heard music/sound parts) appear to be interfering with the control tones' frequency space you can place an EQ or lowpass filter such that frequencies above 17 khz are steeply removed. This will sacrifice some high frequency presence from the mix but the benefit will be a cleaner signal for the AudioStrobe and SpectraStrobe control signals/tones. If using an equalizer, you might want to slowly start cutting frequency content around 13 khz with growing steepness until about 17 khz where the cut should be as steep as your equalizer offers.
There are two main methods of achieving an AudioStrobe or SpectraStrobe mix:
- Use a tone generator to generate sine wave tones at the appropriate frequency and volume that match the duration of your entire mix and place these into an audio track within your DAW.
- Use smaller forward-looped samples of each tone frequency, place them into a sampler (software or hardware) and trigger their natural note (usually C-3) using MIDI or equivalent to add the tones to your mix
If you are more comfortable working with larger bounced audio tracks in a multi-track project, choose method 1. If you would rather keep file size to a minimum and are more comfortable working with digital samplers, choose method 2.
Note: With either technique, the SpectraStrobe reference tone must have the appropriate panning dynamics applied otherwise the AVS device will not recognize your recording as SpectraStrobe and will fall back to AudioStrobe instead.
The DAW you use will determine the best ways to to manipulate the color control tracks' volume and panning to control the LED signals. This can often be achieved through track automation by editing control envelopes of the track's volume and panning. Your DAW also might allow you to apply LFO control over track volume and panning. Some DAWs will let you use audio from audible channels to drive the volume level or panning balance of your LED color tracks. It might be possible to combine some or all of these techniques to creating complex, modulated signals.
The following is a list of resources that can be used in conjunction with the information presented in this document to create AudioStrobe and SpectraStrobe content using digital audio software.
A .zip file of the resources directory can be found in the releases section.
The tones directory of this repository contains four 1 second looped samples that can be loaded into a digital sampler and played back using MIDI note C-3. These tones are used by the provided templates in this repository.
Note: Ensure that your sampler has forward-looping of the samples enabled otherwise you will only get 1 second of audio playback from the tones and then they will stop. Also these tones are produced at max volume, so you will want to want to add a -24 dB adjustment to their output, either on the sampler or the audio track the sampler is mixed into. Ideally each of these tones would be output to their own, separate stereo tracks (four individual stereo tracks, one per tone). The SpectraStrobe reference tone does not have proper panning applied so you will need to figure out how to replicate this using the software or sampler of your choice; the provided templates do have this panning applied and should work "out-of-box".
The templates directory of this repository contains example templates for creating SpectraStrobe recordings with various DAWs. The templates are designed for use with the Kasina from Mind Place. You should be able to open the template projects in their corresponding DAW, study the configuration, and begin to add your own content following the example they provide. These templates should work as-is, you will just need to either use the Kasina's USB audio device driver as the output in your DAW or use a standard soundcard output into the Kasina's AUX input.
Ableton Live is a popular modern commercial DAW which specializes in live performance and playback. This template uses the built-in sampler instrument (one per track) to implement the reference and color tones and uses the AutoPan effect on the reference tone track to achieve the specified panning requirements to be detected as a SpectraStrobe recording.
Renoise is a lesser known DAW that is based on older computer tracker programs. This is a contemporary implementation of a tracker as a DAW which has its roots in the computer demo scene. Although it has a steeper learning curve for most, once you understand its workflow and shortcuts it is an extremely fast way to create music, similar to using a word processor to compose.
Renoise has slightly more interesting LFO and modulation options out-of-box compared to Ableton Live and is considerably more affordable. It works on Windows, MacOs, and Linux. Renoise offers a free restricted version with the most notable missing features being: ASIO support on Windows, and rendering to .wav disabled. The paid license is quite affordable and paid license upgrades come very infrequently.
Cymatic SpectraStrobe is a VST effect used to generate and control AdobeStrobe and SpectraStrobe signals that can also modulate the signals with audio input.
Cymatic LFO is a VST effect that is designed to pair with CymaticSpectraStrobe. It can also be used as a general audio/panning modulating LFO. It offers a variety of waveforms and generates frequencies from 0 → 100 hz. Use this plug-in to modulate a control tone's brightness or panning. Put multiple instances of the plug-in in series to create complex modulations.
Cymatic Tone Generator is a VST instrument with variable waveforms, base frequency, beat frequency, and a handful of other waveform adjustments. It can generate a primary fundamental tone between 0 hz to 1000 hz and a secondary tone at an offset from the fundamental of 0 hz to +100 hz. It is possible to generate and mix smoothly between monaural and binaural tones.
Cymatic SpectraStrobe is an Audio Units effect used to generate and control AdobeStrobe and SpectraStrobe signals that can also modulate the signals with audio input.
Cymatic LFO is an Audio Units effect that is designed to pair with CymaticSpectraStrobe. It can also be used as a general audio/panning modulating LFO. It offers a variety of waveforms and generates frequencies from 0 → 100 hz. Use this plug-in to modulate a control tone's brightness or panning. Put multiple instances of the plug-in in series to create complex modulations.
Cymatic Tone Generator is an Audio Units instrument with variable waveforms, base frequency, beat frequency, and a handful of other waveform adjustments. It can generate a primary fundamental tone between 0 hz to 1000 hz and a secondary tone at an offset from the fundamental of 0 hz to +100 hz. It is possible to generate and mix smoothly between monaural and binaural tones.