Hi everyone!
I’m doing a sound-design and audio-analysis experiment with the Syntakt, and I’d like to understand its filter behaviour in a little more detail, particularly how the FREQ and resonance parameters affect short percussive sounds.
I’m interested in comparing recordings made with controlled settings and looking at how the spectral content, resonant peak and decay change across different parameter values. I don’t want to assume that the relationship between a parameter value and the resulting sound is linear, so having a few consistent recordings would be really helpful.
Would anyone with a Syntakt be willing to help record a small set of samples? The main test should be fairly quick once the pattern and track are set up. The additional tests below are optional, so even a partial contribution would be appreciated.
1. Main test: FREQ sweep with UT IMPULSE
For the primary set, could you use the following setup?
- Machine: UT IMPULSE.
- Filter: low-pass.
- Resonance: a relatively high setting, ideally high enough to make the filter’s resonant character apparent, but without unwanted clipping or distortion.
- FREQ: set to the exact value requested for each sample.
- LFO and filter envelope: disabled, along with any other modulation affecting the filter, if possible.
- Delay and reverb: off.
- Other effects: off or disabled wherever possible.
- Sequencer: one isolated trigger per recording, with no other active steps or tracks contributing to the sound.
- Recording: direct audio output, with no external processing, EQ, compression, normalization or noise reduction.
Please keep the machine, track, pattern, resonance, trigger conditions and recording gain unchanged throughout this first set. The only parameter that should change between recordings is FREQ.
Suggested FREQ values
If possible, the full set would be:
0, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 104, 112, 120, 127
If that is too much work, this smaller set would already be very useful:
0, 16, 32, 48, 64, 80, 96, 112, 127
If you have time for a few extra points, these would help fill some gaps:
19, 22, 54, 111
These are the Syntakt parameter values, not frequencies in Hz. Please include the exact FREQ value in each filename or in a separate list.
For each setting:
- Set FREQ to the requested value.
- Trigger UT IMPULSE once.
- Record the complete sound, including the initial transient and decay.
- Leave a little silence before and after the sound if convenient.
- Save the recording without editing or processing it.
Around one or two seconds per recording should be enough, provided the complete sound and decay are captured.
2. Optional test: low versus high resonance
It would also be useful to compare the effect of resonance independently of FREQ.
If possible, please repeat a smaller subset of the FREQ sweep at two resonance settings:
- One relatively low resonance setting.
- One high resonance setting, as used in the main test.
A useful subset would be:
16, 32, 48, 64, 80, 96, 112
The exact resonance values would be ideal, but if the parameter display is not convenient to document, a screenshot or a short note describing the settings would still help.
For each recording, please change only FREQ within that subset, then repeat the same subset at the other resonance setting. Keep all other settings unchanged.
This should help show how resonance affects the spectral peak and the overall shape of the filter response across the FREQ range.
3. Optional test: resonance sweep at a fixed FREQ
As a separate, very small experiment, it would be interesting to keep FREQ fixed and change only resonance.
For example:
- FREQ: 64, or another convenient middle value.
- Resonance: low, medium and high.
Please use the same trigger and recording conditions for all three samples.
The purpose is to compare how resonance changes the sound when the FREQ setting itself remains constant. If you can provide the exact parameter values, even better.
This test is separate from the main sweep, so there’s no need to combine the two sets into one complicated pattern.
4. Optional test: repeatability
If convenient, could you record three to five repetitions at a few representative FREQ settings, such as:
16, 64, 112
Please do not change any settings between repetitions.
This would help establish how consistent the transient and decay are from one trigger to another, and whether any small differences are part of the sound itself or simply recording variation.
There is no need to repeat every value in the full sweep.
5. Optional test: filter envelope
I’d also be interested in a simple comparison of the filter with its envelope modulation disabled and enabled.
If possible:
- Keep the machine, FREQ, resonance and trigger conditions fixed.
- Record one sample with the filter envelope modulation disabled.
- Record another with the envelope clearly affecting the filter.
- Include the relevant envelope settings, such as amount and decay, if available.
This test is optional and separate from the static FREQ measurements. The main sweep should remain unmodulated so that the results are easier to compare.
6. Optional reference recording
If the machine and track configuration allow it, a reference recording with the filter bypassed or with a very low resonance setting could also be useful.
The aim would be to compare the source transient with the filtered version under otherwise identical conditions.
If a true filter bypass isn’t available for this machine or configuration, there’s no need to work around it. A clearly documented low-resonance recording is perfectly fine as a reference.
7. Recording format
WAV would be ideal, preferably:
- 48 kHz sample rate, if convenient.
- 24-bit or 16-bit PCM.
- Mono or stereo direct output, whichever is easiest, provided the same format is used consistently.
- No normalization or post-processing.
If your usual recording setup produces another sample rate or bit depth, that’s absolutely fine. Please just mention the format.
A ZIP containing the recordings would be ideal, with filenames such as:
freq_016_res_low.wavfreq_016_res_high.wavfreq_064_res_high_take01.wavfreq_064_res_high_take02.wavfreq_064_res_low.wavfreq_064_res_mid.wav
The filenames are only suggestions; any consistent naming scheme is fine.
8. A few details that would help
If possible, please include:
- Syntakt OS/firmware version.
- Exact FREQ and resonance settings.
- Any filter-envelope or LFO settings that remained active.
- Whether the recordings were made directly from the Syntakt’s outputs.
- Sample rate and bit depth.
- Any other relevant machine or track settings.
- Any observations about particular values where the sound changes noticeably.
Please don’t worry if you cannot provide every detail. A smaller, carefully controlled set with clearly identified parameter values is more useful than a large set with uncertain settings.
9. What I’m particularly interested in
The main questions are:
- How does the sound’s spectral content change as FREQ increases?
- Does the resonant peak move smoothly or are there noticeable changes in behaviour?
- How does resonance affect the peak and the overall response?
- Is the transient and decay consistent across repeated triggers?
- Does enabling the filter envelope produce a clearly different time-dependent response?
I realise that the strongest spectral peak isn’t necessarily identical to the filter’s cutoff frequency, especially with high resonance. I’m therefore mainly interested in the raw recordings and the exact settings, rather than asking anyone to measure or estimate frequencies in Hz.
Thanks very much to anyone willing to help! Even the main sweep with the smaller set of FREQ values would be greatly appreciated. The additional tests are entirely optional, and I’d be happy to share some general observations from the analysis with the community afterwards.
