Synthesis Technology Hyperion Stereo Effects Processor Now Available For Preorder

Texas-based Synthesis Technology has launched a Kickstarter project to fund production of their new E520 Hyperion Stereo Effects Processor for Eurorack modular synthesizers.

The E520 Hyperion is a mono/stereo in, stereo out effects module. Powered by a 480MHz ARM Cortex M7 CPU, the Hyperion has 64MB on on-board SDRAM for recording up to 6 minutes of stereo audio, at 48KHz sample rate.

The Hyperion’s algorithms fall into 3 categories:

  • Time Domain Effects – delay, chorus, phase shifting/flanging, frequency shifting and others.
  • Frequency Domain (Spectral) Effects – the unique Spectral Crusher, vocoder, time-stretching and others.
  • Utility – frequency and level measurements

There are 25 effects algorithms included with the E520, with the capacity for more to be added in the future, as internal memory allows.

Each algorithm can have up to 4 PARAMETERS under voltage control. The initial value of a PARAMETER is set by a panel knob, and then external CVs of up to +-5V at 10KHz can be applied. Most algorithms can also employ feedback which can be modulated by external CVs as well. Typical PARAMETERs are delay time, LFO speeds, pitch shift amount among others.

The row of ‘soft keys’ are usually 1-3 for algorithm selections (such as range of Delay Time) while the fourth is a BYPASS function. Some algorithms use the TAP TEMPO button for tapping in Delay Time, LFO Speed or similar functions. Other algorithms repurpose the TAP TEMPO as a ‘FREEZE’ or ‘HOLD’ function. The SYNC IN jack can be used like the TAP TEMPO functions, but driven by a clock or Gate signal.

The Hyperion promises to offer effects capabilities that go beyond most current Eurorack/modular effects modules, based on the Spin FV-1.

“Unlike the current Spin FV-1 parts, now over 6 years old, our 480MHz ARM H7 can do both time and spectral effects,” explains SynthTech founder Paul Schreiber. “You no longer need a PC or Mac to run these spectral algorithms. An example is the former NI Spektral Delay, which we can implement in a slightly different way, mainly our UI with a small LCD.”

“The Spin part is 32KHz sample rate, we are 48KHz,” adds Schreiber. “The Spin is limited to 3 CV inputs, we have 8 on the E520. The Spin has about 2 seconds of audio buffer memory, we have 5 min at 48KHz and over 20 minutes at 8KHz.”

Schreiber has developed the module with a core team that has previously worked on SynthTech’s E370 and E352 modules: Schreiber, DSP coder Eric Brombaugh, and composer & synthesist Robert Rich.

“I also have a large team, about 15 people, who will be beta testers and providing demos,” adds Schreiber. “One early beta tester is Russell Brower, [who is] the former head composer at Blizzard (World of Warcraft) and before that, [was] at Disney.”

Hyperion Effects

SynthTech has announced a variety of effects for the Hyperion:

Time-Domain Effects

  • Resampling Mini Delay – This effect is based on the E580 Resampling Mini Delay and encompasses or extends all of its functions. Stereo inputs are mixed down to a single channel, resampled (the sampling rate may be reduced by a variable amount through a band-limited interpolation process), pass through a delay line with a fixed and variable tap, and are then resampled to two separate outputs. Separate L/R feedback controls allow fine adjustment of recirculation on both the fixed and variable delay. Three mode settings emulate various classical delay characteristics.
  • Stereo Clean Delay – The Stereo Clean Delay is a non-resampling delay (the sampling rate is constant throughout the processing) which always runs at the full 48kHz rate and introduces no interpolation artifacts into the signal while the delay length is stable. This is a true stereo effect in that there are separate delay lines for Left and Right channels but they are processed identically (same delay length, same filtering, etc.). A switchable resonant 4th-order lowpass filter is available in the feedback path if desired. Ping-pong feedback between the two channels may be enabled.
  • Stereo Resampling Delay – The Stereo Resampling Delay is resampling delay which allows control of both the delay line length (for granular cross-fade effects) as well as sample rate (for smooth delay changes and FM effects). This is a true stereo effect in that there are separate resamplers and delay lines for Left and Right channels but they are processed identically (same delay length, same filtering, etc.). A switchable resonant 4th-order lowpass filter is available in the feedback path if desired. Ping-pong feedback between the two channels may be enabled.
  • Prime Dual Tap Delay – The Prime Dual Tap Delay is an emulation of the Lexicon Prime Time T93 rack-mount delay. It is a resampling delay which allows control of both the delay line length (for granular cross-fade effects) as well as sample rate (for smooth delay changes and FM effects). As with the original, this is a mono effect in that there is one delay line whose input is an equal mix of Left and Right channels and separate output taps for Left and Right channels with independent delay lengths that are selectable from a pre-determined set which includes delays that have relatively prime ratios. Controllable fourth-order lowpass filters are available in the feedback path, if desired. Buffer lengths “well beyond” that of the original Prime Time effect are available and the two outputs may be mixed down to mono if desired.
  • Stereo Reversing Delay – The Stereo Reversing Delay is a reversing and resampling delay which allows control of both the delay line length (for granular cross-fade effects) as well as sample rate (for smooth delay changes and FM effects). This is a true stereo effect in that there are separate resamplers and delay lines for Left and Right channels but they are processed identically (same delay length, same filtering, etc.). A switchable resonant fourth-order lowpass filter is available in the feedback path, if desired. The unique features are that this delay can be run forward or backward and it can be frozen, to keep playing the current delay contents without changing.
  • Deflector Shield – The Deflector Shield is a recreation of the original E560 Deflector Shield module. It is a monophonic effect similar to the original Bode frequency shift which applies a variable frequency offset to any input signal. As the original, there are three distinct operating modes, as well as a unique carrier morphing feature that provides unusual output harmonics. Although the input is mono, there are two outputs which allows simultaneous up/down shift as well as true quadrature outputs. Variable feedback is available as well as linear or exponential frequency CV response.
  • Stereo Frequency Shifter – The Stereo Frequency Shifter is similar to the Deflector Shield but is a true stereo effect, the Left and Right channels are completely independent but are processed with identical parameters. Similar frequency shifts, carrier morphing and modes are available but the output modes are limited to variations of Up and Down shifting. Variable feedback is available as well as linear or exponential frequency CV response.
  • Stereo 24-stage Phaser – The Stereo Phaser is an all-pass phaser with selectable number (6, 8, 18, or 24) of all-pass stages and an integrated LFO with variable frequency and depth. It is a true stereo effect: the Left and Right channels are completely independent but are processed with identical parameters. Variable feedback is available as well as selectable LFO waveform and optional phase offset between the Left and Right channels.
  • Stereo Pitch Shifter – The Stereo Pitch Shifter is a granular pitch shifter with +/-2 octave range, variable grain size and chaos offset. It is a true stereo effect: the Left and Right channels are completely independent, but are processed with identical parameters. Variable feedback is available as well as selectable shift quantization to an equally tempered scale.
  • Stereo Shimmer Reverb – The Stereo Shimmer Reverb is a reverb combined with a granular pitch shifter in the feedback path that can be used to introduce a variety of ambient, soaring textures. It is a true stereo effect: the Left and Right channels are processed independently but share the same reverb loop so that sounds in one channel will reverberate in the other, but from a different part of the process so that stereo imaging does not collapse. Variable feedback is available as well as selectable shift amounts to allow for a wide range of moods, dark, chaotic and ethereal. Infinite mode is controlled via the Tap Tempo input and preserves the sound in the reverb by disabling inputs, filters and internal attenuation, extending its duration for hours.
  • Stereo Looper – The Stereo Looper is a time-domain looper with resampling that provides up to 5.8 minutes of looping duration at the full sample rate and approx 45min at the lowest rate. It provides the usual feature set of recording, overdubbing and playback.

Spectral Effects

Spectral effects are processed in the frequency domain using the Fast Fourier Transform. All spectral effects in the E520 use a 1024 point complex FFT to provide two x 512-point real transforms at a rate of 2.7ms which provides a good compromise between time and frequency resolution, while allowing full stereo processing for most effects. Frequency resolution is 512 bins of 48Hz each.

  • Spectral Crusher – The Spectral Crusher is a multi-effect which provides a number of unique ways to manipulate sound in the frequency domain. Six different algorithms are available, each of which crushes audio in a different way.
  • Spectral Time Machine – The Spectral Time Machine is a combination of a looper, pitch shifter and time stretcher. A large (5.8 minute) memory records and loops sounds which can be played back continuously or looped with variable speed and pitch. Due to the processing overhead, this is a mono effect. Left and Right channels are mixed 50/50 prior to input. Output has pseudo-stereo at some settings of the Phase Blur control.
  • Spectral Delay + Pitch Shifter – The Spectral Delay + Pitch is a stereo effect with independent delay and pitch shift applied to the Left and Right channels. Pitch shift quantization to equal tempered half tone intervals is available independently on each channel and a fixed value phase blur may be applied to both channels simultaneously.
  • Spectral Delay – The Spectral Delay is a stereo effect which delays each spectral bin by different amounts. Similar to the NI Spektral Delay plug-in, everal modes are provided to vary the bin delays in different ways. Left and Right channels have independently controlled delay parameters.
  • Variable-Band Spectral Vocoder – The Spectral Vocoder is a mono effect which superimposes the amplitude response of one signal (the Modulator, in the Left channel) upon another signal (the Carrier, in the Right channel) which can (among other things) be used to create talking instrument when speech is used as the Modulator signal. Various amplitude and frequency manipulations are provided to improve or distort the intelligibility of the resulting sounds, including gain adjustments, compression, enhanced sibilance (S sounds) and frequency resolution reduction.

Schreiber says that he expects the Hyperion’s capabilities to grow over time, similar to some of his earlier modules. “Like both the E352 Cloud Terrarium and E370 Quad VCO, the E520 Hyperion uses the same, easy and quick (now under 3 seconds) firmware update system. The E370 Quad VCO, in particular, is very different now, 2 years after release, than what was presented during its KickStarter.”

“We expect and encourage our early KickStarter customers to suggest algorithms to include,” he adds. “We feel this new H7 hardware platform has enough memory and raw CPU speed to allow years and years of upgrades, always free and with a lifetime warranty.”

Audio Demos:

Pitch Shifter: voice in, then pitch-shifted +400cents, then -300cents.

Looper: Source is lap steel guitar, play by Robert Rich (some reverb on original). Using the Looper with some subtle, then not so subtle pitch shifting (especially the very end).

Resampling Mini-Delay Delay. Source is NI Una Corda (prepared piano). Similar to the Looper example above, but using a 50Sec delay + Tape Mode with Wow/Flutter. Pitch changes via altering the Sample Rate.

Stereo Resampling Mini-Delay. Voice input, purposely over-the-top modulation.

Resampling Mini-Delay: E370 Saws. Tap Offset: 50%, Delay Time: 1.3x (synced), Warble: 12%, In/Out Mix 50/50

Deflector Shield: Digitakt source. InMix: 50/50, Carrier: S/H, Feedback: 80% positive

A more comprehensive selection of audio demos is available at the SynthTech site.

Pricing and Availability

Production of the SynthTech E520 Hyperion Stereo Effects Processor is being funded via Kickstarter project, and it is available to project backers starting at $539 USD. See the project site for details.

Update: The Hyperion Kickstarter has met its funding goal. Stretch goals include assignable CV routing, programmable LFOs, independent left/right effects and more.

19 thoughts on “Synthesis Technology Hyperion Stereo Effects Processor Now Available For Preorder

  1. This seems like a no-brainer – it’s less than the cost of a Z-DSP and a single cartridge.

    I’d love to see Valhalla’s algorithms on this, though. They’re a little grainy on the Z-DSP, but the desktop versions are some of the best reverbs that I’ve heard.

    1. “He’s been adamant from the start that this will not have any reverbs.”

      And yet it will ship with a reverb and Schreiber says “We expect and encourage our early KickStarter customers to suggest algorithms to include”.

      Did you miss that?

      1. “As far as the actual algorithms: the emphasis is on non-reverbs. It’s not a reverb because there are 100s of plug-ins, Spin-based modules (ZDSP, etc), and honestly, a good reverb represents years of work, and in cases like Eventide/Lexicon decades of work. And reverb by it’s nature is not ‘happy’ with fast-changing, dynamic variables (which is why Spin-based CV inputs are such narrow bandwidth). We have bigger fish to fry.”

        “It has exactly 1 SPECIAL case of reverb, where ‘reverb’ is not the main sound. That is called ‘Shimmer’, and that means there are pitch shifters in the reflections/feedback loops. Listen to the 3rd “Chanting Monks” demo, where you hear the ‘soaring to the heavens’ effect.

        Yes, there has to be a ‘room’ to ‘reflect’, so it does have the most basic reverb, called Schroeder 4-corner algorithm. This is NOT what is in a Lexicon or Eventide. :deadbanana:

        When I was saying ‘no reverb’, I mean what is expected: plates/halls/gated/blah blah blah.

        Shimmer has a ‘reverb component’ but it’s minimal at best.”

        These are quotes from Paul S. I think it’s fair to say this is not the module you should get if you are hoping for reverbs

  2. Can any of you Synthtopia readers answer this: I’m based in Europe (Ireland specifically). If I back up a Kickstarter project that includes shipping a final product from the US, will I have to pay TVA tax on top of the final displayed price (which already includes shipping cost)? Can’t find a definite answer anywhere…

    1. usually yes. However Kickstarter is not a store, and you are not buying but rather funding a project. I’ve heard of a manufacturer, charging only for the parts when shipping based on that idea and not the actual price.

    2. Yes, unfortunately, customs are always the buyer’s headache. You ARE buying, because you’re giving money to receive a product.
      Getting in touch with the vendor is probably the best coarse of action – you gotta know for sure.

      1. 2 opposite answers confirm to me that there’s definitely a lack of clear information on the subject… @Fedor Tkachev, your suggestion to directly contact the manufacturer is probably the best option indeed, so I’ll do that and will post whatever reply they provide here if anyone is interested in knowing as well.

          1. Synthesis Technology have got back to me confirming the commercial status of backing up their campaign. So yes, TVA must be paid by Europeans. I believe this to be the case with every other Kickstarter campaign.

            1. I believe at muff’s it was mentioned by many that this was not the case with Killpatrick’s Phenol. Perhaps someone can confirm this here as well.

                1. Paul has said that the full price will be on all shipments so expect to pay VAT plus duties on shipments to Ireland. (Coming from someone who purchased the E370 and lives in Dublin).

  3. btw I am torn between this and zoia. The 520 has a lot of cv which is a plus but it is not clear to me whether all these algorithms can be combined or you just run one each time.

      1. thanks, I like some of the spectral stuff that I see and it would be nice to see them combined with time based effects, but if it doesn’t do fx combinations I don’t see the point of the screen. Taking up so much real estate for what really? Anyway I know it is still very early.

  4. Paul’s description of the reverb situation is precise and enlightening. Its a mini-lesson in the side considerations and disciplines involved. I kind of regret not being a modular player at times. I’d be all over this as someone who sees ‘effects’ as an important module of its own. That 6 minutes of sample time is like Robert Rich in a box.

  5. Addressing come comments:

    1) There is 1 algorithm with reverb as PART of the overall effect, called Shimmer. But we are not Lexicon. Or Eventide. We are not “reverb experts”. And like I said before: it’s very difficult to modulate reverb algorithms and reverb cannot be “improved” by us. Rather, what we CAN improve on has been shown. Just listen to our pitch-shifters. Or the Spectral Crusher. Or the Stereo Reversing Delay.

    2) Everyone seems to think that having effects in a chain like some crazed guitar player’s pedalboard is “cool”. But the issue is this: if every algorithm has 4 CVs already used, and you want to chain 2, then from where do 8CVs magically appear? The hardware still has 4 CV ins (the Mix and Feedback are global). So now what?

    People complain about ‘menu diving’, well… arbitrary effects chaining and parameter re-assignment seems to be a menu-diving nightmare. And then we have to save that in the patch scheme. And so on.

    What we can do is have PRESET chained effects with fixed parameter mapping. And this we plan to offer. And as mentioned, there is a Stretch Goal to have different effects on L/R channels, so 1 patch cord later and there’s a chain (in mono, yes).

    We’re not a quad-core 1.2GHz ARM with 4GB SDRAM. We are a smaller ARM with 64MB sample RAM running on ‘bare hardware’ and for < $600 and <300ma of +12V. We simply do not have the ability to arbitrarily chain like VSTs. What we offer is FULL voltage control. Something very hard if not impossible for any VST. From Lexicon. Or Eventide 🙂

    For people with E352 or E370s: tell here in the comments what the module could do when you got it, and NOW after years of our free support what it can do. In the case of the E370 KickStarter, we since have added like 9 new features and improvements. The same will be true for the E520. All the demos, all the UI screens, all you see and hear are only 13 weeks of coding effort. And we are like 27 weeks from shipping 🙂 There is much to do, add, improved WITHOUT trying to out-Lexicon Lexicon. And we want our backers to be actively involved in this process.

  6. Looks and sounds great, cool to see that the Kickstarter already blew past the (relatively modest) goal. Here’s hoping they hit $100K by the end and get all the stretch goals into the first production units.

Leave a Reply