Expert Sleepers Announces USAMO (Universal Sample-Accurate MIDI Output)


Expert Sleepers has announced USAMO (Universal Sample-Accurate MIDI Output) – a new utility devices that generates a sample-accurate, jitter-free MIDI output from an audio input.

The audio signal is generated by the USAMO software, a plug-in which runs as a virtual instrument (AU/VST/AAX, Mac & Windows) in your DAW. The DAW sends MIDI to the plug-in; the plug-in translates that to audio and sends it to the USAMO hardware, via an output on your computer’s audio interface. The USAMO then reconstructs the MIDI and outputs it on a standard 5 pin DIN MIDI socket.

The USAMO software also generates MIDI clock, Song Position Pointer etc.

Because the MIDI signals are generated and transported as audio, they are guaranteed to be sample-accurately synced with your audio, and free of the jitter often associated with computer-generated MIDI.


The USAMO uses the same technology previously available from Expert Sleepers Silent Way and the ESX-8MD module, but now in a standalone box, and from an analogue audio connection (the ESX-8MD requires a Eurorack housing and a digital audio connection).


  • Input: 1/4″ audio jack socket, accepts balanced or unbalanced signals
  • Output: standard 5 pin DIN MIDI socket
  • Controls: ‘Trim’ pot
  • Indicators: ‘Power’, ‘Clip’, ‘Activity’
  • Power requirements: 9V DC, external PSU
  • Power connector: 2.1mm barrel socket, positive or negative tip
  • Aluminium box, 108x80x44mm (4.25″ x 3.15″ x 1.73″) (1U high, approximately quarter rack width)

Expert Sleepers plans to launch the USAMO at the 2015 NAMM Show, in January 2015.

32 thoughts on “Expert Sleepers Announces USAMO (Universal Sample-Accurate MIDI Output)

  1. How does this compare to devices like the ERM Midi Clock which just generates its own timing internally? I’m looking for a master clock as Ableton sucks if you want to sync with external hardware.

    1. Fantastic device. Was looking for a one-trick-pony thing for ages (ACME, Karma and innerclock can do much but are expensive).
      ERM has such a device in the pipeline as well, so I hope pricing will be competitive, around 150$.
      @jonathan: u wouldnt want to slave ableton to an external midi clock either.
      syncing via computer as master and audio pulse to clock is the only feasible way to integrate a computer.

    1. It do more than Syncgen, it send midiclock (like syncgen) + midi data (note, CC) !! If you want to clock and send midi data with Syncgen, you need to use an classic midi interface and merger …

      I am curious for the price of course but also to know if it will be easy to stack multiple USAMO. I want to use one + midi thru to midiclock my Elektron machines, and one + midi thru to sync and send notes to my synths.

  2. I have 6 external synths and drum-machines, with just one midi output on this box, how would this help me? Would I connect the Out to my existing midi interface – or can it only control one external machine?

        1. If you daisy chain all your MIDI devices, you will probably hear a ‘propagation’ effect, where the machines at the end of the chain are trailing those at the beginning. It does depend somewhat on how the MIDI thru sections of each device are designed. As others have mentioned, a decent MIDI hub simplifies something like that. And in the case of the MIDI hub you can (usually) route internally.

  3. So, this does the same thing as an opto-isolator. Only instead of electric analog or digital signal to light signal, it converts MIDI data to sound data and back to MIDI data. Interesting. Isn’t there something called SMPTE timecode? Doesn’t that work any more? I’m not knocking this product, just wondering if it is better or necessary.

      1. Those would be any inputs labeled ‘SMPTE’, most likely on video gear, less likely on audio gear. If you ever want to do a decent quality live video, you will be using SMPTE. Or if you ever have to do any sound design for ‘film’.

        SMPTE has rock solid timing, too. I don’t understand why MIDI was created to be so sloppy, CV was/is a far better control schema.

        1. Yeah, I was being facetious. SMPTE is fine for its intended video purposes, but how it could be proposed as a credible stand-in for the subject of this post is quite beyond me.

          MIDI may not be as solid as any number of similar/competing clock/automation standards, but its obvious trump card is that it’s built into everything.

    1. not at all.

      “The DAW sends MIDI to the plug-in; the plug-in translates that to audio and sends it to the USAMO ”
      might be confusing.

      the problem is that computer/DAWs are not able to generate a tight midi clock on their midi out interface (old ataris were better at that).
      the problem only arises if you work with external gear since the DAW compensates for internal jitter, internally a DAW is always sample accurate.
      this is not the case for the emitted midi clock signal.

      this device follows the original innerclock solution to let the DAW output a special audio signal to the outer world and convert this audio signal to midi clock.

      the difference in tightness is huge.

      1. You don’t need a master midi in, at least with the MidiexpressXT it just detects incoming clock in any port and sends it out of the rest.

  4. the people that things these amazing product is unnecessary or can be replaced with a USB interface clearly doesn’t sens or fill the “jittering” artefacts in MIDI…. Is like trying to sell a colourful HD Display to a blind person.

    For me “jitter” in MIDI from DAWs is a torture and these product is amazing. ¿Remember ATARI, old Commodore and Old Macintosh’s use to have better MIDI timing using serial MIDI interfaces.?

    1. You are right! MIDI was tight on my Atari ST, and it wasn’t as good, but not too terrible on Cubase running on OS 7 using a MIDI interface connected to the serial part. Current USB MIDI is generally abysmal, and is compounded by processor synthesis that uses buffers and very often doesn’t properly synchronize MIDI real time input to the right position in the buffer.

      Solutions involve direct synthesis and time tagging all MIDI data at the interface, and that data respected by the DAW.

  5. I’m totally amazed by the comments in this thread, which I agree with. Usually when I say the same I get downvoted and am told that jitter and latency problems don’t exist and are in my imagination. Jitter and latency problems are severe problems modern software synthesis, and that includes hardware synths that generate output by filling buffers with samples rather than direct digital or analog synthesis.

  6. Innerclock and the Ploytec GM5 should have been good solutions up till now-the ESI staff uses Ploytec (e.g. M4uXT) and that has been pretty good for us. But guess what? I still like to fire up my Atari and Cubase score for that old school feel and super-tight 16ths:-)

Leave a Reply