Home       Sync-Shift       Sync-Lock       Sync-Split       Mark-I       Litmus       Clock Watch       ICS Lab       Contact   
   Q&A       Sync       Pandora's Box       Jukebox       History       Travel       Friends       Analogic-ACS       iSync   
 
 




  Home  
  Sync-Shift  
  Sync-Lock  
  Sync-Split  
  Mark-I  
  Litmus  
  Clock Watch  
  ICS Lab  
  Contact  
  Q&A  
  Sync  
  Pandora's Box  
  Jukebox  
  History  
  Travel  
  Friends  
  Analogic-ACS  
  iSync  



asa

An extension of the Litmus Page but well worth a link of it's own considering its significance.

Midi Clock. 

Simple in concept and mostly taken for granted its importance is often overlooked. We simply connect up, press play and enjoy the ride.

Or not.

Some clocks are better at keeping good time than others and if you use Midi Clock in your sequencing setup you need to know if your master tempo Midi Clock source really is up to the job or not.

By modifying a standard Midi cable with a capacitor and a resistor across the active and ground pins and then connecting these to a 1/4" audio jack, Midi data pulses are converted to audio clicks which may then be recorded to any DAW software application to check timing stability.

 

The maths is very straight forward:-

Midi Clock runs at 24 pulses per rhythmic quarter note.

At 120 BPM this equates to 2880 Midi Clock Pulses per minute.

A DAW recording sample rate of 48 kHz captures 2,880,000 samples of continuous Midi Clock pulses per minute.

At 120 BPM there should be exactly 1000 samples between each Midi Clock pulse when recorded at 48 kHz.

[Remember that for 4/4 time - a single rhythmic 16th division equates to 6 x Midi Clock Pulses.] 

The above screenshot shows 16 recorded Midi Clock 'pulses' as audio clicks into Sound Forge 9 (48kHz/24 bit) with precise sample markers placed at the exact front of each recorded pulse.

 

This screen shot is at maximum zoom showing the first marker placement at the very front of the pulse sample.

 

Same zoom setting but showing 2nd Marker at the same position of the pulse waveform.

 

The above screenshot shows two sequential markers and the exact number of samples between them.

 

 

 

 AKAI MPC-3000/Vailixi 3.50 Software Update

Recorded Midi Clock Intervals in samples: Port A Output

[1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000]

Maximum Midi Clock Error between any two sequential pulses = Nil Samples.

Maximum deviation away from 120 BPM/1000 sample target = Nil Samples.

 

Friend-Chip Mark 1 SMPTE Reading Clock

Recorded Midi Clock Intervals in samples: Midi Clock Output

[1000/1000/1000/1000/1000/998/1002/1000/1000/1000/1000/1000/999/1001/1000/1000]

Maximum Midi Clock Error between any two sequential pulses = 4 Samples (0.08ms)

Maximum deviation away from 120 BPM/1000 sample target = 2 Samples (0.04ms)

 

Innerclock Systems Sync-Shift Mark II

 

Test 1: MPC-3000 Midi Clock conversion to Din Sync

[1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000/1000]

Maximum Midi Clock Error between any two sequential pulses = Nil Samples.

Maximum deviation away from 120 BPM/1000 sample target = Nil Samples.

Conversion process maximum added jitter = Nil Samples

 

Elektron Machine Drum SPS-1UW MKII - OS Version 1.50F

 

Test 1: Pattern Mode, Internal Sync.

[999/1000/1002/998/1000/1000/1000/1000/1000/1003/998/1000/1000/1003/997/1000]

Maximum Midi Clock Error between any two sequential pulses = 6 Samples (0.125ms)

Maximum deviation away from 120 BPM/1000 sample target = 3 Samples (0.063ms)

 

 

Intel P4 2.4 1GB Ram/XP Pro SP2 32 Bit/ Live 6.0.7/FA-101 Driver 1.02

This Test: New Project, No Audio Tracks, No Midi Data - Midi Clock Output Only

[962/971/948/1453/947/982/951/948/960/973/958/952/974/955/1439/948]

Maximum Midi Clock Error between any two sequential pulses = 506 Samples (10.54ms)

Maximum deviation away from 120 BPM/1000 sample target = 453 Samples (9.44ms)