Korg Forums

[salomao]

I use an external keyboard to play my kronos (the sounds I assign to midi channel 2).

I use a MIDI cable connected in "midi out" port of the external keyboard and the "midi in" of my kronos.

I wonder if there is the chance of using a second midi external controler, since the "midi in" of my kronos is already occupied.

Is there any solution for this?

[voip]

Use a MIDI merge box to combine multiple MIDI sources into one data stream.

.

[psionic311]

You can use a 2nd MIDI controller into the Kronos using USB. For example, these will work:

Akai Advance
Akai MPK Mk2
Novation Remote SL
Arturia Keylab Mk2

As long as that other MIDI controller is USB class compliant, it should be plug and play.

[Xenophile]

Here's an old thread with a huge list of compatible USB controllers.

http://www.korgforums.com/forum/phpBB2/viewtopic.php?t=72768

[Hector Space]

Here’s some options that I’ve tried and that work:-
MidiSport 2x2 has merge and can be connected to the Kronos either by midi din or usb. If you use one of the Kronos usb A type ports this effectively gives you 3 midi din inputs ( 1 Kronos midi din plus 2 on the midisport)
Another route is to use a keyboard that has usb midi. These can be plugged directly into the Kronos’s usb A type ports. I’ve used the Casio Privia PX5s, a Roland RD2000 and a Kurzweil Forte all work.

A small point to remember: midi is a fairly slow serial data stream. This becomes obvious when you try to transmit real-time key information on several channels at once down a midi din cable. You get noticeable delays or latency on your note on events. Especially if you’re a pianist and your playing handfuls of notes to strick time (eg boogie woogie) ! Some keyboards are better at managing these timing issues than others.

You’d have thought using USB to USB to transfer midi would be much faster than DIN but often it doesn’t always appear to be that way.

This is because keyboards like the Casio Privia PX5s process midi at the DIN speed for both midi DIN and midi USB regardless of their zone settings.

I’ve found the best solution when using the Kronos as a sound engine connected to external keyboards is to transmit on only one channel and to use midi usb to midi usb.

Using a single channel and doing the zone and layer splits inside the Kronos means there’s less of a serial data traffic jam between keyboard and sound engine so things happen quicker. Similarly usb to usb for 1 channel is faster than midi din. Even if the transmitter only works at DIN processing speed when sending usb midi, at the receiving end the Kronos is faster at processing usb midi.

[Hector Space]

Here’s some options that I’ve tried and that work:-
MidiSport 2x2 has merge and can be connected to the Kronos either by midi din or usb. If you use one of the Kronos usb A type ports this effectively gives you 3 midi din inputs ( 1 Kronos midi din plus 2 on the midisport)
Another route is to use a keyboard that has usb midi. These can be plugged directly into the Kronos’s usb A type ports. I’ve used the Casio Privia PX5s, a Roland RD2000 and a Kurzweil Forte all work.

A small point to remember: midi is a fairly slow serial data stream. This becomes obvious when you try to transmit real-time key information on several channels at once down a midi din cable. You get noticeable delays or latency on your note on events. Especially if you’re a pianist and your playing handfuls of notes to strick time (eg boogie woogie) ! Some keyboards are better at managing these timing issues than others.

You’d have thought using USB to USB to transfer midi would be much faster than DIN but often it doesn’t always appear to be that way.

This is because keyboards like the Casio Privia PX5s process midi at the DIN speed for both midi DIN and midi USB regardless of their zone settings.

I’ve found the best solution when using the Kronos as a sound engine connected to external keyboards is to transmit on only one channel and to use midi usb to midi usb.

Using a single channel and doing the zone and layer splits inside the Kronos means there’s less of a serial data traffic jam between keyboard and sound engine so things happen quicker. Similarly usb to usb for 1 channel is faster than midi din. Even if the transmitter only works at DIN processing speed when sending usb midi, at the receiving end the Kronos is faster at processing usb midi.

[voip]

There was an interesting thread a while ago regarding latency, where the actual delays were measured:

http://www.korgforums.com/forum/phpBB2/viewtopic.php?t=116380

A real acoustic piano has quite a considerable delay between the application of finger pressure on the keys and the hammers hitting the strings. Something in the region of 20-30 milliseconds, when playing staccato forte, based on this article:

http://www.speech.kth.se/music/5_lectures/askenflt/measure.html

The delay would be expected to be considerably larger when playing pianissimo

It takes standard MIDI (31.25kbps) around 1 millisecond to send a note on message.

It takes sound around 12 milliseconds to travel 4 feet.

[Hector Space]

The latency issue is not just about the delay caused by playing a single note. Midi is a serial communications system. Meaning that every note played at the same time (and being a pianist this may easily 6 or 7 notes) has to be turned into a sequential stream (one note after the other) each note requires 3 bytes of data (note on command, note number, velocity value) at 31.25kbaud each byte ( including the extra start and stop bits) takes about 300uS so a single note will take 1 mS. 7 notes will take 7mS. Since the receiver cannot process a note until all 3 bytes are received, minimum delay will be the time taken for the transmitting device to scan the note and convert it into midi data (this will depend on how good the scanner is and how many notes it’s scanning) the midi transmit time + the time taken by the receiver to decoded the note, assign it and start the audio engine. Typically this will be about 3 or 4 mS. So once you get 7 notes you’re reaching anywhere from 14mS to about 25mS delay between the first note playing and the last note.

Whereas a consistent delay for all notes would be easily manageable, the flamming or slurring effect caused by the transmission mechanism is not. Nor is it always the same. When playing a real piano the delays are driven by a consistent parallel mechanism, ok a typical piano action is slower in the bass than the treble but it is consistent! And a proper piano action delay will consistently vary with touch, not the number of notes played!!

A typical midi controller with gradedhammer action suffers from midi latency, wobberly hammer delays and inconsistent scanner velocity measurement. Ok generally technology has slowly caught up with the pianist so now after 40 years of trying a keyboard like the Roland RD2000 with it PHA-50 keybed is extremely good at behaving like a real piano keyboard.

The midi latency issue was known from the start back in 1983 but for simple one note synth lines it wasn’t a big deal. Indeed they even created a system called running status which effectively allows the transmitter to only send the command byte once at the start of a string of note data. This reduces the delay for each note by a 3rd but it can only be used when transmitting on ONE channel and ONE command type. Hence the reason why I stated that it is better to use the Kronos's combi mode to do all the zone splitting and layering rather than sending multiple zone data on different channels via midi to the Kronos.
QED