midi keaboard lag

[MookleFlop]

hey there. im trying to run this basic midi controller (the e-mu x board) threw reason (v. 2.5) but im getting really bad lag times, to the extent that it makes it impossible to play to a beat. ive got it hooked up threw usb. its the old, slow version, 1.1 i think. im running xp w/ up to date service packs, amd athlon processor, 1.24 ghz, 512mb ram. on reason's cpu monitor it doesn't get about one or two notches when i play. is it the slow usb ports?
thx
-pat

[xjscott]

It turns out that's a really hard question to answer.

I have MIDISport USB MIDI interface and it has fine performance on a slow USB 1 port. I also have a Behringer BCR2000 that also connects via USB interface and has knobs and its own a MIDI port - but there is a huge amount of latency when using the BCR as a MIDI interface for an external MIDI keyboard.

My conclusion is that the Behringer's microcontroller is running some sort of embedded basic or forth or something rather than being coded in assembly or C.

I have seen this delay in MIDI in certain other recent MIDI products as well.

I also have concluded that the USB drivers on the M-Audio MIDI interface are quite good. But this does not necessarily extend to their whole product line, I know of one of their keyboard products which seems to have strange timing issues just between the keyboard and the unit's standard 5 pin MIDI interface.

Back in the good old days all synths had good timing. But nowadays I think a lot of these companies don't have actual musicians working in design and prototyping and they simply are not aware there is even a problem because they are not sensitive to timing issues. The same can be said for some companies making audio input devices that are sensitive to RF noise because they have decided to save some money and forgo the analog brick wall filter. If there was an actually musician working at such a company would have screamed bloody murder and such a product would not go out the door.

Technically, USB has non-deterministic timing, but it is good enough as long as the driver is well written and there isn't a bunch of other stuff on the same USB controller, like running USB audio on the same port or such, which can affect latency as the MIDI has to wait for the audio packets to pass through before getting a slot.

Although I have even seen bad MIDI timing on Firewire devices, which are deterministic so that is kind of astonishing.

I would try running your controller with no other USB devices and see if that helps, also see if you have the latest driver, and if still running into problems, try a sanity check of installing onto a completely clean PC with just the operating system and absolutely nothing else and see if that helps.

These bad timing situations can be very tricky to solve.

By the way, I assume you know about buffer settings in Reason's preferences, and this is not just an issue of having a large buffer and the high latency that comes with it. And that the latency is not of the audio coming out of Reason into your system, or that perhaps there is some audio routing driver that adds latency (On the Mac there is this audio routing thing called Soundflower that adds a lot of latency to the entire system, as will it if you allow various screen recording software to intercept audio system wide.) Assuming that you have already narrowed this down to MIDI input latency specifically and that's why you are asking about that.

[MookleFlop]

yes! the buffer worked, i was not aware of that. easy fix. when i have it real low the sound quality seems to suffer but i found a happy medium with no apparent latency. thank you!

[xjscott]

Check out pages 18-19 of the Reason 4 Operation Manual.pdf that came with the program, also chapter 12 "Optimizing Performance" starting on page 142. This will explain a lot about the buffer sizes and the reasons why they affect timing and performance, and these issues affect all software synthesis programs so it's worth becoming familiar with.