Lag when controlling expression knob via midi

[pauliusmm]

There is a very big lag when controlling h9 exp knob via midi. Is there a way to fix this?

I control my axe fx via midi all the time and there is no lag.

[pauliusmm]

Any word on this? I want to use the expression pedal from my midi controller, but it is unuseable now, because of lag.

[gkellum]

My colleague, Nick, is on vacation now. I'll have to talk to him about it when he gets back. 

[pauliusmm]

Thanks, looking forward to this!

[nickrose]

There may well be a lag using Midi, probably depending on your controller. Tell us what you are using, with what settings, and we can look into it. But, it may well be beyond our control.

[pauliusmm]

I use liquid foot 12+ to control axe fx ii and h9, i tried liquid foot directly connected to h9 and it is the same. I have no problem with axe fx, there is no lag, so its not a controllers fault. Have you tested controlling h9 wah algo via midi expression pedal?

[string6]

Any update on this one ?

[MarcoR]

I also use a Liquid Foot +
12 + and have the same issue with an expression pedal controlling the the
factor pedals (not H9) via midi.

For example, using the
expression pedal to control PitchFlex responds slowly; it goes from 0 to 100
moving the pedal slowly but if you move it quickly, it just jumps to 0 or 100.

Same behavior using the expression
pedal through a midi solution box (no LF involved).

So I do not believe it is
the Liquid Foot. Also, the Liquid Foot controls my Lexicon PCM 80 without the
lag.

Using an expression pedal
connected to the pedal jack on the factor pedals works as expected. Controlling
the switches via midi CC works perfectly but trying to control any of the knobs
or the pedal has the lag.

[nickrose]

I'm going to guess that the LF sends a flood of controller messages when the pedal is changed, each one of which has to be processed by the 'factor. This probably takes time (display update, etc) and that causes the lag.I don't see an actual fault.

Because the PCM80 and the 'factor do different things on receipt of the message, the response time will be different. The solution would be to only respond to the latest CC and ignore earlier ones, but that is not a trivial mod, so not expected soon.

[MarcoR]

I'm going to guess that the LF sends a flood of controller
messages when the pedal is changed, each one of which has to be processed by
the 'factor. This probably takes time (display update, etc) and that causes the
lag.I don't see an actual fault.

Maybe
a lag is not the best way to describe it. It is more like only a certain number
of events can be processed at a time.

As
I said:

using the expression
pedal to control PitchFlex responds slowly; it goes from 0 to 100 moving the
pedal slowly but if you move it quickly, it just jumps to 0 or 100.

With the pitchfactor, in
the scenario I described it appears the values between 0 and 100 are skipped if
the pedal is moved to fast. There is no lag, it just doesn’t process the values
between 0 and 100.

With the Timefactor,
controlling the mix knob for example, there is a noticeable lag between the
time you stop moving the expression pedal and when the mix changes. It is harder
to hear the linier change with the mix than with the pitch so I can’t be sure
each value is processed.

The solution would be to only respond to the latest CC and ignore
earlier ones, but that is not a trivial mod, so not expected soon.

That
sound like what the Pitchfactor is already doing and that’s not desirable.

Nick,
is there any way you guys could test this? Even if you won’t be able to fix it,
it would be nice to have a reason so we can understand the problem better.
Maybe it is something that was just overlooked?

Most
users are likely just using the pedal jack and not trying control the pedal
with midi CC messages but I think more people will be using midi with the H9
and this issue may continue to come up. I would also be curious to know if the Eclipse
has this issue.

[nickrose]

We'll have to save up for a Liquid Foot 12+ and try it out.

The jumping from 0 to 100 is probably due to processing all the MIDI messages before updating the display, so this may not mean much – you don't say that this is audible.

The lag on controlling the mix knob is more problematical.

[MarcoR]

nickrose:

We'll have to save up for a Liquid Foot 12+ and try it out.

The jumping from 0 to 100 is probably due to processing all the MIDI messages before updating the display, so this may not mean much – you don't say that this is audible.

The lag on controlling the mix knob is more problematical.

You don't need a Liquid Foot 12+, same behavior with a Midi Solutions Pedal Controller. Give it a go with what you have available.

It is audible in both cases with controlling the pitch in the pitchfactor and the mix on the timefactor.

Using the pedal input on the factors works perfectly with an
expression pedal so I’m curious to see what happens if try to control the
Pitchfactor from the Timefactors expression via midi. If there is still a lag or sent
values not being processed, there is a problem that needs to be fixed. I'll do some more testing with the Space pedal as well. 

[nickrose]

I'll try MIDI Solutions…..

The pedal input will work better because it only measures the input when it is ready and does not have the "rubbery" message queueing issues that you are seeing with MIDI.

I would expect Space to be the same as the others – this part of the code is common.

[MarcoR]

Well Nick, I was mistaken… I tried the MIDI Solutions again and it works perfectly. I will contact FAMC about the Liquidfoot controller and see what they say. Thanks for looking into it anyway.

[MarcoR]

This is a shame… I can
control every other piece of gear I have via midi with an expression pedal from
the Liquid Foot without any problem except all the factor pedals and
Space.  But there is no problem controlling
the factors via midi with an expression pedal using a Midi Solutions Pedal
Controller. What can it be with the Eventides and the LF that doesn’t allow
them to respond properly?

[nickrose]

Huh – just ordered the MIDI solutions part – didn't want to spend the money on the LF. Maybe they will send me one if they are listening….

My guess is that it does not filter the pedal input much, and sends a flood of messages that clogs the Factor.

[MarcoR]

Sorry about that, I tried to post my results before you wasted your time.

I have a help ticket submitted with FAMC so they are looking into it also.

Nick, I have two Liquid Foot controllers, one is being sent to FAMC for another issue but I would be happy to send it to you to try after I get it back if you wish to investigate further.

[brock]

To everyone having MIDI pedal lag problems:

Wouldn't it be helpful to hook up a MIDI Monitor application? Then you can see what any given pedal is transmitting, for diagnostic purposes.  Many of them feature a MIDI Thru setting.  Insert a laptop or desktop computer between the MIDI pedal and the Eventide stompbox, and launch the MIDI Monitor application.

You can Google 'MIDI Monitor' for something that will fit your O/S and computer setup.  I've used MIDI-Ox (an old app), Bome's SendSX, and TenCrazy's MFX PortDiag (within Sonar), among other Windows applications.  I'm sure there are MIDI Monitors that launch within other common DAW applications.

Even if you don't 'read' hexidecimal, or know exactly what the MIDI messages are, you'll still be able to see any large volume of messages, or possible pedal 'jitter'.  You could copy / paste the the results into a forum post, and someone here would be able to translate.

You may just discover something new about your MIDI pedal.  For example, my FCB-1010 transmits 4 program changes (if configured that way), then a MIDI CC message, followed by the fifth PC message.  That can be used to configure Bank Changes before a Program Change (an undocumented feature).  It's always nice to know exactly what your pedal is sending, or check the configuration with visual feedback.

As for the MIDI Solutions Pedal Controller:  They make great products for unique situations.  I have several, but never had the need for the Pedal Controller.  But a line from their product description always intrigued me:

Quote:

The Pedal Controller accurately tracks the position of the pedal or potentiometer connected to it. Small fluctuations are eliminated when the pedal is in a stationary position and no extraneous messages are sent out. When the pedal is moved slightly, the Pedal Controller will send an update even if its position has only changed by a single value. Many control wheels or knobs only update themselves every two or every four values, resulting in coarser control, and sometimes zipper noise.

[MarcoR]

Hey Brock,  thanks for replying!  Yes, I used to use Midi-Ox on an XP machine. I will have to check out the others you mentioned.

FAMC has said that the next firmware an editor should address the issue with the factor pedals,Space and the H9. They have added a new option  "Force Zipper" .

"It will do exactly what it sounds like- force the pedal to send all
intermediate values between last value and the current. This shouldn't
be used in normal operation, but perhaps will help for those having 'H9'
type issues."

I will report back once to becomes available and I've given in a try.

[MarcoR]

Well, FAMC released 4.06 Firmware with a zipper setting for the expression pedals:

Quote:

NEW: Expression Pedal: Zipper Parameter. When turned ON, a CC type
expression pedal will automatically send all sequential values from the
last position to the current position instead of jumping. If the
previous position was value 2 and a very quick change brings the value
to 127, the zipper parameter will transmit every CC# between 2 and 127
automatically. Normally this would be turned OFF.

It did not fix the issue. I was skeptical it would as the expression pedal from the liquid foot already worked correctly with my other midi gear.

The ball is back in Eventide's court. There is something in the way Eventide reads the CC messages that is different from other manufactures that causes this problem with the factor pedals and the H9.

Nick, my offer still stands, I will send you a LF+ JR+ controller to test if you want to look into this further.

I will try some MIDI Monitor programs to see what's different between what the LF sends and Midi Solution sends. Maybe there is still a chance FAMC can do something if they know what to look for. They are really good about quickly releasing updates to fix issues their customers find.

[MarcoR]

Liquid Foot With Expression Pedal Eventide Issue: http://youtu.be/7xa02NMmRoQ

[nickrose]

MarcoR:

Nick, my offer still stands, I will send you a LF+ JR+ controller to test if you want to look into this

I'd be happy to. Send an email to support@eventide.com, and I'll send you a FedEx number to use.

[MarcoR]

Nick, I got your Fedex info. I'm going to hold off on sending it to you for now as it looks like FAMC is narowing down on the issue.

Here is the latest:

Quote:

Spoke to Jeff. He said what you want is the opposite of Zipper. It appears the Eventide can not process CC's fast enough and doesn't have its own anti-zipper. Seems odd. Solution- DO NOT use zipper. 

Jeff suggested a different idea- called partial zipper. He said he will program it in. This would send intermediate values, but spaced apart so the effects unit can grab it with space between values. 

Jeff said he believes that since the LF+ is so accurate w/ the pedal- it is capable of processing ALL pedal values. Most equipment are designed with midi buffer and zipper checks to handle this perfectly- which is why most equipment will work just right all the time. 

For devices with latency, or a device that makes MIDI a low priority in operation of unit, then perhaps it's too much to be real-time. So he can dumb it down. Its not good to be substandard, but perhaps this can help you. 

By nature, the LF+ will not allow zipper, it always reports real-time position, so we will add a little "extra" and call it Partial zipper. At least that is what Jeff said. 

That should help with the H9 device he is guessing. 

He said he would do exhaustive testing if Eventide sends him a test unit. He can study the issue and test various "types" of output to make it work perfect. Ask Eventide if they want to send him a timefactor or something. 

He said he would work with them with "sync" as well. Wow! He usually tells manufacturers who call him that he isn't interested at the time, so I guess he likes this particular company. 

If you guys could send them a box that would be great. If not, I can send my Pitchfactor but I think they should get an H9 for testing as well.

It is truely a great thing to have these two companies working together like this!

[nickrose]

Can you ask "Jeff" to contact me via support@eventide.com ?

Note that some CCs can take real time to process, depending on what they do, so there will always be limits on how much you can send. But, I'd be interested in knowing if this is the problem.

[MarcoR]

Nick,

I dropped the ball on this
issue… last I heard, Jeff was still looking at it with the H9. I did
want to report that I recently got an Eclipse and it does not have the issue; I
can use the Expression pedal connected to my Liquid Foot controller to control
any parameter on the Eclipse perfectly! I’m Whammy-ing up a storm. The Eclipse
is also replacing my Timefactor, Pitchfactor and Space now.

[jamongrande]

I just picked up a LF+ Mini and am having the same problems with Pitchflex. I've tried using the Zipper parameter as well to no avail. Other devices aren't having the same issue with quick changes between heel and toe. That said, I'm only 5 days into my LF programming, so I may need to try out a couple of other methods. I don't have the same issue when using a midi app slider though. Strange. I'll try to post more once I've done some more experimentation. 

Joe

[MarcoR]

jamongrande:

I just picked up a LF+ Mini and am having the same problems with Pitchflex. I've tried using the Zipper parameter as well to no avail. Other devices aren't having the same issue with quick changes between heel and toe. That said, I'm only 5 days into my LF programming, so I may need to try out a couple of other methods. I don't have the same issue when using a midi app slider though. Strange. I'll try to post more once I've done some more experimentation. 

Joe

Have you reported the issue
to Jeff at FAMC? Are you using an H9? Jeff was looking at this for me but was
having trouble recreating the issue because he has an H9 to test with where I
was using a Pitchfactor. He requested the exact steps to re-create the problem
with an H9. If you could provide that to him, I'm sure he’ll look into it again.
He’s literally released firmware updates to fix bugs the same day I’ve reported
them in the past.

[jamongrande]

No, I'm running a Pitchfactor as well. I'll check in on the FAMC soon; too much preset level LF programming to do before I get into those details. 

[spaceJam]

Hi,

What's the status on this issue?

I'm thinking in getting the LF+ Mini to control a Space, Pitchfactor (and possible H9) and that could be a 2nd strike.

[jmargulies]

hi – we've been discussing what i believe is the same issue here:

http://forum.eventide.com/cs/forums/t/10162.aspx

the response is not at all encouraging. i'm experiencing the problem using a Dunlop expression pedal converted to MIDI CC with a DB11 Switch Dr. 

i have also replicated the problem sending MIDI from Ableton Live – this makes it very easy to document exactly how bad the problem is. I will post pics/video as soon as I have time. 

as far as I can tell, Eventide's position seems to be that expecting the H9 to process a dense stream of MIDI is beyond reason, even though it's something that many far less powerful devices can do.  

j

[spaceJam]

I read your thread and a few others as well. It's pretty clear to me
that the MIDI implementation on these Stompboxes are behind other stompboxes. If these Eventide Stomboxes didn't sound so good, I'm pretty sure a lot of us would probably be on a different forum right now.

But here we are, reporting bugs, sharing our experiences, trying to help other users while we patiently wait for a solution to basic things that must to be tested before put it them on a product that will be on the market.

[LA Keys]

@spacejam:

I've posted some new and interesting results about this in this thread:

http://forum.eventide.com/cs/forums/t/10162.aspx

At this point I can see that there's maybe a few ways to alleviate the proble and come up with a working solution.

LA

[MarcoR]

As I mentioned on the first page, the Midi Solutions Pedal Controller works perfectly with the Eventide boxes for expression control via midi and they're pretty cheap. It's still a shame these pedals don't respond well with the amazing Liquid Foot controllers. I was working with FAMC on this issue but I ended up selling my factor pedals after getting an Eclipse. The Eclipse doesn't have as bad of an issue processing the midi messages (although it's not flawless) so I did not pursue the support request. I would encourage Liquid Foot users to contact FAMC about the issue as there is more likely a chance to receive a solution from them… 

[jmargulies]

It's great that there's a MIDI pedal that works. But this isn't really an issue with the Liquid Foot controller. It can be replicated with lots of other MIDI control devices, including automation sent from a DAW. 

[jmargulies]

The reason the MIDI Solutions Pedal controller works is almost certainly it's slow processing speed. A close friend of mine who designs and implements custom performance solutions for a big Vegas show said they had to dump their MIDI Solutions Event Processors because they were too slow for the needs of the show…funny that in this situation that same slowness is a feature!  

[nickrose]

MarcoR:

 It's still a shame these pedals don't respond well with the amazing Liquid Foot controllers. 

I sent Mr Foot a pedal to look at some months ago, but have not heard from him, so am not as hopeful as you.  I'll try and get it back, along with a pedal and hope to see what is happening.

The Eclipse and Factors have similar MIDI implementations. The fact that MIDI Solutions works fine and you say the Eclipse is not too bad, suggests to me that the issue is more complex than some believe. I'll report back when we know something.

[brock]

jmargulies:

The reason the MIDI Solutions Pedal controller works is almost certainly it's slow processing speed. A close friend of mine who designs and implements custom performance solutions for a big Vegas show said they had to dump their MIDI Solutions Event Processors because they were too slow for the needs of the show…funny that in this situation that same slowness is a feature!  

That wouldn't be a fair conclusion to draw.  The Event Processor converts up to 10 MIDI message events in a series protocol, with up to 32 event processes in the EV Plus (including processing of variables). Three keyboardists playing with polyphonic aftertouch.

I use an EV Plus under my 'board, fed through a MIDI router to my PitchFactor and/or H9.  I've seen what you're reporting when feeding a Stompbox with – for example – a cycling MIDI LFO, or a worn out encoder sending jitter.  Not so much from a DAW [Sonar], MIDI keyboard controllers in general, or an FCB 1010 (often processed by the MIDI Solutions Event Processor Plus).

Come to think of it, I don't recall having those results using "double" precision 12-bit pitch wheels.  But your experiences will have me revisiting just how hard I can push MIDI through the Eventide 'boxes.  In the home studio, I usually have a MIDI Monitor open in the background.  It should be easy enough to correlate the data throughput with visual & audible feedback.

[jmargulies]

fair enough! 

maybe it's just doing a more clever job of paring down the data stream when the pedal is moved quickly? this issue has gotten me very curious about the behavior of various MIDI controllers, and MIDI monitor results are showing that there are quite a few that send a more limited data stream when the control is moved quickly (i.e. jumps of 3 or 4 in the CC value). 

[jamongrande]

I posted this to the other thread, but figured I would add here in case others were interested. 

Of interest to Factor users, I reverted back to v3.5 software (I don't use H9 control, but rely on multiple midi controllers). I, able to get better response when adjusting multiple parameters simultaneously with the expression in the older software. The PitchFlex still jumps, though I don't have to move quite as slow. 

Given that I'm not using H9 control and can't say I hear a significant difference in the v5 algorithms' sounds (that was the update I was really hoping for), I might keep 3.5 for now. 

the next workaround will be to send the expression stream through another device that will translate it into the PF expression control. Perhaps that little bit of latency will improve the PitchFlex response. I'll post my results once done. 

joe

[Author]

Posts