PDA

View Full Version : Can't get damper pedal to work right



brucesiegel
01-26-2009, 04:20 AM
Hi,

It's been a few months since I've used my Garritan Steinway, and I'm having problems setting it up. I use it as a midi plugin in Protools on my Mac, and my controller is a Roland FP7. The damper pedal on the FP7 works fine when I play the keyboard all by itself, but when I use it as part of my Protools/Garritan setup it doesn't function properly.

Notes sustained by the pedal don't release exactly when I release the pedal. Instead, they linger for a full second or so. (In contrast, notes sustained without using the pedal, release instantaneously.)

Makes for VERY muddy pedal changes, to say the least.

Interestingly, the pedal icon next to the keyboard on the Garritan interface behaves properly. It releases precisely when I release the pedal.

Right now, I'm using "under lid" and everything else turned off (sustain resonance, sympathetic, ambience, tone).

I'd love some help with this!

Thanks,
Bruce

electone2007
01-26-2009, 06:02 AM
You might have a progressive kind of pedal on your FP7? You should set it up to be an on-off pedal.

brucesiegel
01-26-2009, 01:18 PM
Hi Electone,

Thanks for the reply, but that's not it. I set the pedal to on-off and I still get that horrible one-second delay before the sound quits. Anybody else have a suggestion? This is driving me crazy.

I didn't used to have this problem. I recently did a clean install of Mac OS 10.4.11 on the drive where I run ProTools. Could that have something to do with it?

OK—an update. I just got the pedal to work right when I use Garritan as a stand-alone app. Now I need to find out what's happening when I try to use it as a ProTools plugin.

Bruce

brucesiegel
01-27-2009, 04:51 AM
For anyone who's interested, I managed to solve this problem (thanks to Mirai, who posted a similar distress message on this forum), by disabling the lower MIDI transmission channel on my keyboard.

Life is good again.

Bruce

rickpaul
01-27-2009, 02:13 PM
It's been a few months since I've used my Garritan Steinway, and I'm having problems setting it up. I use it as a midi plugin in Protools on my Mac, and my controller is a Roland FP7. The damper pedal on the FP7 works fine when I play the keyboard all by itself, but when I use it as part of my Protools/Garritan setup it doesn't function properly.

Notes sustained by the pedal don't release exactly when I release the pedal. Instead, they linger for a full second or so. (In contrast, notes sustained without using the pedal, release instantaneously.)

Makes for VERY muddy pedal changes, to say the least.

Interestingly, the pedal icon next to the keyboard on the Garritan interface behaves properly. It releases precisely when I release the pedal.

Right now, I'm using "under lid" and everything else turned off (sustain resonance, sympathetic, ambience, tone).

I can't help here, but I have noted the same thing with the Basic version of the Steinway, which I just installed last week. I see it not only in SONAR, but also in the standalone version of the instrument. My controller is a Roland Rhodes MK-80, with an on/off sustain pedal. I tested other virtual pianos (e.g. Best Service Galaxy Steinway 5.1, NI Akoustik Piano), and they aren't seeing the same issue, so I know it isn't a problem with what is getting to the instrument from hardware or the layers in between the hardware and the softsynth.

I reported this issue, and I know Gary and Jeff are aware of it and planning to look into it in the not-too-distant future. Jeff mentioned thinking he had an idea on what might be causing it, but hadn't been able to look into it in detail yet due to being on the road.

rickpaul
01-27-2009, 02:13 PM
For anyone who's interested, I managed to solve this problem (thanks to Mirai, who posted a similar distress message on this forum), by disabling the lower MIDI transmission channel on my keyboard.

Bruce, can you elaborate on what you mean by this?

Thanks,

Rick

brucesiegel
01-27-2009, 07:42 PM
Hi Rick,

Here's the email Mirai sent me. I don't think he'd mind if I quote him here:

"I solved the problem by disabling the other two MIDI channels my keyboard was sending on. Usually, they split the played notes in upper and lower hand and then even use a third channel for something else. I just deactivated these channels. So, now my controller only sends the signals using one channel. This solved all issues for me."

On my own FP7, I only had one channel to disable. My keyboard's MIDI panel now shows these settings:
Upper Tx ch: 1
Lower Tx ch: off

This seems to have totally resolved the issue for me.

Bruce

rickpaul
01-27-2009, 08:09 PM
Here's the email Mirai sent me. I don't think he'd mind if I quote him here:

"I solved the problem by disabling the other two MIDI channels my keyboard was sending on. Usually, they split the played notes in upper and lower hand and then even use a third channel for something else. I just deactivated these channels. So, now my controller only sends the signals using one channel. This solved all issues for me."

On my own FP7, I only had one channel to disable. My keyboard's MIDI panel now shows these settings:
Upper Tx ch: 1
Lower Tx ch: off

This seems to have totally resolved the issue for me.

Thanks for clarifying, Bruce.

I'm afraid my keyboard only sends on one MIDI channel in the first place, so that won't help here. I also have that set up to be channel 1, which is the same as yours, and tried setting SONAR up to force anything coming into the associated MIDI track to channel 1, and to only take input from one MIDI port on the one channel, too, just to rule out anything funky going on elsewhere. (I do have another MIDI port with another controller on it, but that has been turned off while I've been trying the Steinway out.)

It's not clear to me why this sort of thing would make a difference anyway, but it's great that you've got a workaround for your uses.

I should probably clarify that I'm running the Basic version (latest official release, which is 1.02) on Windows XP Home SP3. I'm not sure how that may relate to the other versions of the software, other than that the Basic version only has the under-the-lid perspective and 16-bit samples.

Thanks again for the quick response.

Rick

brucesiegel
01-27-2009, 09:18 PM
Sorry to hear you couldn't resolve the issue, Rick. Our situation is evidently different, though, because my sustain pedal was working properly in the standalone version even before I changed the MIDI settings. I too have the Basic version of Garritan, by the way.

Bruce

electone2007
01-27-2009, 10:08 PM
Two ideas:

1) Maybe set the sequencing software to receive on just one channel instead of "Omni".
2) Look into "Event List" or Staff View to see what MIDI data is being sent and where the pedal releases are effected (P and * in Sonar).

rickpaul
01-28-2009, 03:10 AM
1) Maybe set the sequencing software to receive on just one channel instead of "Omni".

As I mentioned in my reply to Bruce earlier, I tried forcing SONAR to not only send just MIDI Channel 1 to the Steinway VST, but also to only accept input from the specific MIDI port and MIDI channel (also channel 1) that my controller was on for the relevant MIDI track. It didn't make a difference.

Note, too, that I'd initially observed this behavior in the standalone version of the Steinway.


2) Look into "Event List" or Staff View to see what MIDI data is being sent and where the pedal releases are effected (P and * in Sonar).

I just took a look at the test case I sent to Garritan to demonstrate the issue. There is one thing that looks somewhat funky to me, which I've noticed before, but which also behaves just fine with the other software pianos I tried with it. In particular, many, but not all, times there is a sustain pedal event, there is actually an extra pair of events.

For example, say there is an initial sustain pedal on event, before any other events have occurred. Instead of just having one instance of CC#64 with the value 127, it has a quick sequence of CC#64 with 127, then CC#64 with 0, then CC#64 with 127 again, for example with on the order of 3 ticks (in 960 ppq) separating the events. In the case of a sustain pedal off event, it has a quick sequence of CC#64 with value 0, then CC#64 with value 127, then CC#64 with value 0. And some of these might additional quick event pairs, all in rapid sequence.

As I noted, this doesn't happen all times, so I suspect it could be a relatively natural phenomenon with on/off pedal movements, where the pedal goes quickly back and forth through the mid-point more than once enroute to its final destination.

I just tried a very simple test where I pencilled in a chord that got held out partway over four short notes (e.g. sixteenth notes) at a quarter note pulse. The chord cut out around the same time as the second short note. I then tried writing a sustain on event just after the start of the chord and first short note, then a sustain off event. At that point the sequence played back as I would expect, which is to say the chord and second short note cut out with the sustain off and the two remaining short notes sounded staccato (the reason for writing in the short notes). Next, I added a quick sustain on and off just prior to the above-mentioned sustain off, with three ticks separating each event. This time, the chord and first short note rang longer than expected, as did the additional short notes.

This suggests that the extra events are triggering some special behavior, probably repedalling. This would probably be legit given the quick pedal movements, EXCEPT that the sustain off should then kill the repedalling on the chord and second short note, and we should never hear the longer release samples for the additional short notes. So maybe the sustain off is being missed once it has gotten itself into repedalling mode (at least until such time as some longer breaks between sustain pedal movements come along to reset things)?

The other thing I should note is that, while it might seem that, at least for a sequenced performance, a workaround might be to manually edit the sequence to remove any unintended sustain events, actually editing my original test sequence this way fixed some of the parts that weren't behaving as expected, but not all of them. (And keep in mind, all these parts behaved as expected on Akoustik Piano and Galaxy Steinway 5.1.) Thus, it would seem that some normal pedalling (i.e. without extra events being added by a finicky on/off boundary's being crossed multiple times in one pedal movement), which just happens to occur quickly, can also trigger this behavior.

Of course, the ability to tediously edit this kind of thing in a sequenced part won't help for a live performance, be it through the sequencer or in the standalone mode. However, hopefully this extra information will help in finding a fix as I'm pretty sure the behavior is a bug.

Rick

brucesiegel
01-28-2009, 03:55 AM
Correction: I said I have the basic version. Actually I have the standard version.

domi
01-28-2009, 05:07 AM
Hi all

I exactly noticed the same problem (about 1 second delay before damper off stops the notes) wirh my basic Steinway, used in standalone mode.

I was using a midi setting from a MAXMASP application, where pedal information is duplicated on midi channels 1 and 2. As someone pointed, I found that sending pedal information on the single channel 1 solves the issue.

I am pleased with that workaround, but there are some case where you really need multi channels pedal information. I hope this will be solved in a future release.

Domi

rickpaul
01-28-2009, 05:19 AM
I was using a midi setting from a MAXMASP application, where pedal information is duplicated on midi channels 1 and 2. As someone pointed, I found that sending pedal information on the single channel 1 solves the issue.

I think I finally understand why the duplicate channel thing makes a difference based on the above reply. It isn't because of transmitting on multiple channels, but rather because these keyboards are likely duplicating the pedal information, which the Steinway likely effectively processes in Omni mode, as if everything were on a single channel. Thus, it gets to a scenario similar to what I was seeing, where there are tightly spaced extra pedal events, which trigger the repedalling, but then it doesn't shut off appropriately on sustain pedal off events.

I suspect those of you who are finding turning off the extra channel may also find that certain types of pedalling action, especially when using the pedal quickly in fast passages, may still trigger this behavior. It just won't be as constant as it was when the keyboards and/or applications were intentionally generating the extra events on a separate channel.

Just to clarify (or restate?), in my case all the sustain pedal events were on channel 1. However, there were extra events in rapid succession owing to the way my sustain pedal and keyboard controller trigger events in response to my specific foot motions.

Rick

Curt
01-28-2009, 06:33 PM
I wasn't seeing this problem before but I just installed the beta update that Jeff gave us and now I'M seeing it too! I'm running the Pro version in standalone.

I don't really like the idea of tampering with how my controller is sending data. I mean it wasn't an issue before - it isn't an issue in Ivory. It just showed up after I installed the update???

I really do like the update though!

Jeff Hurchalla
01-29-2009, 10:05 AM
It sounds like a regression (a bug introduced by something you've changed) and it will be one of the first things I look at when I'm back this weekend. It's pure speculation until I can look at it, but it may be that a sustain pedal that sends only 0 or 127, or that somewhat randomly sends sustain information, is what triggers this. Using a continuous response sustain pedal, I hadn't experienced it. I don't think I would have missed hearing this if it were present on the systems I had been using at home and at NAMM, or that all the testers would have missed it, unless it's some specific combination of circumstances that triggers it. I'll definitely look at it starting this weekend, and hopefully be able to reproduce the bug without too much trouble.

electone2007
01-29-2009, 05:24 PM
Don't know if this will help, but if you look at the keyboard thingy at the lower part of the GUI, the keys stay pressed down even after note off with the sustain pedal down. Didn't notice this before the update.

Curt
01-29-2009, 05:34 PM
I noticed this too. I figured it was a "feature" of the update. :)

Curt
01-29-2009, 06:32 PM
It sounds like a regression (a bug introduced by something you've changed)
To my knowledge the only thing I changed was installing the update?

but it may be that a sustain pedal that sends only 0 or 127, or that somewhat randomly sends sustain information, is what triggers this. Using a continuous response sustain pedal, I hadn't experienced it.
I use the stock pedal in continuous mode that came with my Roland FP-4.

Good luck reproducing it! Hopefully a fix won't be far behind!

Jeff Hurchalla
01-30-2009, 10:36 AM
I should have been clearer and not said - something "you've" changed - but instead - something "a developer" (me) has changed. I was typing too quickly :)

Curt
01-30-2009, 06:17 PM
OK, you had me scratching my head on that trying to figure out what I did!!

Don't work too hard this weekend. ;)

r2m3
02-13-2009, 09:31 PM
I'm also having trouble with the damper pedal. My setup is a Roland HP-207 connected via USB to a Vista machine running GS Professional. I've tried all versions, including the latest beta. My two problems:

1. Notes sustain for about a second longer than they should after the damper pedal is released. I was able to work-around this (below).

2. Proportional pedaling does not work at all.
I will note that I regularly use Ivory and pianoteq, neither of which exhibits the above problems. This suggests the problem is with GS and not my keyboard and computer setup.

Regarding the first problem, I was able to get notes to release correctly after releasing the damper by changing a setting in the HP-207's menu under "Damper Pedal Part" to "RIGHT" only. The Roland manual says this function is related to when you split the keyboard with two tones, you can specify if either or both the left and right parts should be affected by the damper pedal. Somehow changing the setting to RIGHT fixed the damper sustain in GS. I was thinking this problem may be a glitch with the HP-207's MIDI commands somehow duplicating sustain info (like the channel issue others reported earlier in the thread) that confuses GS causing notes to sustain longer than they should. However, Ivory and pianoteq work fine...

The above workaround works well but is a bit of a nuisance as I have to reset the damper setting in the menu each time I turn on the keyboard.

The second problem with proportional pedaling is extremely annoying. I'm certain the Roland HP-207 has continuous pedaling support as it proportional pedaling works flawlessly in its own generated sounds as well as other software synths. The pedal in GS appears to work in either full ON or full OFF mode - I have to fully release the pedal for notes to dampen at all (half-pedaling doesn’t work – full sustain remains on until I completely release the pedal). Additionally, there's an ugly sonic artifact of the preceding cord that occurs when you release and quickly depress the pedal.

The manual and specs all say proportional pedaling is supported, however I have can't get it to work. I had been waiting for the next update to fix this issue (as it has been rumored in other posts), but seeing as the beta didn't resolve any of my problems I wanted to ensure these issues were heard.

Thanks for the great work so far. I really hope there's a way to fix these issues soon, as I can’t use GS without proportional pedaling for much of anything right now.

electone2007
02-14-2009, 01:01 AM
The manual and specs all say proportional pedaling is supported, however I have can't get it to work.

I purchased the Steinway because of this particular feature that was mentioned in the ads and even painstakingly defined in the manual (though I don't agree with the definition at all). I know they're very nice guys, but $399 is not something to sneeze at either...

Sigh...

I still await the update...

Larry G. Alexander
02-14-2009, 06:55 AM
I can't get proportional pedalling to work in my Basic edition either. It's either full on at 127 or full off at 0 - there is no in-between.

Pingu
02-14-2009, 08:25 AM
I'm a bit of a newbie on the Steinway. I've done little more than plug it in and check it made a sound. I have a progressive pedal, and when I pressed it the picture of the sustain pedal on the GUI got progressively darker, so I assumed things were fine, although I didn't listen too closely.

Is the progressive pedalling not working for anyone?

Curt
02-14-2009, 09:13 AM
I had been waiting for the next update to fix this issue (as it has been rumored in other posts), but seeing as the beta didn't resolve any of my problems I wanted to ensure these issues were heard.In fact, the beta made the pedal release issue even worse!! Almost a full second worse.



Thanks for the great work so far. I really hope there's a way to fix these issues soon, as I can’t use GS without proportional pedaling for much of anything right now.
Yeah, you're not alone there. I too have to hold off using it for any recordings. It's killing me too as I can't wait to get working with it.

Curt

rickpaul
02-14-2009, 03:26 PM
I'm sure Jeff will chime in sometime soon with the status of his exploration (and hopefully fix) of this (these?) issues, but here are a few tidbits I'm reasonably sure I understand that may clarify some of the confusion mentioned in the last few posts.

First, one of the problems with the pedaling, specifically the one relating to sustain hanging on when it shouldn't, seems to have come into the picture with the Basic version of the Steinway, which I believe (from memory) has a version stamp of 1.02. The earlier Pro version (and maybe the in-between version?) had an earlier version stamp, so did not have that specific problem. However, there was also a beta posted for one or both of those that was also 1.02-something, and that did introduce that same problem with the other version(s). Basically, at least from what I'm grokking on this, something Jeff fixed in 1.02 likely introduced this specific problem.

Second, I don't know, one way or the other, about the proportional pedaling thing since I don't have a continuous sustain pedal. However, if it was working in the 1.01 (I think) version for the Pro and other versions, and is not in the Basic and new beta of the other version(s), then it would stand to reason that problem might be related and possibly introduced with the 1.02 changes.

Third, some people have been able to workaround SOME of the symptoms with this "sticking sustain" thing by turning off one channel of sustain pedal sending on controller keyboards that duplicate sustain pedal messages on multiple channel. This isn't actually a 100% reliable workaround, but may alleviate some, or even most, symptoms of it, depending on the nature of the piece you're playing, because the problem appears to have something to do with the frequency of sustain pedal messages. Thus, a piece where they are relatively spread out, as opposed to having quick offs then ons, probably wouldn't trigger the issue. On the other hand, if you're playing a fast piece with a lot of quick off and on action, you'll probably still see the problem. Additionally, if you happen to have a sustain pedal and controller like my combination of those (Roland Rhodes MK-80) that sends extra off/on pair messages in some cases (probably related to crossing the threshold between full 0 and full 127 somewhere in the middle of the pedal's action, where it may oscillate back and forth quickly with certain types of foot motion), albeit only sending this info on one channel, then you will also likely see the problem, though you may be able to manually edit out the extra messages to alleviate it. That editing will not, however, avoid the issue with (intentionally) quick off/on action.

I believe the nature of the problem is that the quick on/off action is (properly) triggering repedalling to occur, but the second "off" in the pair of quick movements of the pedal (or duplicates in the case of a controller sending on multiple channels) is not properly turning this mode off. Thus, the key to avoiding it is to avoid quick off/on motions -- i.e. where the next sustain pedal on comes very shortly after a sustain pedal off.

While I don't know anything about the proportional pedalling, based on the notion that some here are suggesting a problem in that came along at the same time, perhaps it may be something like, once you get in the repedalling mode, further proportional pedalling action breaks down until once the repedalling mode is turned off again? That could be easy enough to test by checking what is in your MIDI data and making sure any extra off/on commands (or data on a second MIDI channel, which would also create duplication) are deleted, while also being careful to avoid the quick off/on motions. If it does turn out to be the case, it might provide a workaround, albeit one that would be a pain (i.e. tedious) in practice, until such time as a fix is available.

Rick

electone2007
02-14-2009, 05:13 PM
I can't get proportional pedalling to work in my Basic edition either. It's either full on at 127 or full off at 0 - there is no in-between.


You have to have a continuous type of sustain pedal.

electone2007
02-14-2009, 05:15 PM
I'm a bit of a newbie on the Steinway. I've done little more than plug it in and check it made a sound. I have a progressive pedal, and when I pressed it the picture of the sustain pedal on the GUI got progressively darker, so I assumed things were fine, although I didn't listen too closely.

Is the progressive pedalling not working for anyone?

No.

It only works on the pedal image in the GUI but does not translate to correct behaviour in sound.

In fairness to the developers, they implemented the proportional pedaling to affect only the soundboard resonance. Try it with the resonance volume way up. That is also how they defined proportional pedaling in the manual, i.e. to affect the soundboard resonance.

I completely disagree with this implementation because in a real piano, and in other pianos vsti's such as pianoteq, and also in my Clavinova, proportional pedaling affects the decay of the note itself, not just the soundboard resonance.

I've written Gary about this a month after I purchased the Steinway. That was about six months ago.

I want to know from Jeff if this will be fixed in the update. If not, then I won't have to be coming here everyday in this forum just to check. It becomes tiresome after a while. I hope the developers understand how I feel about this and I apologize if I come across as too demanding.

Curt
02-15-2009, 08:14 AM
I don't know if THIS is related to the same issue...

With no pedal at all - if I play octaves in the left hand in a quick chromatic run starting at, let's choose A1 and play six or seven steps up (A, A#, B, C, C#, D, D#) and then stop and listen... there is an incredible amount of hiss and noise remaining that lasts for about the same time that the "hanging sustain" issue lasts - one to two seconds.

This is with the beta 1.02 build....

Curt

r2m3
02-15-2009, 07:42 PM
In fairness to the developers, they implemented the proportional pedaling to affect only the soundboard resonance. Try it with the resonance volume way up. That is also how they defined proportional pedaling in the manual, i.e. to affect the soundboard resonance.

I completely disagree with this implementation because in a real piano, and in other pianos vsti's such as pianoteq, and also in my Clavinova, proportional pedaling affects the decay of the note itself, not just the soundboard resonance.

Interesting. Thanks for sharing that. I can confirm by turning up "sustain resonance" to 99 I could hear the effect of proportional pedaling on soundboard resonance, but not on note decay. So GS is reading continuous pedal data.

I'd like to echo that the current implementation feels very unnatural to me. It is inconsistent with most good digital pianos and software synths I've played, and of course real acoustics. I'm certainly not knowledgeable about how proportional pedaling should be implemented, but it seems a quick fix might be to simply continuously decrease the volume of sustained (and not pressed) notes proportionately to the lowest continuous pedal output, untill the pedal is full off at 0. I believe that's what I hear in most digital pianos.

I very much hope Jeff or someone can explain the reasoning behind the current implementation of regulating only the soundboard resonance or confirm that a fix including note decay is forthcoming.

electone2007
02-15-2009, 08:09 PM
I'm certainly not knowledgeable about how proportional pedaling should be implemented, but it seems a quick fix might be to simply continuously decrease the volume of sustained (and not pressed) notes proportionately to the lowest continuous pedal output, untill the pedal is full off at 0.

I think what others do is send the continuous pedal output to modify the ASDR envelope proportionately.