SFZ Mute Groups

You name it!
Post Reply
wduplessis
Newbie
Posts: 3
Joined: Sat Mar 17, 2012 1:04 pm

SFZ Mute Groups

Post by wduplessis » Sat Mar 17, 2012 1:39 pm

Hi Everyone!

This is my first post here, so I would just like to say WOW! LinuxSampler is absolutely awesome! I spent about a month trying to find a piece of software that would allow me to use my Roland electronic drum kit to trigger the drumkit samples I downloaded. Finding LinuxSampler was a revelation - what an amazing tool!

The reason for this post is that I have encountered a minor problem with mute groups under SFZ. I am not sure whether this is just me or the system, but any help would be greatly appreciated!

The problem is with muting hi-hats. The idea is that the hi-hat sound should be muted when another hi-hat sound is played. This is especially important with open hi-hat sounds which should only sound until the next note is played or the pedal is depressed. As I understand the SFZ specification at http://www.cakewalk.com/DevXchange/article.aspx?aid=108, a sound should be muted when a region with the group=??? equal to the off_by=??? value of the currently playing sample is triggered. However, this is not happening, and I am not sure why. This is a problem because electronic drum kits typically only have two notes (some only one) for different degrees of hi-hat openness, but many sample sets have 3 or more degrees of openness. I am using version 2331 of LinuxSampler and its libraries from the SubVersion server.

An example of one of the SFZ files I have set up is given below for the excellent Big Mono kit samples at http://www.analoguedrums.com/details-bm.php. The hi-hat bow on the drum kit triggers either note 42 or note 46 depending on the level of openness, and note 44 for the hi-hat pedal. The sample set in this example has 3 degrees of openness, so the two notes available are clearly insufficient to trigger all the sounds. The value of MIDI controller 4 sent by the kit is determined by the degree of openness of the hi-hat. The sample groups for notes 42 and 46 are identical because I am not sure where the transition between the notes happens.

Thanks for your time!
Warren


<group> key=42 locc4=65 hicc4=127 pan=0 loop_mode=one_shot seq_length=2 group=7 off_by=7
<region> sample=Wav/ZildMstrsnd-DynClsdLH01.wav volume=+42.52 lovel=1 hivel=8 seq_position=1
<region> sample=Wav/ZildMstrsnd-DynClsdLH02.wav volume=+37.01 lovel=9 hivel=16 seq_position=1
<region> sample=Wav/ZildMstrsnd-DynClsdLH03.wav volume=+33.61 lovel=17 hivel=24 seq_position=1
...
<region> sample=Wav/ZildMstrsnd-DynClsdLH16.wav volume=+6.20 lovel=120 hivel=127 seq_position=1
<region> sample=Wav/ZildMstrsnd-DynClsdRH01.wav volume=+40.19 lovel=1 hivel=8 seq_position=2
<region> sample=Wav/ZildMstrsnd-DynClsdRH02.wav volume=+35.47 lovel=9 hivel=16 seq_position=2
<region> sample=Wav/ZildMstrsnd-DynClsdRH03.wav volume=+30.81 lovel=17 hivel=24 seq_position=2
...
<region> sample=Wav/ZildMstrsnd-DynClsdRH16.wav volume=+5.75 lovel=120 hivel=127 seq_position=2

<group> key=46 locc4=65 hicc4=127 pan=0 loop_mode=one_shot seq_length=2 group=7 off_by=7
<region> sample=Wav/ZildMstrsnd-DynClsdLH01.wav volume=+42.52 lovel=1 hivel=8 seq_position=1
...
<region> sample=Wav/ZildMstrsnd-DynClsdRH01.wav volume=+40.19 lovel=1 hivel=8 seq_position=2
...


<group> key=42 locc4=1 hicc4=64 pan=0 loop_mode=one_shot group=7 off_by=7
<region> sample=Wav/ZildMstrsnd-DynSmiOpn01.wav volume=+33.48 lovel=1 hivel=8
...
<group> key=46 locc4=1 hicc4=64 pan=0 loop_mode=one_shot group=7 off_by=7
<region> sample=Wav/ZildMstrsnd-DynSmiOpn01.wav volume=+33.48 lovel=1 hivel=8
...

<group> key=42 locc4=0 hicc4=0 pan=0 loop_mode=one_shot group=7 off_by=7
<region> sample=Wav/ZildMstrsnd-DynOpn01.wav volume=+44.45 lovel=1 hivel=8
...
<group> key=46 locc4=0 hicc4=0 pan=0 loop_mode=one_shot group=7 off_by=7
<region> sample=Wav/ZildMstrsnd-DynOpn01.wav volume=+44.45 lovel=1 hivel=8
...

<group> key=44 group=7 off_by=7
<region> sample=../../nothing.wav

wduplessis
Newbie
Posts: 3
Joined: Sat Mar 17, 2012 1:04 pm

Re: SFZ Mute Groups

Post by wduplessis » Sat Mar 24, 2012 12:09 pm

Hi Everyone,

I seem to have found my problem... Well, most of it anyway!

Believe it or not, the main problem was that the last line was indented (which was not clear from my post - sorry!). I prefer to indent my <region> statements to make the files easier to read, but this is not supported by the SFZ standard as far as I can tell. This meant that the final group with key=44 was not being activated because, according to LinuxSampler, it had no sample. Without indentation, closing the hi-hat issues note 44 which now mutes all sounding hi-hat notes - pretty much what I was looking for. So now my kit mappings are working! Yippee! I will try to post some mappings for free drum kits available on the internet in the next couple of weeks.

The only additional thing that would be nice would be to have each <group> mute each <group> with the same group=ID to allow hi-hat notes to be muted without fully closing the hi-hat, but this does not seem to be supported by the SFZ standard. (This would be useful because it would allow a "tight" hi-hat sound where pressure over that required to close the hi-hat is applied to the pedal - this is supported by Roland's VH-12 hi-hat trigger.)

Cheers!
Warren
Last edited by wduplessis on Sat Mar 24, 2012 3:33 pm, edited 1 time in total.

wduplessis
Newbie
Posts: 3
Joined: Sat Mar 17, 2012 1:04 pm

Re: SFZ Mute Groups

Post by wduplessis » Sat Mar 24, 2012 3:32 pm

Me again...

My one e-drum kit is now working perfectly. But I am looking to sell it, so it is (hopefully) not going to be around for much longer...

However, my main kit only ever sends note 46 for the hi-hat bow, so the sound is never muted. The MIDI CC4 signal is correctly sent depending on how open the hi-hat is, so I can trigger the different levels of openness. On the down side, only having one note to work with means I cannot mute open hi-hat sounds with closed hi-hat sounds. (I realise that this is indicative of a problem with the kit, but I was still hoping that I could get everything to work together.)

Does anyone know whether it is possible to allow different groups on the same note to mute each other?

Thanks again!
Warren

Post Reply