Sound Editing Section

[Falko]

Hi
Just stumbled upon this page today. Been researching different drumkits as wanted to get rid of my ddrum4se and purchase either alesis strike or 2box speedlight, but this little piece of software saved me from doing this. Excellent work! I just tried it and made my first .mid for dd4. Since I'm currently away from my kit I cannot test it yet on module , but with some trial and error I'm sure I will get it. I have some questions regarding velocities and config file. I plan initially to convert samples from 2box. I load .dsnd file into Drumit Manager (DSoundManager) and export with this tool to .wav so I have all the velocities in separate .wav files :-) I see that the .cfg file (in Begin-Layers section) contains 8 fields for Gain Velocity, does this mean ddrum only discerns 8 hit velocities? And for each velocity I put different sample/Layer (will choose obviously from these .vaw files with trial and error best ones). Is this the right way I am thinking ? If not, please tell me how to configure this .cfg files so that when I have 8 .wav samples of snare drum from ghost notes to fully loud , which sections of config file to edit and how. There is really no explanation in manual for this section

Thanks for your effort ! 

[kenjwright]

Hi and welcome!

Yes, the user guide does not go into much detail about parameter adjustments. Will add that to the todo.

You are correct that the dd4 is limited to 8 velocity zones, basically 128/8 = a range of 16 values per zone. So range would be close to below in blue with respect to mapping from the 2box.

The value is either FF (on) or 00 (off). On means the zone is enabled, off = disabled for that layer. 
-------------------------------------------------------------------------------
 Sample  HiHat Gain Velocity 
   ^     <---> <-soft-----------hard-> 
   01 02 03 04 05 06 07 08 09 10 11 12
-Begin-Layers-
L1 00 00 00 00 FF FF FF FF FF FF FF FF
   |           |  |  |  |  |  |  |  +- 122-127
   |           |  |  |  |  |  |  +---- 116-125
   |           |  |  |  |  |  +------- 109-120
   |           |  |  |  |  +---------- 100-113
   |           |  |  |  +-------------  90-106
   |           |  |  +----------------  75-97
   |           |  +-------------------  53-85
   |           +----------------------  01-70
   +---------------------------------- Sample# (00-09)

Overlapping zones between layers will result in multiple samples being played simultaneously which may or may not be what you want! (i.e, possible phase cancellation, etc.) The samples (wav files) listed in the cfg file are assigned 0-9 in the order read for a maximum of 10 samples. The sample is bound to a layer in the first parameter value after the L1-A id, i.e, L1 00 = layer #1, sample #0, L2 01 = layer #2, sample #1. Any sample can be assigned to any layer.

Hope this helps. Let me know if you need more clarity.

Cheers!

Ken

[Falko]

Ok so as I understand to achieve my goal - 8 .wav files of snare from ghost notes to fully loud would be something like this:
-------------------------------------------------------------------------------
Sample HiHat Gain Velocity
^ <---> <-soft-----------hard->
01 02 03 04 05 06 07 08 09 10 11 12
-Begin-Layers-
L1 00 00 00 00 FF 00 00 00 00 00 00 00
L2 01 00 00 00 00 FF 00 00 00 00 00 00
L3 02 00 00 00 00 00 FF 00 00 00 00 00
L4 03 00 00 00 00 00 00 FF 00 00 00 00
L5 04 00 00 00 00 00 00 00 FF 00 00 00
L6 05 00 00 00 00 00 00 00 00 FF 00 00
L7 06 00 00 00 00 00 00 00 00 00 FF 00
L8 07 00 00 00 00 00 00 00 00 00 00 FF

and then in SAMPLE FILES section specify each .wav file from soft to loud ?

If my thinking is ok with this, is it also possible to expand the above to alternate for example two highest velocities (f.e. through sequences?), just to add a little bit more realism and less machine gunning. This is already done in VST libraries - you hit maximum velocity and one sound plays, hit max again and different sound plays. This would obviously require loading these two extra samples in SAMPLE FILES section.
I imagine it to be something like this:
Begin-Layers section looking like above, extended further with two extra sample lines:
L7 09 00 00 00 00 00 00 00 00 00 FF 00 <-- sample 9 on old L7 layer since this is the one I would like to alrernate, so layer L7 would alternate sample 09 and 06
L8 0A 00 00 00 00 00 00 00 00 00 00 FF <-- sample 10 on old L8 alternating samples 07 and 0A
but not sure if this is possible, these two above lines would probably only replace previous L7 and L8 instead of adding alternation? So maybe incorporate sequencers ?
And Gain Position field - is it also 00 = off and FF - on ?
Thanks for effort !

[kenjwright]

It seems I lied a bit about the velocity ranges ;-) Just did some testing and the zone ranges are not equal divisions but rather logarithmic so the lower velocities have a wider range and the higher velocities a narrower range. There is also an overlapping cross-fade between zones as they transition from one zone to another.

I have added DUMP_118-speak_velocity to the Sound Files directory on the DropBox site that demonstrates what you are wanting to configure. The sample files are spoken numbers, 1-8 (brilliant borrowed idea from @krillo !!). Load this into your module and assign to one of your pads to hear each velocity zone spoken as you hit from soft to hard. The config file shows the parameter setup. Regarding the sequencing, look at the VARIATION parameters. There are 10 possible variations that can be defined. Variations have two configuration lines each. The first line (VLx) toggles which layers are active for that variation. There must always be at least one layer toggled on (on = any value greater than 00). The second line (VSx) sets the sequencing. Any layer that is "on" (not 00) will be included in a random sequencing of the active layers. In DUMP_118 I set VL1 to have layers 1-8 active, but no sequencing (all 00). For VL2 I set layers 1-10 as active and assigned sequencing to layers 8,9,10 (and assigned sample 0 to layer 9 and sample 1 to layer 10). With variation set to 1 on the module (the default) you will hear the transition between zones, with hardest hits only repeating the word "eight". But with the variation dialed to 2 you will hear randomly spoken "one",'two","eight" with repeated hardest hits. To differentiate between a normal "one" and "two", layers 9 & 10 have the pitch raised by 6 half steps.

Note that the anti-machine gunning built into the module itself is controlled by the filter1&2 parameters. ddrum4edit sets these to all zeros by default but these can be altered to suit. However although the parameter definitions are documented in the config file, the actual values used are a bit of a mystery. I have spent some time playing with these but have not yet come up with any clearly defined recipes.

Regarding the Gain Position parameters (positional sensing); yes, FF = active and 00 = off.

Cheers!

Ken

[Falko]

This is freaking amazing it's possible to achieve this ! Unfortunately tomorrow I have a travel day so can't test but weekend is going to be probably big testing time :-) can't wait
I cant see the dropbox link, probably I need to register to this forum?
Speaking of Gain Position field, not sure about the limitation of number of samples in one file, but would it be also possible to do all the above for, say, 5 slots from center of the pad, and then another 10 different samples for the 3 edge zones of the pad? By editing Gain Position fields and then making one single .mid file with positional sensing snare?

[kenjwright]

This should(?) work; www.dropbox.com/sh/1xd477h1b0hqu0j/AAC0D8h99OjexEMXzq-M3XJoa?dl=0

You only have the 10 max layers and 10 max sample files to work with so any combinations that you can cook up just need to fit within those.

[Falko]

Coming from the coding background I can see your coding inclination by that semicolon after "work" word above ;-)
Just had a look at config file from DUMP Speak Velocity. Just to be clear since I'm still away from my drumset - in the VARIATIONS section you typed
VL1 01 02 03 04 05 06 07 08 00 00 (01 = Layer enabled in this variation)
VS1 00 00 00 00 00 00 00 00 00 00 (01 = Layer enabled for Sequencing in this variation)
VL2 01 02 03 04 05 06 07 08 09 0a
VS2 00 00 00 00 00 00 00 01 01 01

Not sure why there are higher numbers than 01 here. As I uderstand, anything other than 00 would enable the layer, so in fact it wouldn't matter if you replaced all these numbers 01 02 03 .. 0a with only 01s, am I right? So this would produce same result:

VL1 01 01 01 01 01 01 01 01 00 00 (01 = Layer enabled in this variation)
VS1 00 00 00 00 00 00 00 00 00 00 (01 = Layer enabled for Sequencing in this variation)
VL2 01 01 01 01 01 01 01 01 01 01
VS2 00 00 00 00 00 00 00 01 01 01

Or does this matter?
Thanks !

[kenjwright]

Ya caught me! Yeah, just me making layer visibility a bit easier in the config file (cut/paste from another file in fact). Any non-zero works. The factory sound files use 01 but deep in the code somewhere just translates to a boolean so either way works fine.

Cheers!

Ken

[kjob3022]

merry xmas to you Ken and congrats on an awesome volume of work.I last saw this in 2020 and back then it was just starting to cook,now look at what we have from you.I downloaded this from dropbox onto a MacMini but I am lost even thru trying to figure out the pdf parts so if you could walk me thru how to use terminal for edit.Do i need sysex too now and do i hook my module up with the old midi cables from ddrumtool days?I feel like such a chump asking you for your help on this after what you've given us but any help you could for a regular,thanks again,Ken

[kenjwright]

Dec 26, 2022 17:05:08 GMT -5 kjob3022 said:

merry xmas to you Ken and congrats on an awesome volume of work.I last saw this in 2020 and back then it was just starting to cook,now look at what we have from you.I downloaded this from dropbox onto a MacMini but I am lost even thru trying to figure out the pdf parts so if you could walk me thru how to use terminal for edit.Do i need sysex too now and do i hook my module up with the old midi cables from ddrumtool days?I feel like such a chump asking you for your help on this after what you've given us but any help you could for a regular,thanks again,Ken

Hi Ken! Happy holidays to you!

ddrum4edit is indeed a console based app so are you needing general help with use of the terminal? If so, some of that is covered in an earlier post within this thread. 

For the MIDI cable, yes same as before. I find the iconnectivity products work really well. For the Mac there are no drivers required.
The old ddrum4Tool still works fine for MIDI transfers using Windows. For the Mac I suggest SysEx Librarian which supports both .sys and .mid file types which ddrum4edit can produce.

To edit an existing ddrum4 file, first use the -e export switch; ./ddrum4edit -e pathToFilename.mid. This will create a new configuration file Filename.cfg file in the current directory. You can then use any text editor (TextEdit works fine) to adjust the parameters as desired and then create a new .mid file from this using ./ddrum4edit -c Filename.cfg.

That should get you started.

Cheers!

Ken

[kjob3022]

Thanks a lot again Ken,this is really helping

[kenjwright]

Greetings All! And Happy New Year!! ddrum4edit v1.01.00 is now available!

This version provides some new features, enhancements, and bug squashing.

Highlights for this release (since 1.00.29);

• Force a resample to 44100Hz if source wave file sample frequency is > 44100.
• Added --layer toggle to apply layer parameters to extract and listen functions. This feature applies the embedded gain, pitch, and decay values in the sound file to the output to match how the ddrum4 module plays the sounds (excluding the realtime dynamic filtering).
• Added --aiff and --riff toggles to explicitly define the extract file format when only one or the other is desired (the default is both).
• Reading/writing of System & Pallet settings is now supported. Like Kits, the System & Pallet settings can now be dumped and exported to, and maintained in text based configuration files.
• Smarter file recognition to automatically determine the various file types now used as input for printing, exports, configurations, etc.
• The listen feature (--listen) now also includes support for playing AIFF, RIFF, and DSND (2Box) files.
• ddrum4edit can now convert a ddrum4 sound file directly to a single zone 2Box dsnd file (--dsnd).

Cheers! and Enjoy!

Ken

Download Link for User Guide and ddrum4edit executables.

[kjob3022]

Hi,Ken can you tell me if the downloaded ddrum3 files are ready to be put in a ddrum4 module with Iconnect and sysex dump from download folder in finder? I can't really figure out the ddrum4edit book as I'm not very good with terminal stuff,but I did manage to get everything from dropbox to finder and there are alot of ddrum3 samples hopefully ready as is for a ddrum4 module,also I saw a ddrum4 for sale 189.00 on reverb on today if you know someone who needs one ,thanks again Ken....KenOB

[kenjwright]

Jul 12, 2023 22:31:12 GMT -5 kjob3022 said:

Hi,Ken can you tell me if the downloaded ddrum3 files are ready to be put in a ddrum4 module....

Hi Ken. What folder are these dd3 files in? Could you provide a sample file name or maybe a screen shot? If they have a .mid extension then they are likely in the ddrum4 format and should load fine. Just want to be sure I understand which files you are referring to.

Cheers!

Ken

[kjob3022]

Hi Ken, this is in finder downloads. ddrum4edit folder- click to clavia folder- click to ddrum4se folder-click to factory library 3.0 folder- click on that to all the catagories here is one: SNRE_142 HAYMAN 14X8 met.mid 19KB MIDI file. There are also other samples in the Sound library folder like SNRE_203DW 14X5snr dd3.2spl.mid 32kb.KICK_110Bonkickdd3.mid so there are some stuff with dd3 on them and others without.Thanks for the help,KenOB