Go Back   Cockos Incorporated Forums > REAPER Forums > REAPER Q&A, Tips, Tricks and Howto

Reply
 
Thread Tools Display Modes
Old 07-03-2018, 10:34 AM   #201
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
Tod, what do you mean by 'simulate the change'? You mean like, change the decay from below 3/4 open down to before 1/2 open starts, or something else? That might be a good workaround, I think.
Right, that's one way and probably the best to do it, going from 3/4 to 1/2 open isn't that big of
change and could be easily setup.

Then do the same with the rest of the the value slots.

Actually as I think about this, you wouldn't even need any value slots in this case. Just put enough
articulations and simulations together to make it happen for 0 to 127.
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-03-2018, 10:42 AM   #202
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Another thing I think that might be helpful would be to provide a separate volume per openness sample range. I notice that when playing some sampled hi-hats there is a volume jump between ranges, which is offputting. For example, opening up the pedal might sound like I hit the hi-hat pad harder when I didn't, because the more open sample set is louder (or the velocity curve is way off). It seems like tricky stuff to work out, figuring out if it is a velocity or volume issues. But you can't properly investigate it and dial it in unless you have controls for both.
__________________
The media are misleading the public about...pretty much everything.
brainwreck is offline   Reply With Quote
Old 07-03-2018, 10:49 AM   #203
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Quote:
Originally Posted by Tod View Post
Actually as I think about this, you wouldn't even need any value slots in this case. Just put enough
articulations and simulations together to make it happen for 0 to 127.
I think that is a matter of e-hat pedal openness; not about the sound of the sample sets. Also, it kind of depends on how smooth the transistion is between sample sets after applying decays. The smoother the transition can be made to sound, the less need for differentiating ranges. But also, I think these samplers are internally using CC4 ranges as specific MIDI notes for triggering each sample set.

All this talk is making me want a more full-featured drum sampler, where I can decide how things should work for myself. I might just have to take a good look at Kontakt, although I would really like to not chain myself to that ball. I think jsfx is probably capable enough for producing a useful drum sampler, but some of the humps are steep (the gui hump especially).
__________________
The media are misleading the public about...pretty much everything.

Last edited by brainwreck; 07-03-2018 at 10:56 AM.
brainwreck is offline   Reply With Quote
Old 07-03-2018, 10:49 AM   #204
minerman
Human being with feelings
 
minerman's Avatar
 
Join Date: Jun 2009
Posts: 369
Default

Quote:
Originally Posted by Tod View Post
Okay, you've got me to thinking brainwreck, there might be a way to better simulate this.

Lets say we have 6 basic settings. Just because there are only 6 settings doesn't mean we can't use more
articulations and it also don't mean we can't have slightly different variations over the full 128 values
with changes on each value.

For example, lets say we have a 3/4 open for the most open and lets assign a total of 30 CC4 values for it.

3/4 open: 127 to 118
1/2 open: 117 to 108
1/4 open: 107 to 98

While the 3/4 is being played we will simulate the changes down to the 1/2 open, using the full 10 values.
This simulation would move down to the 1/2 open so that the change to the 1/2 open would be as seamless as
possible. Then continue down to the 1/4 open, then on to the next assigned value and do the same thing.

In other words use the full 128 values from open to very tight. That should be more like using a pedal on
a real hat.

Just a thought.

EDIT: Actually this should go from 0 down to 127 instead of 127 down to 0, so take that into account.
This is basically what SD3 does with the incoming CC values, it allows the user to set where each degree/level of open/closed articulations starts/stops...
(I know I already posted this pic, but just for reference):

On the right, you can see the "Tight" articulation is selected (highlighted with a blue dot), & the number (126) is where it stops, so in short, "Tight" CC goes from 127 - 126...

On the left, "Closed" is selected, & stops at 80, so "Closed" is set from 125 - 80...

Each level of openness has the ability to start/stop where the user wants, which I think is great...

I don't have the CC in SD3 actually set like that, I just put this up for reference, but this is pretty much how Tod's Kontakt drums will be when he gets it squared away & will be a really good addition IMO...

On my end, the FSR with Karbo's suggested tweaks, & SD3's CC4 tweak-ability should get me where I wanna be with some trial/error...

Thanks again guys...

EDIT:
Quote:
Originally Posted by brainwreck View Post
Another thing I think that might be helpful would be to provide a separate volume per openness sample range. I notice that when playing some sampled hi-hats there is a volume jump between ranges, which is offputting. For example, opening up the pedal might sound like I hit the hi-hat pad harder when I didn't, because the more open sample set is louder (or the velocity curve is way off). It seems like tricky stuff to work out, figuring out if it is a velocity or volume issues. But you can't properly investigate it and dial it in unless you have controls for both.
I already have this sorta set in SD3, I thought there would be a slight volume difference between full closed & a slightly open hh, the slightly open would be a little louder, or I thought so anyway...This is another little thing the user could adjust to make their e-kit a bit more realistic...
minerman is offline   Reply With Quote
Old 07-03-2018, 11:03 AM   #205
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 22,159
Default

Quote:
Originally Posted by minerman View Post
I think the resistor/foam will do the trick, can't thank you enough man...
The pleasure is mine, I keep doing these tests, eventually, I'll build a desktop drum machine/brain/pads that is actually worthwhile lol.
karbomusic is online now   Reply With Quote
Old 07-03-2018, 11:19 AM   #206
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
Another thing I think that might be helpful would be to provide a separate volume per openness sample range. I notice that when playing some sampled hi-hats there is a volume jump between ranges, which is offputting. For example, opening up the pedal might sound like I hit the hi-hat pad harder when I didn't, because the more open sample set is louder (or the velocity curve is way off). It seems like tricky stuff to work out, figuring out if it is a velocity or volume issues. But you can't properly investigate it and dial it in unless you have controls for both.
Actually normalizing the samples would take care of that, with the added benefit of allowing the velocity
intensity do it's main job. You get the same timbre regardless of whether you normalize or not.

I think the most important aspect would be that when you record and edit the samples, to be very aware of
this. Also during the recording it would be advantageous to setup the pedal with some kind of physical
strap or what ever, so that you can get a good fixed and controlled positioning.
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-03-2018, 11:24 AM   #207
minerman
Human being with feelings
 
minerman's Avatar
 
Join Date: Jun 2009
Posts: 369
Default

Quote:
Originally Posted by karbomusic View Post
The pleasure is mine, I keep doing these tests, eventually, I'll build a desktop drum machine/brain/pads that is actually worthwhile lol.
I've got a Yamah DTX400 module & power supply I'd give you if you'd want it Karbo...

It's not top of the line or anything, but you could build yourself a small kit with it, even a "finger kit" using really small pads/buttons/whatever you'd want...

You wouldn't have to use pads, you could use anything with a piezo to send the trigger info to the module...

Lemme know...
minerman is offline   Reply With Quote
Old 07-03-2018, 11:27 AM   #208
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 22,159
Default

Quote:
Originally Posted by minerman View Post
I've got a Yamah DTX400 module I'd give you if you pay the shipping on it Karbo...

It's not top of the line or anything, but you could build yourself a small kit with it...

Lemme know...
Oh, that's OK but thank you for such a wonderfully generous offer. I don't need one that bad, it's just I keep gathering expertise, it might be fun to build small one from scratch.

Right now I'm tied up learning the Maschine MK3 I bought a few weeks back. It's pads will do me fine for now.
karbomusic is online now   Reply With Quote
Old 07-03-2018, 11:32 AM   #209
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Quote:
Originally Posted by Tod View Post
Actually normalizing the samples would take care of that, with the added benefit of allowing the velocity
intensity do it's main job. You get the same timbre regardless of whether you normalize or not.
But if you normalized the samples you would still need to apply a volume per each sample set at a given velocity level so that every sample set per velocity level doesn't have the same output level. In other words, there absolutely should be volume differences per velocity value (and velocity values don't change output volume). But at the same velocity level across openness ranges, I feel that more open samples can sound too loud, along with the change in timbre. I think that having control over volumes per openness levels would be helpful to me.
__________________
The media are misleading the public about...pretty much everything.
brainwreck is offline   Reply With Quote
Old 07-03-2018, 12:18 PM   #210
minerman
Human being with feelings
 
minerman's Avatar
 
Join Date: Jun 2009
Posts: 369
Default

Quote:
Originally Posted by karbomusic View Post
Oh, that's OK but thank you for such a wonderfully generous offer. I don't need one that bad, it's just I keep gathering expertise, it might be fun to build small one from scratch.

Right now I'm tied up learning the Maschine MK3 I bought a few weeks back. It's pads will do me fine for now.
It's all good Karbo...The offer still stands as long as I have the module, & as long as you're interested...

I actually had a 2 module setup, but simplified it not long ago...The extra module gave me a couple 3-zone options I don't have with the DTX502, plus the ability to choke another cymbal, but I realized since I edit the midi afterward anyway, there's no need for that many pad/cymbal inputs (I had the ability for like 20 pad/cymbals, maybe more)...The DTX502 allows 4 extra pads/cymbals ( 3 tom inputs & an input on the kick pad tower), which is plenty...
minerman is offline   Reply With Quote
Old 07-03-2018, 02:25 PM   #211
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
But if you normalized the samples you would still need to apply a volume per each sample set at a given velocity level so that every sample set per velocity level doesn't have the same output level. In other words, there absolutely should be volume differences per velocity value (and velocity values don't change output volume). But at the same velocity level across openness ranges, I feel that more open samples can sound too loud, along with the change in timbre. I think that having control over volumes per openness levels would be helpful to me.
Actually velocity changes are volume changes, they are linked directly to the volume. It's the velocity intensity
that controls how much the velocity changes the volume. The intensity is measured in percentages (%).

If you have 9 velocity layers that are normalized to 0.0dB, an intensity of 0% will not affect the the volume and all
the samples will have the same volume. However if the intensity is set at 75%, then the velocity will greatly affect
the volume of each sample and the volume will be very even from loudest to softest, no sudden or noticeable jumps or
changes in volume.

On the other hand, if the samples are not normalized, then you will get the noticeable changes in volume, no matter
where the intensity is set.

I showed this before but here it is again. Notice the orange 70% line and the blue 80% line. At a velocity level of 8,
the 70% line is at approximately -28dB and the 80% is close to -36dB. I actually find 75% works pretty well for
normalized drum samples.

Now if I have 32 or more layers, I do prefer to not normalize, if for no other reason than to satisfy those folks
who think that normalizing will somehow do something mysterious to the samples. And with 32 layers you really have
no need to normalize.



So getting back to these hi-hat samples, I see no reason for not normalizing, even if you do have 32 or more layers. The
reason being the pure logistics of using the combination of samples and simulated samples. Also the velocity intensity
will be much more effective as I mentioned above.

EDIT: The graph I show are the velocities with a linear curve. With a little imagination you might be able to see how
different curves might affect that graph.
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-03-2018, 06:31 PM   #212
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Is using normalized samples the norm with drum samplers?
__________________
The media are misleading the public about...pretty much everything.
brainwreck is offline   Reply With Quote
Old 07-03-2018, 08:19 PM   #213
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
Is using normalized samples the norm with drum samplers?
Hi brainwreck, that all depends, but I'd have to say no. However, there are a lot of
ifs and buts in that question.

First I have to say I'm mainly speaking from my own experience, but I do have some
experience with this and I think many of those who have done it as much as I, will agree.

You've already asked one of the main questions, why do the samples have different volumes
from sample to sample? This is a big problem and can only be fixed by recording lots of
samples at different levels or normalizing them.

So let's just record lot's of samples, right? Well that can make sense for Snares, Hats
(foot, closed, & tight), Kicks, and toms, but for cymbals not so much. Well except for
Ride, it requires all the velo layers you can get.

Some developers go ahead with 16 or less layers on cymbals, or even all the other kit
parts. If they don't normalize them the transition between the samples will be quite
noticeable. Still some of these drum libraries are quite popular, especially if they
are free.

And this leads me to one my other observations, many of the younger folk, or maybe not
so younger folk, are simply setting all the velocities of the midi notes to 127. The
only help for them is provide lots of RRs.

Another important factor is that the average person playing a keyboard or pads, will
play an average velocity range of of 65 to maybe 118, and I do know that's true of
myself. I'm just mentioning this to put it into equation.

So you can kind of squeeze the main velocity layers into the 65 to 118 velocity region
and get better dynamics? I think this is true but so far I haven't totally explored it.

Something else that many might not know or understand. Getting 32 or more different
samples out of a drum kit with a meaningful disparity is not so easy.
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-04-2018, 07:11 AM   #214
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Yea, I guess having all the samples normalized allows for better control of the transfer function between velocity and audio level, where with samples left at their dynamic levels you don't have as much flexibility in shaping that.

Quote:
Originally Posted by Tod View Post
And this leads me to one my other observations, many of the younger folk, or maybe not
so younger folk, are simply setting all the velocities of the midi notes to 127. The
only help for them is provide lots of RRs.
It seems to be a popular trend, as part of the oversimplification of music today, where for example, drums tend to sound looped super-squashed. And I think that sampled drums probably plays a big role in all that.
__________________
The media are misleading the public about...pretty much everything.
brainwreck is offline   Reply With Quote
Old 07-04-2018, 07:15 AM   #215
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Tod, what I meant to say above on velocity not being volume is that velocity is just a control mechanism, and how that mechanism is implemented is independent of volume. In other words, velocity can be used to control volume, or it can be used to control anything else, independent of raw sample volumes.
__________________
The media are misleading the public about...pretty much everything.
brainwreck is offline   Reply With Quote
Old 07-04-2018, 07:33 AM   #216
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
Tod, what I meant to say above on velocity not being volume is that velocity is just a control mechanism, and how that mechanism is implemented is independent of volume. In other words, velocity can be used to control volume, or it can be used to control anything else, independent of raw sample volumes.
Right brainwreck, in Kontakt, and I think most VSTis, it a modulator, but I can script it to do other things as well.
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-06-2018, 08:05 AM   #217
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Hey Tod, if/when you get to 'simming' the inbetween velocity states of a sampled hi-hat, I would be curious to hear how it sounds.
__________________
The media are misleading the public about...pretty much everything.
brainwreck is offline   Reply With Quote
Old 07-06-2018, 10:12 AM   #218
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
Hey Tod, if/when you get to 'simming' the inbetween velocity states of a sampled hi-hat, I would be curious to hear how it sounds.
Yeah, I'd definitely post it back here. The problem is that I don't have all the samples to
try it yet. If I had the samples, even if they are just one velocity per articulation, I could
give it a try. Here are the articulations I think I need.

Open 3/4
Open 1/2
Open 1/4
Open 1/8 or sizzle
Closed
Tight
Very tight

In the real world I think I would use more articulations, but I think this would be enough to see
what we can do.
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-16-2018, 05:42 AM   #219
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

I was thinking about the sampled hi-hat issue a little last night. For say, 5 channels (hi-hat mic, stereo overhead mics, stereo room mics) at 16-bit 44.1k and a length of (just for example) 5 seconds for each sample (could be shortened up, especially on more closed hits); 16 openness levels, 16 velocity ranges per openness level, and 4 variations per velocity range, we're looking at over 2 GB's worth of samples for the hi-hat sample files alone. Would that be a performance problem? It sure would signicantly increase loading times, but would switching among that much data in realtime (along with the samples for the other drums) cause performance problems? I know nothing about these things. I wouldn't think it would be a problem because I have had multiple sampled kits (at multiple GB's) loaded up ready to play for comparison, but I was only playing one kit at a time. I guess routing MIDI to simultaneous kits might give a better idea of whether it would be an issue.

Also, recording, editing, and setting up 1024 samples just for the hi-hat would be a long and tedious chore. I'm sure some scripting could be worked out to help with that, though.

I think that something like 16 openness levels is needed to help with greatly reducing the blatant openness level switching. I'm not so sure on the number of velocity levels per openness level. Chances are, that could be reduced. And I think 4 variations per velocity level would be pretty good. Overall, the amount of data might be widdled down to something like 1/2 of what I calculated above for worst case.

There is quite a bit to think about for sampling a hi-hat at an acceptable level of detail.

For myself, where my priority is on just playing without the annoying blatant switching (recording not being a concern), I could probably just mono mic a hi-hat for sampling and use the end product for that purpose, later taking things further if desired. Better yet, maybe someone already offers up highly sampled hi-hats that would meet the requirements. Anyone know of something out there?
__________________
The media are misleading the public about...pretty much everything.

Last edited by brainwreck; 07-16-2018 at 06:08 AM.
brainwreck is offline   Reply With Quote
Old 07-16-2018, 10:22 AM   #220
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
I was thinking about the sampled hi-hat issue a little last night. For say, 5 channels (hi-hat mic, stereo overhead mics, stereo room mics) at 16-bit 44.1k and a length of (just for example) 5 seconds for each sample (could be shortened up, especially on more closed hits); 16 openness levels, 16 velocity ranges per openness level, and 4 variations per velocity range, we're looking at over 2 GB's worth of samples for the hi-hat sample files alone. Would that be a performance problem? It sure would signicantly increase loading times, but would switching among that much data in realtime (along with the samples for the other drums) cause performance problems? I know nothing about these things. I wouldn't think it would be a problem because I have had multiple sampled kits (at multiple GB's) loaded up ready to play for comparison, but I was only playing one kit at a time. I guess routing MIDI to simultaneous kits might give a better idea of whether it would be an issue.

Also, recording, editing, and setting up 1024 samples just for the hi-hat would be a long and tedious chore. I'm sure some scripting could be worked out to help with that, though.

I think that something like 16 openness levels is needed to help with greatly reducing the blatant openness level switching. I'm not so sure on the number of velocity levels per openness level. Chances are, that could be reduced. And I think 4 variations per velocity level would be pretty good. Overall, the amount of data might be widdled down to something like 1/2 of what I calculated above for worst case.

There is quite a bit to think about for sampling a hi-hat at an acceptable level of detail.

For myself, where my priority is on just playing without the annoying blatant switching (recording not being a concern), I could probably just mono mic a hi-hat for sampling and use the end product for that purpose, later taking things further if desired. Better yet, maybe someone already offers up highly sampled hi-hats that would meet the requirements. Anyone know of something out there?
Hi brainwreck, it's hard to know for sure how many samples would be needed, but I'm thinking that some simulation might work better to try to cover the openness
range then trying to do it with the samples themselves. It might also be easier to control. I think 6 to 8 pedal levels going from open to sizzle, not including sizzle, might
work. If you were going to have RRs, I think 3 would suffice, odd numbers of RRs seem to work better then even numbers of RRs. Then maybe 2 to 3 from sizzle to
slightly closed. I think the number of layers should increase as the hat closes so that closed to tightly closed will have 32 or more layers. Also the number of RRs should
probably increase too, the the closed to tightly closed should have at least 5RRs or even 7RRs.

The actual number of layers that gives maximum performance can be difficult to determine, but I think 32 works pretty well. The most important factor is what level
and how well the samples are played going from -32dB to -1dB, most folks don't understand the difficulty in getting a good 32 usable levels. We learned a lot with
SMDrums 1 so we have a lot better idea how to approach SMDrums 2.

Regarding the differences you can hear as the intensity of the dynamics goes from soft to loud, the main thing you hear is the difference in volume with a little added
difference in timbre. This is why I think normalizing is so important for the hi-hat, normalizing totally smooths out the volume differences and gives the velocity
intensity the ability to function at it's best. You can still hear slight changes in the timbre, but it's not as noticeable.

Regarding the CPU load as the number of samples increases, I think the main factor is the RAM, not so much the CPU, it will probably be the same, regardless of the
number of samples. Normally you would have the same number of FX on a smaller kit as you would a larger kit, which is a factor in CPU.

You're right brainwreck, there's a lot to think about when sampling a hi-hat.
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-16-2018, 10:53 AM   #221
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Tod the thing that I am noticing most between openness levels is the timbre change, not volume change. But I think it causes a response from me to change hit dynamics. Say I'm playing with a medium hit strength, completely closed, and I begn to open the pedal. The timbre doesn't gradually change. It suddenly changes. That may sound subtle to some people, but it doesn't to me. So then with the sudden timbre change, I begin reactively velocity level hunting with my hit strength to adjust for what my ears think they should be hearing (in vain). And the same goes when going the other way around when playing open samples toward closed samples.

Regardless of any previous statements I have made, I'm realizing that part of all of this is unraveling what is going on technically vs. what I am perceiving, which makes jsfx so handy. It allows for me to isolate things and look closer at what is happening according to how granular I want to look at things. For example, just being able to stay within an openness range and within a specific velocity range and being able to adjust by a single increment of either is very helpful in figuring out what a sampler is actually doing.
__________________
The media are misleading the public about...pretty much everything.

Last edited by brainwreck; 07-16-2018 at 10:59 AM.
brainwreck is offline   Reply With Quote
Old 07-16-2018, 03:07 PM   #222
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Just outside of Glacier National Park
Posts: 11,826
Default

Quote:
Originally Posted by brainwreck View Post
Tod the thing that I am noticing most between openness levels is the timbre change, not volume change. But I think it causes a response from me to change hit dynamics. Say I'm playing with a medium hit strength, completely closed, and I begn to open the pedal. The timbre doesn't gradually change. It suddenly changes. That may sound subtle to some people, but it doesn't to me. So then with the sudden timbre change, I begin reactively velocity level hunting with my hit strength to adjust for what my ears think they should be hearing (in vain). And the same goes when going the other way around when playing open samples toward closed samples.
Okay, that could be a combination of any number of things. You'd have to know exactly how they were recorded, then know the way in
which they were edited and programmed. The first thing is to record them to capture the dynamics as good as possible, and there are a lot
of considerations when you do that. Then the way you edit them and program them will depend a great deal on how they were recorded.

Quote:
Regardless of any previous statements I have made, I'm realizing that part of all of this is unraveling what is going on technically vs. what I am perceiving, which makes jsfx so handy. It allows for me to isolate things and look closer at what is happening according to how granular I want to look at things. For example, just being able to stay within an openness range and within a specific velocity range and being able to adjust by a single increment of either is very helpful in figuring out what a sampler is actually doing.
What exactly is this jsfx analyzing?
__________________
Kontakt Vid Tutorials->Create Outputs / Create Templates -|- SMDrums Free drums -|- Elk Video Productions -|- Tod's Music
Tod is offline   Reply With Quote
Old 07-17-2018, 05:30 AM   #223
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Tod, hearing the switching among sample sets is dead easy. You don't need to know anything about how the samples were recorded or any other information to hear it. You just input some fixed MIDI values to a sampler and listen, adjust, listen again, repeat.

I'm talking about being able to knock up jsfx plugins for specific purposes, such as the previously posted cc4 slider for looking at how any given sampler responds to control change value ranges. And the same goes for looking at how a sampler responds to velocity. Jsfx is very handy to have available.
__________________
The media are misleading the public about...pretty much everything.
brainwreck is offline   Reply With Quote
Old 09-03-2018, 09:28 AM   #224
brainwreck
Human being with feelings
 
brainwreck's Avatar
 
Join Date: Jul 2006
Location: The 90's
Posts: 19,403
Default

Some thoughts up to this point on playing electronic drums...

I have completely abandoned using the internal module samples. The latency is significantly lower than when using a drum sampler, but the module's builtin samples are terribly uninspiring to play. If someone asked me about drum modules, I would say to not even consider the internal sounds as a factor. The only real considerations for a drum module to me are good triggering, full MIDI resolution (including hi-hat pedal control), and plenty of inputs. And if someone asked me about the TD-11 module specifically, I would say to skip it, since the module does not output full MIDI resolution and the inputs are pretty limited.

Sensitivity of triggering isn't great with electronic drums. It is adequate. Playing ghost notes and buzz rolls leaves something to be desired, where varied hits at very low playing force come out pretty much the same as one another. This is down to the module. Relative to that, mesh heads are more sensitive than rubber pads, but when playing softly, the bounciness of mesh heads is much more apparent to me, along with the weakness of triggering sensitivity of the module. And yes, sensitivity can be adjusted, but it is always a tradeoff with crosstalk from other pads.

Hi-hat sensitivity is less adequate. Lowest strength hits aren't detected at times, so if you play ghost notes inbetween normal strength hits, you might be met with a blank space here and there.

On sensitivity on the whole, I think that use of a rack is a bad idea, because it tends to require that sensitivity be traded off for avoiding crosstalk. Where if every kit piece were mounted on it's on stand, this wouldn't be an issue. So if someone asked me, I would say to diy the 'kit' from a real drum kit or similar, where no kit piece is mounted to a common rack.

And on modules, given the expense of one of the better Roland modules for triggering only, I would say that something like Megadrum (trigger module only) looks much more attractive with many more inputs and features for around the same price as an older used Roland module.

On latency. I really would avoid USB audio interfaces for the purpose of playing through a drum sampler, simply because latency tends to be higher than for any other connection type. PCI(e) would likely be best, whether it be a lower cost M-Audio or higher cost RME card. Using a USB interface is doable, but the latency is defintely felt as a sluggishness.

Just some thoughts so far, in case someone else comes along thinking about e-drums. And I think the recommendations above are where I will head toward, eventually, i.e., diy kit, Megadrum, and PCI(e).

Even given the above considerations, I think that an electronic kit for triggering is a much better MIDI interface than pads/keys.
__________________
The media are misleading the public about...pretty much everything.

Last edited by brainwreck; 09-03-2018 at 09:33 AM.
brainwreck is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 05:29 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2018, vBulletin Solutions Inc.