Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
Posted: Thu Dec 03, 2009 4:37 pm Post subject:
I like this demo:
Note that he relies on other drum controllers for MIDI output.
Besides, somebody here will find a way to hack one to transmit OSC or talk to Chuck or some such thing. There's too much talent here for that to not happen. _________________ Current and recent work on Soundcloud
Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
Posted: Thu Dec 03, 2009 4:40 pm Post subject:
And now for a completely different "demo" - I'm a fan of Jet Daisuke and his quirky gear demos. Cute way to combine two interests - music and scantily clad girls - into a single project:
Note that he relies on other drum controllers for MIDI output.
Besides, somebody here will find a way to hack one to transmit OSC or talk to Chuck or some such thing. There's too much talent here for that to not happen.
Hmmm... I'm afraid that with these new fangled thingies, theres a small PCB with a few proprietary chips inside, and no way to extract any of the interesting data/signals. The most interesting part being the decoding and processing of the sensor data which is most probably firmly locked up inside those chips.
What is it with this MIDI denial trend in controllers, anyway...
The drum sounds nice and looks as if it's fun to play though.
Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
Posted: Sun Dec 06, 2009 11:12 am Post subject:
DrJustice wrote:
GovernorSilver wrote:
Note that he relies on other drum controllers for MIDI output.
Besides, somebody here will find a way to hack one to transmit OSC or talk to Chuck or some such thing. There's too much talent here for that to not happen.
Hmmm... I'm afraid that with these new fangled thingies, theres a small PCB with a few proprietary chips inside, and no way to extract any of the interesting data/signals. The most interesting part being the decoding and processing of the sensor data which is most probably firmly locked up inside those chips.
What is it with this MIDI denial trend in controllers, anyway...
The drum sounds nice and looks as if it's fun to play though.
DJ
--
Ah, so maybe I was a tad optimistic about the hackability of this newest version of the Wavedrum.
What other controllers do you want to have MIDI but do not have it?
There are three acoustic transducers on the Wavedrum, One just under the metal logo, below the controls, and two underneath the rim, at about 2 o'clock and 10 o'clock. There is also a pressure sensor beneath the rim.
With that in mind, let's first consider the Wavedrum's DSP synthesis. When you strike the Wavedrum, it's not delivering a note-on message, or using polyphony as in other trigger-based products. Instead, the DSP synthesis algorithms in the Wavedrum are always running, and just waiting to be excited. The acoustic signal from the transducers is fed into the DSP algorithm, and treated as the foundation for whatever synthesis model is currently being used. That's why the acoustic models react so realistically... The actual sound of your performance is what's being manipulated, not just the fact that you hit the drum at a certain spot, at a certain velocity.
On the other hand, consider the PCM content that coexists with the DSP algorithms. When a strike is "heard" by the transducers, it can then trigger the PCM instruments. In other words, a volume threshold triggers the instrument. This is data that COULD be translated to MIDI... as it was on the original Wavedrum. The difference here is that the original one didn't have any PCM content, and instead fed a note message to the MIDI OUT port.
As previously mentioned, the original Wavedrum's MIDI implementation did not allow MIDI playback. The MIDI IN port was only used for program changes, modulation, and system exclusive functions... No note messages of any kind. This is simply because as described above, the Wavedrum's DSP synthesis relies on your acoustic performance to generate its sound, and that can't be delivered by a MIDI message. Again, MIDI IN could theoretically trigger the PCM content... but then so much of the Wavedrum's expressive potential would be ignored.
Lastly, the price difference can't possibly be ignored. The new Wavedrum costs less than 1/4 of the price of the original. Korg is really thrilled to offer some of its most sought-after, best-sounding technology to a MUCH wider range of musicians. We don't want to compromise that by adding a costly feature that can only do a portion of what our users want it to do.
Yeah, with the various products there are various justifications. Still, it's quite possible to make useful MIDI implementations, for the wavedrum and other products. But hey, it's their design decisions. I'll vote with my wallet
Joined: Jan 18, 2005 Posts: 4145 Location: Sweden
Audio files: 371
G2 patch files: 100
Posted: Sun Dec 06, 2009 2:45 pm Post subject:
Actually, the motivation for the wavedrum is kind of ok - having the sensor interacting directly with the DSPs. The MIDI stuff would only get in the way. What I would like for this particular device would be some way to write my own DSP code, or at least being able to download more models that are created in the future.
What I would like for this particular device would be some way to write my own DSP code, or at least being able to download more models that are created in the future.
If the wavedrum transmitted its sensor data by MIDI CC or NRPN streams, which is feasible and IMO not unreasonable, then we could write code for it on our Macs/PCs, as well as using it with sequencers
Joined: Jan 18, 2005 Posts: 4145 Location: Sweden
Audio files: 371
G2 patch files: 100
Posted: Sun Dec 06, 2009 11:54 pm Post subject:
DrJustice wrote:
Antimon wrote:
What I would like for this particular device would be some way to write my own DSP code, or at least being able to download more models that are created in the future.
If the wavedrum transmitted its sensor data by MIDI CC or NRPN streams, which is feasible and IMO not unreasonable, then we could write code for it on our Macs/PCs, as well as using it with sequencers
DJ
--
Or maybe just offer audio outs direcly from the sensors, enabling us to hook them up to whatever exciter we'd want. That sounds like hackable thing...
Yeah, with the various products there are various justifications. Still, it's quite possible to make useful MIDI implementations, for the wavedrum and other products. But hey, it's their design decisions. I'll vote with my wallet
DJ
--
Will you vote "yes" with your wallet if Korg released a MIDI Wavedrum, but for $2500? That was the price of the original Wavedrum with MIDI Out, and MIDI In (but only for certain MIDI messages as Rich described), according to this Keyboard Museum entry:
Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
Posted: Mon Dec 07, 2009 11:54 am Post subject:
Antimon wrote:
Or maybe just offer audio outs direcly from the sensors, enabling us to hook them up to whatever exciter we'd want. That sounds like hackable thing...
/Stefan
I like your thinking.
It addresses the problem of how to translate the use of brushes on the drum head into control messages for MIDI or other protocol. This is a problem that has never been solved on drum controllers. Somebody claimed you can use brushes on Roland V-Drums but that proved to be false.
Similarly, try holding a drumstick against the drum "skin" and bowing it - you will probably get nothing out of the V-Drum or any other drum pad. You'll probably get something from a Wavedrum. _________________ Current and recent work on Soundcloud
Will you vote "yes" with your wallet if Korg released a MIDI Wavedrum, but for $2500?
No, an $1875 increase for a MIDI port is ridiculous. The claim that was made in the "justification" that MIDI would impact the price greatly is a bit rich. It certainly doesn't mean a jump from $625 to $2500. Id's say a jump of around $25 wold be enough to cover for it. The hardware would cost a couple of dollars and the added development would be minimal. They could have used a USB interface too, or ethernet even, although development cost would be slightly more than for MIDI.
If we were to tap the sensors directly, we'd most probably need a bit of electronics and maybe DSP to condition the signals, and it would not give sequencing (playback) capabilities. By ending up only using the shell and sensors to achieve limited functionality, the point of getting a Wavedrum starts to become lost perhaps?
Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
Posted: Mon Dec 07, 2009 1:08 pm Post subject:
DrJustice wrote:
GovernorSilver wrote:
Will you vote "yes" with your wallet if Korg released a MIDI Wavedrum, but for $2500?
No, an $1875 increase for a MIDI port is ridiculous. The claim that was made in the "justification" that MIDI would impact the price greatly is a bit rich. It certainly doesn't mean a jump from $625 to $2500. Id's say a jump of around $25 wold be enough to cover for it. The hardware would cost a couple of dollars and the added development would be minimal. They could have used a USB interface too, or ethernet even, although development cost would be slightly more than for MIDI.
DJ
--
My guess is Korg will introduce a Wavedrum with MIDI within 3 years, for $150 more than the WD-X. And of course it will be mostly MIDI out. I have no idea how MIDI In (for sequencing) would be achievable on this instrument. If I wanted to a sequence with jazz brushes, I'd have to add a robot arm with brush, right? The basic sound of the instrument is from a real drumhead and pickups, not a digital synth engine residing on a chip.
My guess is Korg will introduce a Wavedrum with MIDI within 3 years, for $150 more than the WD-X. And of course it will be mostly MIDI out.
There you go
Quote:
I have no idea how MIDI In (for sequencing) would be achievable on this instrument.
The sensors generate a stream of data when you wack the drum. This data is fed to the DSP based drum modelling algorithms. By recording that stream (using the hypothetical MIDI port) and then playing it back, so that the Wavedrum can feed it to it's drum modelling algorithm again, you should get a fully sequencable Wavedrum.
Quote:
The basic sound of the instrument is from a real drumhead and pickups, not a digital synth engine residing on a chip.
As I understand it, the sensors signals drives DSP based drum synthesis algorithms. Korg says "The acoustic signal from the transducers is fed into the DSP algorithm, and treated as the foundation for whatever synthesis model is currently being used...". Thus we do have a digital synth engine on a chip making the actual sounds.
Joined: Apr 26, 2004 Posts: 1349 Location: Washington DC Metro
G2 patch files: 1
Posted: Mon Dec 07, 2009 2:07 pm Post subject:
DrJustice wrote:
Quote:
I have no idea how MIDI In (for sequencing) would be achievable on this instrument.
The sensors generate a stream of data when you wack the drum. This data is fed to the DSP based drum modelling algorithms. By recording that stream (using the hypothetical MIDI port) and then playing it back, so that the Wavedrum can feed it to it's drum modelling algorithm again, you should get a fully sequencable Wavedrum.
I have doubts about this approach. A lot of data would have to be thrown out to translate the audio volume into the 127 possible values of MIDI Velocity. If you have a good formula for scaling amplitude to MIDI Velocity values, you might get somewhat listenable results upon playback. 0-127 is supposed to represent the entire dynamic range from no sound at all to the roar of a airplane, but there are many, many fine dB values between, say MIDI Velocity=60 and 61 that the average listener can easily pick out.
I suspect though that for most end users, recording Wavedrum audio into Ableton Live clips might still sound much better than relying on MIDI playback via the MIDI In port from a sequencer.
DrJustice wrote:
Quote:
The basic sound of the instrument is from a real drumhead and pickups, not a digital synth engine residing on a chip.
As I understand it, the sensors signals drives DSP based drum synthesis algorithms. Korg says "The acoustic signal from the transducers is fed into the DSP algorithm, and treated as the foundation for whatever synthesis model is currently being used...". Thus we do have a digital synth engine on a chip making the actual sounds.
DJ
--
There are two sound engines on the WDX. One is the PCM sample player which you describe above. The other is the acoustic sound of the drum head, which is used as input into the DSP processor - this is basically effects processing. The PCM sample player could probably be triggered effectively by MIDI sequence playback. But if I put a stick on the drumhead and bowed it, this would not be captured effectively by the sequencer.
Drumming with brushes vs. sticks would be indistinguishable to the PCM engine, I suspect. Your solution does not address the brush problem.
You have a Roland VG-99 or similar product right? Try recording a guitar performance on the VG-99 with the VG-99's audio and MIDI sound source (triggered by the guitar over MIDI), on separate tracks. Then play each track to your friends and have them try to guess which was generated by MIDI and which was generated by real guitar audio processed through a sophisticated DSP engine. _________________ Current and recent work on Soundcloud
Joined: Jan 18, 2005 Posts: 4145 Location: Sweden
Audio files: 371
G2 patch files: 100
Posted: Mon Dec 07, 2009 2:21 pm Post subject:
What I've pictured in my head is sensors placed on various parts of the drum head and rim that, when hit, shake around on the material they are attached to, and describe that shaking as data fed into the DSP. I'm mostly thinking about how I like to use the string oscillator on my NMG2, or how you use any Karplus-Strong algorithm I guess - it isn't something you activate by triggering a gate, the sound is shaped by an impulse fed into the system. (I know this probably isn't Karplus-Strong, but some other excitor-input algorithm that I don't know what it's called).
This impulse would be sampled at audio-rate, and then sent over MIDI somehow (44100 messages per second, double the amount if you want 16-bit resolution). The DSP on the wavedrum is constantly listening to the impulse, you could gate the impulse so you don't send data when the impulse input is below a threshold, but still I don't see a usable solution sending this stuff over MIDI.
Sure, you could do a simpler kind of sensor that just says "I'm hit", with the old velocity sensing that MIDI does so well, but don't we already have a ton of these kinds of devices?
I'm reading my own dreams into how this thing should work, it may be possible that the sensors don't work the way I think. I think it's a cool thing that a device tries to go beyond note-on and single axis velocity sensing. That's a big problem with midi: lack of data attached to note on messages.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum
Forum index » News... » Gear News
FAQ
Calendar
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
Chat Room
[/disappointment]
the WaveCrapDrum 
