Why do drum modules use "modeling" instead of "real" samples
-
stephenpaulharper stephenpaulharper https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=73921
- KVRian
- 778 posts since 4 Jul, 2005 from Atlanta
Wasn't sure where to post this, but figured I'd get the best info here. I have no agenda, other than being an Edrummer looking for an explanation from someone with more technical expertise than myself.
What exactly is the technical/physical drawback to packaging a real acoustic drum sample library - BFD, DFHS, Addictive Drums, etc. in a unit roughly the size, shape and weight of say - a Roland TD-20 module, which is filled with modeled sounds?
I use a combo of BFD and Addictive Drums with an Ekit via PC for recording, and gig with an acoustic kit, so this isn't really an issue for me, but I'm very puzzled as to why the Edrum market doesn't pursue modules loaded with real sounds. More and more drummers are switching from their modules to drum software, which basically turns your module into a very expensive TMI. I know there are things like BFD Turbo Receptor, but those things are heavier, more cumbersome, and harder to navigate than your average drum module, IMHO.
Any explanations or theories would be greatly appreciated. BTW, many of the Edrum forums are sponsored by companies with an agenda, so it's a little hard to get a truly objective opinion sometimes.
Thanks.
A
What exactly is the technical/physical drawback to packaging a real acoustic drum sample library - BFD, DFHS, Addictive Drums, etc. in a unit roughly the size, shape and weight of say - a Roland TD-20 module, which is filled with modeled sounds?
I use a combo of BFD and Addictive Drums with an Ekit via PC for recording, and gig with an acoustic kit, so this isn't really an issue for me, but I'm very puzzled as to why the Edrum market doesn't pursue modules loaded with real sounds. More and more drummers are switching from their modules to drum software, which basically turns your module into a very expensive TMI. I know there are things like BFD Turbo Receptor, but those things are heavier, more cumbersome, and harder to navigate than your average drum module, IMHO.
Any explanations or theories would be greatly appreciated. BTW, many of the Edrum forums are sponsored by companies with an agenda, so it's a little hard to get a truly objective opinion sometimes.
Thanks.
A
-
- KVRAF
- 1743 posts since 3 Dec, 2004
The alesis SR-16 is the only one I can think of off the top of my head...but then again I used to own one 
http://www.alesis.com/product.php?id=60
http://www.alesis.com/product.php?id=60
my sig will go here
-
Stupid American Pig Stupid American Pig https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=4753
- KVRAF
- 7065 posts since 25 Nov, 2002 from not sure
Its expensive to have several gigs worth of drum sample ROM in every module mainly.
-
- KVRAF
- 4735 posts since 18 Jul, 2002 from London, UK
Yep, you pretty much need a fairly high end embedded PC, plus trigger hardware, plus a large volume of HD or solid state storage. The big hardware companies with the chops to pull this off seem to be generally a bit distrustful of embedded PC technology (partly for price reasons), aside from Korg who don't do drums.
The other issue here is that Roland has a fairly wide reaching patent on the mesh head technology used in their edrums, so you'd either have to sell the brain as an after-market add-on, or use some other technology which may have disadvantages. Basically you've got a choice between rubber pads, which are cheap and don't feel so great, and wouldn't be balanced costwise when paired with an expensive PC-brain, or something like RET's "VPT" e-drums, which play well but are more expensive to make than decent mesh pads, and less transportable (though still better than the real thing).
OTOH, a full RET system with trigger I/O, hard drive and BFD software comes to a shade under $4,000, which even by the time you add a decent laptop and some audio I/O, is less than the $6.5k a TD-20 system goes for, and comparable to the $4.5k for a TD-12.
The other issue here is that Roland has a fairly wide reaching patent on the mesh head technology used in their edrums, so you'd either have to sell the brain as an after-market add-on, or use some other technology which may have disadvantages. Basically you've got a choice between rubber pads, which are cheap and don't feel so great, and wouldn't be balanced costwise when paired with an expensive PC-brain, or something like RET's "VPT" e-drums, which play well but are more expensive to make than decent mesh pads, and less transportable (though still better than the real thing).
OTOH, a full RET system with trigger I/O, hard drive and BFD software comes to a shade under $4,000, which even by the time you add a decent laptop and some audio I/O, is less than the $6.5k a TD-20 system goes for, and comparable to the $4.5k for a TD-12.
This account is dormant, I am no longer employed by FXpansion / ROLI.
Find me on LinkedIn or elsewhere if you need to get in touch.
Find me on LinkedIn or elsewhere if you need to get in touch.
-
- KVRAF
- 2139 posts since 15 Jul, 2003 from ex-NJ, PA
-
- KVRist
- 357 posts since 16 Jun, 2006 from Stockholm - Sweden
The Roland modules call their sounds "modelled" but I'm pretty sure it's just samples. But not nearly enough of them. They certainly don't sound too good considering their price.. (The "modeling" seems to be pitch and EQ mostly)
www.xlnaudio.com - Creators of "Addictive Drums"
Demo version available for download!
Check out the new Jazz Paks too
Demo version available for download!
Check out the new Jazz Paks too
- KVRAF
- 19156 posts since 13 Feb, 2003 from Vancouver, Canada
AFAIK, all those drum modules use samples too. They don't sound as good as the VST modules becuase they don't use as many samples/velocity layers/sample-swapping, etc.
-
stephenpaulharper stephenpaulharper https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=73921
- KVRian
- Topic Starter
- 778 posts since 4 Jul, 2005 from Atlanta
Is this Lars? Anway, speaking of which, for some strange reason I've been unable to register my XLN drums - I registered at the site a few mos back to try the demo, but can't get in to the user area anymore - any advice? I've had the thing for almost a month and am still using the demo.XLN Audio wrote:The Roland modules call their sounds "modelled" but I'm pretty sure it's just samples. But not nearly enough of them. They certainly don't sound too good considering their price.. (The "modeling" seems to be pitch and EQ mostly)
And no, I think those samples are definitely physical models - one listen to any of the toms or kicks would tell you that. They even have their own proprietary (?) process of modeling, called COSM. They explain the process in detail on their web site - under the link "why our drums sounds are crap."
No offense Angus, but shouldn't you be hard at work on my copy of BFD2 instead of answering my stupid questions? BTW, RET may be next on Roland's list of people to go after. They've already shut down X-drums, and G-drums are on shaky ground. Plus those real heads defeat the "silent" drumming option which is important to many Edrummers. Also that Alesis I/O seems to be lacking in many areas according to several testimonials I've read.Angus_FX wrote:OTOH, a full RET system with trigger I/O, hard drive and BFD software comes to a shade under $4,000, which even by the time you add a decent laptop and some audio I/O, is less than the $6.5k a TD-20 system goes for, and comparable to the $4.5k for a TD-12.
In any case www.vdrums.com would be a good place for both of you guys to check out if you haven't already. Both of your products are discussed in much detail from an Edrummers perspective, and it seems the most agenda-less of all the Edrum forums.
As for the mixosaur thing, they've managed to make it unusable for Edrummers - gotta control the hh with the mod wheel (?). Plus that's $1000 software AND requires Kontakt 2. My suspicion is that it's appropriately named, as it will indeed go the way of the dinosaur.
Anyway, thanks for all the answers.
A
-
- KVRist
- 357 posts since 16 Jun, 2006 from Stockholm - Sweden
Sure is! Email support[at]xlnaudio.com with your full name + email, and I'll have a look!algodon wrote: Is this Lars? Anway, speaking of which, for some strange reason I've been unable to register my XLN drums - I registered at the site a few mos back to try the demo, but can't get in to the user area anymore - any advice? I've had the thing for almost a month and am still using the demo.
www.xlnaudio.com - Creators of "Addictive Drums"
Demo version available for download!
Check out the new Jazz Paks too
Demo version available for download!
Check out the new Jazz Paks too
-
- KVRAF
- 4229 posts since 9 Apr, 2003 from Right here, in front of my computer...
-
stephenpaulharper stephenpaulharper https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=73921
- KVRian
- Topic Starter
- 778 posts since 4 Jul, 2005 from Atlanta
That's interesting beej, and I don't doubt you, but on one of their many treatises on COSM, they talk about using COSM for amp modeling in the manner you describe. But I'm almost positive I read in their literature that no actual audio samples were used for the drums - I could be mistaken. Whatever the process, they manage to make "realistic kits" that sound like cheap synth patches. I've got nothing against Roland - use plenty of their gear and I love my drums themselves, but their drum samples really are a joke, IMHO.
In any case, my respect for Spectrasonics just went down the tubes.
Thanks again for the scoop.
A
In any case, my respect for Spectrasonics just went down the tubes.
Thanks again for the scoop.
A
-
- KVRAF
- 4229 posts since 9 Apr, 2003 from Right here, in front of my computer...
Yep, one of Roland's COSM processes is amp modelling.algodon wrote:That's interesting beej, and I don't doubt you, but on one of their many treatises on COSM, they talk about using COSM for amp modeling in the manner you describe.
I've never read that in V-Drum literature. Roland never mentioned (afaik) that the drums are physically modelled, it's just that the "V-Drum" and "modelling" and "Coo, you can change the shell type" etc made magazine reviewers assume the drums were not samples and were models, and this became the perception, even though it was wrong. And they downplay saying the drums are samples as well.algodon wrote:But I'm almost positive I read in their literature that no actual audio samples were used for the drums - I could be mistaken.
I agree. The drums are decent for a drum machine, but will be dissapointing if expecting a "real kit" experience. Things get 1000% better when you rig up a kit to properly play some of the better drum libraries out there, with room ambience and so on. The Roland sounds are the weakest part of their V-kits imho. I sold my V-Custom kit.algodon wrote:Whatever the process, they manage to make "realistic kits" that sound like cheap synth patches. I've got nothing against Roland - use plenty of their gear and I love my drums themselves, but their drum samples really are a joke, IMHO.
-
- KVRAF
- 4735 posts since 18 Jul, 2002 from London, UK
"modelled" and "sampled" are not mutually exclusive in any case.. a lot of people seem to take modelling to mean Karplus-Strong style efforts, but a model complex enough to recreate an identifiable real drum (not just "snare", but the sort of thing proper drum geeks are in to - "1972 Ludwig Black Beauty, tight tuning, loose wire, clear head, nylon-tipped medium gauge sticks, left hand strike, slightly overweight drummer with moderate to high blood alcohol, trace blood THC, played in a medium sized wood panelled room on an overcast day") you're going to need to do automated parameter capture + analysis on the sound anyway, which is really just another word for sampling + (re)synthesis.
If we wanted to put extra marketing spin on BFD, we could hype several of the synthesis features as "modelled".. you can model the effect of damping of a drum by putting a decay envelope on it - it might not be the most /accurate/ model, but it's a model nonetheless and one which sounds a sight better than some of the woodwind models I've heard over the past year or two
If we wanted to put extra marketing spin on BFD, we could hype several of the synthesis features as "modelled".. you can model the effect of damping of a drum by putting a decay envelope on it - it might not be the most /accurate/ model, but it's a model nonetheless and one which sounds a sight better than some of the woodwind models I've heard over the past year or two
This account is dormant, I am no longer employed by FXpansion / ROLI.
Find me on LinkedIn or elsewhere if you need to get in touch.
Find me on LinkedIn or elsewhere if you need to get in touch.
-
stephenpaulharper stephenpaulharper https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=73921
- KVRian
- Topic Starter
- 778 posts since 4 Jul, 2005 from Atlanta
Hey Beej,
This is from www.sweetwater.com, but I know I've seen it elsewhere, so I assumed it was a Company line:
"If you've never experienced Roland's COSM technology, you might actually be wondering what all the fuss is about. Essentially, COSM is a process that allows virtual instruments to be built without any samples being involved. Highly complex mathematical formulas (and some top secret Roland magic) produce sounds that are so real, you'll be totally blown away".
I was blown away alright - right in the direction of the nearest Guitar Center, where I bought BFD! The TD modules are such a sham, because they're the most expensive part of a kit, but yet they're the most useless. Hopefully, somebody will wake up one day and make a decent TMI that works well with all kinds of triggers and has plenty of inputs for expansion. They'd make a mint.
I'll have you know, I rarely drink, don't have a panelled room, and I doubt my THC levels would be considered "trace". BTW, that money I've set aside for BFD2 is starting to burn a hole in my pocket. Jamstix 2 is starting to look pretty sweet...
Maybe I misunderstand "modeling". I assumed modeled instruments started life as a machine generated tone/waveform, that was then tweaked to the point of resembling the real thing. I know there are things like Saxlab and such that use a combination of modeling and real samples, but would never have guessed that an actual recording of a real drum was involved in any way in the Roland sounds.
That's why I come here - to learn.
Thanks, guys.
Steve
This is from www.sweetwater.com, but I know I've seen it elsewhere, so I assumed it was a Company line:
"If you've never experienced Roland's COSM technology, you might actually be wondering what all the fuss is about. Essentially, COSM is a process that allows virtual instruments to be built without any samples being involved. Highly complex mathematical formulas (and some top secret Roland magic) produce sounds that are so real, you'll be totally blown away".
I was blown away alright - right in the direction of the nearest Guitar Center, where I bought BFD! The TD modules are such a sham, because they're the most expensive part of a kit, but yet they're the most useless. Hopefully, somebody will wake up one day and make a decent TMI that works well with all kinds of triggers and has plenty of inputs for expansion. They'd make a mint.
Nice way to describe some of your customers Angus.Angus_FX wrote:...slightly overweight drummer with moderate to high blood alcohol, trace blood THC, played in a medium sized wood panelled room on an overcast day...
Maybe I misunderstand "modeling". I assumed modeled instruments started life as a machine generated tone/waveform, that was then tweaked to the point of resembling the real thing. I know there are things like Saxlab and such that use a combination of modeling and real samples, but would never have guessed that an actual recording of a real drum was involved in any way in the Roland sounds.
That's why I come here - to learn.
Thanks, guys.
Steve


