11 Jun
2004
11 Jun
'04
7:25 p.m.
Hi,
I have been discussing the design of a drum sample
library for Hydrogen with Comix (Hydrogen author),
Bill Bailey (my studio partner and a drummer) and
Bobby Josland (drummer). I don't think any of us has
experience with designing a high quality sample
library. We've got some questions and hope to get
feedback in the form of "this is my experience" and "I
don't know but imagine...."
The objective is to create a high quality sample
library that has the potential to "sound" like a live
drummer. Hydrogen has features that will help achieve
sound and feel. From the perspective of the library
I'm only concerned with "sound."
Snare Drum
For the sake of argument, assume there are three
primary sounds; 1, 2 and 3 and that each sound has
three volume levels; low, mid and high. This means
there needs to be nine snare drum samples--the reality
is that we'll have many, many more than that. The
objective is to sequence a part that has the dynamics
of a live player--snare roll or whatever.
Are three volume/dynamic samples for every single
sound enough for a sequence to achieve or get close
enough to the dynamics of a live player? Does anyone
know how the commercial drum libraries are produced?
Maybe we need five volume/dynamic samples for every
sound.
Stereo Image
I'm interested in knowing what people think is the
best solution for achieving live sounding stereo
imaging. I've spent several days recording test sounds
and conlcuded that some instruments are probably best
treated as stereo samples. Dry (no reverb) snare drums
can potentially be mono tracks but I suspect cymbals
must be stereo.
The cymbal recording strategy will be to form a
somewhat natural semicircle and possibly use an X/Y
mic pattern. When sequencing ride cymbals they need to
bleed from the listeners Hard left side across the
stereo field and into the Hard right side. If we don't
treat cymbals as stereo images they could fail to
sound like a live player when sequenced.
Snare drums might achieve live sounding stereo imaging
from 1 in 2 out reverbs. In addition to dry mono snare
drums, we intend to produce a number of reverb
processed (wet) stereo samples.
Anyway, I'd appreciate any thoughts on these issues. I
don't expect production to begin for another two or
three weeks and that will be somewhat dictated by the
usability of Ardour. With the push for Ardour 1.0 I'm
preoccupied with testing and not using it.
Incidentally, the production will be done in our
studio which has accoustically tuned rooms and with
some high quality signal chains; Neumann -> Avalon,
and some average chains. The overheads are SM81 which
run about $600.00 U.S. a pair. They aint great but
they are good enough. Those signal paths will run
through a Soundtracs Solo consol and insert into a
Tascam Dm-24. The drums will sound great because Bill,
Bobby and I know what we're doing. If we don't know
what we're doing someone should shoot us--we've got
about 90 years of combined experienced with playing
and recording drums.
ron
__________________________________
Do you Yahoo!?
Friends. Fun. Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/
11 Jun
11 Jun
7:41 p.m.
On Fri, 11 Jun 2004 14:17:06 -0700 (PDT)
R Parker <rtp405(a)yahoo.com> wrote:
[many good points snipped]
From my point of view [as a weak drummer, producer of
some electronic music, a little better jazz guitr player, and consument of all kinds of
music] there's always one important thing missing in every "virtual" drummer
i heard, except for those which were loop based [and even loop based drumsequences have a
characteristic and easily distinguished sound]:
The interplay between the different drums: Each drum you hit, makes sound. This sound
reonates in the other parts of the drumset. If you hit the bass drum, the snare drum will
produce sound, too..
Also i think talkover is very important for a live drum sound. If you mic a bd and a
snare, the bd mic will always pick up the snare and the other way around.
Also very important is the interplay of the drum stick with a still resonating drum
cymbal. A ride cymbal will sound different when hit from a silent state, than it will
sound when hit while it's already ringing..
For rides and crash cymbals, the physical movement of the cymbal is very important, too..
If you hit a symbol on one side, it will swing in the air [i mean large scale movement,
not the vibration that produces the primary sound]. Everytime you hit a cymbal it will
travel a different path thus endlessly modifying the sound which reaches the mic..
If all these things aren't present, it won't sound like a real drumset.
Florian Schmidt
--
Palimm Palimm!
8:04 p.m.
Hi Florian,
--- Florian Schmidt <mista.tapas(a)gmx.net> wrote:
On Fri, 11 Jun 2004 14:17:06 -0700 (PDT)
R Parker <rtp405(a)yahoo.com> wrote:
[many good points snipped]
The interplay between the different drums: Each drum
you hit, makes sound. This sound reonates in the
other parts of the drumset. If you hit the bass
drum, the snare drum will produce sound, too..
No doubt. Earlier today I sat down and hit the kick
drum and thought what on earth has happened to that. I
immediately realized the snares were turned off. :)
Bill and I will discuss these issues before we start
printing.
An interesting challange with controlling bleed is the
requirements of different genres of music. With Pop,
it's common for me to get most of the bleed, ambiance
and room from the overheads and seriously limit bleed
in the other instruments. However, with Jazz that
aproach is simply wrong.
This definitely requires consideration. I've thought
about direct micing one instrument at a time while
leaving the kit and a room mic at the other end of the
room. That mic would pick up the room and subsequent
sounds from the kit and be a seperate track that could
be used at the discretion of the end user. This
probably wouldn't work for everything but it's one
thought.
Good stuff Florian.
ron
Also i think talkover is very important for a live
drum sound. If you mic a bd and a snare, the bd mic
will always pick up the snare and the other way
around.
Also very important is the interplay of the drum
stick with a still resonating drum cymbal. A ride
cymbal will sound different when hit from a silent
state, than it will sound when hit while it's
already ringing..
For rides and crash cymbals, the physical movement
of the cymbal is very important, too.. If you hit a
symbol on one side, it will swing in the air [i mean
large scale movement, not the vibration that
produces the primary sound]. Everytime you hit a
cymbal it will travel a different path thus
endlessly modifying the sound which reaches the
mic..
If all these things aren't present, it won't sound
like a real drumset.
Florian Schmidt
--
Palimm Palimm!
__________________________________
Do you Yahoo!?
Friends. Fun. Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/
7:51 p.m.
R Parker wrote on Fri, 11-Jun-2004:
I have been discussing the design of a drum sample
library for Hydrogen with Comix (Hydrogen author),
Bill Bailey (my studio partner and a drummer) and
Bobby Josland (drummer). I don't think any of us has
experience with designing a high quality sample
library. We've got some questions and hope to get
feedback in the form of "this is my experience" and "I
don't know but imagine...."
What licensing scheme are you looking at? Is the idea for this
to become the new default acoustic set for H2, or something to be
purchased? Freely modifiable and redistributable or not?
Looking forward to it....
jlc
8:23 p.m.
Hi Jesse,
--- Jesse Chappell <jesse(a)essej.net> wrote:
R Parker wrote on Fri, 11-Jun-2004:
We'd like to follow in the traditions of Bill Gates
and Steve Jobs. Ah, really, really just kidding. :)
Comix has asked for an appropriately licensed sample
library for Hydrogen. He and I have discussed both our
interests. Neither of us wants to restrict anyone's
right to produce and freely distribute derivitive
products in the form of songs or audio tracks of
whatever type.
However, I don't have any motivation to allow for
profit enterpizes to include the library with their
for sale products. Well, not without written
permission which very well might have to be purchased.
I also can't see any reason to allow anyone to modify
and redistribute the library without written
permission.
I have no reasons to restrict any other Open Source
License project from including the unmodified library
for distribution with their product.
Please do hammer on these concerns. We definitely want
to get the licensing right.
Ultimately, my motivation is just an act of gratitude
to this community.
Is the
I have been discussing the design of a drum
sample
library for Hydrogen with Comix (Hydrogen
author),
Bill Bailey (my studio partner and a drummer)
and
Bobby Josland (drummer). I don't think any of us
has
experience with designing a high quality sample
library. We've got some questions and hope to get
feedback in the form of "this is my experience"
and "I
don't know but imagine...."
What licensing scheme are you looking at? idea for this
to become the new default acoustic set for H2,
It's intended to be the new default set for H2.
ron
or
something to be
purchased? Freely modifiable and redistributable or
not?
Looking forward to it....
jlc
__________________________________
Do you Yahoo!?
Friends. Fun. Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/
11:10 p.m.
On Fri, 2004-06-11 at 18:14, R Parker wrote:
You might be interested in using the creative commons license. You can
find the details here: http://creativecommons.org/
What licensing
scheme are you looking at?
We'd like to follow in the traditions of Bill Gates
and Steve Jobs. Ah, really, really just kidding. :)
Comix has asked for an appropriately licensed sample
library for Hydrogen. He and I have discussed both our
interests. Neither of us wants to restrict anyone's
right to produce and freely distribute derivitive
products in the form of songs or audio tracks of
whatever type.
However, I don't have any motivation to allow for
profit enterpizes to include the library with their
for sale products. Well, not without written
permission which very well might have to be purchased. 
8:04 p.m.
On Fri, Jun 11, 2004 at 02:17:06PM -0700, R Parker wrote:
Are three volume/dynamic samples for every single
sound enough for a sequence to achieve or get close
enough to the dynamics of a live player? Does anyone
know how the commercial drum libraries are produced?
Maybe we need five volume/dynamic samples for every
sound.
I dunno how many you'll need, but I think the two hardest
things in the kit will be hi-hat and snare, because they
sound so different depending on how you play them.
Stereo Image
I'm interested in knowing what people think is the
best solution for achieving live sounding stereo
imaging. I've spent several days recording test sounds
and conlcuded that some instruments are probably best
treated as stereo samples. Dry (no reverb) snare drums
can potentially be mono tracks but I suspect cymbals
must be stereo.
+1.
The cymbal recording strategy will be to form a
somewhat natural semicircle and possibly use an X/Y
mic pattern.
Ever tried M/S?
I haven't (don't have suitable mics)
but I've heard some nice overheads done that way.
Snare drums might achieve live sounding stereo
imaging
from 1 in 2 out reverbs. In addition to dry mono snare
drums, we intend to produce a number of reverb
processed (wet) stereo samples.
I tend to think that the sound of the whole kit includes
the various bleeds and resonances: e.g. toms and snare
bleeding into overheads.
I'd be inclined to keep the mic positions constant once
you get your sound locked down, and leave at least the
overheads mixed in to the close mics when you're recording
individual hits. This may help the whole thing "gel".
--
Paul Winkler
http://www.slinkp.com
9:48 p.m.
On Fri, Jun 11, 2004 at 05:55:43PM -0400, Paul Winkler wrote:
I dunno how many you'll need, but I think the two
hardest
things in the kit will be hi-hat and snare, because they
sound so different depending on how you play them.
... errr, and cymbals. Drum machine cymbals are always bad.
Cymbals are just hard to represent via samples, they're
too darn complex.
--
Paul Winkler
http://www.slinkp.com
11:10 p.m.
On Friday 11 June 2004 18:39, Paul Winkler wrote:
} ... errr, and cymbals. Drum machine cymbals are always bad.
} Cymbals are just hard to represent via samples, they're
} too darn complex.
Why do they have to sound like cymbals? {Granted this is for a specific
package... take it as a general question} Digital is supposed to be an avenue
for progress no?
{In other words... "7 zillion and one available sounds and you gotta' have a
simulation of something some bronze age guy managed to come up with when he
screwed up the kings dinner plate order?}
--
If I had saxophones / Big baritone, cleanin' up the muddy breaks
If I had Saxophones / I could get some recognition from
that Mobile Alabama DJ {J.Buffet}
12 Jun
12 Jun
7:11 a.m.
lördagen den 12 juni 2004 03.00 skrev RTaylor:
On Friday 11 June 2004 18:39, Paul Winkler wrote:
} ... errr, and cymbals. Drum machine cymbals are always bad.
} Cymbals are just hard to represent via samples, they're
} too darn complex.
Why do they have to sound like cymbals? {Granted this is for a specific
package... take it as a general question} Digital is supposed to be an
avenue for progress no?
{In other words... "7 zillion and one available sounds and you gotta' have
a simulation of something some bronze age guy managed to come up with when
he screwed up the kings dinner plate order?}
Yep :^)
/R
--
http://spamatica.se/music/
2:37 p.m.
--- Robert Jonsson <rj(a)spamatica.se> wrote:
__________________________________
Do you Yahoo!?
Friends. Fun. Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/
l�rdagen den 12 juni 2004 03.00 skrev RTaylor:
Well, this is where you would route H2:playback or
specimen:playback to pd:input or something to that
affect. One of the synthesis tools is jackified right?
With a stable Ardour just around the corner I'm almost
ready to go beyound the beating of dinner plates. But
I'm a little scared.
ron
On Friday 11 June 2004 18:39, Paul Winkler
wrote:
} ... errr, and cymbals. Drum machine cymbals are
always bad.
} Cymbals are just hard to represent via samples,
they're
} too darn complex.
Why do they have to sound like cymbals? {Granted
this is for a specific
package... take it as a general question} Digital
is supposed to be an
> avenue for progress no? {In other
words... "7 zillion and one available
sounds and you gotta' have
a simulation of something some bronze age guy
managed to come up with when
he screwed up the kings dinner plate order?}
Yep :^)
/R
--
http://spamatica.se/music/
13 Jun
13 Jun
3:19 a.m.
On Saturday 12 June 2004 11:28, R Parker wrote:
} --- Robert Jonsson <rj(a)spamatica.se> wrote:
} > lördagen den 12 juni 2004 03.00 skrev RTaylor:
} > > Why do they have to sound like cymbals? {Granted
}
} Well, this is where you would route H2:playback or
} specimen:playback to pd:input or something to that
} affect. One of the synthesis tools is jackified right?
}
} With a stable Ardour just around the corner I'm almost
} ready to go beyound the beating of dinner plates. But
} I'm a little scared.
Frankly, I'm not all that interested in re-creating cymbals for the sake of
re-creating cymbals. I'd never use them for what I do anyway.
What if linux audio were to utilize some sort of {FM?} synthesis sample
module {small, loadable by anything and capable of generating a specific
range of sounds. Not pre-recorded but generative... tunable, etc... over time
individual modules would get optimized to a high degree... "presets"
basically but for single sounds.} that could work across the platform.
Samplers {like postgresql {gnumeric, open office, etc...} with some audio
specific code} could function as librarians... to standardize tunings across
a given bank of samples, to set up dependencies and interactions between
given sound modules, etc, etc... with apt {synaptic, aptitude, dselect, etc}
you could set up a server somewhere and interchange libraries and modules.
...Anyway... given some time it could make the need for standardized "sample"
libraries a thing of the past.
} > > {In other words... "7 zillion and one available
} > sounds and you gotta' have
} > > a simulation of something some bronze age guy
} > managed to come up with when
} > > he screwed up the kings dinner plate order?}
--
If I had saxophones / Big baritone, cleanin' up the muddy breaks
If I had Saxophones / I could get some recognition from
that Mobile Alabama DJ {J.Buffet}
3:23 a.m.
On Saturday 12 June 2004 04:02, Robert Jonsson wrote:
} lördagen den 12 juni 2004 03.00 skrev RTaylor:
} > {In other words... "7 zillion and one available sounds and you gotta'
have } > a simulation of something some bronze age guy managed to come up
with when } > he screwed up the kings dinner plate order?}
}
} Yep :^)
http://www.zildjian.com/
--
If I had saxophones / Big baritone, cleanin' up the muddy breaks
If I had Saxophones / I could get some recognition from
that Mobile Alabama DJ {J.Buffet}
7560
days inactive
7562
days old
linux-audio-user@lists.linuxaudio.org
12 comments
7 participants
participants (7)
-
Doug James -
Florian Schmidt -
Jesse Chappell -
Paul Winkler -
R Parker -
Robert Jonsson -
RTaylor