1 Mar
2005
1 Mar
'05
11:11 a.m.
Hi,
I have some questions and ideas about workflow using Ardour and Jamin
and wanted to see what other people's experiences have been in this area.
Here's the workflow I want to put into place:
Say I have 3 or 4 songs that I want to have segue into each other (i.e.,
like on most prog rock concept albums where each song flows into each
other rather than individual tracks that are separate). Each song is
recorded in a separate session, since each has different
instrumentation, tempo, atmosphere, etc. Maybe I want to have some wind
or sound effects or something also as part of the segue, so one song
fades out to wind while a second song starts while the wind is still
blowing (yeah, yeah, I know, how cliched).
Now, from here, maybe my next step would be to do initial mixes of these
tracks to stereo WAV files, and then I build a super session and import
the WAV files and build a playlist with the tracks, so I can get flow
between the tracks the way I want, set crossfades, seques, etc. Then
perhaps I master these tracks via Jamin so I can get a good sound level
balance between the tracks, and then record this output back into the
Ardour session as a final mixed and mastered sequence, and thence onto
the DAO CD mastering (where I probably would put in track markers with
zero gaps).
Or, would I perhaps want to do some pre-mastering of the individual
multitrack sessions first, and then mix down to stereo WAV and then
build a playlist out of the pre-mastered tracks, then back through Jamin
for final audio mastering, recording this back into Ardour, and then
finally off to the CD mastering step?
-- Brett
1 Mar
1 Mar
12:26 p.m.
--- "Brett W. McCoy" <idragosani(a)chapelperilous.net>
wrote:
Say I have 3 or 4 songs that I want to have segue
into each other (i.e.,
like on most prog rock concept albums where each
song flows into each
other rather than individual tracks that are
separate). Each song is
recorded in a separate session, since each has
different
instrumentation, tempo, atmosphere, etc. Maybe I
want to have some wind
or sound effects or something also as part of the
segue, so one song
fades out to wind while a second song starts while
the wind is still
blowing (yeah, yeah, I know, how cliched).
Now, from here, maybe my next step would be to do
initial mixes of these
tracks to stereo WAV files, and then I build a super
session and import
the WAV files and build a playlist with the tracks,
so I can get flow
between the tracks the way I want, set crossfades,
seques, etc.
One track/playlist with all stereo mixes on it. At
this point you can easily build multitrack seques and
route them and stereo mixes to the Master bus which
routes to JAMin. Then add a second Ardour bus, name it
jam-return, and route JAMin into it and route
jam-return to your playback ports.
Then
perhaps I master these tracks via Jamin so I can get
a good sound level
balance between the tracks, and then record this
output back into the
Ardour session as a final mixed and mastered
sequence,
An alternative is to Export Range rather than record
into a new track. It eliminates the record pass and
Export/freewheel is faster than realtime. The drawback
is that after leaving realtime priority for freewheel
the realtime priority is lost. You have to restart
jackd to get realtime back. It sucks bad but isn't a
big deal in this case because you've produced the
final product.
Having produced the final product, I like to visually
view and listen to it using Rezound. It has excellent
visual clip indication. It is not acceptable to have
even one clip. To address that concern, you can set
the JAMin limiter to -0.02.
If you plan to produce an .ogg then you'll need to
hold the amplitude around -0.05 because oggvorbis
encoding, much to my dismay, increases amplitude and
will create clipping in an otherwise perfect file.
As a last note, you could use the LADSPA OSC plugin to
automate scence changes in JAMin. This works but for
several reasons is really more a proof of concept
implementation. Reguardless, it is awesome and
eventually internal OSC should be implemented into
Ardour. Great stuff!
That's a couple things to consider. You really do have
the right production concepts.
ron
and thence onto
the DAO CD mastering (where I probably would put in
track markers with
zero gaps).
Or, would I perhaps want to do some pre-mastering of
the individual
multitrack sessions first, and then mix down to
stereo WAV and then
build a playlist out of the pre-mastered tracks,
then back through Jamin
for final audio mastering, recording this back into
Ardour, and then
finally off to the CD mastering step?
-- Brett
__________________________________
Do you Yahoo!?
Take Yahoo! Mail with you! Get it on your mobile phone.
http://mobile.yahoo.com/maildemo
2:11 p.m.
New subject: [linux-audio-user] Midi Cards
How can I keep my MIDI from sounding like a cheesy video game?
I suspect its my card. What card would give me excellent/sample quality
midi or playback?
Chaz Kiser
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 266.5.1 - Release Date: 2/27/05
3:09 p.m.
New subject: [linux-audio-user] Midi Cards
On Tue, 2005-03-01 at 13:11 -0500, ewf wrote:
How can I keep my MIDI from sounding like a cheesy
video game?
I suspect its my card. What card would give me excellent/sample quality
midi or playback?
If you have an emu10k1 based card and are using the hardware wavetable
synth, the single best thing you can do is get a decent soundfont. The
ones that come with the Windows driver all suck.
Lee
2 Mar
2 Mar
8:09 p.m.
New subject: [linux-audio-user] Midi Cards
How can I keep
my MIDI from sounding like a cheesy video game?
I suspect its my card. What card would give me excellent/sample
quality midi or playback?
If you have an emu10k1 based card and are using the hardware
wavetable synth, the single best thing you can do is get a decent
soundfont. The ones that come with the Windows driver all suck.
8:27 p.m.
New subject: [linux-audio-user] Midi Cards
On Wed, 2005-03-02 at 19:03 -0500, LinuxMedia wrote:
This is my standard answer to that... hehe... this is the best soundfont
I've heard....
http://inanna.ecs.soton.ac.uk/~swh/fluid-unpacked/
It's 141 Megs. I (think) some hardware wavetable synths have limited
capacity. If so use a soft synth like fluidsynth.
I believe the emu10k1's hardware synth is limited to 128MB. It would be
very useful if someone could trim that 141MB down to 128.
Lee
How can I keep my MIDI from sounding like a cheesy
video game?
I suspect its my card. What card would give me excellent/sample
quality midi or playback?
If you have an emu10k1 based card and are using the hardware
wavetable synth, the single best thing you can do is get a decent
soundfont. The ones that come with the Windows driver all suck. 
9:32 p.m.
New subject: [linux-audio-user] Midi Cards
Lee Revell wrote:
I'd really like to rearrange that soundfont and get rid of
(to me) some crap... and perhaps try to improve some of the
sounds with better samples... but every time I have tried
to use smurf over the years it just dies on me.
Are there any other tools, especially CLI ones, that could
pull apart a sf2 soundfont and put it back together ?
A CLI mode sf2compiler would be excellent... then some parts
of the procedure could be easily batched/automated.
One solution is to simply use LinuxSampler (for decent sounds)
but the likeyhood of finding free and useful gig files is near
or less than zero.
--markc
http://inanna.ecs.soton.ac.uk/~swh/fluid-unpacked/
It's 141 Megs. I (think) some hardware wavetable synths have limited
capacity. If so use a soft synth like fluidsynth.
I believe the emu10k1's hardware synth is limited to 128MB. It would be
very useful if someone could trim that 141MB down to 128. 
3 Mar
3 Mar
2:19 a.m.
New subject: [linux-audio-user] Midi Cards
On Wednesday 02 March 2005 08:31 pm, Mark Constable wrote:
Lee Revell wrote:
I'd really like to rearrange that soundfont and get rid of
(to me) some crap... and perhaps try to improve some of the
sounds with better samples... but every time I have tried
to use smurf over the years it just dies on me.
Smurf is working here. Let's see if we can get a consensus on "crap",
or a least see which way the wind is blowing.
http://inanna.ecs.soton.ac.uk/~swh/fluid-unpacked/
It's 141 Megs. I (think) some hardware wavetable synths have limited
capacity. If so use a soft synth like fluidsynth.
I believe the emu10k1's hardware synth is limited to 128MB. It would be
very useful if someone could trim that 141MB down to 128.
Are there any other tools, especially CLI ones, that could
pull apart a sf2 soundfont and put it back together ?
A CLI mode sf2compiler would be excellent... then some parts
of the procedure could be easily batched/automated.
One solution is to simply use LinuxSampler (for decent sounds)
but the likeyhood of finding free and useful gig files is near
or less than zero.
--markc
2:46 a.m.
New subject: [linux-audio-user] Midi Cards
John Check wrote:
Start at 128 and work back up is what I would suggest.
The only ones I could imagine ever using, perhaps, are
.Reverse Cymbal
.Synth Drum
.Kalimba
.FX 4 (atmosphere)
but only if they were worthwhile decent samples to start
with and that is unlikely.
097 FX 1 (rain) 105 Sitar
098 FX 2 (soundtrack) 106 Banjo
099 FX 3 (crystal) 107 Shamisen
100 FX 4 (atmosphere) 108 Koto
101 FX 5 (brightness) 109 Kalimba
102 FX 6 (goblins) 110 Bagpipe
103 FX 7 (echoes) 111 Fiddle
104 FX 8 (sci-fi) 112 Shanai
PERCUSSIVE SOUND EFFECTS
113 Tinkle Bell 121 Guitar Fret Noise
114 Agogo 122 Breath Noise
115 Steel Drums 123 Seashore
116 Woodblock 124 Bird Tweet
117 Taiko Drum 125 Telephone Ring
118 Melodic Tom 126 Helicopter
119 Synth Drum 127 Applause
120 Reverse Cymbal 128 Gunshot
Another issue is the licensing of the samples and the
soundfont as a whole but just getting a <128Mb file would
be a great start.
>>I believe the emu10k1's hardware synth is
limited to 128MB. It would be
>>very useful if someone could trim that 141MB down to 128.
# asfxload fluid.sf2
sfxload: no memory left
But it mostly seems to work but I've never bothered to
hunt down which instruments are not loaded.
I'd really
like to rearrange that soundfont and get rid of
(to me) some crap... and perhaps try to improve some of the
sounds with better samples... but every time I have tried
to use smurf over the years it just dies on me.
Smurf is working here. Let's see if we can get a consensus on "crap",
or a least see which way the wind is blowing. >Are there any other tools, especially CLI ones,
that could
>pull apart a sf2 soundfont and put it back together ?
Anyone ?
--markc
6:59 a.m.
New subject: [linux-audio-user] Midi Cards
On Thu, 2005-03-03 at 16:46 +1000, Mark Constable wrote:
John Check wrote:
I can load the entire sounfont without any error with my audigy 1 OEM (I
have 256MB of RAM).
I've tried and some programs don't play any instrument (around 10% are
not assigned). With qsynth I can play all the 127 instruments instead.
I've read some time ago that the Live card is limited and can load
soundfont untill to fill half of the system RAM, while the audigy can
load any soundfont size limited only by the free RAM. Maybe that was
only another false creative's claim (as the 24bit on the audigy 1).
- Antonio
>>I believe the emu10k1's hardware synth
is limited to 128MB. It would be
>>very useful if someone could trim that 141MB down to 128.
# asfxload fluid.sf2
sfxload: no memory left
But it mostly seems to work but I've never bothered to
hunt down which instruments are not loaded.
4:01 p.m.
New subject: [linux-audio-user] Midi Cards
On Thursday 03 March 2005 05:59 am, AntonioI wrote:
On Thu, 2005-03-03 at 16:46 +1000, Mark Constable
wrote:
Same on my SBLive! The samples >128Mb just aren't loaded. The order
in which they are stored in the file would determine what's missing.
John Check wrote:
I can load the entire sounfont without any error with my audigy 1 OEM (I
have 256MB of RAM).
I've tried and some programs don't play any instrument (around 10% are >>I believe the emu10k1's hardware synth
is limited to 128MB. It would
>> be very useful if someone could trim that 141MB down to 128.
# asfxload fluid.sf2
sfxload: no memory left
But it mostly seems to work but I've never bothered to
hunt down which instruments are not loaded. not assigned). With qsynth I can play all the 127
instruments instead.
I've read some time ago that the Live card is limited and can load
soundfont untill to fill half of the system RAM, while the audigy can
I have nearly a GB of RAM, so the logic is inconsistent with my observation.
load any soundfont size limited only by the free RAM.
Maybe that was
only another false creative's claim (as the 24bit on the audigy 1).
I can buy that, but that's the way it is with music stuff. Mackie 1604
anybody?
- Antonio
3:56 p.m.
New subject: [linux-audio-user] Midi Cards
On Thursday 03 March 2005 01:46 am, Mark Constable wrote:
John Check wrote:
Start at 128 and work back up is what I would suggest.
The only ones I could imagine ever using, perhaps, are
.Reverse Cymbal
.Synth Drum
.Kalimba
.FX 4 (atmosphere)
but only if they were worthwhile decent samples to start
with and that is unlikely.
097 FX 1 (rain) 105 Sitar
098 FX 2 (soundtrack) 106 Banjo
099 FX 3 (crystal) 107 Shamisen
100 FX 4 (atmosphere) 108 Koto
101 FX 5 (brightness) 109 Kalimba
102 FX 6 (goblins) 110 Bagpipe
103 FX 7 (echoes) 111 Fiddle
104 FX 8 (sci-fi) 112 Shanai
PERCUSSIVE SOUND EFFECTS
113 Tinkle Bell 121 Guitar Fret Noise
114 Agogo 122 Breath Noise
115 Steel Drums 123 Seashore
116 Woodblock 124 Bird Tweet
117 Taiko Drum 125 Telephone Ring
118 Melodic Tom 126 Helicopter
119 Synth Drum 127 Applause
120 Reverse Cymbal 128 Gunshot
So these are "crap" or not? Easiest solution is just whack it into to parts
"melodic" and "other", but trimming 19MB worth of samples may be
really easy.
>>I believe the emu10k1's hardware synth
is limited to 128MB. It would be
>>very useful if someone could trim that 141MB down to 128.
# asfxload fluid.sf2
sfxload: no memory left
But it mostly seems to work but I've never bothered to
hunt down which instruments are not loaded.
I'd
really like to rearrange that soundfont and get rid of
(to me) some crap... and perhaps try to improve some of the
sounds with better samples... but every time I have tried
to use smurf over the years it just dies on me.
Smurf is working here. Let's see if we can get a consensus on "crap",
or a least see which way the wind is blowing.
Another issue is the licensing of the samples and the
That's _the_ issue as far as doing anything redistributable..
According to the comment field:
DO NOT REDISTRIBUTE ANY OF THESE SAMPLES.
I don't see an email address but I'll see if I can get a hold of the author
for clarification. There's no telling where he got 'em, it may be CYA.
The copyright holder for the soundfont appears to be an individual, which
raises some questions as far as having the resources WRT acquiring the raw
samples in the first place.
Making patch sets and variations from different sources is expected. Exactly
what would constitute fair use of the raw samples would depend on sample
length.. Loops of a few cycles aren't likely to be issues. Full bar rhythm
loops OTOH....
soundfont as a whole but just getting a <128Mb file
would
be a great start.
Ignoring bit/sample rates, bigger soundfont size doesn't count for much. It
takes less time/skill to produce.
Are there any other tools, especially CLI ones,
that could
>pull apart a sf2 soundfont and put it back together ?
Anyone ?
--markc
1:39 a.m.
New subject: [linux-audio-user] Midi Cards
On Wednesday 02 March 2005 07:27 pm, Lee Revell wrote:
On Wed, 2005-03-02 at 19:03 -0500, LinuxMedia wrote:
This is my standard answer to that... hehe... this is the best soundfont
I've heard....
http://inanna.ecs.soton.ac.uk/~swh/fluid-unpacked/
It's 141 Megs. I (think) some hardware wavetable synths have limited
capacity. If so use a soft synth like fluidsynth.
I believe the emu10k1's hardware synth is limited to 128MB. It would be
very useful if someone could trim that 141MB down to 128.
Lee
That's what I'd heard but I just loaded it into a fairly early revision
Platinum. All I've done was check patches 0 and 127, but one can at least
load 128MB worth and use parts of it.
How can I keep my MIDI from sounding like a cheesy
video game?
I suspect its my card. What card would give me excellent/sample
quality midi or playback?
If you have an emu10k1 based card and are using the hardware
wavetable synth, the single best thing you can do is get a decent
soundfont. The ones that come with the Windows driver all suck.
1:34 a.m.
New subject: [linux-audio-user] Midi Cards
On Wednesday 02 March 2005 07:03 pm, LinuxMedia wrote:
This is my standard answer to that... hehe... this is the best soundfont
I've heard....
http://inanna.ecs.soton.ac.uk/~swh/fluid-unpacked/
It's 141 Megs. I (think) some hardware wavetable synths have limited
capacity. If so use a soft synth like fluidsynth.
Rocco
What's the deal with the licensing on that? Where did it originate?
I'll break out some patches to have a look see, but if we can get a definite
call on on the legality I'll shave it down to 128MB. Breaking it into
separate Music & FX wouldn't suck. Wouldn't be GM, but that's a lowest
common
denominator anyway.
How can I keep my MIDI from sounding like a cheesy
video game?
I suspect its my card. What card would give me excellent/sample
quality midi or playback?
If you have an emu10k1 based card and are using the hardware
wavetable synth, the single best thing you can do is get a decent
soundfont. The ones that come with the Windows driver all suck. 
2 Mar
2 Mar
4:15 a.m.
New subject: [linux-audio-user] Midi Cards
ewf wrote:
How can I keep my MIDI from sounding like a cheesy
video game?
"Cheesy video games" would indicate that you're currently using an
OPL3 FM synthesizer chip.
I suspect its my card. What card would give me
excellent/sample
quality midi or playback?
The only supported wavetable chip is Emu10k1 (i.e. Soundblaster Live
or a non-LS Audigy).
You could use a software synthesizer or some external MIDI
synthesizer, too.
Regards,
Clemens
7207
days inactive
7209
days old
linux-audio-user@lists.linuxaudio.org
14 comments
9 participants
participants (9)
-
AntonioI -
Brett W. McCoy -
Clemens Ladisch -
ewf -
John Check -
Lee Revell -
LinuxMedia -
Mark Constable -
R Parker