[linux-audio-user] Recommendations for audio software needed

LinuxMedia linuxmedia4 at netscape.net
Sat Jan 10 06:05:09 EST 2004


>>> Rosegarden does need JACK for audio, yes.  If you only want MIDI
>>> and notation -- and when I say MIDI I include the ability to
>>> drive ALSA soft synths such as Timidity, qsynth or Fluidsynth --
>>> then you can build it happily without JACK or run a JACK-built
>>> version without running jackd. >>> Rosegarden does need JACK for audio, yes.  If you only want MIDI
 >>> and notation -- and when I say MIDI I include the ability to
 >>> drive ALSA soft synths such as Timidity, qsynth or Fluidsynth --
 >>> then you can build it happily without JACK or run a JACK-built
 >>> version without running jackd.

Although (and correct me if I'm mistaken) doesn't Rosegarden (I'm 
talking about Rosegarden-4) allow syncing with other programs via 
jack_connect? I know that JACK can be challenging and I know that 
jack_connect is still new (I think). And so the question would be "what 
multitrack recorder to sync it up to"? I know that ecasound is (not) for 
the "average user", but now there's "TKeca" which is a front end to 
ecasound (see the links below).

Since this gentleman is asking about software I'll explain ecasound and 
the gui front end "TKeca". Ecasound is a command line multi-track 
recorder and effects processor that has many other  features (too many 
to mention and I would probly leave some out). It's very stable, 
powerful, no-nonsense command line progam that you can use "TKeca" as 
the gui interface (links below). I've used TKeca and it is very strait 
forward and (like ecasound) pretty stable. TKeca starts up quickly and 
is pretty intuitive and easy to use.

I'm not sure if ecasound is difficult to set up or not because I've been 
using it for so long that I know a lot about how it does it's business. 
But I can tell you that I've *never* had a problem compiling the 
program. If I install the (stated) dependencies, it always compiles and 
installs correctly (unlike a lot of programs). And you don't (have to) 
install some of the programs that ecasound uses (unless you want) I'm 
talking about the progams that do things like converting to mp3, or 
converting a midi file to a WAV (via timidity soft synth). So there's a 
bunch of program that you (may not) even need when installing ecasound. 
  I found that just recoding WAV's is the best because you can do 
"seeks" with them anyways ("jump" to a specific time in the song, or 
proper fast forward or rewind functioning). File formats like mp3 don't 
allow for proper "seeks" (and possibly maybe not even fast forwarding or 
rewinding). TKeca takes advantage of all this. And there's no extra 
components (other programs) to install just to record WAV tracks. 
Ecasound (and therefor TKeca) take advantage of the LADSPA effects plug 
in system. So the user can put 4 effect on any track at any time. The 
sliders to adjust the effects work in real time. In other words, you can 
change the depth of a reverb effect as you track is playing so that you 
hear the changes in real time.

The only thing that I can say negative about TKeca is that it doesn't 
monitor the effects as you record. But when you stop recording and play 
a track, you can add up to four effects per track.

TKeca also has a "roll-in" mechanism. It allows for specified amount of 
seconds of playing before the actual recording starts. Another words, If 
you record (for instance) the vocal part that goes "My big legged momma 
loves a man that aint me" and you sang out of key on the word "momma". 
Well you don't want to just press record over the part that needs 
fixing, because it's too abrupt. Most people would like to hear the 
parts that leads into the spot that needs recording over. It will 
"pre-roll" into the specified part you want to record over. And when it 
gets to that part, will start recording at that point only.

On the last version I tried, the roll-in mechanism worked, but you had 
to start from the beginning of the song. I hope that was fixed in the 
latest version.

I only used some basic parts of the (new) version of TKeca, so I can't 
vouch for it completly. But I would be glad to be of assistance if you 
want to install both of those programs.

This email is getting long. But I wanted to share some of my experiences 
with muse. It not only has several "built in" soft synths, but two 
(perticulary) will provide a complete set of sound. There's actually two 
soft synths in the "first" catagory of soft synths. I'm talking about 
soft synths that use sound fonts. They are "fluidsynth" and "Fluid". I 
don't know if these programs are related, but I did a "side-by-side" 
comparison of the sound of both of them and found that "fluid" sounded 
much better than "Fluidsynth". And if you download the sound font 
"FluidR3 GM.SF2" from the link below, I think you will be happy with 
what you find. It is 141 Megs but it's worth every bit of download time 
because they are "sterio panned" "real" instruments. The "end users" 
don't even have to worry about doing any panning. I put a couple of 
intruments on a few tracks and they were panned beautifully together. I 
really like this font.

The other advantage of the softsynth "Fluid" (instead of "Fluidsynth") 
is that once you load a soundfont into "Fluid" and then save muse's 
configuration, the soundfont will load automatically into "Fluid" when 
you start muse again. I could'nt get "Fluidsynth" to do this. I had to 
reload the font every time I started the program. And if you want to 
make this easy on the end user, you may want to use the "Fluid" soft synth.

The second catagoy of softsynths that come with it are the buzzy 
"synthy" sounds that are electronically produced. In fact the "vam" 
softsynth has an interface for some pretty good control over the sound. 
And not only can you move sliders in real time to get all kinds of cool 
"buzzy" sounds while your playing, but I even was able to use the 
sliders on some already pre-recorded midi tracks as they were playing.

So you have real intruments and you have buzzy (synth sounding) intruments.

I get the impression that the "live recoding" capababilities (using a 
microphone and "real" instrument as opposed to midi recording) isn't 
very good at this time with muse. And that's why I suggested TKeca. And 
if you can find a midi recorder that syncs with jack_connect, then you 
have the best of both worlds.

Actaully, here's a bunch of links...

   The "FluidR3 GM.SF2" sound font...
http://inanna.ecs.soton.ac.uk/~swh/fluid-unpacked/

   ecasound home page...
http://www.eca.cx/welcome.html

   TKeca download page...
http://sourceforge.net/project/showfiles.php?group_id=63518

   While we're on the subject of programs built on top of ecasound. This 
next one is an "effects box" (my terminoligy) that uses all the 
functionality of ecasound. It's called Ecamegapedal...
http://www.wakkanet.fi/~kaiv/ecamegapedal/

I think these are good programs. And if you have the patience, if you 
get jack running well, it opens up the possibility of using jack_connect 
to sycronize programs. I (think) that there's (not) a lot out there that 
takes advantage of it. One of my biggest hopes was to install the latest 
version of the "Hydrogen" drum machine because the version that came 
with SuSE 9.0 doesn't have the jack transport functions (for 
syncronizing with other programs). I spent a whole night satisfying all 
the dependencies. It built without a hitch, but it just wont start up 
properly. I'm now using Hydrogen 0.8.0-80 (that came with SuSE 9.0 
distribution) and I really like it. Maybe I just don't have a lot of 
experience with linux drum machines, but these are the first good drum 
samples that I've heard. I've never been satisfied with any drum sample 
that I've heard prior to using Hydrogen. And you can get "drum kits" 
online (I haven't downloaded any yet). But Hydrogen is not only easy to 
use, but the manual is easy to understand also.

You know, I really avoided jack because it just added another layer of 
problems. But I connected Hydrogen up to it and the latency (the amount 
of time from pressing the keyboard to the time the sound came out) 
disappeared. And I was using the lowest buffer size when (not) hooked up 
the jack and couldn't get good response time until I hooked it up to 
jack. In fact I get immediate responce time.

There's a program called qjackconnect that allows for a point and click 
connecting of programs and devices when using jack so you don't have to 
go into the shell and type cryptic command to hook up programs and 
devices (and then not hook up correcty after all the effort). And since 
I'm speaking of connections. The one that hooks up things like your midi 
keyboard to synths (and things like that) is called "kaconnect". What I 
like about it, is that it not only shows (in the windows) what there is 
to connect together but (for me) it makes it easy to understand how all 
these cryptic things actually work. And (I'm assuming) it will make 
"aconnect" (command line program that kaconnect runs) easier to 
understand when it comes to adding it into scripts so that these devices 
will automatically connect apon starting up programs. Prior to 
kaconnect, I couldn't visualize what a 65:0 or a 192:0 was (or why there 
was a 65:0, 65:1, 65:2 and... 65:3). Both kaconnect and qjackconnect 
puts them in front of you in two boxes ("from" ports and "to" ports). 
You just connect them with mouse clicks.

I'm becoming a bit long winded, but what I'm trying to say is, you could 
get along well without jack, but I'm just wondering if (after a while) 
it will become difficult to (not) use jack. I know that some of the 
newer versions of programs I've seen are building as "jack only" apps. 
So I finally gave in and did the research nessesary (and it was painful)

I was wondering if this would be a good time for us to use this 
situation to get this gentlman up and running with jack. Maybe we can 
all just "push thru" and get him set up. That way we can have a thurough 
conversation about setting up jack. I'm sure I'll learn a lot also. It's 
not that jack is that complex. It's just finding the information 
nessesary. You know... "were is the information... is it relivant to 
what I'm doing... what is this other aspect they are talking about... do 
I even need it... where can I find that information..."

Rocco

Although (and correct me if I'm mistaken) doesn't Rosegarden (I'm 
talking about Rosegarden-4) allow syncing with other programs via 
jack_connect? I know that JACK can be challenging and I know that 
jack_connect is still new (I think). And so the question would be "what 
multitrack recorder to sync it up to"? I know that ecasound is (not) for 
the "average user", but now there's "TKeca" which is a front end to 
ecasound (see the links below).

Since this gentleman is asking about software I'll explain ecasound and 
the gui front end "TKeca". Ecasound is a command line multi-track 
recorder and effects processor that has many other  features (too many 
to mention and I would probly leave some out). It's very stable, 
powerful, no-nonsense command line progam that you can use "TKeca" as 
the gui interface (links below). I've used TKeca and it is very strait 
forward and (like ecasound) pretty stable. TKeca starts up quickly and 
is pretty intuitive and easy to use.

I'm not sure if ecasound is difficult to set up or not because I've been 
using it for so long that I know a lot about how it does it's business. 
But I can tell you that I've *never* had a problem compiling the 
program. If I install the (stated) dependencies, it always compiles and 
installs correctly (unlike a lot of programs). And you don't (have to) 
install some of the programs that ecasound uses (unless you want) I'm 
talking about the progams that do things like converting to mp3, or 
converting a midi file to a WAV (via timidity soft synth). So there's a 
bunch of program that you (may not) even need when installing ecasound. 
  I found that just recoding WAV's is the best because you can do 
"seeks" with them anyways ("jump" to a specific time in the song, or 
proper fast forward or rewind functioning). File formats like mp3 don't 
allow for proper "seeks" (and possibly maybe not even fast forwarding or 
rewinding). TKeca takes advantage of all this. And there's no extra 
components (other programs) to install just to record WAV tracks. 
Ecasound (and therefor TKeca) take advantage of the LADSPA effects plug 
in system. So the user can put 4 effect on any track at any time. The 
sliders to adjust the effects work in real time. In other words, you can 
change the depth of a reverb effect as you track is playing so that you 
hear the changes in real time.

The only thing that I can say negative about TKeca is that it doesn't 
monitor the effects as you record. But when you stop recording and play 
a track, you can add up to four effects per track.

TKeca also has a "roll-in" mechanism. It allows for specified amount of 
seconds of playing before the actual recording starts. Another words, If 
you record (for instance) the vocal part that goes "My big legged momma 
loves a man that aint me" and you sang out of key on the word "momma". 
Well you don't want to just press record over the part that needs 
fixing, because it's too abrupt. Most people would like to hear the 
parts that leads into the spot that needs recording over. It will 
"pre-roll" into the specified part you want to record over. And when it 
gets to that part, will start recording at that point only.

On the last version I tried, the roll-in mechanism worked, but you had 
to start from the beginning of the song. I hope that was fixed in the 
latest version.

I only used some basic parts of the (new) version of TKeca, so I can't 
vouch for it completly. But I would be glad to be of assistance if you 
want to install both of those programs.

This email is getting long. But I wanted to share some of my experiences 
with muse. It not only has several "built in" soft synths, but two 
(perticulary) will provide a complete set of sound. There's actually two 
soft synths in the "first" catagory of soft synths. I'm talking about 
soft synths that use sound fonts. They are "fluidsynth" and "Fluid". I 
don't know if these programs are related, but I did a "side-by-side" 
comparison of the sound of both of them and found that "fluid" sounded 
much better than "Fluidsynth". And if you download the sound font 
"FluidR3 GM.SF2" from the link below, I think you will be happy with 
what you find. It is 141 Megs but it's worth every bit of download time 
because they are "sterio panned" "real" instruments. The "end users" 
don't even have to worry about doing any panning. I put a couple of 
intruments on a few tracks and they were panned beautifully together. I 
really like this font.

The other advantage of the softsynth "Fluid" (instead of "Fluidsynth") 
is that once you load a soundfont into "Fluid" and then save muse's 
configuration, the soundfont will load automatically into "Fluid" when 
you start muse again. I could'nt get "Fluidsynth" to do this. I had to 
reload the font every time I started the program. And if you want to 
make this easy on the end user, you may want to use the "Fluid" soft synth.

The second catagoy of softsynths that come with it are the buzzy 
"synthy" sounds that are electronically produced. In fact the "vam" 
softsynth has an interface for some pretty good control over the sound. 
And not only can you move sliders in real time to get all kinds of cool 
"buzzy" sounds while your playing, but I even was able to use the 
sliders on some already pre-recorded midi tracks as they were playing.

So you have real intruments and you have buzzy (synth sounding) intruments.

I get the impression that the "live recoding" capababilities (using a 
microphone and "real" instrument as opposed to midi recording) isn't 
very good at this time with muse. And that's why I suggested TKeca. And 
if you can find a midi recorder that syncs with jack_connect, then you 
have the best of both worlds.

Actaully, here's a bunch of links...

   The "FluidR3 GM.SF2" sound font...
http://inanna.ecs.soton.ac.uk/~swh/fluid-unpacked/

   ecasound home page...
http://www.eca.cx/welcome.html

   TKeca download page...
http://sourceforge.net/project/showfiles.php?group_id=63518

   While we're on the subject of programs built on top of ecasound. This 
next one is an "effects box" (my terminoligy) that uses all the 
functionality of ecasound. It's called Ecamegapedal...
http://www.wakkanet.fi/~kaiv/ecamegapedal/

I think these are good programs. And if you have the patience, if you 
get jack running well, it opens up the possibility of using jack_connect 
to sycronize programs. I (think) that there's (not) a lot out there that 
takes advantage of it. One of my biggest hopes was to install the latest 
version of the "Hydrogen" drum machine because the version that came 
with SuSE 9.0 doesn't have the jack transport functions (for 
syncronizing with other programs). I spent a whole night satisfying all 
the dependencies. It built without a hitch, but it just wont start up 
properly. I'm now using Hydrogen 0.8.0-80 (that came with SuSE 9.0 
distribution) and I really like it. Maybe I just don't have a lot of 
experience with linux drum machines, but these are the first good drum 
samples that I've heard. I've never been satisfied with any drum sample 
that I've heard prior to using Hydrogen. And you can get "drum kits" 
online (I haven't downloaded any yet). But Hydrogen is not only easy to 
use, but the manual is easy to understand also.

You know, I really avoided jack because it just added another layer of 
problems. But I connected Hydrogen up to it and the latency (the amount 
of time from pressing the keyboard to the time the sound came out) 
disappeared. And I was using the lowest buffer size when (not) hooked up 
the jack and couldn't get good response time until I hooked it up to 
jack. In fact I get immediate responce time.

There's a program called qjackconnect that allows for a point and click 
connecting of programs and devices when using jack so you don't have to 
go into the shell and type cryptic command to hook up programs and 
devices (and then not hook up correcty after all the effort). And since 
I'm speaking of connections. The one that hooks up things like your midi 
keyboard to synths (and things like that) is called "kaconnect". What I 
like about it, is that it not only shows (in the windows) what there is 
to connect together but (for me) it makes it easy to understand how all 
these cryptic things actually work. And (I'm assuming) it will make 
"aconnect" (command line program that kaconnect runs) easier to 
understand when it comes to adding it into scripts so that these devices 
will automatically connect apon starting up programs. Prior to 
kaconnect, I couldn't visualize what a 65:0 or a 192:0 was (or why there 
was a 65:0, 65:1, 65:2 and... 65:3). Both kaconnect and qjackconnect 
puts them in front of you in two boxes ("from" ports and "to" ports). 
You just connect them with mouse clicks.

I'm becoming a bit long winded, but what I'm trying to say is, you could 
get along well without jack, but I'm just wondering if (after a while) 
it will become difficult to (not) use jack. I know that some of the 
newer versions of programs I've seen are building as "jack only" apps. 
So I finally gave in and did the research nessesary (and it was painful)

I was wondering if this would be a good time for us to use this 
situation to get this gentlman up and running with jack. Maybe we can 
all just "push thru" and get him set up. That way we can have a thurough 
conversation about setting up jack. I'm sure I'll learn a lot also. It's 
not that jack is that complex. It's just finding the information 
nessesary. You know... "were is the information... is it relivant to 
what I'm doing... what is this other aspect they are talking about... do 
I even need it... where can I find that information..."

Rocco




More information about the Linux-audio-user mailing list