i promised myself rosegarden

[pradagio]

i really like rosegarden (rg) though still quite new to it.
i've used musescore, then went to lilypond (emacs and frescobaldi).

i would appreciate knowing if my modus operandi is valid | workable | sensible, for composing musicals:

1. use rosegarden to enter midi from keyboard for all parts -> quantize to produce playable notation -> generate lilypond file for pdf.

2. enter audio for voice parts and sequence them into the existing midi (muting the midi voice), so we get a playable number.

i've tried recording audio through rg, but keep getting "out of processor power" error after a few seconds (possibly because rg doesn't want to hear my voice), so i gave audacity a shot and it didn't seem to mind - plan to give ocenaudio a try too. i can then import the wav files into rg.

is this a feasible plan for my purposes of creating musicals - particularly vocal scores?

[Basslint]

I think Rosegarden is not the best program for recording audio. You could:
1) Use Rosegarden (or Musescore) for the score, export the MIDI to a file and load it into Ardour, where you can easily do the audio recording
2) Use MusE to do everything, since it includes a score editor (which is however not as advanced as Rosegarden's) and better audio recording capabilities than Rosegarden (but not as good as Ardour's)

[pradagio]

thx basslint!
i see you've posted a fair bit about muse ... this thread being the most interesting to me:
viewtopic.php?f=61&t=23215

i will start with muse and see how it compares to rosegarden.
appreciate the tips too!

it also appears that the rosegarden audio recording out of processor power has nothing to do with audio recording. i just got the same error playing back a simple midi file. also, the cpu usage is mostly less than 5% and the mem is around 20% (i have several other programs open). so i'll take the matter up on the rosegarden mail list if it continues.

[folderol]

This is weird.
I've been using Rosegarden for years, and the only time I ever see that message is when something else is stealing CPU power.
Having said that I don't record audio directly into Rosegarden. Instead I usually record it into Audacity (while playing MIDI into Rosegarden. Then do basic cleanup of the audio, save it, then load it into Rosegarden, and line everything up.
Incidentally, Rosgarden is currently being very comprehensively worked on by some new(ish) devs, and a number of long-standing bugs have been fixed.

[pradagio]

@folderol
thxs for this info about rg.

i joined the rg list and have found only one thread with that error message and i didn't exactly relate.
being on archlinux, i should usually get the most recent updates so the issue may go away on it's own.
on the other hand, the problem may be at my end:
1. i am using the standard archlinux kernel - not a low latency
2. pipewire-jack (not pure jack)

neither you nor basslint use rosegarden to record audio, so i'll look at audacity too - ocenaudio is pretty good, but doesn't have the depth audacity does ... then again, i probably don't require such depth being a noob. also, some people aren't happy with the new privacy changes coming up with audacity:
Audacity Alternatives: It’s Time to Get a Better Audio Editor
though that was nearly a year ago and was followed a few days later with an apology:
Audacity Developers Apologize, Revise Controversial Privacy Policy

the latter article mentions tenacity which might be neat to try too. i have very little experience with audio recorders so this exploration is fun.

right now i'm just trying to find out exactly what rosegarden is. i thought it was a daw, but then i saw basslint's post:
"Rosegarden does not try to be a DAW, sadly. If they did, they would implement LV2 at the very least."

so i'm rethinking how it should be utilized - i guess as
1. a pure midi sequencer
2. a preliminary notation creator (fine tuning done in lilypond directly)

then go to some audio recorder and daw to finish.

[folderol]

At the risk of being burned as a heretic, nobody actually needs a DAW at all.
Yes, for many they are convenient, but at the same time they force you to work in their way.
Being somewhat a maverick, I want to do it my way, so I stitch just the parts I want together in the way I want them. My commonest setup is:
MIDI keyboard to Rosegarden.
Rosegarden to Yoshimi (creating new tracks in Rosegarden and loading instruments in the associated parts).
Yoshimi audio to Jack.
Jack to Audacity and Audio Out.

[Babarosa]

At the risk of being burned as a heretic, nobody actually needs a DAW at all.

I surely will not burn you. To my mind there is too much fuss about the technic and numbers.
Today you get cheap multitrack recorders including effects providing quality you dreamed of 40 years ago (I am 57). You can record your singing, instruments, mix them, apply effects and export without any computer.

[pradagio]

At the risk of being burned as a heretic, nobody actually needs a DAW at all.
Yes, for many they are convenient, but at the same time they force you to work in their way.
Being somewhat a maverick, I want to do it my way, so I stitch just the parts I want together in the way I want them. My commonest setup is:
MIDI keyboard to Rosegarden.
Rosegarden to Yoshimi (creating new tracks in Rosegarden and loading instruments in the associated parts).
Yoshimi audio to Jack.
Jack to Audacity and Audio Out.

@folderol
well i'm all for heresy ... prad is actually short for pradeviant.

i am trying to understand your setup - it is a completely new and intriguing idea to me (not that i understand how to use a proper daw, either, since i haven't tried one, yet).

step 1
midi -> rg (that's easy)
but

step 2
rg -> yoshimi, i don't understand.
yoshimi is a software audio synth, so it can create sounds as i see from their website (their demos sound fabulous!)
why though? i thought that's what i have fluidsynth for.

unless, the new tracks you create in rg are using "synth plugins" made through yoshimi (instead of using "general midi device" which would be fluidsynth)?

if so, then

step 3
yoshimi -> jack
would make sense to me (because it's actually going through rg still).

however, why would you

step 4
jack -> audacity
pipe things into audacity at this point instead of just to audio out?
oh unless you are adding more audio recordings ... in which case, audacity is sort of acting like a daw?

so the convenience of a daw would be that you can skip steps 3,4 (because you daw it) and go directly to audio out?

i appreciate these new possibilities and the guidance, since the only thing i'm so far familiar with is using lilypond to create notation and midi files.

[folderol]

taking it from your step 2
rg is sending MIDI to Yoshimi as a standalone module, not a plugin. The rg tracks can be set to whatever part in yoshi that you want, not necessarily in numerical order, and two tracks could be sent to the same part.

Also yoshi calculates the sound in real-time, it's not just playing samples, and you can make extensive changes to the controls while it is playing.

I send to both Audacity and the the main out, because I want to be able to hear it whether I'm actually recording or not

I don't generally multi-track in Audacity. Yoshimi combines all parts to the main L/R out, so a single pass catches the lot.

[pradagio]

taking it from your step 2
rg is sending MIDI to Yoshimi as a standalone module, not a plugin.

...

Also yoshi calculates the sound in real-time, it's not just playing samples, and you can make extensive changes to the controls while it is playing.

...

I don't generally multi-track in Audacity. Yoshimi combines all parts to the main L/R out, so a single pass catches the lot.

ok!! this is all really interesting.
i'll give your setup a serious try this week and let you know.
thx again!

[pradagio]

@folderol

i have yoshimi working with rg now! pretty nice! and lots more sounds than the soundfont i'm using.
it can apparently work as a plugin too, though not with rg because it is lv2

1. is it preferable to have something like this work as a plugin or 'plugged in' via jack (or pipewire)?

2. i have to go to yoshimi to set the instruments to match the segments - with a plugin, i presume i can stay in rg?

as i see it, reasons for using yoshimi instead of timidity|fluidsynth other than
a. fluidsynth crashes rg (for me anyway)
b. more sounds
c. as you wrote, generated in real time so you can make changes (haven't gotten this advanced, yet)

i am also trying linuxsampler as a plugin, but haven't figured out how to yet.

[folderol]

Don't forget you can change Yoshimi patches from Rosegarden with the standard program and bank change CCs - although you do need to know what you are looking for

[pradagio]

Don't forget you can change Yoshimi patches from Rosegarden with the standard program and bank change CCs - although you do need to know what you are looking for

yoshimi is excellent! many thx!!
i've also received comments from the rg-users list that strongly suggest not using plugins - it seems i'm not the only one who hasn't been able to get them to work.

i was having a terrible keyboard syncing with the metronome with timidity (not fluidsynth though), but there are no issues at all with yoshimi!

i have yoshimi working both as ALSA preferred or JACK preferred.
both seem to work fine, though i have to make an additional connection with JACK whereas everything just works with ALSA preferred (i'm using pipewire).

EDIT: then ted felix from the rg-user list reminded me that JACK is for audio not midi!

so i will just stick to ALSA preferred in the settings for yoshimi.

[Michael Willis]

i am also trying linuxsampler as a plugin, but haven't figured out how to yet.

Running linuxsampler as a plugin is kind of trouble, you might have better luck with sfizz if you're wanting to use SFZ-format instruments.

[pradagio]

thx michael, but i couldn't get it to come up in rg.
the link shows that it is lv2, but rg doesn't support lv2, so may be that's why.

@folderol
i have things working now pretty nicely between yoshimi and rg.

i downloaded this yoshimi.rgd after reading this thread which addressed an issue i was having with rg resetting anything i did in yoshimi to whatever it felt like (well that's what it felt like):
Rosegarden - Yoshimi issue

i put the yoshimi.rgd into
~/.local/share/rosegarden/library/
where there are a huge number of rgd files for various instruments and synthesizers (too i guess).

then in rg, i went into
Studio > Manage MIDI Devices
where i
1. deleted the default general midi bank (though it keeps coming back, hence i have to open a template), because i don't really need it and never could do anything with it either
2. created a new device (click New in topleft box), then click Bank, then click Import (bottom left box).

the import button gave me access to the yoshimi.rgd file and eventually i had the banks and instruments showing up in the instrument parameters box from where i could assign them to various tracks.

i found though that i had to set the channel dropdown to 'fixed' or all the tracks would change to the same thing when i set a single track. every instrument i set for an rg midi track, is reflected in yoshimi too! (however, that is true only within the bank that is chosen on yoshimi if the instrument number exists - i can't seem to change the bank in yoshimi from rg.)

i midied out the melody, and a piano accompaniment for the first 4 bars of twinkle, twinkle little star - hey, jane taylor and mozart (even though he didn't really write the melody) would be pleased. then i recorded audio in rg, singing out both (though not simultaneously) a tenor and baritone part - neither jane nor wolfgang would have been pleased, but rg didn't crash!! (may be because yoshimi was doing the midi heavy lifting instead of timidity|fluidsynth?)

so, assuming this is the a|the correct way to do these things, all is well indeed!