Discussion:
[jOrgan-user] "Sound quality" comparison FluidSynth versus GO.
Erik De Schrijver
2017-06-18 15:20:02 UTC
Permalink
The expression "Sound quality" is used here (for lack of a better term)
to make a subjective judgment of how close to real pipe sounds the
reproduction of sampled or synthetized organ sounds comes using sample
players.

The current forum discussion focuses (among others) on fluidsynth using
sf2 soundfonts versus GrandOrgue using wav files.
To get some feel about how both sounds i made (with very little effort)
the comparison between a sf2 soundfont played with fluidsynth, and the
wav files used to make this sf2 file using GrandOrgue. I used for this
the excellent versions of the Balzan organ i have available, namely
Balzan_v1.1_jO3.20 for jOrgan and Balzan_Wet_v1.1_GO_0311726, and let
them play the midi file G.deMacque-Conzonance.mid .
The sf2 file has samples for every note, and this is also the case for
the GO sample set, which has 3 releases per note. I think that the GO
samples have been used to make the sf2 soundfont; perhaps Mark Bugeja
would like to confirm or deny this.

If my assumption is correct, then differences one could hear are
primarily due to the performance of the playback engines, respectively
fluidsynth and GO.

If you happen to have both jOrgan and GO running, you can do the same
comparison with very little effort.
Probably your ears are better than mine. The more ears do the comparison
the better to level off individual preferences.

I would be interested to read what your findings are; not to bias
anybody i shall keep my observations for myself until hopefully more
ears have made their own appreciation.

All the best.

Erik.
Dr Mark Bugeja MD
2017-06-18 17:39:22 UTC
Permalink
Well, to answer Erik, I will allow Panos to speak from beyond. I traced an
email from him (Nov 2013) addressed to me, Robert (restorer) and Hugo (local
professional organist) and this excerpt from it would be the relevant reply:

"I managed to go down in memory usage of just 266MB only, while in the
meantime keep ALL PIPES and sounds/fx ie FULL compass ranks and so
preserving the historic idiom and unique character of the instrument.

Main rule was to be around 2.5 sec of each pipe (GrandOrgue version will
have full 9sec of each, but will raise the memory to 3GB minimum..). PITA
was to loop in such a way so the beatings of the pipes (where bein, like in
mixture) remain with the same ..."tempo" ect ect

There are switches for 434, 440 and 415 Hz and another for Valotti/Equal,
plus more.

I'm extremely satisfied with the result. There may be purists out there
after release, sayin that some pipes sound brighter some duller, or louder
less loud ect ect BUT this is original, at least as captured by Mark, who by
the way did an excellent pro recording.

As you both know the instrument better than me, you must verify this, so if
it is not correct I can repair (Brighter>duller ect)."

Mark



--
View this message in context: http://jorgan.999862.n4.nabble.com/Sound-quality-comparison-FluidSynth-versus-GO-tp4665213p4665215.html
Sent from the jOrgan - User mailing list archive at Nabble.com.
John Reimer
2017-06-18 22:15:44 UTC
Permalink
Post by Erik De Schrijver
i made (with very little effort)
the comparison between a sf2 soundfont played with fluidsynth, and the
wav files used to make this sf2 file using GrandOrgue.
If you happen to have both jOrgan and GO running, you can do the same
comparison with very little effort.
Eric,

I think that's a great idea. It has encouraged me to install GO on my organ
console laptop. I already have an old version of GO on my other laptop,
which is the one I do all my organ development work on. I didn't realise
that GO also has a recorder built-in. One consideration in making the sort
of comparison you suggest is that we arrange the jOrgan version to have a
similar reverberation time to what is apparent when playing the GO version,
and of course setting it up WITHOUT the Fluidsynth reverb but using one of
the excellent software programs now available. Presumably many of us are
already using one.

John Reimer




--
View this message in context: http://jorgan.999862.n4.nabble.com/Sound-quality-comparison-FluidSynth-versus-GO-tp4665213p4665216.html
Sent from the jOrgan - User mailing list archive at Nabble.com.
g***@gmail.com
2017-06-19 02:15:49 UTC
Permalink
I am reading this thread with interest.

Firstly I must declare that my work to date with jOrgan has been to learn
and emulate the full console and relay functions of the local pipe organ
instrument with which I am closely involved. The sound of the virtual
instrument has been secondary to the function. In saying this, I have made
many adjustments to the Compton.SF2 soundfont developed by synthesis by the
late Bruce Miles to drag as much as I can from it.

I am now planning to take the Christie virtual organ project to the next
level which will require high quality recorded samples and will need a
sound engine to faithfully reproduce the samples. I am pondering my options
while trying to keep other constraints in mind.

I have only taken a quick look so far at Grandorgue. It would appear it
would handle the sound generation from samples very well, but does it have
the power of jOrgan for console and relay functionality, particularly for
Theatre Organ?

With both jOrgan and Grandorgue being open source software projects, would
it be feasible for the strengths of both packages to be brought together to
create what to me would seem like the perfect solution?

I realise that right now it is possible to use the two packages linked by
MIDI. What would it take to have a seemingly single software package, like
the incorporation of Fluidsynth in jOrgan, but Grandorgue and jOrgan.

Is there anyone here involved in the Grandorgue forum who would care to
offer some thoughts about this proposal?
Is there any perception in the Grandorgue community of a need for greater
flexibility and capability in building the virtual console?

My end goal is to be able to circulate a package that could be enjoyed by
as many people as possible. This would mean it would need to be easily
installed and provide tonally pleasing functionality straight away. For
example it could play a demo MIDI file on first opening.

I won't get on my soapbox just now about what could be done to jOrgan to
make it easier to install and get running.

The Sourceforge jOrgan site says there have been 90 downloads this week.
Where are all of these users? Did they get it working?
The Sourceforge Grandorgue site says 300 downloads this week.....

Regards
Rick
Post by John Reimer
Post by Erik De Schrijver
i made (with very little effort)
the comparison between a sf2 soundfont played with fluidsynth, and the
wav files used to make this sf2 file using GrandOrgue.
If you happen to have both jOrgan and GO running, you can do the same
comparison with very little effort.
Eric,
I think that's a great idea. It has encouraged me to install GO on my organ
console laptop. I already have an old version of GO on my other laptop,
which is the one I do all my organ development work on. I didn't realise
that GO also has a recorder built-in. One consideration in making the sort
of comparison you suggest is that we arrange the jOrgan version to have a
similar reverberation time to what is apparent when playing the GO version,
and of course setting it up WITHOUT the Fluidsynth reverb but using one of
the excellent software programs now available. Presumably many of us are
already using one.
John Reimer
--
http://jorgan.999862.n4.nabble.com/Sound-quality-comparison-FluidSynth-versus-GO-tp4665213p4665216.html
Sent from the jOrgan - User mailing list archive at Nabble.com.
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
jOrgan-user mailing list
https://lists.sourceforge.net/lists/listinfo/jorgan-user
Loading...