What's new

Playing on Mari Samuelsen W/ SWAM Violin

Sorry mate. The fact that I have now played SWAM instruments on a freakin’ iPhone and a 24 key Roli Seaboard for an entire day has made me realize one thing: physical modeling IS the future ;)

"Physical modeling IS the future"

Yes, yes, yes.....

Full orchestras where every single player is at our full command awaits us in the near future.

There is the problem of the EULA. Since a real violin has no EULA, I believe the government needs to create a law that says that physically modeled instruments can be used just like a real one, meaning no EULA restricting you from creating your own sample labrary using the virtual instrument.

It will all happen anyway, as eventually there will be free open source physical instruments out there.

:)
 
Full orchestras where every single player is at our full command awaits us in the near future.
There's a difference between instruments and players.

It's a good thing physical modeling doesn't require any actual skill to coax out a musical performance.

Since a real violin has no EULA, I believe the government needs to create a law that says that physically modeled instruments can be used just like a real one, meaning no EULA restricting you from creating your own sample library using the virtual instrument.
All of the synthiness of PM along with the lack of flexibility of pre-recorded samples? Count me in! ;)

It will all happen anyway, as eventually there will be free open source physical instruments out there.
There have been free, open source physically modeled instruments available for years.
 
There's a difference between instruments and players.

It's a good thing physical modeling doesn't require any actual skill to coax out a musical performance.


All of the synthiness of PM along with the lack of flexibility of pre-recorded samples? Count me in! ;)


There have been free, open source physically modeled instruments available for years.

Hello David,

Yes, thanks for this.

I should have said instruments, not players, as virtual AI type players are perhaps decades away. :)

Do you view the SWAM demo video in this thread to have Synthiness to it? I am so blown away by the SWAM demo that I cannot tell the difference between it and the real recording.

When you say that "There have been free, open source physically modeled instruments available for years.", I am assuming that none of these are close to the near perfection of SWAM violins?
 
Do you view the SWAM demo video in this thread to have Synthiness to it? I am so blown away by the SWAM demo that I cannot tell the difference between it and the real recording.
Sorry, I was being silly and snarky, and I guess that didn't translate into text well. :P

The value of physically modeled instruments is that they're expressive. They generally sound quite good, but if you're going to sample an instrument, you'd want to sample the real thing, no? After all, in the process of sampling, you the flexibility of physical modeling.

But so there's no misunderstanding, I've got a SWAM and AM instruments, and love playing them.

When you say that "There have been free, open source physically modeled instruments available for years.", I am assuming that none of these are close to the near perfection of SWAM violins?
Yes, but... there's so much to unpack here.

"Perfection" is a strong word. There's certainly room for improvement in the sound of the SWAM violins, especially when compared to the Samplemodeling strings. Both instruments have their strengths and weaknesses. Neither is perfect, but I certainly sound much, much better playing either of them than I do playing on my real violin. :cool:

You mention "free and open source" almost as if all a developer has to do is type in the equation for how a violin makes sound, and wonderful instrument will result.

I believe the underlying technology SWAM's physical modeling is waveguide synthesis, and that's been around for a while. It's not people are hiding the technology from the rest of the world.


Yamaha used it in their VL-1 synthesizer. Lots of people have released VSTis that use this technique. Chet Singer did some really cool (and free) string instruments for Reaktor.

But physical modeling is often an extreme simplification of reality. Creating a playable physically modeled instrument is more art than science. I've written my share of free and open source software, but if I coded a physically modeled violin, I'd probably want something back for the effort.
 
Last edited:
"Perfection" is a strong word. There's certainly room for improvement in the sound of the SWAM violins, especially when compared to the Samplemodeling strings.

So what I don't get here is that if the SWAM demo from this tread sounds indistinguishable from a real violin, then how can there be need for improvement, or how can the Samplemodeling version sound better if the SWAM is already at an indistinguishable from reality level? Obviously when I use SWAM it sounds no where near as good as when Luigi plays it, but that is not the SWAM software's fault.
 
So what I don't get here is that if the SWAM demo from this tread sounds indistinguishable from a real violin, then how can there be need for improvement, or how can the Samplemodeling version sound better if the SWAM is already at an indistinguishable from reality level?
Not everyone will agree that:
  • The SWAM demos from this thread is "indistinguishable" from a real violin.
  • The sound of the SWAM violin is the sound they want their violin to sound like.
  • The SWAM violin has every articulation they'd ever want.
But you're not obliged to share their opinions. :cool:

The only thing that's really important is that SWAM works for you with the music that you're making. Making music is far more productive than arguing with people on the internet.

Obligitory xkcd reference:

duty_calls.png
 
Hello again David,

I played just the audio from the above demo to various clever music people I work with and they could not tell the difference.

I obviously get that some or lots will disagree, but what about yourself. Do you hear a difference?
 
I obviously get that some or lots will disagree, but what about yourself. Do you hear a difference?
Caveat: I'm just computer programmer who happens to have an interest in music as a hobby. There are many people of this forum much better qualified to give their opinion.

It's a bit hard to watch and listen to the video, because the two don't quite sync up - there are visual differences in the timing performance, and the audio is clearly in a different space - both how it's miked, as well as the room it's in.

That's not meant as a criticism, because the performance that Luigi gets is very good.

Closing my eyes and listening: the cues that indicated this is a "real" instrument are there - especially the scratch of the bow attacks, as well as the ringing of the strings when the bow lifts off. The feeling that this is being performed on strings comes through, but the resonating wood chamber? Not so much. The sustained tone is slightly synthetic, but that might be the reverb or just expectation bias.

If I heard the performance without it being labeled as a VI violin, it's highly unlikely that I'd think anything other than it was a violin - and performed well, at that.

But there have been plenty of VIs that have fooled my ear, so what does that mean?

I've got both the SWAM and Samplemodeling strings. No one's complained about the sound of the SWAM violin.

But when I've done an A/B comparison for people, the Samplemodeling violin always gets selected as the better sounding of the two. And once I've played one versus the other, people can hear right away the difference between the two instruments.

So as good as the SWAM strings are, there's certainly room for improvement.
 
Last edited:
Top Bottom