LinuxMusicians

Hello everyone.

I'm working on a new, simple open-source sampler for Linux. If you're interested in having a look, the website is here: http://www.johnnylee.org/jlsampler/.

Any feedback is welcomed.

Cool, one of the first real-time audio projects I've seen that is done with Go. How's the overhead and performance?

The performance is reasonable. Running on an i7 laptop with an Alesis iO2 express over USB I can run with 2 buffers of 64 frames without too much trouble. That's 2.7 ms latency at 48 kHz. I haven't run into any problems with polyphony yet. The real question is the automatic garbage collector, but so far it hasn't been a problem, and I'm planning to do some optimization in that regard when I have some time.

Interesting... but if I understand well, you seem to build a new format for samples. Would it be too hard to use the sfz format ? There are already a lot of exisitng libraries in this format.

It's true that the format is different, and I understand the desire for backward compatibility. The choice was driven by the desire for simplicity. There are several features that would be necessary to support sfz files that would require a substantially different architecture.

That being said, I'm willing to convert sfz format samples if it's possible.

JohnnyLee wrote:It's true that the format is different

Oh, please, not again

JohnnyLee wrote: The choice was driven by the desire for simplicity

isn't sfz simple enough?

This is an open source project, so feel free to download the source and have a go.

There may be other projects supporting sfz. Actually having a new (working) sampler is a good thing. After all, nobody complains about DrumGizmo sample format for instance, as long as it sounds good.

>After all, nobody complains about DrumGizmo sample format for instance, as long as it sounds good.[

I do
I like many drumgizmo ideas, but since it doesn't support sfz samples, I still use linuxsampler instead.

>There may be other projects supporting sfz

Where are they?

>Actually having a new (working) sampler is a good thing.

We already have some .

JohnnyLee wrote:This is an open source project, so feel free to download the source and have a go.

sorry, but I don't have time to hack just another sound brand new sampler now, and even if I had any, I'd better do something with lv2 version of linuxsampler, for example.

We already have specimen/petrifoo, drumgizmo, samplv1, Drmr/Hydrogen, Jackbeat (ok, last two have sequencers onboard, but anyway). No one supports sfz.

Of course we need a better sfz sampler, and I think that there is a project cooking actually. I won't say more, because the devs may not want fell some pressure about that. JLSampler is a new project : why not ?

I won't say more, because the devs may not want fell some pressure about that.

good new, everyone!

JLSampler is a new project : why not ?

Any new project is ok, but NIH never sleeps, we should remember that :]

Just look at this page and count unmaintained projects, or those which were released 6-7 or more years ago.

lazyklimm wrote: NIH never sleeps

What's NIH? http://en.wikipedia.org/wiki/NIH_%28disambiguation%29

Keep up the good work Johnny Lee!!!

falkTX wrote:It's not like I could help them by contributing or taking over the project...

I'm generally for cooperation and converging projects too, rather than fragmenting resources and user base and wasting previous efforts' fruits, though I also understand if someone wants to do something that's already been done for the educational experience (i.e. I'm reading up on implementing my own FFT, no chance it will beat fftw). Its good to share these pet projects too, but developers shouldn't be surprised if their new software doesn't shake the status-quo, or is met with some resistance as this one has. And they will hopefully consider the best way to use their work to benefit the community, and consider merging their efforts with established/abandoned projects. Thats my opinion.