The Song Fight Bulletin Board

Well, this sounds super cool!

http://arstechnica.co.uk/information-te ... -and-daws/

Google Chrome gains MIDI support, enables Web-based synths and DAWs
Over 30 years after its creation, MIDI gets a modern makeover.

Glory hallelujah!

I wouldn't call this a "modern makeover," but I would call it an interesting step towards cloud-based everything. Lack of MIDI was the last reasonable objection I had to the idea of trying to do everything on a Chromebook-like device.

I mean I still have no desire to give up local processing and high-bandwidth low-latency storage for any reasonably-sized project but this at least makes for an interesting case for cloud-based DAW software that'll make it easier for folks to get started out as musicians, and I think anything that makes music more accessible is a good thing.

Yeah, it seems like a good step toward something else. But not at all immediately useful. I don't see any advantage to working on a browser as a replacement for a DAW... but perhaps there are some interesting opportunities for remote collaboration in the future.

Nah, remote doesn't instantly benefit from it being web-based either, software has to be written with collaboration in mind either way. I just think that it's a great way of reducing the barrier to entry though - now all you need is a USB MIDI controller and a web browser and you can start making music with an online-hosted DAW (such as the one linked to from the article).

I think anyone who wants to get serious about music will still likely want to find a free/cheap local DAW (like Reaper or GarageBand/Logic) but this definitely makes it easier to get started. Plug in your M-Audio keyboard, open a web browser, and go. No need to install a DAW, find an instrument plugin, understand how to route MIDI -> plugin, etc. Granted a local DAW could be built that does the same thing but it seems to me like having the browser already take care of that abstraction layer for you makes building it easier too.

Actually, come to think of it, I have a bunch of experimental DAW workstation ideas I'd love to build but don't want to deal with the hassle of platform GUI/MIDI/audio frameworks. So this makes actually building the software I want to build way the heck easier too.

One potential nice thing about a Web DAW would be if they could move all the CPU/memory intensive work out into the server side. Then you could do beefy DAW work (e.g. run a zillion plugins) on a non-beefy device, like a cheap Chromebook or whatever. If not for writing code and making music I'd be just fine with a Chromebook that costs 1/10th what my computer did.

Yeah, exactly. Doing automatic offline/background track freezes, and WebRTC streaming for realtime changes. I don't think the latency would ever get low enough for direct manipulation of automation/fiddlyknobs to feel perfect but it could get pretty close, at least.

Although then the curmudgeon in me just thinks, why can't a local DAW do a better job of predictively rendering tracks?

It could pave the way for something like Google Docs, and how it turned boring word processing into a multi-user collaborative affair. Of course a DAW is way more resource intensive than a word processor, but I could see a MIDI-only DAW that could skirt a lot of latency and bandwidth issues.

MIDI-only meaning virtual synths and no audio tracks, or as in it has to control external, locally-present hardware? Because the former would be great, the latter not so much.

Personally I think the MIDI protocol itself needs to see some major changes, because it's ridiculously limited for a lot of things, and the various extensions to try to get around those limitations are limited and inconsistent across vendors. Like, per-key aftertouch, or per-note expression controls. Even every damn softsynth seems to have its own interpretation of how to do those things.

Like, MIDI does a great job of encapsulating about 98% of how classical pianos are played. The rest is just hacks. Of course, after 30 years of entrenchment and extension and kinda-sorta standardization, it's probably what we're stuck with for the foreseeable future.