W3C home > Mailing lists > Public > public-audio@w3.org > April to June 2012

Re: MIDI web synth (was Re: MIDI proposal implemented)

From: Chris Wilson <cwilso@google.com>
Date: Wed, 13 Jun 2012 19:18:10 -0700
Message-ID: <CAJK2wqWKjbiD9OnSiVHYfByWbk6UYjmyhSq5DHHyhf+13aFtMQ@mail.gmail.com>
To: "Tom White (MMA)" <lists@midi.org>
Cc: public-audio@w3.org
Actually, it was easier to just deploy a synth skeleton I've been working
on.  http://midi-synth.appspot.com/ has a very fast attack time.  (no UI at
all though.  :)

On Wed, Jun 13, 2012 at 12:49 PM, Tom White (MMA) <lists@midi.org> wrote:

> **
> If the "loopback app" uses the same data path (up to the output) as the
> MIDI web synth app, then yes (but "select input" is not needed because the
> input should be the same)...
>
> The idea is to *only* change the output device, while keeping *all* of the
> other code unchanged.
>
> Thanks,
>
> - TW
>
>  ------------------------------
> *From:* Chris Wilson [mailto:cwilso@google.com]
> *Sent:* Wednesday, June 13, 2012 11:40 AM
> Tom, do you just want a loopback app (select input, select output)?  I can
> try to toss one together and post it.
>
> On Wed, Jun 13, 2012 at 10:02 AM, Tom White (MMA) <lists@midi.org> wrote:
>
>> **
>> I agree it is better, but there is still enough delay to make live
>> playing difficult... but perhaps that is the response time of the synth
>> itself, I don't know... if I could use the same interface to select an
>> external synth then I could isolate the software synth and determine the
>> delay from the API...
>>
>> - TW
>>
>>  ------------------------------
>> *From:* Tom White (MMA) [mailto:lists@midi.org]
>> *Sent:* Monday, June 11, 2012 4:01 PM
>> *To:* public-audio@w3.org
>> *Subject:* RE: MIDI web synth (was Re: MIDI proposal implemented)
>>
>>   I'll check it out sometime this week, and let you know if it performs
>> noticeably better for me...
>>
>> - TW
>>
>>  Incidentally, I converted this demo over to use the Jazz MIDI plugin
>> (instead of the Java-based MIDIBridge), and latency seems better to me
>> (startup time is unnoticeable now).  Unfortunately, I don't think I'll be
>> able to build a Web MIDI API shim on top of Jazz, because it only seems to
>> support short messages (i.e. no system exclusive, and possibly no system
>> real-time either).  But for note and continuous controller support, it's
>> much more responsive.
>>
>>
>> On Wed, Jun 6, 2012 at 11:34 AM, Tom White (MMA) <lists@midi.org> wrote:
>>
>>> **
>>> Chris,
>>>
>>> FWIW, I tested this on my Thinkpad T420s (and Chrome 19.0.1084.52 m) and
>>> there was obvious latency (compared to using the onscreen keys).
>>>
>>> (I didn't want Vilson's note to give you the sense that everything works
>>> perfectly in the world <g>).
>>>
>>> Tom White
>>> www.midi.org
>>>
>>>
>>>
>>>  ------------------------------
>>> *From:* Vilson Vieira [mailto:vilson@void.cc]
>>> *Sent:* Wednesday, June 06, 2012 9:10 AM
>>>  Chris, awesome, tested with a Novation Launchpad in Mac and it worked
>>> perfectly. I didn't perceived latency.
>>>
>>> Cheers.
>>>
>>>     On May 31, 2012, at 12:04 PM, Chris Wilson wrote:
>>>>>
>>>>> > I found a few spare moments to graft MIDIBridge on to Aike's Web
>>>>> Audio synth.  Try it out at http://midiwasynth.appspot.com/, check
>>>>> out the code at https://github.com/cwilso/webaudiosynth.  Note that
>>>>> MIDIBridge can take a few seconds to grab the MIDI device, so I put a popup
>>>>> dialog in to let you know it's ready to go.
>>>>>
>>>>
>
Received on Thursday, 14 June 2012 02:18:41 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 14 June 2012 02:18:46 GMT