VirtualMIDISynth 2.x - Release Candidate 2 released

Not related to 2.0, but can you add sflist support? I have 10-soundfont chains which are difficult to switch between, by changing soundfonts one-by-one.

Hi,

as you expect I wait for beta3 release.

To be shure living all errors behind me I do a freas install on a new SSD with:

WIN 10 build 11099 / Sonar Platinium / Asio4all / VMS2-beta3.

Hardware ASUS SabertoothX99 / i7-5930k / 32 GB Ram / ROG Poseidon GTR780 / Realtek® ALC1150 HD Audio (internal)

The phenomen now is: can't get work, that VMS plays with Asio together.

VMS alone plays midi notes, Asio alone plays audio. But routing midi to a syth that using Asio VMS stops working.

When I close SONAR and start it again I get "Error opening MIDI stream - Error code: 4"

What i done in the first step was looking in debug report, and saw that ->midi: wdmaud.drv and midi1:wdmaud.drv was not correct so I changed this in the registry (x86 and X64) -> no success.

I tryed out all possible combination of Sonar, Asio4all and VMS (changing to MME,WDM and back; Sample Rate; installing more then 1 VMS instanz; buffer sizr; resampling rate and so on) -> no success.

So now my hope is that you can help me with two fault registry keys:

*** HKCU\Software\Microsoft\Windows\CurrentVersion\Multimedia\MIDIMap ***
Error reading registry key
The key MIDIMap does not exist, what should there be ????
 
*** HKCU\Software\Microsoft\ActiveMovie\devenum\{4EFE2452-168A-11D1-BC76-00C04FB9453B}\Default MidiOut Device ***
Error reading registry key
The key Default MidiOutDevice does not exist, what should there be ????

Please help me to find a solution for my problem.

Thanks in advanced

CID

Many thanks for the quick reply. I'm using a Terratec TS22 PCI sound card. Checked its driver, and its up to date.

Attached is the Asio info dialog I get from the sound card's control panel. The samples/buffer can be increased, but that leads to higher latencies.

Does this info translate into a value for VMS's Additional Buffer, or do I still have to do some trial and error? :-)

KaleidonKep99 wrote:

I think you're using BASS_CONFIG_UPDATETHREADS for the autoupdating system, because I noticed it slows down when there are too much notes, instead of skipping them.
You should consider switching to BASS_ChannelUpdate or BASS_Update, to get a performance boost with Black MIDIs.

Automatic update is already disabled, each of the (up to) 4 devices has its own update thread that calls BASS_Update() independently.

KaleidonKep99 wrote:

Switching to BASS_ChannelUpdate (And setting BASS_CONFIG_UPDATETHREADS to 0) solved the problem.

Your driver is (I suppose) an in-process driver, like VMS1. It means that it runs in the same process as the MIDI client (the player), so the communication between them is lightning fast.
VMS2 has an out-of-process IPC channel to let the player and VirtualMIDISynth.exe communicate each other.

Even if it's really fast (as I can see) it will never be as fast as an in-process direct function call, ever.

One thing I could try in next version is to increase the IPC buffer between the processes, which is supposed to fill up when a BlackMIDI reaches its "burst points".

Thanks for your technical debugging anyway ;)

notator wrote:

Just thought I'd re-report that I'm still getting "VirtualMIDISynth #1 already in use" messages from Chrome, as originally reported in the previous thread.
Shall I file a bug with Chrome? Or is this something you want to solve in VMS?

The only thing I can do (and I've done) on my side is to allow chrome being blacklisted.

Open the Options tab of Configurator while Chrome is locking VirtualMIDISynth driver, then follow these steps:

Sadly this is an all-or-nothing setting, we can't allow Chrome to load VMS only once or force it to release it when needed...

No need to open a new bug in Chrome, just subscribe to the existing one: https://code.google.com/p/chromium/issues/detail?id=497573

notator wrote:

Is there any way VMS could detect stutters, and so find the best/lowest value for Additional Buffer on the system in use?

No, sadly I haven't found any.
BASS docs have some recommended values for its minimum value and they are already applied by VMS (note that the setting name is "Additional buffer", just because it only increases the internally calculated one).
The final "safe" value depends on a lot of things like OS, drivers versions, sound card buffer size, system speed, ...

notator wrote:

Attached is the Asio info dialog I get from the sound card's control panel. The samples/buffer can be increased, but that leads to higher latencies.
Does this info translate into a value for VMS's Additional Buffer, or do I still have to do some trial and error? :-)

That buffers are from the ASIO side of the driver, and could or could not be the same buffer used for the standard interface (non ASIO).
It seems to me that trial and error is the only way to go.