cmASIO feature in Thetis

[w3ub]

Well, two it is!

Thanks Bryan!

Doug

[tony77]

Bryan
just today noted your comment re the ASIOblocknum key value and Saturn/G2 systems.
I've been running a G2 for several months now, and a couple of days ago made the switch from VAC to cmASIO. Brilliant, a great upgrade for those of us who like to be able to switch seamlessly from sideband to digital (mid-QSO as it were!). The work you've put into this is greatly appreciated!

Yesterday however during a long SSB session on 15m I noticed a couple of brief RX audio dropouts - audio just completely stopped for about 1-2 seconds, so now I'm wondering about your suggestion that the default ring-buffer setting of 5 may be too low for a G2? Can you expand on this at all?

My audio hardware revolves around a Steinberg UR22C interface...the Yamaha-Steinberg ASIO driver set up easily and seems to run perfectly, and the audio performance, ergonomics, and build quality of the UR22C sure works for me.

73,
Tony ZL3ES

Hi Doug,

Yes, the default ring-buffer size is 5 sample blocks. The ASIOblocknum registry key value is there for situations where 5 is not adequate. If you reduce it too low, you'll get obvious audio artifacts as audio samples are being lost.

73!

Bryan W4WMT

PS - I should add that the default of 5 was reached empirically over a year's stretch of testing by volunteers. This was done before the introduction of the Saturn board and the Anan G2 models. There is early evidence that, for the Saturn, the default of 5 may be too low,

[Michael]

It appears that I do have a similar problem as already reported here.

I am using a Behringer UMC202HD audio interface together with a Anan G2 and the latest version of Thetis.
cmASIO is providing green light after adding the UMC ASIO name to the registry. All set to 64 - 48000 but I do not get audio. Audio is only available through VAC.
I deleted all Voicemeeter software and drivers assuming that they cause some confusion but the issue remains. Apparently, I am one of the unlucky ones with this issue as it seems the same setting is working fine for others. Maybe I overlooked it but is there a solution?

I guess a solution would be to use speakers and a mic directly connected to the audio interface but currently cannot check as I miss the required adapter plugs. How have others solved the issue? Just staying with VAC and not using cmASIO? Any advise is appreciated.

73,
Michael, DF2RQ

[w-u-2-o]

Michael,

Please post screenshots of the output of hostsample.exe (just the Behringer part), the registry entries you made, and the settings in the Behringer control panel.

[Michael]

Hi Scott,

here it is:

UMC ASIO Driver
Drücken Sie eine beliebige Taste . . .

ASIO driver: [0] = UMC ASIO Driver
asioVersion: 2
driverVersion: 1337
Name: [UMC ASIO Driver]
ErrorMessage: No ASIO Driver Error
ASIOGetChannels (inputs: 2, outputs: 2);
ASIOGetBufferSize (min: 8, max: 2048, preferred: 64, granularity: -1);
ASIOGetSampleRate (sampleRate: 48000.000000);
ASIOOutputReady(); - Not supported
AsioCanReportOverload Not Supported
ASIOGetLatencies (input: 136, output: 168);
channel: 0 isInput: 1 isActive: 1 channelGroup: 0 type: 18 name: In 1
channel: 1 isInput: 1 isActive: 1 channelGroup: 0 type: 18 name: In 2
channel: 0 isInput: 0 isActive: 1 channelGroup: 0 type: 18 name: Out 1
channel: 1 isInput: 0 isActive: 1 channelGroup: 0 type: 18 name: Out 2

ASIO Driver started successfully.

[Michael]

UMC Control is showing:
UMC202HD 192k

Sample rate 48000hz

64 Sample - Safe Mode (not sure what Safe Mode is)

Driver version: 5.57.0
Release Build

[w-u-2-o]

So far, so good. What's in your registry?

Also, without Thetis running, please use Windows to monitor your Behringer mic input and let us know if you can hear it OK.

[Michael]

Without Thetis running the UMC control shows UMC inactive. It looks like that the device is being used whatever driver and VAC input/output I use (virtual cable, Windows driver and mic and speaker, etc..). With defined blocksizes of 64 all devices run smoothly. With Thetis and cmASIO active also the UMC controller shows UMC active. However, I only have audio (computer speakers) when VAC is on and when the computer speakers are selected.

Unfortunately I cannot use the Behringer input/output as I do not (yet) have the required plugs/adapter plugs.

Maybe it works as it is supposed to work and I wrongly assumed that audio should be available at the computer and VAC could be used for e.g. digital modes only.

[Michael]

Scott,

forgot to send you the registry entries. Here they are:

[img]

Screenshot%202024-11-10%20131012.png

[/img]

Screenshot 2024-11-10 131012.png (29.26 KiB) Viewed 18850 times

[w-u-2-o]

I guess a solution would be to use speakers and a mic directly connected to the audio interface but currently cannot check as I miss the required adapter plugs. How have others solved the issue? Just staying with VAC and not using cmASIO? Any advise is appreciated.

You know, I totally missed this part of your original post. But when you posted:

Unfortunately I cannot use the Behringer input/output as I do not (yet) have the required plugs/adapter plugs.

All became clear

It's a very simple problem: you need to run your mic and speakers via the UMC202. That's why nothing is working. Because cmASIO completely bypasses both VAC in Thetis and the audio I/O on the ANAN. Until you get a mic and speakers on the UMC202 you'll have to use a different method. More on this below...

Maybe it works as it is supposed to work and I wrongly assumed that audio should be available at the computer and VAC could be used for e.g. digital modes only.

Yes, that's the way it's supposed to work If you want to use cmASIO for both voice and digi modes it's easy to do with the Voicemeeter audio mixer software from VB Audio. Install Voicemeeter Potato, assign the UMC as output A1 and input A1, then assign cmASIO to use one of the Voicemeeter virtual ASIO channels (B1, B2 or B3). Digi mode software can go to one of the other virtual channels in the normal fashion. Then mix and route in Voicemeeter to your heart's content.

[Michael]

Thanks, Scott - yes - now it is much clearer. I wrongly assumed the UMC feeds back the signal to the speaker in the computer via the USB connection. I will try to to mix the signal with Voicemeeter (it is a powerful software but I find it very cryptic).

Thank you again for the support. Always appreciated!

73, Michael, DF2RQ

[GW4NNO]

Hello Keith and thanks for that info. I did try WSAPI some time ago but didn't realise I had to use loopback to get an input to show. However, I just use the VAC I/O for digital modes anyway so I think I'll leave my mic input as it is through the hardware line in on the back of the Anan 7000DLE mark 11.

[KC2QMA]

Hello about to go down the cmASIO path and have a few questions.

1. How does cmASIO handle multi I/O interfaces like an 8in 8out unit or others?
I know you only need a min 1 in 2 out interface but was just curious what can be done.
2. Is all that's needed it to edit the registry?
3. Is there a list of cmASIO supported interfaces or can you use basically any ASIO interface?
4. Will there be UI control of cmASIO added to Thetis setup window in the future?

Thanks
John

[w-u-2-o]

Hello about to go down the cmASIO path and have a few questions.

1. How does cmASIO handle multi I/O interfaces like an 8in 8out unit or others?
I know you only need a min 1 in 2 out interface but was just curious what can be done.

I believe it will only entertain the first two channels on a multi-channel interface.

Of course you can fix this by assigning cmASIO to something like Voicemeeter or VB Audio Matrix and patching/routing in that software to whatever hardware channels you want. I do this with Voicemeeter and my Presonus interface. I've measured it and Voicemeeter adds a measly 4 ms of latency.

2. Is all that's needed it to edit the registry?

Almost. You need to ensure your interface is running with a 64 sample buffer size and 48KHz sample rate. That size and rate are the only ones that cmASIO supports. You may also want to use Bryan's hostsample.exe program to positively identify the interface names.

3. Is there a list of cmASIO supported interfaces or can you use basically any ASIO interface?

If it supports 64 sample buffer size and 48KHz it should work.

4. Will there be UI control of cmASIO added to Thetis setup window in the future?

I believe that Richie plans this eventually.

It's easy to keep several registry keys available so you can just quickly rename them to switch interfaces or turn cmASIO off. For instance, right now the key ASIOdrivername in my registry is assigned to Voicemeeter, but I have another key sitting there called disabledASIOdrivername and that points to the Presonus. When I want to switch I just rename them and restart Thetis.

[W1AEX]

Just a quick note of an unexpected Thetis experience after I finally fixed my broken Windows Update on the station computer so that I could move from Windows 11 v21H2 to Windows 11 v23H2 yesterday. I suppose that there really are no unexpected experiences with windows updates as they can be quite an adventure of multiple rabbit holes. Anyway, cmASIO has worked flawlessly with Windows 11 v21H2 at the default settings, however, after the update I ran into audio dropouts on receive and transmit. Just random blips every second or so. If I opened up another program or a new webpage in a browser it would go off the rails for a few seconds. Initially, I played around with the Thetis DSP buffer settings but that had no effect. So... I updated my UMC202HD driver and that made no difference. I went to the cmASIO page and raised the Block Number from the default 5 to 6 and worked my way up to 10 blocks where things were better but still not perfect. I went up to 15 blocks and that was 90% clean. I jumped to 20 blocks and it has been 100% clean no matter what other applications I opened. Maybe that solved the problem? I still wonder if this is compensating for something else in the system that is no longer optimal.

So, I am just curious if anyone else has run into a similar experience with Windows 11 v23H2.

I am holding off on letting my station computer jump to v24H2 even though it started nagging me to do so right after the v23H3 update was completed. Same nagging push to v24H2 is coming up with all of the other the computers here. It looks like as of January 17, 2025 Microsoft has resolved most of the 24H2 compatibility issues (not sure if they fixed the voicemeeter issues) that prevented a big rollout so they are pushing it out widely at this time.

73, Rob W1AEX

[K1LSB]

Rob,

This is prolly a stupid question but did you double-check the audio bitrate (48kHz) of all relevant drivers after the Windows update?

Mark

[W1AEX]

Definitely not a stupid question Mark. I did check that the UMC driver was still set for 48k and it was. As far as I know that's the only thing in the path as my mic is plugged into the UMC202HD and the audio runs from there into a USB port of the computer for cmASIO to handle in Thetis. I might have missed something else though so I will trace it all out!

Thanks! Rob W1AEX

[ea3aqr]

Check in the registry that the throttling index tweak is correctly applied.

After a few Windows updates, it reverts to its default values.

[N1ZLE]

Hi Rob, just a though here but I've run into weird issues in the past with Microsoft updates replacing the USB Chipset drivers (Realtek in this case) with Microsoft (windows) versions. This caused similar issues to what you're describing. In my case re-installing the correct Motherboard (MB) chipset drivers solved the weird USB issues I was having.
Good luck!

73
Pat

[W1AEX]

Thank you all for taking the time to hit all the right spots. Mark, I went through and checked every device in the SOUND settings to make sure everything was set for 48k. Jordi, I hit the Network Throttle Index Tweak and corrected that setting that I had neglected. Pat, I downloaded all the latest ASUS ROG Maximus X Hero motherboard drivers and installed the whole package. Then I rebooted and set Thetis with the cmASIO default Block Number setting of 5...

...and everything is running great again!

I'm sure all those steps were necessary and I thank you all for the assist! I'm going to write all this down and keep the notes for the inevitable day that is coming soon when 24H2 is pushed all by itself!

73, Rob