cmASIO feature in Thetis

[w-u-2-o]

Scott,

If we were to experiment with the ASIOblocknum value, how would we know if/when we're getting too small?

Thanks,
Mark

It will be plainly evident to your ears. Sound quality will become poor if the value is set too small. On powerful computers it's conceivable the minimum setting of 2 is achievable. I could only get it to run well at 3 here. It has been stable for days on that setting for me.

BREAK

This begs the question: have you measured latency of cmASIO, Bryan? I'd guess that it's a hair better than a well optimized VAC-ASIO setup?

[w-u-2-o]

I thought I'd give this a try and it worked straight away using my UMC2020. No problems noted. My only question is what the purpose of this is? Is it simply to eliminate having to use VAC to get the benefits of ASIO drivers (not to belittle that time saver!) or is there something else that I don't readily see? I notice that latency seems to be the same with this method as I saw with VAC.

Thanks,

Joe W4WT

The advantage is the ability to monitor RX2 along with RX1 through the ASIO sound device. VAC can now be dedicated to digital modes.

One has to look way back to the original, pre-Apache, Hermes days to understand the designer's intent and why the openHPSDR audio architecture wound up the way it did.

Way back when the designers were concerned that the audio streams had to remain in the same clock domain as the ADC. This is, of course, not at all required. There appears also to have been a secondary ergonomic consideration in that the designers desired a box that looked and felt more like a conventional radio whereby all audio connections were made to the box and not the computer. With both of those ideas firmly in mind, the software (originally a variant of PowerSDR) was designed with an emphasis on the audio stream that went to the audio CODEC chip in the box. VAC was only seen as a method for digi mode software and voice keying. The design was therefore optimized for that, with a separate VAC interface for each receiver.

As the popularity of fully virtualized audio (audio routed through computer interfaces) increased, the shortcomings of the VAC architecture for that became more apparent. This is even more true with the advent of openHPSDR derivative designs that are "black boxes", for example the Hermes Lite designs.

[G3ZQH]

Now I'm progressing (thanks to Ernst and Scott)! I have SSB PC audio (headset) using these cmASIO settings for Voicemeeter but I cannot connect to JTDX or WSJT-X for digimodes. Will explore further tomorrow.

I'll be very happy to have a more stable audio link since I have always had problems with running VAC in Thetis.

73 Dave

Assign your digi software to one of the other virtual channels in Voicemeeter. Then click the appropriate "send" buttons in Voicemeeter to make the connections.

For example, if Thetis is on B1 via cmASIO, then put JTDX on B2. After that, activate the B2 send on B1, and the B1 send on B2.

Still having problems here with cmASIO implementation on my 200D. I thought I had it working but this was false- only the PC mic audio came through but no RX audio. Also not able to connect with digimode software. In 2.10.3.5 pre3 I cannot get VAC1 to function at all in my set up. I suspect there is something not right with my ASIO. Back to 2.10.3.3 and VAC1 is working fine and I have digimodes again. For now, I'll sit this one out and wait for further details.

73 Dave G3ZQH

[Tony EI7BMB]

No success here either. I'll stick with VAC for now and see how it develops, was fun trying

[oe3ide]

No success here either. I'll stick with VAC for now and see how it develops, was fun trying

Also worth mentioning that it only works with 64 buffersize in the asio driver (also thanks to Richie for this info)

73 Ernst

[w-u-2-o]

Also worth mentioning that it only works with 64 buffersize in the asio driver (also thanks to Richie for this info)

73 Ernst

Oh wow. So I just got lucky. I just happened to have Int32 and buffer size 64 set in my Presonus/Voicemeeter config. I would have been very frustrated otherwise!

ASIO buffersize should follow the driver setting automatically like all other audio software does. Having it fixed at 64 (or having it be manually adjustable) is not in accordance with how all other audio software works.

[Tony EI7BMB]

Oh thanks Ernst. I'll give it another go

No success here either. I'll stick with VAC for now and see how it develops, was fun trying

Also worth mentioning that it only works with 64 buffersize in the asio driver (also thanks to Richie for this info)

73 Ernst

[Tony EI7BMB]

Still no sound appearing in Voicemeeter . I think the regedit is ok in so far as the entry for Voicemeeter Virtual ASIO is no longer present in the drop down list on VAC1 setup screen in Thetis. Ernst may I ask where you select 64 in the asio driver ? I have 64 selected as buffer size on my MOTU sound card

[oe3ide]

Still no sound appearing in Voicemeeter . I think the regedit is ok in so far as the entry for Voicemeeter Virtual ASIO is no longer present in the drop down list on VAC1 setup screen in Thetis. Ernst may I ask where you select 64 in the asio driver ? I have 64 selected as buffer size on my MOTU sound card

Yes, buffersize should be set in the soundcard driver. In VoiceMeeter you need "default" for the Asio buffersize... should then display in Voicemmeter buf:64 on the A1 main output device.

[Joe-W4WT]

Thanks Scott & W4ATL for the info. I wasn't aware that this enabled RX2 to run through ASIO also now. Nice.

I assume, since we don't have access to any settings, per se, that under/overflows are now a thing of the past? No more "robotic voice" problems as was the case when we had runaways on over/underflows?

I also noted my UMC status panel shows that it automatically set 64 samples under "buffer mode"; it was set to 8 with VAC so that was nice since I didn't know it needed to be set to 64.

Everything seems to be working perfectly here. Nice work! I may even break out VMP and give it a whirl again. I have been using my 7300 for digital for quite a while but maybe now I'll go back to the Anan for digital. We'll give it a go and see how it works.

Thanks to all of our great volunteer "coders" who provide us a "radio" that just continues to get better and better! Ham friends who come over and listen to this Anan 7000 DLE II are just amazed at what they hear and leave telling me they "want one of those!"

It looks like this cmASIO work is another great advancement.

I wonder if this might solve the audio problems the 7000 DLE Mk III people are having?

Thanks,

Joe W4WT

[Tony EI7BMB]

Got it working now thanks Ernst. I've been using digital modes so long now that I completely forgot about "Master AF "

Still no sound appearing in Voicemeeter . I think the regedit is ok in so far as the entry for Voicemeeter Virtual ASIO is no longer present in the drop down list on VAC1 setup screen in Thetis. Ernst may I ask where you select 64 in the asio driver ? I have 64 selected as buffer size on my MOTU sound card

Yes, buffersize should be set in the soundcard driver. In VoiceMeeter you need "default" for the Asio buffersize... should then display in Voicemmeter buf:64 on the A1 main output device.

[W4WMT]

I also noted my UMC status panel shows that it automatically set 64 samples under "buffer mode"; it was set to 8 with VAC so that was nice since I didn't know it needed to be set to 64.

I may even break out VMP and give it a whirl again.

I have been using my 7300 for digital for quite a while but maybe now I'll go back to the Anan for digital.

Hi Joe,

It's not surprising that your UMC automatically set itself to 64 sample buffers when paired with cmASIO. That is standard ASIO SDK compliant behavior. In fact, the only ASIO drivers I have ever seen that don't do this are the Voicemeeter drivers.

As far as giving VMP another whirl, I hope you won't. cmASIO was designed expressly for hardware device drivers, not virtual device drivers. If you're looking for good performance, VMP brings nothing to the table, and greatly burdens the thread that UMC and cmASIO share. And it's a hairy thread indeed!

I hope you do try going back to the Anan for digital. I'm sure you mean to use VAC1 and/or VAC2 for digital. cmASIO is, of course, unsuitable for digital modes. In fact, its primary goal was to act as a permanent replacement for the audio codec chip in the radio, which frees the VACs to be used as dedicated digital mode interfaces (which is what they were designed for in the first place).

73,
Bryan

[w-u-2-o]

I wonder if this might solve the audio problems the 7000 DLE Mk III people are having?

What problems would those be? Nobody is reporting any in the forum.

[Joe-W4WT]

"What problems would those be? Nobody is reporting any in the forum."

Ones like these? I read through numerous problems with buffer over/underruns with the MK III here that were difficult to tame on protocol 2.

viewtopic.php?f=32&t=4811&p=27680&hilit=mk+iii#p27680

[w-u-2-o]

"What problems would those be? Nobody is reporting any in the forum."

Ones like these? I read through numerous problems with buffer over/underruns with the MK III here that were difficult to tame on protocol 2.

viewtopic.php?f=32&t=4811&p=27680&hilit=mk+iii#p27680

There doesn't appear to be anything in that topic that has anything to do audio problems. Everything in that thread is related to RFI.

[Joe-W4WT]

Perhaps I misinterpreted what I read over the course of the last few months or so that the MK III has been available. If it is all RFI issues, then obviously the MK III is just fine.

Joe W4WT

[Tony EI7BMB]

Just noticed the CM symbol at the bottom of Thetis to the left of RX ant. Nice touch

[m0cke]

Hi All

I'm trying to get cmASIO working using Voicemeeter Banana for RX audio so I can use the VAC for IQ to SDC Skimmer

I've used hostsample to check the driver names :-

Voicemeeter AUX Virtual ASIO
Voicemeeter Insert Virtual ASIO
Voicemeeter Virtual ASIO

I then used regedit and insterted the string value ASIOdrivername Voicemeeter Virtual ASIO

I rebooted he PC to be sure

When I start Thetis v2.10.3.5 Pre-Release 4 I get the cm icon on the bottom staus bar of Thetis but it's red, hovering the mouse over this shows the message Issue starting CMASIO. Check driver name in registry

I dont know where to go from here as the name is correct?

Any pointers please?

TIA

Jim M0CKE

[W4WMT]

Hi Jim,

Do you have DebugView from sysinternals.microsoft.com? If so, you can use it to see debug output from Thetis while it is starting which will show what went wrong with cmASIO.

Otherwise, there's a couple of Voicemeeter pitfalls you can check.
Voicemeeter's virtual ASIO driver buffer size should be set to default.
It's sample format type should be set to Int32LSB.
Your actual hardware device driver, you didn't say what you are using, buffer size must be set to 64 and sample rate set to 48000.
Voicemeeter sample rate should follow the hardware driver to 48000, but if it doesn't you may have to force it.

I recommend you first get cmASIO working with your actual hardware device driver (UMC ASIO Driver, Studio 192 ASIO) whatever that may be. Once you've populated the registry key value with its name, then the rest should be completely automatic. Then having assured yourself that everything is working with that configuration, you'll have a better starting place with the Voicemeeter effort. Or you may decide it works fine that way, and you don't really need Voicemeeter after all.

73,
Bryan

[KC2QMA]

So if I understand this correct, the cm ASIO now allows you to assign the mic input in Thetis directly to an ASIO audio interface so you don't have to use VAC ?
Can someone post the Notes about this new feature and what it does in this thread?

[W4WMT]

If you have installed pre-release-4, the instructions for cmASIO are in the Thetis folder.

[oe3ide]

So if I understand this correct, the cm ASIO now allows you to assign the mic input in Thetis directly to an ASIO audio interface so you don't have to use VAC ?
Can someone post the Notes about this new feature and what it does in this thread?

... infos on first post of this thread...

[KC2QMA]

Got It
https://community.apache-labs.com/download/file.php?id=3984

[K1LSB]

The CM-ASIO Guide speaks of a LockMode which "is entered by setting any of the highorder bits in the ASIOblocknum value".

I'm not really clear on what are "highorder bits" in the ASIOblocknum value so here's a snapshot of my Registry entries:

Capture.jpg (30.91 KiB) Viewed 25995 times

Does that snapshot indicate that LockMode is enabled?

TIA,
Mark

[W4WMT]

Yes it does. The lockMode will be enabled.

[W4WMT]

Got It
https://community.apache-labs.com/download/file.php?id=3984

That's an early version of the instructions.
See the Thetis folder after installing pre-release-4 for a pdf of the most recent version

[K1LSB]

Yes it does. The lockMode will be enabled.

Thanks Bryan!

Mark

[m0cke]

Hi Jim,
Your actual hardware device driver, you didn't say what you are using, buffer size must be set to 64 and sample rate set to 48000.

73,
Bryan

Hi Bryan

Hardware is nothing special, either the on board sound card or a pair USB headphones.
Where can I find the buffer setting please as I've looked in the driver and cannot see the buffer setting.

However I have a M Audio M Track Solo coming if a standard card will not work.

Jim

[W4WMT]

Hardware is nothing special, either the on board sound card or a pair USB headphones.
Where can I find the buffer setting please as I've looked in the driver and cannot see the buffer setting.

Hi Jim,

Are you sure that your on board sound card and/or your USB headphones have an ASIO driver? If so, have you figured out what the name of their ASIO drivers are? That would be the first step.

73,
Bryan

[oe3ide]

Hi Bryan!

Thank you very much for implementing cmASIO!!! Awesome job ... and of course also thank you Richie!!
Excellent addition/improvement in Thetis!

73 de Ernst, OE3IDE