Subject:Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Posted by: ctbarker32

Date:9/17/2002 10:17:55 PM

Hi,

I am pretty confused about some of the techinical aspects of Acid 4.

It is my understanding that Acid 4 endorses/uses the ASIO drivers for low latency MIDI/Audio, etc. and has adopted the VST/VSTi for effects and soft synths.

I currently own/use Acid 3, Sound Forge XP, Vegas 3, etc. Up until this point, it seemed to me that Sonic Foundry endorsed/supported the Windows OS defaults for current/future Microsoft OSes including DirectX for effects and WDM for drivers.

The problem I have is I had thought the world was moving towards DirectX/WDM for all things audio under Windows 2000/XP going forward. I have also purchased a competitors product that uses DirectX effects/softsynths and WDM, etc. So, I thought I was all set for future technology updates. I also use the M-Audio Audiophile card that I think currently only supports WDM under Windows 2000/XP?

So, I am confused about the apparent change of direction with Acid 4 in its embrace of ASIO and VST's. Was this a technical change due to perceived technical issues with the the Windows system defaults (WDM & DirectX). Or, was this a marketing issue to allow Acid 4 due differentiate itself from other products in the marketplace.

Has Sonic Foundry published a white paper that explains this issue?

Thanks for any help.

-CB

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: skysurfer

Date:9/18/2002 2:12:05 AM

VSTi is worldwide industry standard (best choice for SOFO) for software instruments and they work best under ASIO (though i think WDM support was also announced for 4.0)
DirectX is industry standard for FX (cf. Waves....) (very good choice too)
Any VST FX run perfect under Acid using a VST-DX wrapper except the bpm-host-sync thing.

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: ctbarker32

Date:9/18/2002 12:13:08 PM

Thanks for the feedback. You say VSTi is a "worldwide standard". I would counter that it is a standard controlled by one company that is not in the platforms business (e.g. OS) but produces apps.

I guess what I was looking for was some intelligent commentary from others (SoFo employees feel free to chime in!) on the technical analysis that was done that led to the result that is Acid 4. From my limited vantage point, SoFo could have just as easily embraced DirectX softsynths and WDM. It would have seemed quite logical from their previous products and standards.

I'm just trying to make up my mind before the end of the month about whether I should upgrade to Acid 4? This ASIO thing for Windows 2000/XP users with M-Audio cards has got me confused. I was previously happy thinking WDM was the future - it certainly is from a Microsoft point of view?

-CB

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: skysurfer

Date:9/18/2002 12:43:53 PM

Sonic Foundry took at 100% the right decision to "embrace" the VSTi platform with DirectX and Asio, both universal audio standards.
This assures the broadest compatibility and the most professional workflow for the Acid 4.0 user.
I am very thankful for this, though the ASIO implementation in A4 is still in a beta stadium.

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: Iacobus

Date:9/18/2002 1:40:37 PM

I can't really comment on the ASIO vs. WDM issue; each has its own strengths and weaknesses. For example, ASIO has lower latency for sure, but doesn't have the multi-device flexibility that WDM has (like routing busses to different audio devices in your system).

For that reason, ACID Pro 4.0 has, and probably always will have, both driver technologies.

Note that the Audiophile 2496 does have ASIO driver support under Windows 2000 and XP. I'm using the Audiophile 2496's ASIO (and WDM) just fine with ACID Pro 4.0 under Windows XP. (These are the beta ASIO drivers, as there was an initial issue with the Delta ASIO drivers and ACID 4.0; not everyone was having the problem, but some were.) Just be sure to head on over to M-Audio's site every now and then to see if they have updated drivers.

My verdict? Go for ACID 4.0. Once you buy into it, you're pretty much entitled to free updates and features (if they pop up) anyway.

HTH,
Iacobus

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: nlamartina

Date:9/18/2002 2:21:00 PM

To add,

There is a handful of DXi's out there, and hunndreds of VSTi's. (1 point for VSTi's). VSTi standard is more mature than DXi (2 points, VSTi). DXi was invented by Cakewalk, if I remember correctly (ewww, gross. 3 points for VSTi). So it's either eat crumbs out of a competitor's hand, or go to a barbeque at the Steinberg house down the street. Okay, poor analogy, but you get the idea.

As for the DX-FX VST-FX thing... 50-50 in my opinion. Both about the same age relatively speaking, relatively same popularity, nearly identical capabilities. Sonic Foundry apps have been supporting DX plugs for years now, so that's more of a practical decision, since wrappers are cheap and work beautifully.

MPOV,
Nick

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: pwppch

Date:9/18/2002 8:44:05 PM

VSTi vs DXi:
At the time I began the development of VSTi support, the DXi standard was at 1.0 and had many hard limitation. The DXi 2.0 spec was in development, but I could not obtain it from Cakewalk at the time. DXi is not an OS (Windows/Microsoft) standard, just like VSTi. DXi is Cakewalks extension to DX, and is just as proprietary to Cakewalk as VSTi is to Steinberg.

When looking at the available choice of instruments, it was apparent that more instruments were available for VSTi and the big name plugs - NI, TC-Works, etc - were available in both standards.

WDM Kernel Streaming vs ASIO:
WDM KS is NOT a standard. It is a back door hack that MS is permiting the few that were lucky enough to get a back door spec to permit using IOCTLs to communicate with the kernel level portion of a WDM base driver. We had originall planned on developing support for WDM KS, but after my intial R&D, I discovered a number of limitations. These limitations along with the fact that it is not a standard means to communicate with kernel drivers, led to the choice of ASIO.

ASIO, while far from perfect, is an established standard for audio i/o on PCs. It also fits into a tighter model of hardware control/sync. The biggest problem with ASIO is the lack of consitency between drivers. There is too much room for interpretation on the driver developers side. (There are also glaring problems in the SDK sample code that lead to many of these problems.)

ASIO though has an SDK. The IHVs that develope ASIO drivers have been doing it for a long time and the drivers are mature.

Sonic Foundry does not exclude the adoption of DXi or WDM KS. For ACID 4.0 we had to make a choice. We chose two standards that are widely known by our market and user base.


No VST FX:
VSTi and VST FX serve two different purposes. They live, from the programs perspective, in two very different places. The the logic to communicate with them in general is similar, but they are not interchangeable.

To integrate VST FX into our FX model we would essentially have developed a wrapper. We felt that the available VST-DX adaptors more than provide a solution to our users. The wrappers have had a lot of man hours put into them and we test with them a great deal. Additionaly. most mainstream plugins are available in both DX and VST versions, so supporting both standards would be redundant.

There is also on fundamental weakness to VST FX. The can introduce a latency/offset when used. There is no means for the plugin to compensate for introduced latencies. The "workaround" is to use a "offset" plugin on all tracks not effected by the latencies introduced by a given plugin. This is not an acceptable solution.

DX has no such limitation, unless the plug in vendor choosed NOT to implement the non-inplace filter model correctly. With DX we can explain clearly to third party vendors how to make DX plugins behave such that there is no latency introduced. With VST, there is no solution to the plug in vendor.

I hope this clears this up.

Peter

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: pwppch

Date:9/18/2002 8:50:05 PM

>>I was previously happy thinking WDM was the future - it certainly is from a Microsoft point of view?
<<

Actually it isn't from MS's view. There is a new standard that is to replace WDM -AV Streaming. While similar, MS will soon be considering WDM legacy. This does not mean that WDM will be going away, but that a new model layered/based on WDM will be developed.

MS is AGAINST any ISV from talking to the kernel directly for stability reasons. If they could determine a means to keep driver developers out of the kernel, they would. Much of the WDM model is actually hidden from the kernel WDM developer. This is by design. MS wants as little third part code in its kernel. Kernel code makes the system unstable.

This does not mean MS will prevent kernel code from being developed, but there will come a day when kernel code will have to be digitally signed before it can be installed. Only this way can MS assure that driver problems will not exist.

Why do you think that MS warns you when 99% of the audio cards are installed? Because they do not want to take the responsability for drivers bringing down the OS.

Peter

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: SHTUNOT

Date:9/18/2002 9:17:57 PM

DX has no such limitation, unless the plug in vendor choosed NOT to implement the non-inplace filter model correctly. With DX we can explain clearly to third party vendors how to make DX plugins behave such that there is no latency introduced. With VST, there is no solution to the plug in vendor----So have any of you guys been in touch with Universal Audio with the development of their DX plugins? IMHO that is the way to go!!! Please give them a kick in the A$$ so that they can hurry and release those drivers so I don't have to use a wrapper in your products. Later.

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: pwppch

Date:9/19/2002 8:16:40 AM

>>So have any of you guys been in touch with Universal Audio with the development of their DX plugins? IMHO that is the way to go!!! Please give them a kick in the A$$ so that they can hurry and release those drivers so I don't have to use a wrapper in your products. Later.
<
Yes, but they are not interested from my discussions with them. We shall see if they get it right or not.

Peter

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: waynegee

Date:9/19/2002 10:31:14 AM

Excellent post, Peter. Thanks

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: ctbarker32

Date:9/19/2002 11:50:09 AM

Thank you to everyone for their input (especially Peter/SoFo). I guess the key takeaway is that there is no ideal solution for audio under Windows. Software and hardware producers do the best they can with what's available. I think I now understand the issues involved and can make a better purchasing decision.

I fully understand the issues of kernel hacks and OS stability. In a previous career/lifetime, I often would assist Windows users with problems and found that many times replacing a buggy display driver with a solid driver suddenly made the OS very stable. One can only hope going forward that the Windows OS evolves to support streaming audio without "kernel hacks" and thus increase app/OS stability.

I am at a bit of crossroads when it comes to OSes. I do not know if I will ever move beyond my stable Windows 2000/SP3 setup. I was disappointed both technically and philosophically with Windows XP. I do not know if "longhorn" or whatever comes next will change my mind. Currently, other OSes are on my radar screen. Some appear to be either more open/flexible or more elegant. Time will tell.

I have downloaded and intend to install the Acid 4 demo. If all goes well, I will purchase the upgrade and continue along the Acid upgrade road.

Thanks again to all that contributed.

-CB

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: Neil_Palfreyman

Date:9/19/2002 1:03:19 PM

...well a quick open search at K-v-R VSTi returned 221 VSTi Synths available under Windows compared to 42 DXi Synths. Is that enough of a "Technical analysis"?

...with my cynical hat on I would guess that the Marketing analysis is more important than the Technical one.

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: Jacose

Date:9/19/2002 1:10:34 PM

wow peter-peter-pumpkin eater--- thats a great explaination.

rock on Sofo

Subject:RE: Acid 4 - ASIO/VSTi in a WDM/DirectX world?

Reply by: Jacose

Date:9/19/2002 1:14:48 PM

...and how many of those Dxi synths are ALSO Vsti's?????