A Programmer's Perspective on the AAX Transition + Q&A

[DaveTremblay]

Originally Posted by DonM

AAX feels like a 180 degree about face from PT9's open strategy to include 3rd party hardware. Supporting VST on the new Avid hardware would have felt much more consistent with that thinking. At this time I am hesitant to invest in the new hardware, software and repurchase AAX versions of all of my plugins.

So can anyone help me not feel as though I stepped into the web of a black widow two years ago and a $9k+ (my estimate to move to PTXI when released) will be worth it.
-D

AAX isn't an about-face from our open strategy. VST and AU have fundamental design limitations that prevent them from being a working solution for us.

1) HDX uses DSPs for hardware accelerated plug-ins. This requires a very specific modular break in the plug-in design to run just the processing part of the plug-in on that chip. Essentially, we run part of the plug-in on the host processor, and part of the DSP. AU and VST don't really work like this. They both have a modular break in their plug-in designs, but they are not in the right spot with the right tradeoffs to work on our high end DSP systems.

2) AU is Mac only, and unlike Logic, we run on both Mac and Windows. VST is really the only "open" plug-in format we could use, and even that one, we could only support in Native only (see point 1).

3) We also have a requirement for really high performance in our real time audio threads. We designed AAX to be really deterministic in that thread to enable really large sessions and high plug-in loads without glitching. VST pushes computation of coefficients into the real time audio thread making it less deterministic. Individual plug-in manufacturer's can take steps to smooth this out, but it isn't a core piece of the design. In AAX, we take care of pushing all that work to lower priority threads and doing all of the thread communication to make it work in a high performance setting.

Enough of the justification... It is ultimately up to the plug-in companies as to their own upgrade policies. Some have announced little to no cost AAX updates, but not all. Upgrade costs will likely vary based on which plug-ins you have... If you want to run VST plug-ins natively in PT, you can always use one of the adapters that converts VST to RTAS.

I don't think you stepped into a Black Widow web by switching to PT10. We're making a lot of strides towards improving our product and our platform recently and I think that will continue. Those sorts of investments on our part will pay off for you in the long run by making you more productive. And if I didn't believe that, I probably wouldn't be spending all this time answering these threads. :)

Dave

[DaveTremblay]

Originally Posted by PiccoliStudio

It's difficult to transit to AAX if there's no AAX plugin available... just the new ones.... what about all other Avid plugins?

It might not be obvious from our website, but we have ported a large number of our existing plug-ins to both AAX Native and DSP. For example, Dyn3, EQ3, Mod Delay, BF-2A, BF-76, Smack, Impact, Maxim, Pultec EQs, and the Reel Tape Suite, among others. Many of them also have nicer new GUIs since we had to rewrite them anyway.

[DaveTremblay]

Originally Posted by Darryl Ramm

Yes which is why they need to start now, even if that is internal only bring-up. I don't know if developers have access to a 64 bit version of Pro Tools or any test/debug frameworks to host/test a 64 bit AAX plugin? Do you? It would be great if they do. Trying to build a 64 bit plugin and seeing what the compiler barfs on is one thing but you do not get far without being able to test it.

Our third party plug-ins developers do have access to a 64 bit AAX SDK and a 64 bit developer build of Pro Tools (severely feature limited) to test plug-ins.

Dave