Gregory Scott from UBK and Kush replied on Gearslutz, to a user commending him for bringing out the UBK1 in DSP format because there are still too few plug-ins in AAX DSP format. Gregory said...

I've not made any secrets about this in the past, and I respect and appreciate the folks I know at Avid, they're awesome people. That said, here are the simple facts: it takes us 2-3 days to do an AAX Native port from AU, and any coder with basic skills can do it.  It took 6 weeks to port the UBK-1 to DSP32, and another 2 weeks to do DSP64. The entire plug-in needed to be recoded, in a new & unique language, in a new & unique framework... twice. Sales were extremely disappointing, we took a big hit. So the reality is we would spend an entire year just porting our existing (modest) catalog over, and all of that work would result in losses, across board. Worse, absolutely nothing else would happen in the meantime, no bug fixes, no new product development. We would sink, fast and hard. It's absolutely out of the question. For a big company with teams of coders and specialists it maybe makes sense... maybe. But probably not, and I think that's what you're seeing. I wish it were different, I would love to support the technology, it's brilliant.

As an AAX DSP evangelist I am very disappointed by this news, but not unduly surprised. In the days of TDM and LE, you needed the power of the TDM system to get a Pro Tools system to perform reliably with no latency issues and high track counts.

With the increase in computer performance, more and more people, especially in the music sector, can work very effectively with Pro Tools vanilla using very low hardware buffer sizes, as we have shown here with our recent tests. Consequently the market for HDX and AAX DSP must have shrunk dramatically, which ultimately must mean a smaller pool for developers wanting to code for AAX DSP.

Is there also a vicious circle at play here? TDM users are hesitant to upgrade to HDX because there aren't enough AAX DSP plug-ins available. So there are less HDX users and so less or a market for DSP developers to get a return from.

However it is strange that a number of developers managed to produce DSP plug-ins very early on, and if my experience jumping to HDX with AAX DSP has been repeated, finding other plug-ins to use that are available in AAX DSP format, then those developers who produced AAX DSP plug-ins first have cornered the market. Consequently with a smaller pool of potential DSP customers, the demand for DSP plug-ins has largely been met, making it harder for developers that came to the party later, to get a return on their investment, because they are left with a very small slice of a much smaller cake.

Even after 2 years or so there are still unanswered questions here. Gregory is not the first developer to make the point about the timeframe and complexity of building AAX DSP plug-ins, especially after we were told by Avid, at the launch of the AAX plug-in format, that one of the key design improvements with AAX over RTAS and TDM was it would be much easier for developers to produce DSP versions, once they had developed an AAX Native plug-in. But based on Gregory's experiences, how did some developers manage to produce both AAX Native and DSP plug-ins so quickly? Were they helped by Avid? If so, why didn't Avid offer that help to all the AAX Developers?  If you are a developer and have any insight into all of this, then we would love to hear from you.