The end of the DSP technology?

It may seem like a boring topic to talk about the world of music plugins, but it's actually quite interesting. In the past, in the world of analog equipment, an effect could only be hardware. However, with the development of computer technology, software appeared that performed the tasks of hardware more cheaply and conveniently. In the early days, these software programs were not of the highest quality, but the industry evolved, and the software gradually caught up with their hardware counterparts. However, there was a problem with the performance of computers, as the better software required more computing power and RAM, and computer technology could not develop as fast as these software programs. This is how DSP (Digital Signal Processing) came into being. This system operated separately from computers, with separate hardware developed to solve the performance issues of computers. The undeniable advantage of the system was that the DSP plugin could handle higher computing capacity, thereby achieving better sound quality. For example, a reverb plugin could calculate its reverberation on a 32-bit DSP, or the software could be designed to calculate a number sequence with a few more decimal points.

Many people considered this process a significant step forward, and it indeed seemed to be so. However, over time, it increasingly seemed to me that this was a dead-end that would not last long. Why do I say this? Relieving the computer workload has simply become pointless. We have such powerful computers today that running a few hundred plugins at once is no problem for them. The differences in quality between DSP and native plugins have virtually disappeared. Both calculate equally well if the software designer has properly programmed the processes.

Today, we have reached a point where the major manufacturer, Universal Audio, which was the first to embrace DSP technology, is slowly backing out of further development of the system, and its software that runs without DSP has also appeared. This means that the focus is shifting back towards native plugins. UAD has realized that there is no longer any economic advantage in selling DSP systems, and is instead focusing on further developing software without hardware. The question is what the future holds for UAD, as they now have to compete with a dozen other software manufacturers unless they come up with something that stands out in terms of knowledge from the crowd. The big name will certainly provide a positional advantage for a while, but this can only be maintained if they come up with something new. I suspect this native software usage will only be a temporary opportunity or an extra option alongside something else. The global economic crisis is knocking on the door, it is easier to develop and sell software than to maintain a complete manufacturing base. But in a few years, UAD will come up with something new, because this situation is more of a step back than forward. It is much more exclusive to say that with UAD hardware, you can produce a much better end product than to say that with this bitset, you will be a King. Of course, DSP thinking does not stop there, and the future's new buzzword is already here. This is FPGA (Field Programmable Gate Array). The technology can be explained in a Sci-Fi-like and somewhat neutral way.

What if we had a special analog hardware that could transform itself into reverb when needed, and at other times into a compressor, always to what is needed? How cool would that be? In essence, this technology produces that, obviously to an achievable extent. In simplified terms, this hardware is a set of programmable semiconductors that physically switch to the appropriate task on a given command. Like a programmable logic matrix. This is not a new technique, it has long existed in other industrial equipment, but it has been unusual for it to be incorporated into an audio technology device as well. This is now also being realized. The advantage, if the manufacturer applies a more complex solution, is that the hardware tries to possess properties similar to those of analog equipment, thus the output sound more closely approximates the characteristics of the signal flow from an analog device.

Sounds amazing, right? However, these are still exploratory steps, small advances. Currently, this technology operates in hybrid mode with traditional DSP solutions. And since FPGA-based effects can only be applied in real-time, sometimes phase-shift-like sounds are produced on stereo tracks. The technology has been available since 2017, but it hasn't spread widely in the midst of the UAD craze. The company Antelope, which uses this technology, may gain some positional advantage while UAD finds its footing, but I wouldn't be surprised if UAD soon adopted a similar solution, just in a more refined form. The slow demise of DSP technology may take decades, just like in the case of ProTools HD, where ten-year-old units are still running on the used market. But it's certain that this technology has already peaked. Pioneering techniques are now being introduced consciously and with great care, so there will be something beyond DSP in the near future, but the real breakthrough will be the introduction of artificial intelligence into the audio technology industry.