ART is based on:
https://www.researchgate.net/public...ontrol_Using_a_Limited_Number_of_Loudspeakers
The algorithm was experimentally developed for car audio where listener positions are well-defined, using 64 measurement positions, for stereo, multichannel and upmixed content.
The only difference in what I am proposing to the current consumer offering of ART is that, if using a microphone cluster like Trinnov's, the position of each microphone in the cluster is fixed and known in terms of distance and height, while the current method uses single or multi-position location estimates. This limits accuracy and consistency of the measured impulse responses and the ability to estimate spatial characteristics, but it is a good practical product and engineering decision, given everything involved.
The full critical range is 15-400Hz IMO. Below 15Hz tonal hearing begins to fall apart. However below 40Hz or so noise is a major issue and unless the SNR is consistently high, the algorithm would try to correct non-signal sounds, which might produce audible and unpleasant support signals or bust your speakers or their amps.
100-400Hz is where SBIR is the most problematic in rooms, and is much harder to address than modal problems, but correction would require high measurement precision and likely a fixed, or at least very narrow, listening position.
20-150Hz is a good compromise. Modal and some SBIR issues are addressed, fixing the kinds of low frequency and sub setup issues that most users struggle to understand. Most cheap acoustic treatments lose effectiveness rapidly below 200Hz, so setting the upper boundary at 150Hz helps home theater planning and installation, or the same for studios, massively decreasing costs since you don't need to invest in pressure-based bass panels, which are large, heavy, expensive and need a lot of space, since you need many for them to be effective.
I don't know how much benefit a microphone cluster would bring. miniDSP has an array it has been selling for years, so I wouldn't rule out integration at some point.
Overall, despite some limitations, I think Dirac ART is a good product with fairly low barriers to entry compared to Trinnov's offering. For someone like me with decent knowledge of acoustics and DSP, DLBC and other automated algorithms were not very useful. ART represents a next step and something fundamentally different, and the Tide16 has a good feature set and finally makes ART somewhat affordable, although I really would like a Tide8.
Like I wrote earlier, I don't mind making compromises in terms of how much I spend, the features of the device, usability and acoustic accuracy as long as the total package and results make sense.
One of the things that really attracted me to the Storm and Bryston devices was the user interface. But miniDSP's is good too, so as far as I'm concerned the value proposition for these very high-priced products became very thin and distant.
helpdesk.dirac.com
