Okay, in short, this is not an issue, yet it is how Plateau and algorithmic reverberators in general can behave in certain conditions. Here’s why:
Algorithmic reverbs use all-pass filters and delays to mimic the sound of a particular reverb by creating a dense series of echoes. These “building blocks” all cascade and feed into each other depending on the type of algorithm, and each have a specific delay time and gain value to achieve a certain reflection pattern and reverb characteristic.
Now, psychoacoustic research states that for a reverb to be stereo, it has to have two outputs that are dissimilar or ‘de-correlated’ so that we perceive a certain amount of width to the sound (Hidaka et al., 1995). Thus, the taps and delay times between the left and right outputs for Plateau in this case have to be different from each other for the auditory image to sound wide.
However, if these delay times are fixed, when you play a sine wave through an algorithmic reverb it will eventually settle into a steady state where a standing wave will form and one output will perhaps be louder than another. This happens in a real room where a wave will interfere with copies of itself, creating cancellations and resonances in different places. This imbalance is frequency dependent, therefore if you play a constantly changing synth sequence into it, the imbalance will not be heard. If anything, it will appear balanced yet fluctuating as the waves interact with each other, much like in a real concert hall. Plus, the balance shifts as you change the Size parameter because it is altering the delay times in the entire reverb, thus changing the resonant frequency. I’ve also tested this with the NYSTHI plate reverb and the same behaviour occurs.
Plateau, and Dattorro’s plate algorithm by extension, gets around the standing wave issue by modulating the all-pass filters subtly so that the resonances and standing waves are smeared out. Try it yourself, slowly crank up the modulation depth and notice how the auditory image seems to become more and more fluctuating. Another work around is it have some dry signal so that our brains resolve the inter-aural time difference (have a look into the Haas effect, and check out Hartmann’s research work in acoustics and localisation https://web.pa.msu.edu/acoustics/ ).
In conclusion, yes it is difficult to create a centrally panned auditory image whilst having wide reverb using wet signal alone, and is not an issue with Plateau.
References:
Hidaka, T., Beranek, L. L., and Okano, T. (1995). Interaural cross-correlation, lateral fraction, and low-and high-frequency sound levels as measures of acoustical quality in concert halls," J. Acoust. Soc. Am. 98(2), 988-1007.
