Just a comment about the motivation for this tool:

This project was motivated by a disastrous recording incident. A friend had been asked to do an on-location recording of a full orchestra and chorus performing The Brahms Requiem — a major work with huge dynamic range and relatively high peak sound pressure levels. He set up a dozen carefully-chosen microphones, placing them in key locations to capture the orchestra, the chorus, the soloists, and the conductor’s comments. He set the gains for what appeared to yield 15 dB of headroom in each channel, and, during the pre-concert run-through, monitored the meters — watching for peaks and meticulously tweaking gain settings as indicated

During the concert, the board’s peak indicator never triggered, and in the loud concert hall where his mixing desk was located, he could not discern distortion in his headphones. Yet when he later started mixing tracks in post-production, he was distressed to discover significant double-digit THD clipping distortion in the chorus tracks. He invited me in to help with triage.

A Claude AI analysis of the distorted tracks pointed to preamplifier saturation. Since gain had attenuated the signal after that saturating pre-amp, those clipped peaks never registered on the board’s metering — not on the VU bars, their peak marks, nor the board’s master peak indicator.  Ouch.

Thus was the motivation for the initial version of this tool. Had he accounted for the 45 dB SPL noise floor of the performance space and understood that the effective dynamic range was more limited than expected, he could have reduced gain at key places in his signal chain, thereby trading off a bit of un-needed Signal-to-Noise ratio in the recording signal path for significantly improved THD.

Version 18 represents a substantial expansion. The tool now includes pull-down menus for all 40 genres from the SPL reference database, an integrated SPL reference panel covering 140 microphone positions and 90 spot instruments, a venue noise floor slider, pre-optimization advisory calculations for the preamplifier section, a gain pin for intentional transformer saturation targeting, and manual entry panels for microphones and preamplifiers not yet in the database. Device counts remain at 103 microphones, 39 preamps, and 21 A/D converter systems.

Daniel,

Appreciate the feedback.

To address the points you raised:
 
1)  "Not sure why you would need AI analysis."   I have an Izotope suite of RX 11 tools that struggled with doing anything useful for Nick's track.  This may have been because the recorded track file still evidenced lots of headroom -vs- banging the rails. I tried Claude AI an experiment, and was impressed at what Claude attempted to do, albeit not perfectly. It's first assessment was a THD approaching 100%.  I pushed back on that and it realized that it was actually measuring the higher-frequency notes in a chord as "distortion" when it was actually distorted program material. The useful piece of Claude's feedback was the insight that distortion was happening well before the A/D in the signal chain - a prior analog gain stage and buffer which was subsequently being attenuated and digitized as something that looked to RX 11 like lots of headroom.

2)  "In practice, you'd need to actually measure SPL at the mic tip" -- Yes and no.  "Yes" - it would be best to have a measurement.  But "No" -  typical SPL levels for spot and over head mics are well established for many genres.  In v18 of the tool, I've adding a much more rigorous pre-processor that develops high quality SPL estimates based on user input of genre (or instrument for spot mics), and mic placement.  This estimate draws from:
      "Katz, B. — Mastering Audio: The Art and the Science (Focal Press, 2002/2015)",
      "Skritek, P. — Handbuch der Tonstudiotechnik (1989)",
      "Dickreiter, M. — Handbuch der Tonstudiotechnik, 7th ed. (De Gruyter, 2008)",
      "Streicher, R. & Everest, F.A. — The New Stereo Soundbook, 3rd ed. (Audio Engineering Associates, 2006)",
      "Woszczyk, W. — Microphone techniques for orchestral recording (AES preprint collection)",
      "Bech, S. & Zacharov, N. — Perceptual Audio Evaluation (Wiley, 2006)",
      "Wittek, H. — Perceptual differences between wavefield synthesis and stereophony (PhD thesis, University of Surrey, 2007)",
      "Rossing, T.D. & Moore, F.R. — The Science of Sound, 3rd ed. (Addison Wesley, 2001): per-instrument SPL reference data",
      "Neumann GmbH application notes — microphone placement and expected SPL for orchestral and ensemble recording",
      "Schoeps GmbH application notes — ORTF, MS, Jecklin disk placement and SPL expectations",

3)  V18 also has a button for manual parameter entry if the microphone, the preamp, or the A/D system has not yet been modeled.  This system uses quite a few parameters to make the math accurate, so manual entry is not trivial in any of these cases.  When that option is used, I'll get a notification so that I can carefully research the provenance of specs for the new item and add a record to its respective JSON data base.

4)  The path analysis associated with the different option buttons (maximize SPL, minimize THD, Balanced) is VERY useful to those of us who use RME equipment with external gear.  In my own case, I've enjoyed using the RME UFX, UCX, and Babyface Pros, not only with their own excellent mic amps, but also with their line inputs to digitize other preamps -[Focusrite ISA , a professional FET amp, and home-grown Seventh Circle Audio J99's, C84's, N72's, and A12b's.]  With several of these, its not  unusual to push gain in the preamp to get a bit more transformer saturation in selected recording situations (female or male vocalists but others as well).  One would, of course, attempt to compensate for the high pre gain and consequent signal output with a lowered gain on the RME gear, or use an external attenuator. 

Here's the crux of the matter:  Users doing such manipulations to "get that desired sound" usually have no clear idea what effect such gain staging has on overall chain SNR, Headroom, THD or Dynamic Range.  They might blame RME for audible distortion when, in fact, to achieve the desired effect, they should have used an attenuator between the pre and their RME.  This application provides that needed clarity, even to the point of recommending a specific level of external attenuation when the optimizer cannot meet the desired path goals.

Herb