Flies as musicians (6b) — Nature-Symphony 71b (Revision 2 — slowed)

(Revision 2 — see further below. I do, however, also recommend Revision 1 as a grotesque fun piece: https://freesound.org/people/Philip_Goddard/sounds/760087/ !)
Nature-Symphony 71b (Flies as musicians 6b: A romp in the shadow of Ralph Vaughan Williams, 2) — using a motif drawn from the sort of scale and chords that tend to characterize music of Ralph Vaughan Williams in works such as his Sinfonia Antartica. — This is a 9-layer work, but whereas in the first revision I much reduced the background sound by use of AI software, for some reason that didn't work out well with this N-S 71b, as I explain further below, and even this version isn't fully to my satisfaction, though it's clearer tha the one it replaces, but with a higher level of background sound that masks some of the very quiet (more distant) flies. That background sound, however, is just breeze in the bracken, and varies, so the flies sound much more natural in this context.

I came to realize that a propeller-driven aeroplane, at a reasonable distance, often produces a tone that is just as amenable to my Nature-Symphony treatment as any self-respecting fly or wind chime, a few are included here. Here the opening aeroplane sounds like a beautifully menacing organ playing, while the flies work hard to increasingly get in on the act. At the end of the work I added a reprise of part of the opening, but with the order of layer offsets reversed, so that the motif is then upside-down, and I raised the pitch of all the layers of that reprise by a semitone, giving a brighter feel to end the work.

This is the second of two versions of this work, using the same recording and the same layer organisation. The one difference is the pitches of the layers, describing different Vaughan-Williams derived motifs. To me they are of equal standing.
___________________________

Why a new version of this one?

(Revision 1)
The original recording, although great for flies and a reasonable number of bees, and some nicely musical aeroplane sound, suffered from the high level of the background sound of the wind in the close-by dense stands of bracken. Because that's inevitably a continuously varying sound, I couldn't sensibly use Audacity's standard function (spectral subtraction) for reducing that, and eventually, clutching at straws, tried using an AI noise-reduction plugin (OpenVino) recently added to Audacity. Although it's allegedly modelled for cleaning-up voices and vocals, I found that this recording was one of those that got effectively cleaned-up by that module. My experience so far is that its effectiveness varies a lot between different natural soundscape recordings, and it appears always to increase dynamic range, which could be a no-no for many purposes. Here the increased dynamic seemed at first to be mostly an advantage, though it did rather underplay the aeroplanes.

I'd already uploaded here a version of this work improved by OpenVino, but Revision 1, which replaced that one, was the same but with two further, same-intensity, applications of that noise-reduction. The clarity so produced was quite dramatic, without obvious distortion or artifacts — just a dynamic range increase.

(Revision 2)
I listened again to the Revision 1 version, and was quite dismayed at what I'd let through. The OpenVino processing was clearly overdone, with too big a dynamic range, making the flies sound ugly and aggressive, and excessively downplaying the aeroplanes. I therefore created a version of this work with two instead of three passes of OpenVino processing. This still makes the flies often sound a little on the grotesque side, but the effect now is more coherent and sounds like a weird orchestra of brass instruments in a natural setting rather than a studio — French horn, trombone, bass trombone, and various sizes of tuba — some loud bass trombone notes sounding like dirtier farts than I've ever heard. There's always room for a bit of fun in my works!

It took me the best part of a week trying out different combinations of method and parameters of the noise reduction, and eventually I had to draw a line and choose one of the versions even though I wasn't all that satisfied with any of the results. This one has more background sound than I wanted (a gentle breeze in the bracken, which was amplified because the original soundscape was very quiet), but my various attempts to reduce that increased the dynamic range too much, so making the flies / bees sound ugly and aggressive again, and losing the aeroplanes too.

This version has the sound slowed-down by an amount that gives pitch a minor third lower than before, but restoring the previous pitch. That has made all the sounds, including notes, just a little longer, which makes the flies sound less aggressive and more musical, with note overlaps more common, so we hear more chords, albeit often very transient. The musical speed hasn't changed, because that's determined by the track offsets, which I didn't change.

___________________________

Unsatisfactory constraints on all the flies Nature-Symphonies are primarily that (a) the sequence of layer offset durations is fixed for the whole work, so every sound will repeat the same quasi-melodic / rhythmic pattern (no rhythmic variation apart from the flies' antics), giving a rather mechanical effect in many places, and (b) such a large number of layers means a quite high sum total of background noise, even after my best endeavours at noise-reduction.

I made the original recording in warm and quite humid weather, as the final recording in a 4¼ hour session on 30 August 2024 on the top of Cranbrook Down (south-west corner of inner perimeter track of Cranbrook Castle, an ancient hill fort), high above the Teign Gorge, Drewsteignton, Devon, UK. The recorder was facing some flowering wood sage plants, which were quite popular with the local bees, and it's the latter that give most of the longer notes and chord effects.

Advisory

Important!
To get the best out of this, with its mass of detail, listen with high-grade headphones.

Arrow shows recording position; the recorder was facing the small spikes of wood sage flowers, just to right of the arrow.

Techie stuff:

Recorder was a Sony PCM-D100, with two nested custom Windcut furry windshields. It was placed on a Sirui carbon-fibre tripod, set at a low height to reduce wind disturbance.

Post-recording processing was to apply EQ in Audacity to correct for the muffling effect of the windshield.

For noise-reduction, after much experimentation I used two passes of Audacity's OpenVino noise-suppression plugin, using the deepfilternet2 model and noise attenuation set to 12, then 6 for the second pass. After that a modest EQ tilt away from the treble was necessary.

Layer pitch shifts (semitones above / below original, ignoring the addition to that value to restore pitch (see below)): +5, +4, 2, -2, -6, -10, -18, -16 -14.
To slow the layers while retaining pitch, I couldn't do that directly, for that too readily produces artifacts (in Audacity and WavePad), so I reduced speed using Audacity's 'Change speed and pitch' function by an amount to give pitch reduction of a minor third (3 semitones), and then adjusted each layer's real-time pitch setting in kHs Pitch-Shifter up by 3 semitones to restore the pitch. (What a faff!)

Layer acoustic: all layers middling foreground in cathedral.
Reprise = all tunings +1 and offset sequence reversed, so the motif is now reversed and sounds as if upside-down.

Note that in this and all the other flies Nature-Symphonies I'm not changing musical speed of any layers; only pitch gets changed (using kHs Pitch Shifter Pro).

Please remember to give this recording a rating — Thank you!