Topic: Tuning the extended low notes

Hello, everybody!

I've been using Pianoteq Standard for more than a year, but until recently I hadn't got into editing the various parameters. Yesterday I discovered that some of the pianos have an extended range, so I incorporated F and G-1 into a piece I'm working on.

However, I find that G-1 seems to sound very out of tune, more so that F-1, which sounds fine. I'm using the Hamburg Steinway D with a slightly modified Prelude preset. I've tried changing the tuning via the detune page of note edit - I can hear a difference when taking it sharp or flat, but can't seem to get it to sound just in tune. I did try experimenting with increasing the string length, reducing the octave stretching (and using the 'Flat' tuning). I suspect that I'm lacking knowledge of how the sound works to know how to make this change.

Are there any other things I could try? Would any of the individual note edits restricted to Pro be useful for me?

Thanks in advance,
Nick

Re: Tuning the extended low notes

My understanding of sound theory and experience with acoustic pianos (though not ones with such low notes) is that there is very little, if any, component from the fundamental frequency and that the brain "infers" it from the harmonics. Moreover, there is incredible amount of inharmonicity in the deep bass, and even more so in those extended low notes. As such, the brain gets confused and cannot correctly identify the fundamental and hence pitch. That is the reason why is basically impossible to play chords in those low octaves, and I suspect the reason for the phenomenon you described.

Assuming pianoteq wants to faithfully emulate an acoustic piano, it should be impossible for you to get those notes well in tune.

Where do I find a list of all posts I upvoted? :(

Re: Tuning the extended low notes

Yes well put - and any individual piano (maker/size/type) will generate dramatically different inharmonic activity below normal bass ranges, just to make this more difficult as a problem.

In general, human ears will hear very low bass as going out of tune as becoming sharp (slightly higher pitches per note as we go ever lower), whereas we hear very high notes going out of tune sounding more flat (slightly lower pitches per note as we go ever higher).

We can take a heavy-handed method to it, and stretch the tuning as dramatically as we want to sweeten things, so listeners either observe the opposite, or something which doesn't bother our brains with "is that in tune?". Pretty hard to objectively say "this heavily edited tuning sounds right" - often if we're diving right into this kind of thing, we might submerge ourselves for days or longer dealing with it (to better understand what's happening), only to find in later re-listens, that we went too far to correct it. It can take weeks for our ears to refresh from long periods of working on strange pitching. But - all interesting, all fun and worth trying and probably many would easily take to it and get it pretty correct for most subjective listeners.

The pianos which physically do have lower than relatively standard bass range each have put sweat into engineering their whole piano, to handle these issues. Given that most dpianos though, have a standard range, they likely don't become well represented in the world of VST instruments - just simply, most users will find it a bother to activate some switch to 'move/play' lower range on their dpianos or MIDI setup (esp. in mid flight during a piece - some love this kind of thing but a vast majority of pianists IMHO want to play real pianos, without fussing with switches). Whereas a piano with a more normal range will be able to be played by all with any dpiano. Just thinking that through, as it may not be thought about much - personally, that above is the reason I don't choose to work lower bass on piano, not just because it presents more sonic problems.

Interesting problem though - and I do believe, right down to a specific piece of music, using any VST it would present a problem to make extra low bass notes sound good to all ears. The ways to improve how they sound could come directly from tweaking Pianoteq in various ways (like in Pro you could manually edit ranges or per note values for various controls, to sweeten the ever more physically realistic inharmonicity down there to make pitch seem more correct at least just 'for the piece').

Using various plugins running in a DAW it might be possible to go beyond just pitch (once that's sounding right), to smooth out all kinds of elements of errant deep bass inharmonicities. In Pianoteq Pro you could do similar by hand per note.. altering overtones in the spectrum. In Pro, you may see "Oh look, there's this tall overtone spike on all the notes" and lowering just those might make extreme bass notes sound clearer.

Some bass plugins for example, can sweeten overtones to be more 'musical' rather than so dramatic (many plugin makers call them names easily recognized as 'for bass'). Many of these process a bass pitch and 'move up' spectral activity in deep bass to higher octaves. (like from bass to low-midrange spectral shifting.. with a musical touch to remove spikes and normalize weird real audio issues).

Beyond those, running the whole piece through other plugins (like console emulators, or master bus processors, so it sound like it's been recorded through a nice studio desk, or further on to tape or through some nice hardware.. unlimited choices for recording things are out there), it may be possible to shine up the audio so listeners won't think "Wow strange hearing those low notes", but instead hear it as a more integral element which doesn't stand out as strange within the context of the whole mix.. unless that uniqueness is what we intend, which can be of course the desired goal too.

Pianoteq Studio Bundle (Pro plus all instruments)  - Kawai MP11 digital piano - Yamaha HS8 monitors

Re: Tuning the extended low notes

To add to dv's and Qexl's great observations, when technicians tune Boesendorfers with extended range (and if memory serves on Stuart and Sons technicians use a similar technique as well), the technician wears gloves to protect the piano strings from corrosion and presses the middle of the string to mute half of the string and then they tune the string based on the first harmonic (1 octave higher), which--if they muted half of the string correctly--will leave the fundamental reasonably in tune.  While that exact functionality is impossible to directly replicate at the moment in PTQ 7, PTQ Pro offers the ability to edit all of the audible overtones of each note, and I suspect that your preset (as Standard honors Pro overtone curves but disallows viewing or editing them) has an out-of-tune overtone that the developers/testers liked or considered "realistic" but for your use has become distracting.  In that case, the only solution would be a new Pro preset with different overtones on the note giving you trouble.  If you modify a preset in Pro and export it, Standard will load it, but you still won't be able to edit it further within Standard.

Broadening the discussion, the extra lower notes on the selected extended range pianos are considered below the threshold of "music" or "tone" vs "sound" or "noise".  As notes lower, we hear less and less of the fundamental and more of the overtones.  If an infrasonic sound has a high enough sound pressure level (SPL), we can feel the oscillating vibrations but not necessarily hear the direct sound, instead we hear the intense overtones or higher-pitched sympathetic vibrations of nearby objects being rattled by the vibration, or we feel the vibration physically and perceive it as sound.  The tam-tam is an excellent example of this, as gongs--which a tam tam can easily be mistaken for--are pitched while the tam-tam is an unpitched percussion instrument that provides a generic low, loud noise, and it makes no practical difference if one orchestra's tam tam or bass drum is slightly larger or smaller, whereas tuned timpani need to constantly be tuned by their performers while on stage or hitting the wrong gong during a performance will be highly noticeable.

When trying to actually tune a sound that low, you have to be extremely accurate in your reading of the sound, as a simple error of 1hz is significantly greater at infrasonic levels than at high pitches.  At C8/c''''' (the highest note on nearly all pianos), the frequency should generally be around 4186hz (or just around 4.2kHz to give an acoustic sense of how slow that oscillation is compared to the electronic communication signal of 5GHz wi-fi), and an error of 1hz will be off by roughly 0.0023889% (that's especially rough as having an error of exactly 1hz is as arbitrary and difficult as being at precisely 4186.009hz for C8/c''''' and we're not even discussing divergent trichord tuning or exotic pianos with tetrachords or aliquots).   An error of 1hz with the low A0/A,, will be 3.636% off, a much higher and audible error.  This is even worse at C0/C,, at 6.116% for one extra (or fewer) oscillation per second.

To combat the complexity of apply hertz units in piano tuning--as even fractions of a herz become very important at the low range and beyond useless at the high range, technicians use "cents" where each semitone is divided into 100 units of relative tune, making for 1200 cents per octave regardless of range or temperament.  So, our first error at the top of the piano was just over two-thousandths of a cent off, which is fantastic accuracy, while the lower errors were 3-4 cents to over 6 cents, which would be audible--especially in the overtones.  When a piano note is really sour, it's 40-60 cents off or more, and anything over 50 cents sounds more like a neighboring note than the actual note.  Most people can hear a difference of 5 cents or more, but technicians should have ear training that is much more precise (or at least a good tuning device that can).   Minor details can change the whole pitch of the instrument.  One hall I worked with tuned the instrument 20 cents sharp (if memory serves) across the whole instrument, so that when they rolled the piano out of the pit and turned on the house lights, the heat in that small recital space from the spot floods would warm the strings enough for the strings to "sag" into tune.  There are even cases where performers will ask for extensive tuning errors, like when that small hall had a performer ask that the A0/A,, be 100 cents flat to play as G# (it was an Ab but generally technicians like to express piano enharmonics in sharps), since the concert hall didn't have a Boesendorfer with extended range to play the Cziffra repertoire that was specifically composed for the Boesendorfer.

Extended range was a direct result of Busoni's attempts to transcribe Bach organ works for piano, as he wanted the full 32' pipe range which goes down to 16.35hz, but the prohibitive complexity of further extending the piano range (and the complexity of tuning and maintenance) meant the innovation has remained sidelined by all but two manufacturer's of pianos (Stuart and Boesendorfer).  With electric pianos, then digital pianos, then virtual pianos, we've gotten to the point that we can functionally emulate that without expending copious resources, but the downsides and intricacies of why piano makers have deliberately avoided those notes for a century and a half still seem to appear in their virtual counterparts.  (Don't even get me started on the disorientation concertizing pianists get when they're suddenly on a longer piano--even with those keys painted black!)

Keep us posted on if you find why exactly this extended range preset isn't working for your needs, and we'll see how we can further help you.

Spotify: https://open.spotify.com/artist/2xHiPcCsm29R12HX4eXd4J
Pianoteq Studio & Organteq
Casio GP300 & Custom organ console

Re: Tuning the extended low notes

Hi, everyone! Thanks for the wonderful replies! I had suspected that the realism of Pianoteq's modelling might make what I'm looking for tricky, so I'm glad that you've offered such a wealth of information. I decided to upgrade to Pro to play around with some of the various settings (I'd been thinking about the upgrade for a while already). I've been trying out small changes of the Spectrum Profile, generally boosting the overtones that produce octaves of the fundamental and cutting those that are creating spikes on an analyser or seem to be far away from the fundamental. I'm not sure that it's made a huge difference yet, but that might be down to my ability to perceive such differences.

I'm calling it a day for tonight, but will read your replies again when I'm fresh over the weekend and experiment further. I'll let you know how I get on.

Re: Tuning the extended low notes

I just realised that I didn't post back here with the results. I ended up boosting overtones 2 and 4 a little, and cut 7, 11, 13, 14 and a couple of higher ones where I noticed a big spike on my EQ. I also flattened the notes by about 30 cents and increased the string length of those low notes a little. I'm not sure that I got them exactly in tune, but they're no longer beyond the pale, and since the track is supposed to create a small feeling of unease, I think that them being slightly out isn't an issue.

This is the track, if anyone wants to hear how it's turned out so far (forum rules seemed to imply that it's ok to share a link, but if not, let me know and I'll delete it): https://soundcloud.com/nick-owenpetch/r...al_sharing

I appreciate all of the advice offered in this thread and will continue to delve into it as I learn to utilise the possibilities that Pianoteq offers.

One final note - the tubular bells in Pianoteq sound amazing!