<HTML> <HEAD> <TITLE>Re: A different temperament / tuning approach</TITLE> </HEAD> <BODY> <BLOCKQUOTE>I like what you are doing, concentrating (almost) exclusively o= n making the octave fifth exactly pure and having that determine stretch. A = couple of thoughts though. 1. They are not nineteenths. Nineteen semitones, yes, but it's a twelfth.<B= R> 2. In the bass, other partials often sound louder -- for example, I often c= oncentrate aurally on the fifth partial because it sings louder to me and th= erefore I presume to the listener. So I wonder whether you might switch to t= he fourth or fifth partial when you descend into the wound strings. TuneLab = Pro defaults to the 6th partial in the bass, but that seems arbitrary. </BLOCKQUOTE> || ||| || ||| || ||| || ||| || ||| || ||| || ||| || ||| jason kanter * piano tuning * piano teaching bellevue, wa * 425 562 4127 * cell 425 831 1561 orcas island * 360 376 2799 || ||| || ||| || ||| || ||| || ||| || ||| || ||| || ||| <BLOCKQUOTE>From: "Richard Brekne" <richard.brekne@grie= g.uib.no> Reply-To: pianotech@ptg.org Date: Fri, 07 Jun 2002 23:19:26 +0200 To: "PTG" <pianotech@ptg.org> Subject: Re: A different temperament / tuning approach </BLOCKQUOTE> <BLOCKQUOTE><TT>Hi folks... Just thought I'd post some interesting comparisons between what RCT came up= with on its OT 3 (Clean) calculated curve visa vi the perfect 19ths tuning = I've been fooling around with. Actually RCT's OT3 liked this arrangement quite<B= R> nicely. I spent some extra time today on a Hamburg D being really accurate = with Tunelab as described yesterday, and when I was done tuning I turned on RCT,= sampled the 6 A's and ran a quick note for note check. In most cases I got = some degree of blush, and in every case there was little (almost no) movement of= the spinner. I didnt get to sort out all the differences between the RCT cents<= BR> offsets table, and the numeric editor in Tunlab... but for your edification= here are the Note TuneLab RCT A5  2.85  2.89 A#5  3.22  3.30 B5  3.60  3.75 C6  3.97  4.19 C#6  4.34  4.73 D6  4.71  5.22 D#6  5.09  5.73 E6  5.46  6.33 F6  6.93  6.99 F#6  8.41  7.74 G6  9.88  8.47 G#6 11.35  9.41 A6 12.83  10.48 A#6 14.30  11.65 B6 15.77  12.97 C7 17.25  14.24 C#7 18.72  15.87 D7 20.19  17.28 D#7 21.66  18.81 E7 23.14  20.64 F7 24.61  22.65 F#7 26.08  25.01 G7 27.56  27.33 G#7 29.03  30.28 A7 30.50  33.71 A#7 31.98  37.70 B7 33.45  41.69 Interesting that the RCT really stretches the top notes on a so called &quo= t;Clean" stretch. This is way beyond the 3rd partial of the octave 5th below, tho th= is is only for the last 4 notes. Otherwise the perfect 19ths starts out just a= little lower, then at F#6 goes sharp of RCT's curve  as much as 3 cent= s at C7, and there after RCT starts to catch up again. These are fundemental (first partial) offsets in both cases, but in Tunelab= they are also the exact frequencies of the 3rd partial  for the octave= and 5th below each note. For example the 3rd partial of A5 is exactly 23.14 the fundemental of E7.   In other words.. how RCT' curve wanders around the actual 19ths for these n= otes by comparing the RCT calculated values to the real frequencies for these as= read in useing Tunelab. RCT's curve starts out by holding the 19th just a t= ad wide, then it gets narrow by as much as 3 cents... and in the end gets wide= by around 10 cents. Not exactly an exponential development. This is a lot of fun and its interesting for me... so I intend to get as<BR= > complete and accurate a comparison of these two approaches to tuning as I c= an. But one thing is already clear. In terms of general stretch attributes.. th= e perfect 19ths is quite viable... and its a very simple thing to accomplish.= RicB Richard Brekne RPT NPTF Griegakadamiet UiB </TT></BLOCKQUOTE><TT> </TT> </BODY> </HTML>