Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum My Piano in EBVT III

Hi Bill, let me check on the Liszt.

I am not a "recording engineer" and as a result, there is learning curve to doing home recording with sophisticated equipment, which I now have. I guess you could say it was the same thing learning to use an ETD to tune.

Continuing to explore the sound files I posted before, I just now re-did "You'll Never Walk Alone". The difference is quite remarkable. There is something called "TPDF Dither"...lol..sounds foreign to me too...This almost always applied to any 24bit recording. I did not have this engaged when making the WAV and MP3's which you hear from my box.net site.

What I am hearing now is a very clean, clear sound, with NO upper end distortion/grunge, and a very coherent stereo image. It sounds almost live. I have just replaced "You'll Never Walk Alone" with the new dithered and normalized version on my original post, but here it is so one does not have to search back through this thread. Comments are welcome! Headphones are a plus.

Another thought...as a result of this clarity in the overall sound, I am hearing more of the effect of EBVT III.

Dithered and Normalized- "You'll Never Walk Alone" http://www.box.net/shared/288rvn3qej

I liked the Misty but sad to say, the piano was already out of tune, not the best representation I have heard. I could still hear the contrasting harmony but the unisons and everything else going a bit haywire were distracting. Same goes for the third recording. We really have to get to the reason why your piano won't stay in tune more than 5 minutes!

LOL...I agree...it's very frustrating!

Here is the last of my recordings...have to get to work and earn some money to pay for this piano!

Mr. Stopper, thanks for your comments, as well as everyone's input. It's been fun. As I mentioned earlier, EBVT III has opened up another dimension for my piano, it's great. There is a place for it in the piano world, just as there is for ET.

Thanks again Bill, and I hope in the future to tackle this issue with the tuning. The problem seems to be mostly in the 5th and 6th octaves. The same 7 or 8 trichord's inner strings go out, and the piano as a whole goes a bit flat.

Enjoy! GP

1. Ampico Piano Roll, Re-cut, "Butterflies in the Rain" played by Frank Miline on the Ampico, #214581 http://www.box.net/shared/ttfgdtv9m0

2. Original Ampico Piano Roll, "Learn to do the Strut" by Irving Berlin, played by Vincent Lopez on the Ampico, #204231E http://www.box.net/shared/zmm6uvdqy8

3. Original Ampico Piano Roll, "Side by Side" by Harry Woods, played by Frank Black on the Ampico, #209001E http://www.box.net/shared/0o5vqxubob

Dear Grandpianoman and Bill Bremmer:

This is an updated note of heartfelt thanks for your labors of love on this project: the traveling, the tuning, the recording, the postings, the encouragement--all of it! As you may know from my other posts, I am a rank amateur pianist with a strong post-mid-life-crisis desire to share my playing with others, for all of the good music can do for the soul. I have been richly blessed by what you have both done and greatly excited by where it seems to be leading other contributors to the thread. I consider myself a happy recipient of spill-over blessings that have been a result of your work, and wanted to share some of the ways it has effected me personally.

Grandpianoman : I have spent hours listening to the various recordings. Aside from enjoying the rich, crystal tones of your piano, you taught me that much of the music that my grandma played was of the fox trot variety! When I hit those, I was ecstatic! Now that I know what it's called, I can look for the scores!!! (I have only four pieces of photocopied music from her, from her day.) I've listened to the recording of "The Continental" over and over again, and it makes me laugh and giggle with delight EVERY TIME!!! You wouldn't happen to have "Up The Lazy River" in your collection, would you?

GPM and Bill : Rhapsody in Blue. I've spent alot of time listening to the first two posted versions. Somehow, when I listened to the Gershwin version the first time, I missed that it was played by Gershwin! So I had to give it a really close second listen. Anyway, after really hawking down on these two recordings, I've concluded that, in this instance, as Bill has said elsewhere, "If you want something done right, you've got to do it yourself!" So I got out my score that used to belong to my dad and started practicing/playing. (When I was in late high school, early college, I asked him, "Do you think I can play this?" He said, "Sure!" I whittled away at it for a while, and then regarded it as too big for me. I'm happy to say that 30 years later, I think it will fit.) This new personal mania with Rhapsody in Blue is a result of my dissatisfaction with both versions posted in the thread--not because of the tuning, of course, but because of the interpretations! (I know, I know! Everybody's a critic! In fact, ohmygodIcan'tbelieveI'mgoingtosaythis, but) In my estimation, they both lack passion and conviction. The Gershwin sounds like he's just throwing out jazz idioms, as if to say, "Here's one," sprinkle, sprinkle, "And, here's one," sprinkle, sprinkle, "And, oh, what the heck, listen to this one," sprinkle, sprinkle. On one listening, I got the impression that he'd played it one too many times, and by the time he got around to recording the piano roll, he was just tired of it! The other one is played very well, but it's too "accurate." It sounds to me like a textbook rendering. Someone recently posted a version in the Member Recordings of the Pianist Corner that's played nicely, but it's done on a digital and his interpretation is way too liquid for my taste. The Adam Carroll version is also nice, as far as it goes, played very thoughtfully and at times tenderly. Thanks so much, GPM, for posting this one, too. But in all of these, what I miss is some VERVE! So, I hope someday to share my own version of Rhapsody in Blue, and believe me, it will be VERVY! (You may have to come to the nursing home to hear it, because that's where I'll probably be by the time it's ready!)

The "Clair de Lune" brought me to tears.

As far as examples of remote keys with multiple modulations, I have wanted to hear Bach's Prelude No. 22, WTC Bk.1 in EBVT III.

Patrick , I really, really enjoyed your Bach/Siloti and appreciated the comparison posting of the Stopper tuning. It did what I wanted to hear out of the Prelude 22. However, if you take requests... No.22 is very "chordal," modulates all over the place minor and major with beautiful, soul-stirring intervals, and seems to me to be a good candidate for the purpose of this project. Also, I just came to realize recently that in my Czerny edition (G.Schirmer 1893) the last chord is B maj., where a more recent Peters edition has the last chord as B min. It certainly changes the meaning of the piece!

So, Grandpianoman and Bill, thank you very, very much for what you started and all you've done on this project. I am still feasting on and studying through this thread. So much of it is over my head, but I'm sponging it up nonetheless, trusting that it will make sense later somehow. What a terrific resource!

By the way, I got a kick out of Jake Jackson's links to Chico Marx, and Nick's recordings of his tunings. Thanks to you guys, too! This is all SO COOL!

With deepest appreciation and gratitude,
Andy Strong
Rockford, IL

Thanks Andy!

I seem to have somehow missed that suggestion earlier in this thread (WTC1 #22). I haven't played that earlier, but I played it through once just a moment ago and it is a beautiful piece, and it does sound great in EBVTIII. I'll see if I'd find the time to practice it, it would indeed be a good example along these lines.

Hi GP,

dither is actually addition of low noise. As strange as it might sound, it's needed to overcome the quite low resolution of .wav and .mp3 files (96 dB max). Especially the sound in the first of the 16 bits (just above the treshold of 0 bits = silence) need something to cling on to

We usually also dither when adding digital processing such as EQ, effects and so on, even if the media stays 24 bit after mixdown. This again could be illustrated as giving the algorithms an easier job, wrapping the last decimals in a forgiving blanket of low-level noise...

Technically speaking dithering is used to overcome quantization problems (such as, for example, in the two cases above).

Remember that if your target media is 16 bit, you can be generous with headroom up to 0dBFS if you record at 24 bits, because there you have about 6 x 24 = 144 dB of dynamic range.

Well --- Bummer! The loaner disclavier upright has been reclaimed; it's not at the school anymore. Back to square one!

Ron Koval

Thanks, Patrick! Actually, I had not made the suggestion earlier--it's just been floating around in my personal wish list as, "Wouldn't this be a great one to hear?" Sorry I didn't use my words better!...

If you would, whenever you're ready, that would be outstanding! I really enjoyed your other posts.

--Andy

Ok Andy, I'll put some work into it and see where it takes me to

Here's the last sound file from my weekend session. It's just a cycling the 5ths in the same spirit that Nick has put up EBVTIII recordings and Isaac (aka Kamin) has done for CHAS:

It's pretty pure and resonant chords, and they extend all over the keyboard. This would be a good example to listen to the pure 12th:s stretch that I used:

cycling the 5ths in EBVTIII

That was fascinating, Patrick! It is something anyone who hasn't been sure about the key color effects of well temperament need to hear. One thing no one has really heard yet is the amazing pipe organ effect. I hope either you or Nick will display a short file when you have the piano tuned very well. Simply hold the damper pedal down, play C1, C2, G2, C3 and then a C Major arpeggio from there to to C8 and let it ring. The resonance will sound amazingly like that from a pipe organ!

For more fun, cut off the striking of the keys and begin the sound file with just the resonance. Slow it down to the maximum and see if you don't really think you are hearing a pipe organ rather than a piano!

Patrick, that sounds wonderful, very easy on the ears, and the harmonies etc. all work together...beautiful!

Thanks for the explanation about dither. The digital recorder I am using uses 1-bit recording. The supplied software, "Audiogate" by Korg, has many ways to convert the 1-bit recording to a WAV or MP3. I hope I set the correct parameters. It seemed to help the recording when I added the dither and normalized it in Audiogate.

Patrick, I have really enjoyed your latest contributions and am trying to get caught up on this thread after a very busy week!

Yesterday I spent time on a customer's piano using the EBVT III and the results were truly outstanding.

I am trying to determine what standard of 'consistency' I can apply to this tuning beyond the temperament octave. I think that equal temperament is measured by consistency and maybe that's just the way I'm used to thinking about tuning.

So it has occured to me that once I am confident of my absolute accuracy in the temperament octave using Bill's instructions, then my own interpretation of what I do next will produce a good standardized result. I am sharing my impressions in case they will help anyone else and also to obtain feedback.

Bottom line is, this temperament does produce a very musical and beautiful sounding piano, when finely tuned. Beyond the temparement octave I have tried varying approaches, and I think I just may settle on this particular formula: temper the octaves and fifths beyond the temperament octave the same.

At first, I tuned fifths pure at the expense (sometimes) of octaves. As I gained consistency in the temperament, it did not seem that there were too many instances where I had to sacrifice either very much. After thinking my approach was perhaps a bit too wild, I then sort of reigned in the octaves to make them more 'correct', but this did not seem to have the same effect.

Therefore, it seems to give equal weight to both and to temper them the same, which is part of the temperament setting process anyways for this style of tuning. I think it makes sense, and will look forward to making what I hope to be some fabulous recordings, once I can get the pianist and recording engineer together again. Thanks,

Nick

Nick,

good to hear from you again. As you might have seen, I've been elaborating a bit on the stretch myself. I used pure 12ths for the last recordings, which happened to work pretty well on this particular piano (the Yamaha G2).

I tuned a small upright yesterday, though, and had to use less stretch not to end up with wobbling octaves. So i think the equal-beating 12/15ths that Bill uses might be the best overall starting point.

I've tried less stretch too, but then the temperament loses some of its sparkling openness.

I'm glad that we are investigating this together, please keep me updated. I will do the same!

Patrick

What happens if you transpose that up or down a half-step?

(Actually, any effect from doing something like that would probably depend more on how well the octaves were tuned rather than the temperament.)

Bill, can we see your temperament for this?:

http://news.bbc.co.uk/2/hi/programmes/world_news_today/8594836.stm

At last we can hear that instrument played, and to whom he can be of some use

It is really more a dulcimer with a keyboard and a half step tuning feature, than a piano (no hammers I question the repetition action)

The justness may be fun ! as all those strings may need a frame, and I have not see one - may be it is underside, as on cymbalums (is dulcimer the name for )

Or may be the strings are more (probably) kind of thick harpsichord strings.

Nick said :
temper the octaves and fifths beyond the temperament octave the same.

What is to be understood, to me, is that that way change the way you listen, from an analytic octave/partial based listening, to a musician like one.

One of the most important thing to get to produce musical tunings.

Then, comparing the 5ths and the octave is not easy, so I'd stay with the 12ths vs double as being easier, and smoother too.

Nick and Patrick, I think you both have the right idea. It was all spelled out in the detailed instructions I sent you. You can't simply tune pure 5ths after going up and down from the temperament octave. What you will want to do is to tune an octave, then compare it to the 5th formed with the upper note of that octave and equalize both the octave and the 5th.

Since the 5ths are of differing sizes, this means that the widths of the octaves will vary too. For example, when you tune G4, the 5th below it is C4. Looking at the Jason Kanter graph of theoretical values, the C Major 5th is -2.64. (1.2 beats per second, theoretically). If you tuned a very pure sounding G3-G4 octave, the C4-G4 5th would sound uncomfortably narrow, beating slightly more than once per second. If you widen the octave slightly, it will improve the 5th.

It is quite a reasonable compromise to make. If you were to make the C4-G4 5th perfectly pure, it would cause the the G3-G4 octave to be overly wide. It would also cause you quite a dilemma when you try to tune D5. If you also tuned the G4-D5 5th pure, the D4-D5 octave would be terribly wide, not an acceptable sound at all.

So, making the octave and 5th equal beating is quite a natural way to tune. In ET, if the temperament octave has a slight width to it to begin with, the ascending octaves and 5ths can all be made to sound exactly alike and that produces a beautifully smooth and clean sound for ET. Once you reach F5, you can make the double octave and the octave and 5th beat exactly alike all the way to the top and that gives you a very nice sound for ET.

Now, for the EBVT III, I use the very same idea but because the 5ths vary in size, the octave widths vary also. Take G#4, for example: The G#3-C#4 4th is tuned pure in the temperament. Therefore, for the G#3-G#4 octave and the C#4-G#4 5th to be exactly the same, the G#3-G#4 octave has to sound very pure. So, you will have a G3-G4 octave with a bit of width in it and a G#3-G#4 octave very pure right next to each other.

This would not happen in ET but it does happen quite naturally when tuning a well temperament with irregularly sized 5ths aurally. No ETD program that I know of can do that. It will dictate some kind of smooth stretch curve or another. With the above example, it may not sharpen G4 enough but it would sharpen G#4 too much. It may not be so bad if all one can do is use an ETD program but it can never be quite as good as tuning the octaves by ear would be.

Tuning the octaves and the 5ths to beat equally creates enough of a dilemma that you would not want to tune octaves any wider than that. As I mentioned before, when you reach D5, you have already sharpened G4, so to get the G4-D5 5th to not beat objectionably, the D4-D5 octave will have to have about a beat per second width in it. That is about as wide as you would want to go. Other octaves near it would not be so wide.

Once you reach F5, you can use the sostenuto pedal to play the F3-F5 double octave. Tune the double octave so that it sounds pure first. Then use the sostenuto pedal to play and hold the A#3-F5 octave and 5th. You should hear some beating. Sharpen F5 slightly until the double octave and octave and 5th sound the same (equal beating). Neither will have a very perceptible beat. They will both sound apparently "pure".

The double octave will still be slightly wide and the octave and 5th will still be slightly narrow but the amount of tempering there will be in each can barely be heard, if at all. This is where the "magic" happens. I believe it is what Herr Stopper and Sr. Capurso are looking for when they tune in ET. I use the same idea when tuning any unequal temperament. The double octaves and octave and 5ths will continue to vary slightly, one to the next, just as the octaves and 5ths did in the upper fourth and lower fifth octaves.

Do this same procedure from F5 to E6. You will find it very easy to do when you use the sostenuto pedal. It is what I have called the "mindless octaves" approach. The PTG Journal will have an article on that in a few days. It is so easy to do, you don't need to really think about it. You are just eliminating beats.

Now, when you reach F6, it becomes even easier. Simply tune F6 pure to A#4 below it. Use the sostenuto pedal and tune a pure interval, as simple as that. After there are no more dampers, you won't need the pedal anymore. Do this for another octave until you reach F7. If you can still hear the beats, you can keep tuning pure 12ths but if you can't clearly hear a beat in that range, you can simply tune the octaves from F7 to C8 to where they sound "sweet" to your ear.

For the Bass, you do a mirror image of what you did in the treble. Compare octaves and 5ths and make them equal beating. Compare double octaves and 12ths and make them equal beating. The lowest octaves can be tuned simply for the nicest resonance.

The "pipe organ" effect is best heard with the long, C Major arpeggio for a few different reasons. The C Major third is the slowest and the M3s and M6s have been tuned as equal beating, so they cancel each other to a large degree, so they are not heard to beat as rapidly as they actually do. The octaves and 5ths and all of their multiples are also equal beating, so they also cancel each other. When C8 is played as the final note of the arpeggio and if it is tuned sharp enough, it will cause all the pitches which have coincident partials to it to be excited one last time across the entire scale.

All of those equal beating intervals are "locked" against each other. This causes the entire large chord which spans the entire keyboard to have a slow, "phasing" sound rather than rapid beats. Instead of conflict, there is reinforcement of the sound. It seems to swell at first, then slowly decay. Every single person to whom I have demonstrated it has agreed that it sounds to them like a pipe organ.

Now, of course, this has very little to do with the actual playing of any music unless the music itself ended that way. I don't know of any music that actually does but it is possible, of course. So, it is simply a side effect of tuning an equal beating well temperament with equal beating octaves and 5ths. It is exciting and fun to listen to, however and amazes everyone who hears it. It confirms for my customers that the piano is really in tune with itself from one end to the other.

When I heard Herr Stopper's tuning at a PTG convention, we tried that same trick and we did hear some of that phenomenon because of the way he tuned the octaves. The full effect was not heard, however because the M3s and M6s were not tuned as equal beating the way they are in the EBVT III. The pipe organ effect is not the reason for tuning the EBVT III with tempered octaves, it is simply an anecdote but an interesting and fun aural experience.

Bill, that is a great way of putting it - making the two intervals (12th/15th) as calm as possible!



Isaac, I agree. You (+ Alfredo indirectly) and Bill have been most important to me in learning how to stretch the piano (you still are... ), and I very much like that you both look at it from a musical point of view, and look for the techniques to get the desired result.

---

I believe strongly in 12th/15ths. To me, the pure 12ths are too wide, the piano doesn't hold harmonically together. This, again, is my own opinion and in no way any claim for truth.

EBVTIII seems to be more forgiving about stretch. I tuned ET with pure 12ths for a recording session last week. I didn't get any angry phone calls, but I met the piano player today. I mentioned that I tuned the piano wider than the last time I tuned for him, and he instantly replied "Well, that I noticed...".

This was for a demo recording, so no big harm done. Still, with close-up mics (like in this case) I'll settle for a more compact stretch in the future.

I recall Ed Foote mentioning something along these lines, that close-up mics (in his tunings for Nashville commercial studios) asked for a more conservative stretch.