 The Clarinet BBoard
|
Author: perryg114
Date: 2007-10-08 01:47
I am starting to rebuild old clarinets and am also trying to get back into playing again after many years. I downloaded a free tuning program from the following link which I found on an old post from this sight.
http://www1.ocn.ne.jp/~tuner/tuner_e.html
The programs are pretty neet although I did find a 3Hz error in the pitch measurement but the tuning fork program is correct and can be used to adjust the tuning program. "A" needed to be set to 437Hz to be in tune with a 440 Hz tone. I figured this out using my guitar tuner. I will eventually get an Korg CA-30.
On two clarinets one a professionally rebuilt Noblet 40 and one I rebuilt I can stay within +-10 cents on most notes and a few are closer to +-20 cents. I found I had to pull the barrel about 1/8 in on both clarinets to get the low register (register key closed) to keep from playing sharp. With the register key open the barrel postion did not seem to matter. I also tried different mouth pieces with about the same results.
I know the clarinet does not play in tune at all notes but what variation is considered normal.
Just a side note,my wife has started playing in a local band and she had to pull the barrel on the Noblet 40 to play in tune as well.
Perry
|
|
Reply To Message
|
|
Author: William
Date: 2007-10-08 15:13
Tom Ridenour (master designer and accoustician of Leblanc's Opus and Concerto clarinets) considers a 2 cents variation--plus or minus--to be "acceptable". He adjusted the scale on my A & Bb set of Concertos to that tolerance and I have little difficulty in playing "in tune" with anyone.
Post Edited (2007-10-08 21:55)
|
|
Reply To Message
|
|
Author: sfalexi
Date: 2007-10-08 16:14
For one thing, make sure you adjust your clarinet correctly. I usually tune the open G by adjusting at the barrel, the clarion G by adjusting at the middle joint, and the C just above the break by adjusting the barrel. I feel it gives me a good "overall" tuning.
I think 20 cents to be too much. Ten cents seems a bit much too. I think I'd be ok with a clarinet that's withing ten cents all the time (+-5 either way).
But that's my opinion which on a matter like this REALLY doesn't mean that much. I have to play on clarinets that can be WAY out and I just have to "deal" with them.
Alexi
US Army Japan Band
|
|
Reply To Message
|
|
Author: perryg114
Date: 2007-10-08 22:23
Considering my playing skills +-10 Cents is probably not that bad. I probably need to get someone who can play better than me and see if they are in tune. I am not 100% sure the tuning program I am using is working correctly either. Thanks for the tuning advice. I will put it into practice.
Perry
|
|
Reply To Message
|
|
Author: interd0g
Date: 2009-07-23 20:18
I found what appears to be an excellent tuning meter on line and for the first time got a good look at my tuning.
The R13 I have just switched to ,varies from -20 to +5 cnts over the range.
With a 5RV mouthpiece I find that I can't pull the pitch very much at all by any changes with embouchure etc., and reed resistance also doesn't affect things much. The tuning is stable.
I didn't realise that instruments would be that far off in factory condition.
|
|
Reply To Message
|
|
Author: ww.player
Date: 2009-07-23 21:01
According to the musical acoustics prof I had in college, the average human ear hears two notes as in tune that vary as much as 2 cents flat or 3 to 4 cents sharp from each other. This means that -3 or +5 will sound out of tune to most everyone. I believe good musicians are a little more sensitive to pitch than this.
Note that virtually everyone is much more sensitive to hearing flat pitches as out of tune compared to sharp ones. I know quite a few trumpet players that illustrate that fact, if you know what I mean. They hear sharp as just bright.
As to acceptable variation, that depends on the player and situation. Once you get beyond 10 cents sharp or flat, you are definitely in the range where it takes some major adjustments in the air/embouchure to tune. Most advanced players will make whatever adjustments are necessary, often without even thinking about it. That being said, I wouldn't want to regularly play a clarinet that has a lot of notes that are more than 12 cents out of tune, especially if any of those are on the flat side. It's a lot easier to lower a sharp note than raise a flat one.
Post Edited (2009-07-24 19:01)
|
|
Reply To Message
|
|
Author: Arnoldstang
Date: 2009-07-24 03:17
One thing to avoid would be a scale with neighboring notes...one very flat and the next very sharp.
Freelance woodwind performer
|
|
Reply To Message
|
|
Author: John Peacock
Date: 2009-07-24 09:47
Attachment: pitch_3bs.png (75k)
I find it's hard to get an accurate reading on pitch with a portable tuner: they can only tell you if a note is way out. But with digital recording, it's easy to make an accurate measurement once you have the sound file on a PC. Here is a plot of the results I get for three Bflats: Buffets from 1982 (R13) and 1936 (pre-R13) and a B+H 1010. The results have been given an arbitrary vertical displacement. The horizontal axis is concert pitch, so 0 means written B, just over the break. The vertical scale is in the sense that up means sharper.
The plot confirms why professionals like R13's: it's generally within +/- 5% of a semitone (5 cents). The exceptions are written top Csharp (sharp) and top E (flat). I was aware of these, and would normally be prepared to lip these notes into pitch, but for the purpose of this exercise I tried not to make such adjustments. The Pre-R13 design does less well in general: it tends to be sharp in the lower register, and has bigger note-to-note fluctuations. The much-maligned 1010 does better than the old Buffet in much of its range, but goes horribly sharp at the top.
I've been meaning to post this plot since Tony Pay's recent posting on playing properly in tune, as opposed to equal temperament. Of course, equal temperament is just a starting point, and you have to use your ears to bend the notes according to key and what your fellow players are doing. But we shouldn't forget how hard it is to even get the machine playing somewhere close to equal temperament.
|
|
Reply To Message
|
|
Author: clarnibass
Date: 2009-07-24 10:10
>> According to the musical acoustics prof I had in college, the average human ear hears two notes as in tune that very as much as 2 cents flat or 3 to 4 cents sharp from each other. <<
This depends on how exactly somone hears those notes. Maybe your accoustics teacher gave you the exact details of the circumstances that people could hear those differences?
Locally, according to one of the people from the university acoustics department (AFAIK, it's one of the best in the world), the area where people can hear a difference of pitch is a frequency that is very close to G above the staff. When hearing two notes one after the other, people can hear about 5 cents difference. Above and below that it is harder to hear a difference. For middle C it is about 8 cents difference. Of course in music it changes depending speed, range, and everything else that affects the way you hear the notes.
|
|
Reply To Message
|
|
Author: David Spiegelthal ★2017
Date: 2009-07-24 13:42
I don't always achieve it on every note on every instrument, but my target is +/- 5 cents, using the trusty little Korg digital tuner set at A-440. Of course some notes can be 'favored' considerably by the performer with little difficulty, but there are other notes (e.g. the throat tones) which are hard to adjust while playing and really need to be 'dialed in' on the instrument itself. Fortunately those are relatively easy notes for the technician to "tune".
|
|
Reply To Message
|
|
Author: FDF
Date: 2009-07-24 16:47
For my knowledge, could someone explain the markings on the Krog CA-30. Does the + or - 20 mean plus or minus 20 cents? And what is the purpose of the triangular pip? Thanks for your help.
|
|
Reply To Message
|
|
Author: Grabnerwg
Date: 2009-07-24 18:38
I think we are all barking up the wrong tree here.
The important thing is to be able to match pitch with another instrument, and/or fit your note into a given harmony to produce the best results. And, as any accomplished clarinetist will tell you, this can be a very different thing than matching your clarinet up against a tuner.
Don't get me wrong, I think tuners are excellent tools. But just stopping the needle on a given pitch or series of pitches, is no guarantee that you can play "in tune" in an ensemble.
Walter Grabner
www.clarinetxpress.com
Summer time sale
|
|
Reply To Message
|
|
Author: David Spiegelthal ★2017
Date: 2009-07-24 18:45
You're 100% correct, Walter, but often while we're playing in an ensemble we can't hear ourselves (especially during those tutti forte passages); and there are those solo entrances where we have to come in on pitch with no external reference to match -- THOSE are the times we need our instruments to be as "inherently" in-tune as possible. An instrument that has been properly tuned in the shop or at home with reference to a tuner, at least gives the performer a fighting chance to play in tune when he/she doesn't have the opportunity to match the other players nearby. I realize I'm preaching to the choir.
|
|
Reply To Message
|
|
Author: mrn
Date: 2009-07-24 20:05
Walter's right. It's whether you can make the instrument play in tune in a given context that counts.
This ties into a thread we had a few weeks ago about "locally just" tuning. Tuning machines are designed to tell you if you are sharp or flat according to equal temperament tuning, the way pianos are tuned. But what is musically in-tune at a given moment in a given ensemble is, most of the time, not equal temperament tuning--in the case of chords, for instance, just intonation often sounds the most in-tune. A just interval can be as much as 15 or (as in the case of minor sevenths) even 31 cents flatter or sharper the equal temperament version of the same interval.
So sometimes, even in a perfect world where everyone in the ensemble plays right on pitch, a little bit of flatness or sharpness is what you want. For example, take an E natural. If you're the third in a C Major chord, you want that E to be about 13-14 cents flatter than equal temperament to make the major 3rd ring true (actually, technically what you want is the interval to be 13-14 cents narrower--the C is not necessarily at A=440 equal temperament, either). So even if you're in tune with the needle, you will still often find it necessary to make significant adjustments like that when playing in an ensemble. I don't usually have to think about it that hard--like most of us, I've been conditioned to make those sorts of adjustments on the fly without thinking about chord names and such--but that's what's going on.
Of course, there is something to be said for having an instrument that is close to equal temperament on all or most notes, because at least that way you don't have to adjust your pitch any MORE than you would from equal temperament (which is, itself, a compromise between different more "natural" tuning systems). But what matters more, I think, is the ability to hear in your mind what in-tune playing should sound like (in the given musical context) and then reproduce that with the instrument. It's like singing.
Post Edited (2009-07-24 21:03)
|
|
Reply To Message
|
|
Author: Koo Young Chung
Date: 2009-07-25 03:52
To FDF:
Those triangles denote diatonic major (or minor) third.
For example equal temperament major third is too sharp ,so you have to
flatten to the triangle marking to achieve 4/3 mathematically correct interval.
For minor third,equal temperament is too flat so you have to raise the pitch to the triangle marking to get correct intonation(9/8).
Post Edited (2009-07-25 13:00)
|
|
Reply To Message
|
|
Author: interd0g
Date: 2009-07-25 19:40
I use the tuner from 'seventhstring.com'.
It gives immediate readout with error in cents plus two visual indicators of the mis tuning condition, and seems immune to the level or quality of the sound and to background noise.
I also like their tuning fork and metronome
|
|
Reply To Message
|
|
Author: John Peacock
Date: 2009-07-25 22:42
David,
"what program did you use to get the pitches measured in your computer?"
I did it by hand: read the .wav file into a fortran program, and wrote something to do Fourier transforms of the data. But after I finished this, I discovered Cool Edit will tell you the frequency of a given note - and the numbers matched pretty well with what I was getting, so I could have saved myself some effort.
|
|
Reply To Message
|
|
Author: clarnibass
Date: 2009-07-26 03:04
>> but often while we're playing in an ensemble we can't hear
>> ourselves (especially during those tutti forte passages);
Also in amplified concerts especially if playing with a rock band (usually it's fine and then their loudest is much louder than your loudest). For some reason the clarinet (and even more the bass clarinet) is the hardest to hear themselves.....
>> and there are those solo entrances where we have
>> to come in on pitch with no external reference to match
Sometimes it's not just entrances. Just one example a film sountrack where you record your part and other parts will be recorded later.
>> Don't get me wrong, I think tuners are excellent tools. But
>> just stopping the needle on a given pitch or series of pitches,
>> is no guarantee that you can play "in tune" in an ensemble.
Yes, but ensemble could be many things. For example, it could be all electronic instruments (i.e. played by a prgram on a computer) that are robotically tuned to (let's say) 440. Sometimes it's like this plus the film soundtrack I mentioned (i.e. you record with electronics and others player are recording later, etc. etc.).
|
|
Reply To Message
|
|
Author: Phat Cat
Date: 2009-07-26 12:25
"...a fortran program..."
INTEGER IHWSTR(3)
DATA IHWSTR/4HHOLY, 9HMAINFRAME, 6HBATMAN/
Ah, the good old days! Punch cards, disks the size of washing machines and printers the size of cars.
|
|
Reply To Message
|
|
Author: mrn
Date: 2009-07-26 15:31
Phat Cat wrote:
> INTEGER IHWSTR(3)
> DATA IHWSTR/4HHOLY, 9HMAINFRAME, 6HBATMAN/
>
> Ah, the good old days! Punch cards, disks the size of washing
> machines and printers the size of cars.
Hey, Fortran's still alive and well. It's still the language of choice for a lot of serious number crunching tasks.
|
|
Reply To Message
|
|
Author: Phat Cat
Date: 2009-07-26 17:40
"Hey, Fortran's still alive and well."
In the Department of Programming Archeology?
|
|
Reply To Message
|
|
Author: John Peacock
Date: 2009-07-26 20:20
Fortran was designed a lot more recently than the clarinet. Both are perfectly adapted for certain classic tasks, Mozart in the latter case, numerical analysis in the former. It's true that fortran tends not to be the first computer language taught to science students these days - probably because the reasoning is that they are more likely to become accountants than stick to doing science. As a result, I get final-year project students who can't execute bread-and-butter tasks like reading in a table of numbers, doing some processing, and writing them out again. It's rather like encountering a player who can flutter-tongue multiphonics, but can't play C major in tune.
|
|
Reply To Message
|
|
Author: Phat Cat
Date: 2009-07-26 20:58
John:
You're right. It's amazing how much trouble recent graduates have answering the standard interview question: In the language of your choice, write a program that reads a text file containing a pair of numbers on each line and displays the resulting array sorted by the first column and then by the second column.
The answer in the vector language q, which we use for a lot of serious real-time number crunching in finance:
raze (flip v) iasc each v:("II";" ") 0: `:/data.txt
|
|
Reply To Message
|
|
Author: Mark Charette
Date: 2009-07-26 21:12
Phat Cat wrote:
> raze (flip v) iasc each v:("II";" ") 0: `:/data.txt
Too many characters ... if I could remember my APL I could get that done in half the number of characters ... and forget how I did it in less time than it took to write.
|
|
Reply To Message
|
|
Author: Phat Cat
Date: 2009-07-26 22:13
"Too many characters"
If counting characters is the game then let's do k, the ASCII successor to APL and the underlying language for q. And who needs that pesky optional whitespace?
(+v)@<:'v:("II";" ")0:`:/qdata/test.txt
My apologies to the OP for hijacking the thread. I'll crawl back into my geek cave now.
|
|
Reply To Message
|
|
Author: Mark Charette
Date: 2009-07-26 22:31
Phat Cat wrote:
> (+v)@<:'v:("II";" ")0:`:/qdata/test.txt
C'mon - who'd use such a long filename???
|
|
Reply To Message
|
|
Author: mrn
Date: 2009-07-27 02:36
Phat Cat wrote:
<<If counting characters is the game then let's do k, the ASCII successor to APL and the underlying language for q. And who needs that pesky optional whitespace?
(+v)@<:'v:("II";" ")0:`:/qdata/test.txt>>
Actually, I use J (Ken Iverson's free ASCII-based successor to APL) sometimes for the kind of thing John is describing. J has libraries for reading and writing WAV files as well as doing Fast Fourier Transforms and plotting graphs. Like APL, though, it's sort of a "write-only" language.
When it comes to pure beauty in programming languages, though, my favorite is Haskell. 
Just for fun, here's what "sort the list by both columns" looks like in Haskell:
import List
sorttwice fname = do
a <- readFile fname
let nums = map (snum . words) $ lines a
putStrLn $ show $ sortBy cfir nums
putStrLn $ show $ sortBy csec nums
where
snum :: [String] -> [Int]
snum = map read
cfir (x:_) (y:_) = compare x y
csec (_:x:_) (_:y:_) = compare x y
Post Edited (2009-07-27 04:02)
|
|
Reply To Message
|
|
Author: John Peacock
Date: 2009-07-27 10:34
To no-one's surprise, there are as varied views on elegance in computing as there are on quality of clarinet tone.
To try to drag the thread back to its origin, I'd be interested if anyone can post plots of pitch results for other clarinets, and if these show the same systematics that I see on an R13: 5 cents of steady sharpening from the bottom up to written B, then a similar flattening down to written Fsharp. Then a steady trend of flattening of about 3 cents through most of the upper register. Leaving aside the probably variable throat notes and top register, these are the main features that I see. I wonder if Tosca does better than R13 in these respects - and what about German instruments?
|
|
Reply To Message
|
|
Author: mrn
Date: 2009-07-27 16:30
John Peacock wrote:
> To try to drag the thread back to its origin, I'd be interested
> if anyone can post plots of pitch results for other clarinets,
> and if these show the same systematics that I see on an R13: 5
> cents of steady sharpening from the bottom up to written B,
> then a similar flattening down to written Fsharp. Then a steady
> trend of flattening of about 3 cents through most of the upper
> register.
I would be interested in knowing how consistent R13s themselves are in this regard. Did you do this experiment with more than one R13?
Also, which fingerings did you use for the altissimo notes? Perhaps some of us can try this experiment ourselves, but it's hard to compare results if we don't use the same fingerings.
|
|
Reply To Message
|
|
Author: John Peacock
Date: 2009-07-27 20:44
mrn wrote:
> I would be interested in knowing how consistent R13s themselves are in
> this regard. Did you do this experiment with more than one R13?
> Also, which fingerings did you use for the altissimo notes? Perhaps some
> of us can try this experiment ourselves, but it's hard to compare results
> if we don't use the same fingerings.
I only tried instruments I own. But, yes it would be necessary to try several R13's before claiming that, say, a Tosca had different characteristics.
As for altissimo, I suspect that varies more with the player. But for the record, what I used for the pitch graph was
Csharp: standard
D: standard (but no RH Eflat)
Eflat: standard
E: standard (except it's flat - so I try to play all LH down + Aflat if I have time)
F: long (I find the standard F very flat and only use it in fast passages)
Fsharp: T xxo ooo plus side Bflat key (I find standard T oxo ooo flat)
G: T oxo xxo +Eflat
From what I've heard, the flat E/F/Fsharp is a common feature (defect) of R13s.
By the way, so there is no misunderstanding, I do agree with other posters who implied that this is all a distraction from just listening and playing in tune. I remember someone who got a tuning machine when they were a novelty, checked their instrument electronically, and who was thereafter 100% confident that tuning problems were never their fault - even when it was obvious to everyone else in the room that they were way out.
But I still think it helps to have a really accurate equal-temperament instrument as a starting point - or (more realistically) to know in advance which notes are likely to need the most attention.
|
|
Reply To Message
|
|
Author: ebonite
Date: 2009-07-27 21:43
>
> To try to drag the thread back to its origin, I'd be interested
> if anyone can post plots of pitch results for other clarinets,
> and if these show the same systematics that I see on an R13: 5
> cents of steady sharpening from the bottom up to written B,
> then a similar flattening down to written Fsharp. Then a steady
> trend of flattening of about 3 cents through most of the upper
> register.
To add one to your sample, there's a similar graph on Clark Fobes' website, also from an R13:
http://www.clarkwfobes.com/articles/TuningtheClarinetforPS.htm
Patrick
|
|
Reply To Message
|
|
Author: clarnibass
Date: 2009-07-28 06:08
>> sorttwice fname = do
a <- readFile fname
let nums = map (snum . words) $ lines a
putStrLn $ show $ sortBy cfir nums
putStrLn $ show $ sortBy csec nums
where
snum :: [String] -> [Int]
snum = map read
cfir (x:_) (y:_) = compare x y
csec (_:x:_) (_:y:_) = compare x y <<
I could reply to that in a logical way..... by letting my cat on the keyboard. However, I don't have a cat.
|
|
Reply To Message
|
|
The Clarinet Pages
|
 |
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
