Melodeon.net Forums

Please login or register.

Login with username, password and session length
Advanced search  

News:

Welcome to the new melodeon.net forum

Pages: [1] 2   Go Down

Author Topic: Tremolo - constant offset VS constant beat frequency VS "tuning on the curve"  (Read 4120 times)

0 Members and 1 Guest are viewing this topic.

Kon

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 165

Happy New Year everybody!

My Erica is due a tuning, for which I intend to send her off to a fettler in the near future - but in preparation, in order to try and figure out what to ask for, I have delved into the weird and wonderful world of accordion tuning. Now I've read some very informative old threads (this one, for example) and understand the basic concepts of how the tremolo is generated and have listened to as many sound examples as I could find to get some idea of what I'd like on that front. I'm fairly certain I'd like the box swing tuned or just slightly wetter (what some people might call demi-swing, but then you've got to be careful using that word because it doesn't always mean the same thing...) - something like 1-2Hz  / 4-8 cents between the two reeds.

Another thing I've flirted with is the idea of Viennese tuning, i.e. tuning the reeds equally to above and below pitch (M-M+) rather than pitch and sharp (MM+), just because it seems a very sensible thing to do to me - but I believe with the low tremolo I'm interested in it might not even make an appreciable difference, and there's no particular reason I would need to sound exactly in tune with A=440 so I've concluded I'm not going to worry about it and go for the standard MM+.

So, that's two questions I've got answered, and so with the rambling preamble out of the way onto the actual question: What I'm stuck on is the question of how to vary the tremolo going up and down the keyboard. As outlined very clearly by Steve here many years ago the options are: to have a fixed offset in cents between the two voices (a bit odd, as low notes will have noticeably less tremolo, and high notes more); to have a fixed beat frequency as advocated by Ian Dedic (fixing the problem with the previous option, but with the result that in terms of musical pitch, measured in cents, low notes will be further apart than high notes); or to tune "on the curve", a compromise between the two.

Now the simple answer is to let the fettler go about their business, and do it according to their preferences and habits, which I'm sure will produce a delightful result. But I'd really like to know and understand what I'm asking for and getting, so the question continues to trouble me. I'm intrigued by the idea of a fixed beat frequency, but also worried it could sound a bit flat maybe, and curious whether it produces any odd acoustic effects when playing two notes simultaneously because of the in-phase or out-of-phase overlap of the two beats. So what I would really like to know is: How do the different options SOUND?

If you have a box which you know is tuned a certain way (constant offset / constant beat frequency / "on the curve") could you please post a sound example below? I'm hoping to compile a bit of an overview to help me make up my mind. Apologies if I am duplicating a previous thread, but I have not been able to find anything which answers this question with sound examples. Any general musings on this topic also welcome! And apologies for the length of the post...

Cheers,

Konrad
Logged
GC Erica

Steve_freereeder

  • Content Manager
  • Hero Member
  • ***
  • Offline Offline
  • Posts: 7511
  • MAD is inevitable. Keep Calm and Carry On
    • Lizzie Dripping

Here's the tuning curve and data I use for Hohners. It is based on the older models and gives a nice mellow tremolo, not too wet. The Viennese 'Dedic' tuning means that the treble end is also in tune with the basses and chords - something which is not achieveable otherwise.
Logged
Steve
Sheffield, UK.
www.lizziedripping.org.uk

Kon

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 165

Thanks for the response Steve! Interesting to see in detail what "the curve" might actually look like.

I have been continuing my reading of old threads and had just stumbled across the fact that "traditional" MM+ isn't properly in tune with itself across octaves nor with the bass end, thank you for pointing that out also. I think that shifts me back to Viennese! Though between the low tremolo I'm after and my uncultured ears I might honestly not hear the difference.

For anybody else wrestling with similar issues, I also found this thread very helpful.

I would still be interested in hearing (preferably solo and indoor) recordings of boxes tuned NOT "on the curve", if anybody has such a thing, but I suspect the thing to do is to go and talk to my local fettler about what I want the box to sound like and what they would advise.
Logged
GC Erica

Steve_freereeder

  • Content Manager
  • Hero Member
  • ***
  • Offline Offline
  • Posts: 7511
  • MAD is inevitable. Keep Calm and Carry On
    • Lizzie Dripping

Here's an example with my C/F 1930s Hohner which I had recently tuned to Viennese tuning according to the curve attached previously.
https://soundcloud.com/steve_freereeder/when-the-king-enjoys-his-own
Logged
Steve
Sheffield, UK.
www.lizziedripping.org.uk

Kon

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 165

Thanks Steve, that sounds lovely! I might choose to go a touch lower on the tremolo for my own box, but I agree that is a nice mellow tremolo, and the "curve" works well to my ear across the range of notes played. A helpful reference point, thank you!
Logged
GC Erica

Theo

  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 14475
  • Hohner Club Too
    • The Box Place

Kon,  with all these choices the differences are greatest with a strong tremolo.  The slower the beat you have the less significant they become.
Logged
Theo Gibb - Gateshead UK

Proprietor of The Box Place for melodeon and concertina sales and service.
See my YouTube channel  for videos on repair and tuning

Kon

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 165

Thanks Theo, that does make sense to me. If I may ask some follow-up questions:

The slower the beat you have the less significant they become.
Would you say there is a point where they (more or less)clearly reach insginificance?

Do you know how French makers and fettlers who commonly tune in the swing/demi-swing range tend to go about this? And what is your own preference when tuning in that range?
Logged
GC Erica

Theo

  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 14475
  • Hohner Club Too
    • The Box Place

I tune to a constant offset from the lowest to about 2/3rd of the way up the treble range, then I taper the offset by ear. 

As for where it makes a difference to have conventional (0/+) or Viennese (-/+) tuning I don’t know where the difference becomes audible, I suspect it varies depending on the sensitivity of the listeners hearing.   I’ve tuned a very large number of Castagnaris based on 1Hz offset and I couldn’t tell the difference between conventional and Viennese tuning with that offset. 
Logged
Theo Gibb - Gateshead UK

Proprietor of The Box Place for melodeon and concertina sales and service.
See my YouTube channel  for videos on repair and tuning

Kon

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 165

That's all very clear and helpful, thanks Theo!
Logged
GC Erica

Graham W

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 277


Do you know how French makers and fettlers who commonly tune in the swing/demi-swing range tend to go about this? And what is your own preference when tuning in that range?

Swing in French terms is often quoted as a tremolo of 1 beat per second at A4 (440), with a range from 0.6 bts at A3 up to 1.8 bts at F6 (To pick a couple of arbitary notes)
In reality, as Theo says, I've seen on more than one newly (french, GC box )swing tuned instrument that the offset appears constant at approx 1 bts up to somewhere around A4 or a little higher, and then tapers pretty progressively up to approx 2 bts at the very high end.

Many (I think probably most) current French makers and fettlers will have studied at ITEMM.fr and be familiar with the methods of Thierry Benetoux, who regularly tutors for courses on accordeon tuning there. He has published a book which contains in depth explanations and theory about tuning systems, complete with tables of values and offsets, curves etc .
It's fascinating, but probably too much info!

Theo's advice is 'real world', works well and gives a lovely sounding result.
Logged

Kon

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 165

Thanks for that extremely informative reply Graham! That adds some helpful context, feel like I'm starting to piece together a bit of a clearer picture.

I had read about ITEMM and realised a lot of current French makers trained there, but it didn't occur to me to ask who teaches them to tune...

Re the Benetoux book:
It's fascinating, but probably too much info!
Sounds exactly like my kind of book! Maybe if I save my pennies...

Though I'm sure you're right it might exceed my requirements.
Logged
GC Erica

Steve_freereeder

  • Content Manager
  • Hero Member
  • ***
  • Offline Offline
  • Posts: 7511
  • MAD is inevitable. Keep Calm and Carry On
    • Lizzie Dripping

I tune to a constant offset from the lowest to about 2/3rd of the way up the treble range, then I taper the offset by ear. 

Even if I am tuning to a calculated curve, I will always do a final check by ear and I will be prepared to deviate from the curve/tuning meter so that it sounds right.

Quote
As for where it makes a difference to have conventional (0/+) or Viennese (-/+) tuning I don’t know where the difference becomes audible, I suspect it varies depending on the sensitivity of the listeners hearing.   I’ve tuned a very large number of Castagnaris based on 1Hz offset and I couldn’t tell the difference between conventional and Viennese tuning with that offset.

The main point about Viennese (-/+) tuning is to ensure that the LH bass and chords are in tune with the RH treble end. I agree that with a 1 Hz tremolo there will be very little audible difference between conventional (0/+) or Viennese (-/+) tuning. I would think that around 2 - 3 Hz tremolo it would start to make a difference and hence an advantage to use Viennese tuning.

Another factor which has not been discussed yet is what to do in the case of a three voice LMM instrument. You have to decide on the relationship of the L voice with the MM voices, which in turn is governed by the action of the stops or switches on the treble end:

(1.) If there is a single stop which just switches the L voice in or out, so that the MM reeds are always sounding, then it is possible to use Viennese (-/+) tuning with the L reeds being tuned at pitch with zero offset and the MM reeds tuned M-/M+. The perceived pitch of the combined M-/M+ should ideally be 'at pitch' and hence sound OK with the L reeds.

(2.) If the individual M reeds can be selected separately, e.g. on Castagnari Mory, Handry, Saltarelle Nuage, etc. giving possible combinations L, LM, LMM, MM, M, then Viennese tuning is not appropriate, because for the LM combination, the reeds must be tuned in octave unisons L0 M0. Otherwise Viennese L0 M- will sound horrible. So in this case conventional tuning has to be used L0 M0 M+.



Logged
Steve
Sheffield, UK.
www.lizziedripping.org.uk

BobM

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 310

I've have my Sandpiper "Viennese" tuned +/- 5 cents both with a bias outwards from the middle A. The instrument sounds more in tune with itself to my ears, and the tremolo is not too thin or overblown, it also has a brass like attack and some complexity when playing RH chords.
I had a PA tuned in a similar way, +7/-5 so I knew what I was getting into, and have no regrets at all. Bear in mind that if playing with others there may be other factors to consider.
Bob.
Logged

Theo

  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 14475
  • Hohner Club Too
    • The Box Place

I've have my Sandpiper "Viennese" tuned +/- 5 cents both with a bias outwards from the middle A. The instrument sounds more in tune with itself to my ears, and the tremolo is not too thin or overblown, it also has a brass like attack and some complexity when playing RH chords.
I had a PA tuned in a similar way,


That would only be a fair test if previously it was freshly tuned to “standard” tuning with the same spread.
Logged
Theo Gibb - Gateshead UK

Proprietor of The Box Place for melodeon and concertina sales and service.
See my YouTube channel  for videos on repair and tuning

BobM

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 310

I've have my Sandpiper "Viennese" tuned +/- 5 cents both with a bias outwards from the middle A. The instrument sounds more in tune with itself to my ears, and the tremolo is not too thin or overblown, it also has a brass like attack and some complexity when playing RH chords.
I had a PA tuned in a similar way,


That would only be a fair test if previously it was freshly tuned to “standard” tuning with the same spread.

I don't understand "fair test". No test required, I asked for a specific rate, but I also knew that there were going to be adjustments towards the extreme ends.
Logged

Theo

  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 14475
  • Hohner Club Too
    • The Box Place

Fair test - a simple way to explain the principle used in science to alter one one variable if you want to make a direct comparison between two sets of data. So if it was a while since the box was last tuned then tuning to either system is going to sound better.
Logged
Theo Gibb - Gateshead UK

Proprietor of The Box Place for melodeon and concertina sales and service.
See my YouTube channel  for videos on repair and tuning

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...

Happy New Year everybody!

My Erica is due a tuning, for which I intend to send her off to a fettler in the near future - but in preparation, in order to try and figure out what to ask for, I have delved into the weird and wonderful world of accordion tuning. Now I've read some very informative old threads (this one, for example) and understand the basic concepts of how the tremolo is generated and have listened to as many sound examples as I could find to get some idea of what I'd like on that front. I'm fairly certain I'd like the box swing tuned or just slightly wetter (what some people might call demi-swing, but then you've got to be careful using that word because it doesn't always mean the same thing...) - something like 1-2Hz  / 4-8 cents between the two reeds.

Another thing I've flirted with is the idea of Viennese tuning, i.e. tuning the reeds equally to above and below pitch (M-M+) rather than pitch and sharp (MM+), just because it seems a very sensible thing to do to me - but I believe with the low tremolo I'm interested in it might not even make an appreciable difference, and there's no particular reason I would need to sound exactly in tune with A=440 so I've concluded I'm not going to worry about it and go for the standard MM+.

So, that's two questions I've got answered, and so with the rambling preamble out of the way onto the actual question: What I'm stuck on is the question of how to vary the tremolo going up and down the keyboard. As outlined very clearly by Steve here many years ago the options are: to have a fixed offset in cents between the two voices (a bit odd, as low notes will have noticeably less tremolo, and high notes more); to have a fixed beat frequency as advocated by Ian Dedic (fixing the problem with the previous option, but with the result that in terms of musical pitch, measured in cents, low notes will be further apart than high notes); or to tune "on the curve", a compromise between the two.

Now the simple answer is to let the fettler go about their business, and do it according to their preferences and habits, which I'm sure will produce a delightful result. But I'd really like to know and understand what I'm asking for and getting, so the question continues to trouble me. I'm intrigued by the idea of a fixed beat frequency, but also worried it could sound a bit flat maybe, and curious whether it produces any odd acoustic effects when playing two notes simultaneously because of the in-phase or out-of-phase overlap of the two beats. So what I would really like to know is: How do the different options SOUND?

If you have a box which you know is tuned a certain way (constant offset / constant beat frequency / "on the curve") could you please post a sound example below? I'm hoping to compile a bit of an overview to help me make up my mind. Apologies if I am duplicating a previous thread, but I have not been able to find anything which answers this question with sound examples. Any general musings on this topic also welcome! And apologies for the length of the post...

Cheers,

Konrad

I assume you've read my posts about why I chose a fixed offset in bps/Hz, other people have different preferences. With only 1Hz-2Hz offset between the two reeds there's not that big a difference between Viennese/Dedic tuning (M-/M+) and standard M/M+, the difference becomes more obvious as the offset goes up, especially for the lower-pitched notes.

Maybe not relevent if you're going for a relatively close tuning but you asked about the sound -- bigger offsets sound louder to the ear all the way up to about 7Hz, which is why old-style opera singers and violin players ended up around here to get the loudest apparent volume when performing unamplified even at the expense of sounding a bit harsh on the ears. To most ears nowadays a melodeon tuned like this is far too wet, though some of the old dance band players (especially in Scotland) weren't far off this.

An often missed key point about tuning is that cents were/are used for describing tuning/temperaments because for the middle of the tuning range this is how the ear percieves absolute pitch error, so it matters for instruments with a single voice -- which means most of them, including all orchestral instruments about which most of the literature on tuning and temperaments has been written. But this isn't what the ear hears with detuning/tremolo on a two-voiced instrument like an MM melodeon, you hear the average pitch as the absolute and the difference as vibrato/tremolo -- and in this case the ear hears bps/Hz as the difference not cents, because it's hearing amplitude changes over time not frequency difference. As the beat rate goes up so does the apparent volume, but this flattens off at around 7bps and just sounds increasingly nasty above this (and actually a bit below this too).

Published research papers found that the above-mentioned ~7Hz opera/violin tremolo rate didn't vary much at all with pitch, which is what you'd expect from this fact. When I tried out M-/M+ detuning experiments (with synthesized reed sounds) to try and get what I found was the best volume/pleasantness compromise I found the same, I preferred around 4Hz with very little variation along the keyboard -- this was a surprise to me at the time because it's not how they're conventionally tuned, but it's definitely what I found by actually trying it for myself instead of just following convention.

I don't know if other fettlers have tried similar experiments, or their tapering is because of being drawn towards "traditional" detuning with slower tremolo on low notes and faster on high notes, either because they think this is best or it's what customers ask for -- can anyone comment on this?

Anyone nowadays with a PC and synthesizer software could easily try similar a similar experiment for themselves -- if you do this (I'd love to hear other people's results), make sure to use detuned reed sounds which are similar to a melodeon, because the results using sinewaves are not the same due to lack of harmonics in the spectrum (reeds sound harsher at big offsets).
« Last Edit: January 03, 2021, 04:12:14 PM by IanD »
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

Kon

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 165

Thanks for the response Ian, that fills in some very helpful detail! Interesting particularly about the difference between single-voice and two-voice pitch perception.

I suspect our considerations about tuning are a bit different, as I don't play outdoors or for dance, so I'm not really looking for a pleasantness/volume trade-off, simply for maximum pleasantness (in my ears) - horses for courses, of course.

What I still don't entirely know is whether I might find a constant offset more or less pleasant than "on the curve" tuning, but I would be choosing the constant offset (or not) for slightly different reasons than you.

I suspect you're right that playing with some synthesised notes would be the most sensible way to explore this in detail, maybe this will motivate me to take on that learning curve!

Cheers,

Konrad
Logged
GC Erica

Winston Smith

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 4066
  • AKA Edward Jennings
    • "Our Luxor B&B" Luxor life, slice by slice.

Didn't Mr Microbot have tuning sound samples on his site, real ones, I mean?
Logged
At last, broken and resigned to accept conformity.
Oh, how I LOVE Big Brother!

Steve_freereeder

  • Content Manager
  • Hero Member
  • ***
  • Offline Offline
  • Posts: 7511
  • MAD is inevitable. Keep Calm and Carry On
    • Lizzie Dripping

Didn't Mr Microbot have tuning sound samples on his site, real ones, I mean?
Yes - there are sound clips of different types of Hohners, but not systematic examples of the entire keyboard range from low to high, which is what you would need to compare tremolo rates across the range.


Edit:
Here is the webpage with the samples.
http://www.michaelrowbotham.com/melodeons.html
« Last Edit: January 03, 2021, 06:04:53 PM by Steve_freereeder »
Logged
Steve
Sheffield, UK.
www.lizziedripping.org.uk
Pages: [1] 2   Go Up
 


Melodeon.net - (c) Theo Gibb; Clive Williams 2010. The access and use of this website and forum featuring these terms and conditions constitutes your acceptance of these terms and conditions.
SimplePortal 2.3.5 © 2008-2012, SimplePortal