Mike's Oud Forums - Powered by XMB

Analysis of Nahhat Bracing

FastForward - 7-7-2010 at 03:35 AM

I found an image of an Abdo Nahhat oud and wanted to determine the relative spacing between the braces, here is what I found (see image).

I assumed that the total length is 49cm.

- String Length is 60.

- Distance between braces around the bridge is not even, the brace towards the neck is closer.

- The width of the area around the bridge is 9.1cm (actual number might be 9cm)

- The braces are not entirely parallel. The layout varies around the bridge and large rosette.

The info is quite interesting. Would love to hear your opinions and feedback.

fernandraynaud - 7-7-2010 at 04:50 AM

Hi, Edward,

Can you clarify: do you know the scale is 600 mm and the body 49 cm? is this a photo of a soundboard taken off the instrument that this Michael Cone published? Or is it shot the from top with braces drawn in? By hand? At what accuracy do you think the braces are drawn? Is left left or is it reversed? What are the curved "braces" on the right of the main soundhole?

For me, and I'm no luthier, an interesting thing is that most every oud I have looked at seems to have a roughly (anyway) similar layout, except for Sukars. It is fascinating having a chance to examine several instruments from the same maker. Sukars lack the little partial braces in the center of the main soundhole, and instead there are braces on both sides of the main soundhole that are parallel to the length and that join the 181.5 and 87.5 braces about halfway between the soundhole and the edge, creating little "chambers". By tapping and recording into an analyzer I can see that Sukar's soundboard this way seems to create more distinct "domains of resonance", resulting in the Sukar's typical greater number of distinct "resonance peaks" (and perhaps its greater volume) compared e.g. to a typical Egyptian.

A curious thing, by the way, about Sukars is that the different models have great differences in feel and timbre, even though the basic design seems to be common. A main difference appears to be the woods used on the neck and the bowl. The "deep-sounding" model 14 with its black walnut bowl sounds very different from the 212 that uses lighter-colored types of walnut(?) in alternating ribs. I'm scratching my head.

I'm not belittling the old Nahats, but there's no question that the modern Boutique Guitar is an amazing powerhouse compared to older guitars. Guitar-making has suddenly enormously progressed in the last couple of decades, a combination of science, better tools and unprecedented devotion, and a lot of novel ideas have appeared that can probably be tried on the oud without destroying its character, for instance many alternate bracing designs that have vastly improved volume and tone balance on Classical Guitars.

There's an interesting series on YouTube, search for "Gourmet Guitars", that features interviews with many prominent luthiers, who discuss their fascinating pet ideas. If I were building instruments, I'd definitely study all of these.

FastForward - 7-7-2010 at 02:21 PM

- I am not Edward, Ibrahim would be more accurate.

- I made an assumption that the length is 49cm. Based on that and from the image the scale can be computed and it is 60cm. Any dimension can be computed using (# of pixels)*49/330.

- the photo is available online, I just googled Nahat bracing and it was one of the images that showed up.

- the soundboard was not taken from the oud. You can determine brace locations by inserting a very bright light inside the oud and taken the image in a dark room. So left is bass, right is treble. Its a right handed oud I believe.

- the curved braces are nothing but the electric cables for the light that was inserted

- I am not too interested in Guitar braces, though I have seen most of the series you referred to.

- I have seen different variations on the bracing where some braces are missing or others are added. I am more into trying to analyze the pattern above and would like to see other examples/patterns.

I have seen Faruk's bracing for some ouds and he has one of the braces slanted and not parallel. There are also other examples. What is more interesting above is that its not just one brace, there are several.

I am also interested in comparing dimensions, such as the area around the bridge, distance between the bridge and the braces around it, etc...

I know the Nahhat pattern in Dr. Oud's book is different from this and would like to see other variations within the Nahhat ouds.

FastForward - 7-7-2010 at 11:22 PM

It seems that Jameel did some work on this oud.

http://www.khalafoud.com/cone_nahat.htm

Jameel can you clarify this and potentially substantiate some of the dimensions assumed/computed?

I am also wondering about the size of rosettes. They seem to bigger after the restoration. I would guess that the image above was taken after restoration.

Also, any chance that you have a sound sample.

Thanks,

Ibrahim...

Jameel - 7-8-2010 at 07:33 AM

Michael Cone did the work on that oud. Nahat bracing is all over the place. It's a secret yet to be "unlocked". The rosettes are original size, as far as I know. I might have a sound sample. I'll do a little digging....

FastForward - 7-8-2010 at 10:06 AM

Thanks for the clarification Jameel. Yes, the Nahat bracing is an interesting subject. Too bad Elias's diaries and notes were thrown away.

The rosettes are definitely a different size on the restored oud. Here is an image I made for comparison. The main reason that made me believe this was the rosette sizes determined from the above image which shows a small rosette with about 4.6cm diameter.

Thanks again Jameel.

farukturunz - 7-9-2010 at 05:29 AM

Quote: Originally posted by FastForward

- Distance between braces around the bridge is not even, the brace towards the neck is closer.

- The width of the area around the bridge is 9.1cm (actual number might be 9cm)

- The braces are not entirely parallel. The layout varies around the bridge and large rosette.

The info is quite interesting. Would love to hear your opinions and feedback.

Thank you FastForward for posting this photo and for firing off some very important questions about very imprortant points on bracing.

There is no need to remark about the importance of Abdo Nahat. To me he is one of the very rare and unusual Oud Constructors who has realised the role of the locations of the braces and tried to arrange the composition of the overtones by differentiating the largeness of the areas lined off by the braces. What a surprising and confusing(!) approach of a great master which do not match with our fascination of SYMMETRY!

Idolizing the symmetry must be the manifestation of our "sense of justice" :applause:

But, Physics and physical phenomena do not "have" any "sense"! They are indifferent and insensible. If we are seeking for any success when dealing with physical objects and trying to make them obey to our will, then we also must be indifferent and insensible against them :xtreme:

I may try to make my remarks more understandable by expanding the physical counterparts of these metaphors.

jdowning - 7-9-2010 at 10:16 AM

It is unfortunate that the restored oud now has enlarged sound holes as the original sound hole diameters may have been a key element in the sound board geometry (it looks as though the original rosettes have also been replaced!). I take it that the image on the right hand side is that of the original oud - more or less as it left the hands of Abdo Nahat.

The early lute makers did not use linear measurements to lay out their sound board geometry but - having determined the sound board profile required - then laid out the rest of the geometry by proportion using only simple dividers. The scribed layout marks can still be seen on some of the sound boards surviving from the 16th and 17th C. Also - and I agree with Faruk Turunz here - that the old luthiers did not feel bound by precise symmetry. Often the exact centre lines are a little 'off', braces set slightly askew, sound hole centres offset a little from the centre line, sound board profile a little asymmetric etc. The old masters were in the business of making fine lutes for profit so made them as quickly as possible (and with great skill) without troubling about what they knew from experience was unimportant detail.
And so it is likely to be the same for the old oud makers - this fine oud by Abdo Nahat being a case in point

While it is not possible to be precise when working with reduced sized images that may also suffer from some degree of optical distortion the following geometry is proposed as one that appears to fit nicely - as closely as I can determine given the above limitations. The geometry is laid out using dividers as follows:

Starting with the profile of the sound board as far as the neck joint A, the overall length A-M is equally divided into 5 parts. The front edge of the bridge K is located at the first part, the centre line of the small sound holes G on the second part, the centre of the large sound hole E on the third part and a brace C on the fourth part. (This also agrees with the classic oud proportions where distance A-E = E-K and MK is half A-E (or E-K). Usually A-E and E-K would also be 1/3 string length and the same length as the finger board - but in this example the finger board is too long to conform to 1/3 string length. So is the neck original or another 'restoration improvement'?).

The sound hole diameter is 1/3 the width of the sound board at the sound hole centre line (this conforms with 15th C lute design geometry of Arnault de Zwolle and later). The diameter of each of the small sound holes is 1/3 the diameter of the large sound hole. The location of the centres of the small sound holes is given by a Pythagorean right triangle of 3:4:5 relative proportions (dimension 4 being distance E-G.
The front edge of the neck block B is located equidistant between the neck joint A and brace C i.e. A-B = B-C.
The below bridge brace L is located midway between the bottom of the sound board M and front edge of the bridge K. i.e K-L = L-M
Brace J is placed mid way between the front edge of the bridge K and Brace H at the widest point of the sound board. i.e H-J = J-K
The below sound hole brace F is placed the same distance above the small sound hole centreline as brace H is below i.e. F-G = G-H.
Two half braces are positioned on the centreline of the large sound hole - on either side of the sound hole and finally a brace D is placed above the sound hole so that distance D-E = E-F

Nahat Oud Geometry (598 x 931).jpg - 97kB

farukturunz - 7-9-2010 at 12:06 PM

Quote: Originally posted by Jameel

Michael Cone did the work on that oud. Nahat bracing is all over the place. It's a secret yet to be "unlocked". The rosettes are original size, as far as I know. I might have a sound sample. I'll do a little digging....

Dear Jameel,

I really appreciate your fastidiousness in choosing the words like "unlocked". I as not being so fastidious as you are most probably might choose a word like "to be inspired of". Any way I have a sincere respect towards your thundering efforts for understanding the great "secrets" of old masters. :applause:

Jameel - 7-9-2010 at 12:22 PM

I like your words "to be inspired" better than "unlocked"! But your appreciation for choosing words carefully is equally inspiring. Thanks Faruk!

Jameel - 7-9-2010 at 12:44 PM

Just a quick reply here. The original oud is not completely original. Note on the pre-Cone oud that the wood around the smaller holes is not a continuation of the face wood. This oud underwent some prior restorations before Michael took possession of it. I'm guessing that the original sound holes were damaged in some way, and the original binding around the holes was as well, so new rosettes were made in the style of Nahat (made by me), the replaced wood around the holes was removed and instead of compromising the soundboard further Michael put a thin binding around the inside edge. I should add that the smaller rosettes I took directly from a Nahat, size and all. I originally made them for an oud I was building, but later didn't use. Michael and I started corresponding about this oud and I mentioned the extra rosettes I had. I sent them to him and they were a perfect fit (after removing the "replaced" wood around the small holes). Given that I took these rosettes from an completely orginal Nahat, I think that's a pretty argument for the current size of the holes.

FastForward - 7-9-2010 at 02:26 PM

Usta Faruk:

Thanks for the comments. True, we are fascinated by symmetry for various reasons. But in the oud case, symmetry does not necessarily equal good tone. I am just fascinated with the approach of the Nahhat.

I have also noticed that on some of your soundboards, you have used slanted braces around the bridge. You and the Nahhat's have different experiences and different tonal requisites which is manifested in the variation of your soundboards bracing from the Nahhat. But all is fascinating and interesting.

---

John:

Very nice and inspiring analysis.

I think the deviation from symmetry is something done on purpose rather than imprecise placement by Nahhat. Looking at the fine woodwork he was doing I wood be surprised that he would err by this much in a much simpler task. Unless, of course, this oud was not made directly by him rather by a new apprentice. It could also be one of the " lower end basic models" that probably were not cared for as much, if such ouds existed! It definitely is not as ornate as many other Nahhats.

Yes, the oud on the right is the original one (apparently it has had some modifications so its not 100% original).

I think that distances A-E , E-K are about equal given the potential visual distortion and measurement errors. The current deviation is about 4mm from even, 19.6 and 20.4cm as opposed to 20cm. I will measure the width of the soundboard at the center of the large rosette to to compare it to the large rosette and will get back to you on this one. I will also measure the other distances and put an image along with the one you provided. Overall, truly inspiring and fascinating decomposition.

---

Jameel:

Interesting story. So I guess the wood added around the small rosettes after the first restoration was oversized. In fact, we can argue upon closer inspection of the small rosettes of the photo on the right that the wood is partially blocking the rosettes.

jdowning - 7-9-2010 at 04:31 PM

Thanks for the factual update on the history of the last restoration Jameel. I had no idea about the prior background only taking the information originally posted at face value.
So, it is safe to say that the oud in its current state may not represent the oud that left the workshop of Abdo Nahat - at least we have no way of knowing for sure.

So what about the sound board that existed prior to restoration? If this sound board was considered not to be 100% original, why, specifically, and on what evidence, is that?

I think that FastForward's suggestion - judged by the oud's lack of decoration and deviation from perfect symmetry (although slight) - that the oud might have been an instrument made by an apprentice could have some virtue (but do we have any written evidence to support this speculation about employment of apprentices by Abdo Nahat - probably not?). However, an apprentice would surely not have been left ,unsupervised , to his own devices but would have been required to follow a geometry dictated by the master? Or, alternatively, was this an earlier sound board (perhaps one of many experimental models, of necessity quickly put together and devoid of unnecessary decoration) made by the master as part of his 'learning curve' to make a better oud? In which case - and either way - a geometrical construction that is straightforward and fits - within a reasonable tolerance - might surely be a good candidate for consideration against a geometry that is otherwise a complete mystery?

FastForward - 7-9-2010 at 05:30 PM

Since the soundboard has had some work, it would be great if we know the extent of repair work that was done. Was the bracing changed? I realize this info may not be available.

Anybody else with an Abdo Nahhat that can post a similar photo so that we can make some comparisons? I will be glad to do the measuring and marking on the images.

An apprentice could very well be a young member of the family, e.g., one of his son at an early age. Such is a typical practice of craftsmen families in Damascus at the beginning of the cenutry. I am not sure this was an experimental oud by Abdo himself, this is oud was made in 1925 and so the learning and experimentation days of Abdo Nahhat himself were most likely well behind him, or maybe not!

FastForward - 7-9-2010 at 10:31 PM

Update,

The large rosette dimension as is (after enlargement) is not 1/3 the width at the center line. One third of the width is 70px, Whereas the large rosettes is 81. BUT, if you reduce the size a little bit, and consistent with what the size would have been based on the other image, then 5px on each side would make the rosette 70px, exactly 1/3 of the width of the soundboard at the center line of the large rosette.

I can see that the distance between the brace in front of the bridge and and the bridge (K-J) is not equal to H-J. The ratio is 3:2 rather than 1:1. At the top near the neck block, I think the distance between the brace above the large rosette and the neck block is divided into 3 equal segments.

jdowning - 7-10-2010 at 05:34 AM

Thanks for the update FastForward. One difficulty is that we are both working from the same low resolution 'fuzzy' image in order to arrive at conclusions about relative proportions.

I see that I forgot to include the brace immediately below the neck block in my analysis!! It does not seem to be a perfect fit but I would accept your conclusion that the intent of the maker may have been to divide distance B-D into three equal parts. This again is a practice found on some of the old lute sound boards.
There is a shadow under the bridge that may be some kind of rigid reinforcement plate (?). If - for the purpose of placing brace J - this was taken as the front edge of the bridge (rather than the front edge of the tie block) then the 1:1 ratio would apply.

Jameel notes that the sound board - as it was prior to restoration - had the (original?) sound holes partly filled in with wood to reduce their diameter. I wonder why someone (most likely then not Abdo Nahat or a family member?) took the trouble to do this. Did they, perhaps, feel that the sound hole diameters were too large and so did not conform to traditional proportions?

Detailed information, measurements and images recording the extent of the restoration work that was undertaken might be very helpful.

jdowning - 7-10-2010 at 09:25 AM

For completeness I have reviewed the proposed geometry to try to find a possible geometrical relationship assuming that the larger diameter sound holes chosen for the restoration are in fact the original diameters. The increase in sound hole diameter does not change the proposed arrangement of the bracing etc as there is plenty of space between brace locations D,F and H to accommodate the holes without having to reposition the braces.

To establish the diameters of the large and small sound holes it is necessary to turn to the Pythagorean 3:4:5 triangle that is used to determine the small sound hole centres relative to the main sound hole centre. The diameter of the small sound holes is then given by 2 units of the triangle and the large sound hole diameter as 5 units. Again, this match has been determined from the low resolution image so may be subject to some error but could be verified more precisely by measuring the restored oud.

The other relationship that seems to fit is that the centre of area of the rectangular plate under the bridge is located mid way between brace H at the widest part of the sound board and the bottom of the sound board M.

Nahat Alternative Sound Hole Geometry (611 x 745).jpg - 75kB

FastForward - 7-10-2010 at 05:21 PM

John, the shadow under the bridge is not a plate, its the bridge itself. If you look at the image of the oud itself rather than the bracing, you will notice that the bridge has a wide base and that the part to which the strings attach is offset to the inside. This typical of the bridges made by the Nahhats. Though, your observation is quite interesting in the sense that this extended bit of the bridge levels off the distance mentioned and provides a 1:1 ratio.

I would still think that the size of the large rosette prior to restoration is almost accurate. In Dr. Ouds book, the size of the large rosette, which is based on a Nahhat, is 11.6cm. So I would still go with the 1/3 of the width soundboard theory.

The 5:2 relationship between the large and small rosettes could very well be true. I hope that some members with a Abdo Nahhat can measure their rosettes and update us on these measurements.

jdowning - 7-11-2010 at 07:31 AM

OK that is good. I was confused by the fuzzy bracing image as the bridge appears to be proportionally much larger than the bridges shown in the before and after restoration pictures of the oud.

To complete this exercise I was curious to find out if there was a simple (non mathematical) geometrical construction that would define the profile of the oud and which might have been used by Abdo Nahat.

The image that shows the bracing was printed onto card and then cut out to give a pattern of the oud profile. This was then traced on to paper as a reference.
For this method of construction the required width of the sound board is divided in two to give the half width X-Y. This is then divided into two equal unit lengths X-A and A-Y. The construction line is then extended by another equal unit Y-Z. These three units form the base of a Pythagorean 3:4:5 triangle N X Z. Point Z is the centre for scribing an arc of radius R1 which defines the curve of the upper section of the sound board (repeated for the opposite side from centre W).

Note that this is the same construction that is used to define the upper sound board profile of ALAMI's Al-Arja oud (See 'Oud or Lute' on this forum). However, unlike the Al-Arja oud, the bottom section of the Nahat sound board is not described by a semi circle but has a flattened profile in the form of an ellipse.

To form the ellipse two 3:4:5 triangles are constructed using unit X-A as the base. Two pins are positioned at A and B and a piece of thin cord is looped around the pins of a length that produces the required curve using a pencil (see attached image). The curve is blended in smoothly using arcs drawn by R1 and R2. (repeated for the opposite side)
It can be seen that the match with the card template profile is very close taking into account possible slight image optical distortions, pencil lead line thickness, reduced scale errors etc.
An 'acid test' to verify this method might be easily done at full scale using an original Abdo Nahat oud (other Nahat ouds may have different profile and geometry) to verify the profile (given that there might still be some slight asymmetry inherent in a hand built instrument profile - as there always is).

In light of the importance of the Pythagorean 3:4:5 triangle in this proposed construction as well as the use of a 3:4:5 triangle to establish the relative dimensions of the larger sound holes (previously posted) it would seem that the larger diameter, post restoration sound holes are also good candidates for being original?

farukturunz - 7-13-2010 at 05:36 AM

Quote: Originally posted by jdowning

To complete this exercise I was curious to find out if there was a simple (non mathematical) geometrical construction that would define the profile of the oud and which might have been used by Abdo Nahat.

Thank you jdowning for giving this geometrical construction. This and similar proposals by laying a method grounding on a "principle" may help searches in standardizing the oud form.
Kind regards

jdowning - 7-13-2010 at 12:22 PM

Thank you farukturunz - I agree with you entirely. As so little (if anything) remains on record about how the great oud makers created their instrument profiles or bracing geometries, the only solution left is to 'reverse engineer' - testing different possible solutions against the surviving instruments to try to find a match. The simpler the geometrical solution and the closer the match (within a mm or so) the more likely it is that a particular solution may have been the one used by a particular luthier.

Many researchers interested in the lute have tried this approach - with some (apparent) success - although there are some surviving lutes by the great masters that currently still defy an obvious solution.

The situation with lutes is, perhaps, a little better than it is with the oud as most of the surviving lutes are in the (hopefully) safe custody of major museums where conservation (rather than restoration) is a paramount objective. Many of the museums - although they restrict access to their instrument collections (conservation) - do offer full scale drawings/ detailed photographs for the benefit of researchers and luthiers.

Most old ouds are probably in the hands of private collectors so the only way to collect important historical data, recorded for posterity about the instruments, would be to persuade the owners to carefully outline trace and measure their oud, analyse the geometry, and post the results for the information of all interested or, alternatively, to post optically undistorted images so that others might attempt an analysis (and then post the results).

There is definitely (from my observations and in my opinion - but not surprisingly!) a very close similarity between the geometries of some old ouds and lutes - although the exact relationships have yet to be examined, established and confirmed through further research. All worthwhile objectives but time consuming to pursue in detail!
Nevertheless each individual could contribute, with a little effort, to what could possibly become an important and accurate historical data base.

Danielo - 7-13-2010 at 12:51 PM

Hi John,

actually a non-insignificant number of these private owners are on this forum ! I think you don't need persuasion, but rather give some clear protocol in order to make the measurements accurately without special tools.. I agree it would be very important to have a database of such data.

regards,

Dan

jdowning - 7-13-2010 at 05:18 PM

Thanks Dan. I will give some further thought to this under a separate, dedicated topic as here - on this thread - we are discussing only one possible geometry of an oud by one important maker. There will, of course, be other variants as there is with surviving lutes. If you are correct about the historical ouds in private ownership then Mikes Oud forum would be an appropriate place - a potential repository - to collect and share this data.
The starting point would probably for each owner to provide good quality, 'full face' undistorted digital images of their instrument(s) together with some basic linear measurement information (string length etc.) and maker, date etc. This would at least allow an initial preliminary assessment of the geometrical possibilities to be made - to then be verified by a more detailed examination by each owner.

Quite busy with other projects at present so will try to get around to it later this year,

jdowning - 7-13-2010 at 05:41 PM

Here is another possible alternative 'lower half' geometry for this Abdo Nahat oud. This (with the potential inaccuracies of a reduced scale analysis) gives results that appear to be close to the original profile but that do not depend upon the additional complication (and potential error) of string and pins for the final result.
The attached image should make the construction clear.

A square ABCD with diagonals AC and BD is drawn with sides equal to the oud maximum width. The upper profile of the sound board and overall length is constructed using the geometry previously posted - based on a Pythagorean 3:4:5 triangle with an arc described by R1 from centre Z etc.

The arc described by R1 and an arc described by R2 from centre N are drawn to intersect at point E.

An arc described by R3 with centre at E and touching the centreline of the sound board is drawn to intersect diagonal BD at F (repeated for the opposite side).

Finally, an arc drawn by radius R4 from centre F - touching and blending into the arcs described by R1 and R2 - completes the profile of the lower part of the sound board.

FastForward - 7-18-2010 at 08:42 PM

I have been busy the last few days and haven't had a chance to read through the comments thoroughly.

Wow John, Every contribution is more astounding than the previous. I second what you say regarding the individual contributions. I think that we should have an extended database of such info. It would be great if it could be hosted here at Mikeouds. I think the first step is to provide a simple set of standardized instructions that the average Joe or Mo can follow to provide the data.

Unfortunately, so much precious info has been lost and there is no way to get it back. Hopefully our combined efforts can provide a wealth of info for the future.

farukturunz - 7-19-2010 at 07:59 AM

Quote: Originally posted by FastForward

Usta Faruk:

I have also noticed that on some of your soundboards, you have used slanted braces around the bridge. You and the Nahhat's have different experiences and different tonal requisites which is manifested in the variation of your soundboards bracing from the Nahhat. But all is fascinating and interesting.

---

John:

Very nice and inspiring analysis.

I think the deviation from symmetry is something done on purpose rather than imprecise placement by Nahhat.

Dear Fastforward,

You are absolutely right in your statement that deviation from symmetri was done on purpose. This is of course a logical inductive reasoning: A great master like Nahhat never could have done an imprecise placement accidentally. He must have done it on purpose...If such is the case...shouldn't we ask "but why?"
His Ouds' fabulous sound Thankfully causes us to think that this deviation may one of the reasons of itself.
Actually we might have reached to a simple implication: As far as the stringing of an instrument can never be symmetrical why would we assume the instrument to be designed in a way inharmonious with the assymetrical character of the stringing manner?
Violins are explicitly assymetrical, bass bar is placed aligning the bass string (Sol) and the sound post is fit up aligning the treble string (Mi)
Likewise an oud's soundboard must be constructed assymetrically.

Best regards

Edward Powell - 6-1-2012 at 01:44 AM

Quote: Originally posted by farukturunz

Quote: Originally posted by FastForward

My opinion on this is perhaps slightly different...
My feeling is that symmetry is a very very good (perhaps essential) starting point... (can you imagine a speaker cone made in a wildly asymmetric shape?)
I think soundwaves are helped by a certain amount of symmetry.

BUT the soundboard would only remain perfectly symmetrical if the instrument only had one string.

Obviously because one side of the SB is treble and the other side bass, then some asymmetric changes will be helpful!

I mention all of this because I have experimented a lot with quite wildly asymmetric shapes of instruments and soundboards, and none of them worked.... it wasn't until I returned to a "more" symmetrical shape that my own experiments started to yield much more positive results.

chaldo - 7-22-2012 at 08:51 AM

Quote: Originally posted by Danielo

Hi John,

actually a non-insignificant number of these private owners are on this forum ! I think you don't need persuasion, but rather give some clear protocol in order to make the measurements accurately without special tools.. I agree it would be very important to have a database of such data.

regards,

Dan

do we have any other bracing measurements for nahats? what about the database Danielo is proposing?

a call to all of Mike`s oud members who restored/received/analysed nahats.. willing to share the bracing measurements & pattern

chaldo

jdowning - 7-22-2012 at 11:42 AM

As far I know I am the only forum member to have attempted analysis of the profiles and bracing of old ouds as well as lutes - information posted on the forum under the topic 'Old Oud compared to Old Lute Geometry'.

Although many forum members may own or have access to old ouds (including Nahats) there has been no response so far to requests for information in a form that will allow accurate analysis of instrument profiles or bracing geometry - with one recent exception (see Yaron Naor's topic 'Repairing a 'Nahhat' Oud').
For this exercise Yaron provided me with an accurate full size tracing of the oud profile (invert the oud on a large sheet of paper and trace around the whole profile including the neck. For this exercise the profile of bowl and pegbox are not required). Onto this tracing he then accurately marked the positions of the braces, nut, bridge, soundhole centre and diameter, neck joint and inside face of the neck block. This assumes of course that the sound board has been removed - otherwise just a tracing of the profile can provide useful information.
Unfortunately, although better than nothing, working from images is potentially prone to error from optical distortions (perspective view, barrel distortion etc) unless the photograph is carefully set up and resolution is high enough to accurately determine essential detail.

I analyse the geometry by first placing the drawing on a draughting board and by establishing the vertical and horizontal axes (through the 'symmetrical' centre line of the sound board and through the widest point) The rest is a matter of trial and error using giant sized dividers (specially made by me from wood) trying to establish conjunct arcs of radii - that are equal to quarter divisions or subdivisions of the maximum sound board width - to match the oud profile. Oud tradition seems to prefer geometries based upon Pythagorean 3:4:5 right triangles.

I would be happy to undertake and report on the geometrical analysis of any old oud on receipt of a detailed full size drawing prepared as described above. Otherwise if anyone wants to try their hand at an analysis I would be more than happy to advise further if requested - provided the final results were to be made available to the world.

FastForward - 7-23-2012 at 12:17 PM

Hi John,

I would love to get a copy of the geometrical analysis/ tracing of the oud/brace locations.

I have traced a Nahhat myself recently and will post the profile. I didn't get the brace locations unfortunately.

Thanks,
Ibrahim...

jdowning - 7-24-2012 at 04:31 AM

Hi Ibrahim
As there is currently no data base for reference we do not know at this point what bracing geometry or geometries may have been used by the Nahat luthiers. This may only become evident with the accurate analysis of many full size drawings that provide the exact brace positions relative to a sound board reference point.

There are two parts to the data collection process that may be treated independantly 1) the external oud profile geometry and 2) the internal brace geometry so the information that you have to hand may be very useful for adding to the data collection of part 1.

Note that this exercise is to some extent reiterative in that ideas about the geometries may change as more evidence becomes available. So, for example, I initially thought that the lower sound board profile of some of the Nahat ouds might have been elliptical curves but I now think that both lower and upper sound board profiles were created in a much simpler fashion from conjunct circular arcs of various radii which can give results very close to a pure elliptical curve. After all why would a luthier go to all of the trouble of creating an elliptical curve - a complicated point by point construction - when an almost identical result can be quickly and accurately obtained using just dividers and a straight edge? Simpler is always better in this respect

FastForward - 7-24-2012 at 09:00 AM

John, check your u2u

jdowning - 7-24-2012 at 12:06 PM

FastForward has kindly sent me the detailed X-Y coordinates of the oud profile that will form the basis of an analysis of his oud. This is a good start but I shall need some additional coordinate and other information in order to attempt a complete geometrical analysis of the oud (excluding the bracing geometry). This will be an interesting project to undertake on this thread to illustrate the level of information required for detailed analysis and the methodology.

In the meantime it would be of general interest if FastForward could post some more general information about the oud - maker, date of construction, number of courses, string length etc. together with some images showing the instrument in general view as well as full face, details of the rosette(s) and bridge details etc

So lets take it from here and see how it goes. Should be interesting

Thanks again Ibrahim for your interest and for investing your valuable time to make this project a possibility.

FastForward - 7-24-2012 at 01:41 PM

I sent you more details.

The oud is a 1921 Nahhat and is not my oud, I only wish!

It has 15 ribs, 6 courses, string length is 60.9cm. The bridge was a typical Nahhat.

I do not have photos with the other requested details.

jdowning - 8-4-2012 at 04:25 PM

Apologies for the delay in reporting progress but I have been busy with non musical project work while the weather is cooperative so have been trying to spread my remaining available time around all of my other interests.

I can say that I have undertaken a very preliminary examination of the Nahat oud profile from the coordinates and measurements provided by FastForward and can confirm that the upper sound board profile appears to almost exactly match a circular arc of 5 basic units radius (4 units = sound board maximum width) identical to the Nahat upper sound board geometry previously posted on this thread, based upon a Pythagorean 3:4:5 right triangle.
The front edge of the bridge tie block is also located midway between the the maximum width (X axis) of the sound board and the bottom of the soundboard confirming one of the two possible bridge positions - the other being 3/4 of a basic unit (again based upon a 3:4:5 triangle construction as shown in the previously posted Nahat construction). The triple sound hole configuration also appears to match a 3:4:5 triangle construction.
A more accurate drawing layout and analysis is now needed to confirm the above.

I also plan to be posting - as time permits - the sound board profile geometry and bracing geometry of a 1936 George Hanna soundboard derived from a very nice and accurately draughted full size drawing of the original by Yaron Naor. String length measured from the drawing is 60.9 cm.

jdowning - 8-9-2012 at 12:04 PM

A more detailed construction has now been completed.

It should be mentioned that there is some scope for dimensional error in laying out a geometry directly from a full sized tracing of a soundboard - compounded when transcribing coordinates measured from a sound board tracing no matter how carefully done. I shall go into more detail later in a separate topic dealing with measuring and recording sound board measurements. Accumulated dimensional errors (including possible slight variances from the original intended geometry at the hands of the luthier) might amount at best to ± 0.5 mm but ± 1.0 mm would be more realistic in practice. This is about the tolerance variation for this project I would estimate.

The attached image is the proposed upper sound board geometry. The geometry is based upon a Pythagorean 3:4:5 right triangle the curve of the soundboard profile being almost exactly described by a circular arc of radius 35 units with center at point E. The width of the sound board CD is 28 units as is the dimension AB.
The centre of the large sound hole is 10 units from B on the X axis. The diameter of the large sound hole is 9 units. The centre line of the two small sound holes is 2 units from the X axis however I have yet to discover how the small sound hole diameter (33 mm) relates to that of the large sound hole. However, the large sound hole diameter locates the inside edges of the small sound holes.

It is interesting to note that 28 is the number of units ('fingers') in the ancient Persian cubit measure of length although in this case the unit chosen (equal to about 12.25 mm) does not match the ancient finger measure of about 22.8 mm. However - there were also subdivisions of the basic finger unit (in the Egyptian Royal cubit measurement system at least) - from 1/2, 1/3, 1/4 ....to 1/16 sub units if I remember correctly. I will have to check further into this but I would not be at all surprised if 12.25 comes out very close to a subdivision of 22.8
Coincidence perhaps?!

Next to describe the lower sound board profile.

Dr. Oud - 8-11-2012 at 11:14 AM

I believe that analyzing the Nahhat brace pattern based on one example is irrelavant and fruitless. I have measured around a dozen Nahhats. Locations, sizes, and shapes are all over the place. I find no common denominator, with mean deviations from 15% to 20%. Even trying to find a common constant that can be applied with a proportional calculation does not work.

Exhaustive studies of the violin master's instruments have resulted in the same inconclusive results. Every Stradivarius violin is different. Great guitars come in all sizes and shapes. Ouds as well, Nahhat, Manol, Haddad, Georges, whoever. The only valid asumption is that the design varies with the materials, and since wood is so variable, the variations are endless. This is why factory made instruments are inferior to hand made, built in fixtures and with machines, they are all the same, and since the material varies, the optimum goal of sound quality is never achieved.

It takes years of building experience to develop a sense of what makes an instrument work, whether oud, violin, guitar or whatever. Why else are the hand made instruments so much better sounding than the scientifically analyzed, fixture-built ones?

Instrument building is an art, using infinitely variable materials, science cannot solve these mysteries.

jdowning - 8-12-2012 at 05:13 AM

Well, of course, trying to discover a common geometrical basis for a Nahat bracing pattern based upon only one example is not going to get us very far but neither is an incomplete list of dimensions taken from 11 ouds made by six members of the Nahat family over a period of 120 years, at least not without that component essential for a geometrical analysis - a precise, full size tracing of the instrument outline.

The earliest records about the oud confirm that they were then built according to well understood proportions and geometry in accordance with a universal harmony " in the image of the science of the Creator" according to the musician philosophers of the 10th C 'Bretheren of Purity' (Ikhwan al-Safa). These early scientific minds also believed in the 'magic' of numbers and their relationship to each other.
If the oud building tradition is an unbroken one then one might expect to find certain fundamental elements in the design of the surviving old ouds - albeit of very recent history. That is the purpose of my investigative work not only as it applies to ouds but also to determine if there is a common relationship with early lute geometry - the European lute supposedly having been developed from the oud.

I do not believe that the Nahats or any of the old oud makers developed their designs in a random fashion (reinventing the wheel so to speak) but followed design concepts dictated by tradition. The precise geometry drawn with only a pair of dividers and a straight edge would be reflected in the mold and the Nahats may have used several molds of different size but all based upon one (or more) basic geometries. Furthermore, there is scope for the dimensions of a bowl to deviate from that of the mold at the hands of the maker as well as after the bowl has left the mold awaiting fitting of the sound board. It is, of course, the shape of the bowl upper face that dictates the final dimensions of the sound board. If the master luthier did not make the bowl (a less critical task assigned to other workers) he would most likely have selected the material, made and fitted the sound board and would have made any necessary small corrective adjustments for any deviations from the original design proportions. All these factors might contribute to dimensional variations in the surviving ouds - assuming of course that they are not fakes and in original condition not otherwise subject to sound board replacement etc.

However, in order to investigate the situation in depth more than a handful of examples will be required. Under current circumstances that is likely to be the biggest challenge!

I cannot comment on the geometrical design of 'figure eight' type instruments like the violin and guitar but as these are not related to oud or lute they are not relevant to this investigation.

I should point out that this current exercise - to examine the geometry of the 1921 Nahhat from dimensions provided by FastForward - does not address the issue of bracing geometry as no details of the bracing layout have been provided. What the analysis does show, however, is that the geometry of the sound board has not been created in a random fashion but almost exactly matches a basic geometry that I have found already in both old ouds and lutes. The upper sound board profile previously posted is based upon the ancient geometry of a Pythagorean 3:4:5 right triangle with the profile simply described by the arc of a circle.

Likewise - as will be seen next - the lower half of the sound board, location of the bridge etc. is a simple geometrical construction created using dividers.
No mystery about it at all!

jdowning - 8-12-2012 at 10:57 AM

The bottom half of the sound board geometry (from the X axis down) is as shown in the attached image. The accuracy against the coordinates given by FastForward - like the top half is about 1mm or less i.e. an almost perfect match full size.

The geometry is self explanatory. The bottom edge of the sound board F is a distance of half the maximum sound board width from B on the X axis. The front edge of the bridge tie block is half the distance BF.
The bottom sound board profile (each half) is a combination of three conjunct circular arcs of radius 42 units with centre at A, radius 35 units with centre at E and radius 12 units centre at 2 units from B on the X and Y axes.

As with all of my proposed geometrical constructions the basic unit is an arbitrary value chosen by the luthier in order to create a design of the required dimensions.

Next to examine the geometry of a 1936 George Hanna sound board from a full size drawing kindly sent to me by forum member Yaron Naor - this time with the bracing detail included.

jdowning - 8-31-2012 at 12:01 PM

I have now completed the geometrical analysis of the full size drawing of the sound board of a 1936 George Hanna Nahhat prepared in meticulous fashion by Yaron Naor which is complete with bracing.

Perhaps Yaron might provide more detail about the sound board and how it came into his hands - for general information?

For clarity, I shall post the geometrical construction in three steps - the basic layout for the sound board profile, the sound holes diameter, bridge location and neck block thickness and finally the bracing geometry.

Step 1 is shown in the attached sketch drawn to scale and should be largely self explanatory.
The geometry is based upon three Pythagorean 3:4:5 right triangles with relative dimensions based upon arbitrary 'finger' units. There are 16 units measuring the maximum width of the sound board with total length AB of 24 units. The upper sound board profile is described by a circular arc of 20 units. The lower sound board profile (8 units deep) has been modified from a semicircle by combining arcs of 6, 20 and 24 units to produce a pseudo ellipse profile.

The centre C of the large sound hole is 6 units from X at the widest point of the sound board. The horizontal centreline of the two small sound holes is two units from X. The relative positioning of the three sound holes C, C1 and C2 is given by two 3:4:5 right triangles as shown.

Next - step 2

G H Nahhat 1936 Step 1 (522 x 600).jpg - 67kB

jdowning - 9-1-2012 at 05:16 AM

I should add that the string length of the oud is 60.9 cm.

The maximum width is 35.4cm (equal to 16 arbitrary units) measured from the soundboard tracing so each 'finger' unit in this case is equal in metric measure to a value of 22.13 mm (compare this to the ancient Persian standard measure of somewhere between 22.5 and 22.8 mm).

Once upon a time the standard 'finger' measure may literally have been the width of a finger (index finger at the middle joint?) no doubt of some king - shah or pharoh perhaps. Apparently measures based upon finger width, palm and arm length etc are still used as an approximate measure in the market places of the Middle East.
It is interesting to speculate, therefore, if some of the Nahhat ouds were custom built to fit the physical dimensions of the hand of a purchaser - the measure of a finger width being the criterion?
The width of my left hand index finger at the middle joint is about 22 mm and I am comfortable with a lute string length of 60 cm - so perhaps this oud was originally custom built for a person of close to my stature?

This might account (in part) for some slight size variations found in the surviving Nahhat ouds whereas the geometrical proportions - based upon units of arbitrary size would remain consistent.

jdowning - 9-1-2012 at 12:07 PM

The attached image shows the second step giving the front edge of the bridge tie block position, the neck joint position, the depth of the neck block and the sound hole diameters.

The neck joint N is located at 1 unit from A. The inside face of the neck block is half the distance from N to a point 10 units from X.

The length of the fingerboard (front edge of nut to neck joint) is equal to the distance from the neck joint N to the centre of the large sound hole C. (Note, however, that in this case the distance from C to the bridge is greater so the 'traditional' proportional division of the string length as 1/3:1/3:1/3 does not quite apply)

The front edge of the bridge tie block is a fifth of the distance NB (23 units) measured from B the bottom of the bowl.
The large soundhole diameter is the distance from the front of the bridge tie block to the horizontal centre line of the small soundholes.
The small soundhole diameter is 1/3 of the large soundhole diameter.

All of these proportions correlate closely to within about a millimeter or less full scale. For example the exact measured diameter of the large soundhole - as made - is 118 mm and that of the small soundholes 40 mm whereas the measured distance from bridge to small soundhole centreline, taken from the full size drawing, is 119 mm. Close enough!

Finally, to follow, the bracing geometry

Correction! The distance of the front edge of the bridge tie block should be a fifth of the distance from the neck joint to the bottom of the bowl NB not AB as shown on the attached sketch

G H Nahhat 1936 Step 2 (574 x 600).jpg - 75kB

Yaron Naor - 9-2-2012 at 02:29 AM

Hi All and John,
This Nahhat oud arrived to me with a new 2years old face made by a luthier here, But the original face was found in the Luthier's shop, so we took it back and than I could measure everything.
It belongs to a person here in Israel, he bought it from someone years ago.
On the ticket is written "1936 George Hanna Nahhat"

Regards,
Yaron.

jdowning - 9-3-2012 at 10:43 AM

The proposed bracing geometry is as shown in the attached sketch.

Brace 6 - is located at the widest part of the sound board (X axis)

Brace 4 - (2 pieces) is located on the horizontal centreline of the large sound hole on each side of the sound hole.

Brace 5 - is located midway between braces 4 and 6.

Brace 3 - is located the same distance from the large sound hole centre C as is brace 5.

Brace 8 - is located midway between the front edge of the bridge tie block and the bottom of the bowl B.

Brace 7 - is located from the front edge of the bridge tie block a distance of 1/4 the distance to brace 6.

Brace 2 - is located at 4/5 NB - the distance from neck joint to bottom of the bowl - measured from B.

Brace 1 - is located the same distance from brace 2 as is
brace 3.

The front edge of the bridge tie block is located at 1/5 NB measured from B.