Mike's Oud Forums

Oud Structure's Question

setarmaker - 3-12-2007 at 10:58 AM

I just needed to know if the end of all braces are attached and glued to the top ribs of an oud. If Yes. Could someone please explain about, how the soundboard resonates then? How does it stay resonating while, playing it? This is what I noticed in an old oud (made by Ali Al-Ajami in1950's) which, came a cross recently and I looked inside of it by using a dentists mirror. Thanks a ton for replying to me.

jdowning - 3-13-2007 at 05:56 AM

It is important for the ends of the bars to be glued to the top ribs of a lute so the same should apply to an oud. The bars provide structural support for the soundboard against collapsing under downward pressure due to string tension loads etc. If the end of the bars become unglued the sound board may start to sink at the edges.
The question of how a braced soundboard vibrates is very complex and I doubt if anyone fully understands or can measure what is going on. Luthiers usually determine the optimum solution that works for them through long experience gained in making many instruments.
The bars are relatively deep in section (for vertical structural rigidity) but are reduced in depth at each end and are relatively narrow in width so can flex in torsion as well as a bit in bending. The soundboard is usually quite thin (less than 2mm) and can flex and vibrate relatively freely between the bars. Also the edges of the soundboard glued to the thin top ribs are often weakened by purfling or banding a situation that will in turn affect soundboard response.
The whole structure of a good responsive instrument is supposed to be light in weight and just sufficiently strong to withstand the string tension without collapsing. On the other hand I recall seeing a film many years ago of an Arabic luthier standing on the inverted bowl of one of his completed instruments in order to demonstrate how strong it was. Interesting (not to mention nerve wracking to watch) but not to be recommended!

Regardless of any theoretical explanations the system does work.

Dr. Oud - 3-13-2007 at 08:16 AM

While the braces in some ouds are tall and thin, there are other designs that work quite well. Gamil George, for instance made the soundboard a bit thicker and the braces a little wider and much shorter. This configuration is used my Shahata today, who's father was ana pprentice in Gamil George's shop. Manol offset the center of the braces toward the treble side and cut down the brace under the smalll roses. His design is used by nearly every Turkish builder. The position of the sound holes and braces can affect the response of the soundboard, but it seems many configurations will work as the variation is nearly infinite. I did a survey of the placement of the tone brace - the one just forward of the bridge that determines the primary resonant frequency node point. Measurements aere taken on 10 Nahats and the placement varied by as much as 6.5% and this was within the same family of makers.

setarmaker - 3-13-2007 at 10:23 AM

Thanks for your quick respond guys. Gluing bracings to the ribs was a storm braining to me. I make different instruments but, Just being a professional Structural engineer for 17 years, when I looked at the back of an old oud’s soundboard, I realized that the design of the soundboard and the bracings are just the way we design structural floor for earthquake purposes e.g. we try to keep the centroid of the floor shape coincide with the center of the gravity of the floor. Scientifically, and practically, this will result in the best vibration mode of the floor in case of earthquake with the least damage. The vibration will mach the received frequencies caused by the earthquake. Thus, from my engineering stand point, size of the sound holes, shape of the soundboard, thickness and height of the bracings are very much important in order to get these two points on top of each other, I believe regardless of luthiers’ skills or wood quality.

Now, it will be a very simple procedure to do this. When you have completed the soundboard sanded down to your desired thickness and sound holes are in place plus all other top works are done, simply put the soundboard on the tip of your index finger to see the equilibrium by trial and error. Then mark this point and measure the distance from any reference point you’d like ( I recommend the center of big sound hole). Since your soundboard is made of a same wood in its entire body, and since the thickness is varied by 0.1 to .4 mm you can logically and practically assume that, the point which you just marked is the both centroid and center of mass (gravity) for the soundboard only.

By adding each bracing on the back, you are moving the center of mass away from center of the soundboards geometry (centroid). This will be solved simply. From this I can tell you why in NAHAT Ouds, you see the bracings heights are a bit different. In my opinion, they have been trying to keep these two centers coincide and this is what the mass of bracings will rule out only.

This is just a thought My friends, just for you and others who may be interested. If you are interested on the topic, I can show you how to find the correct height and location of each bracing to ensure highest vibration and stability of the soundboard regardless of any shape of bracing. ( just needs to be from same family and shape in a certain oud but you can change it for a different shape and get same result).

If you use 8 bracings, 9 or 6 it really doesn’t matter. The last bracing’s location and thickness and height will guarantee the balance. On an old oud that I happen to see, I took all steps above and from the Static formulas, I simply determined the location and height of the last bracing and it was accurate to the original design by zero tolerance. When I say last bracing, it just means the last on you would like to apply. In this case my last bracing was considered as the one under two smaller sound holes.

Thanks a gain, Ahmad

Dr. Oud - 3-13-2007 at 04:17 PM

ok, then how do you explain the asymmetric design of the Manol soundboard? The sound of these ouds is testimony to the efficacy of the design, but the size and shape don't follow your hypothesis. For one thing, wood is not an isotropic material, which changes the science of oud making into an art. hmmmm?

jdowning - 3-13-2007 at 06:11 PM

I am not a structural engineer but would be very surprised if buildings were designed to sustain earthquake loadings based only on a simple, two dimensional, static analysis. Dynamic analysis must surely play a significant, more difficult to determine and likely more crucial part in the design process? And so it must be with an oud, lute or other musical instrument.
Of course, the basic objective of a luthier should be to achieve an overall uniformity and balance in the soundboard geometry and barring as a starting point, as dictated by tradition, but the fine tuning of that basic arrangement - which can make the difference between a superb sounding instrument and one that is barely mediocre, as Richard observes, is surely much more subtle than just arranging for the centroid (the centre of area of the soundboard) to be coincident with the centre of gravity of the soundboard?

setarmaker - 3-13-2007 at 10:18 PM

Good and strong comments on my later posting.
1. I'm not an oud maker, so I probably need to except what you say as an expert. This is fine. but I'm so glad hearing about dynamic behavior of the soundboard. I feel more comfortable talking to someone who's got some knowledge about what I just discussed. What I mentioned was not about how an ouds sound could be enhanced, but just a thought. Asymmetric design of bracing is just perfectly matching what I explained in my previous post. I emphasized that, It does not matter where you put all of your bracings except for the last bracing. That means it could be accidentally symmetrical or asymmetrical dipending on your mass distributions. this is my point.

2. Have you guys ever tried the steps that I have mentioned before in any of your projects? If not, please go ahead and send me an example of one of your ouds. What I need is the average of your soundboard thickness, the number of bracing minus one and all specifications of all of your desired bracings. I may be able to tell you where you need to install your last bracing in both practical and theoretical ways.

3. For the ones who are concerned about Dynamic behavior of the soundboard, I was expecting a different argument! Why no one asks me the question of " how the heck a Luthier would now all about modern structural theories in for example 1800s"? This would explain why I just started with static equilibrium at the beginning. Old master luthiers might not now any thing about all these things but might have learned everything from their
masters in practical ways.

4. The good thing about what I explained before is that you can make the procedure in both mathematical and practical ways. I'd love to explain both ways which will result in a same conclusion. but the dynamic exploration which is a more accurate way is possible only through some mathematical approaches such as matrix and dynamic structures theories knowledge. I'd love to explain dynamic behavior as well, but it may not be a useful presentation for every luthiers.

5. I strongly believe that, an instrument will sound perfect if a good and right old wood, a skilled luthier with an accurate ear, lots of years of experience and a touch of spirit meet at a right time simultaneously. But as I mentioned before, this is just a thought of mine.
you guys give it a try and check it out. It's worth it to try it, if my concept is new to you. If there are 100 points needed to make an instrument perfect, my understanding of bracing arrangement could just be one out of 100.

6. I'm trying to get some sleep now, but I will explain how all this would work in a traditional way with no need of math, physics or structural engineering knowledge.

7. isotropic material issue. I 100% agree that wood is not an isotropic material as you mentioned. Since no instrument is100% perfect ( not talking about exceptional deals), a 2mm thick wood material can be considered somewhat as an isotropic material. for such dedicated member, calculations show an ignorable discrepancy in determining center of mass and centroid. Please check this and let me know if I'm wrong.

8. I'm not new in playing and making lute instruments, but I would never want to put my nose in oud business. All I had was an honest curiosity.

9. Dreams will come true some day if you honestly believe them.


Thanks, Ahmad

Peyman - 3-14-2007 at 06:51 AM

Very interesting discussion. Ahmad, I sent you a u2u (up in the page).

jdowning - 3-14-2007 at 07:44 AM

Don't get me wrong Ahmad I am very interested in your thoughts and the discussion thread on this subject as one never can tell where they might lead in a positive way to the benefit of all.
If you have not put your theory into practice as an instrument maker then I think that you should first make a series of test instruments and then publish the results rather than requesting others to do that work for you. This is your project after all.
Sound enhancement of an instrument (however that may be judged from person to person) is what it is all about as we would all like to know how we can get rid of some of the 'mystery' and consistently build good instruments in a more 'scientific' manner. Of course luthiers - past or present - do not know about modern structural theory - they don't need to in order to make fine sounding instruments based on their experience. In any case, I very much doubt if a mathematical model of oud soundboards can be developed to reliably predict dynamic behaviour so that the data might be simply expressed in a form that can be of use to an instrument maker - but I remain open minded about that. Your expertise as an engineer will be needed to translate the complex mathematical language of matrices etc. into plain written language that we can all understand so I look forward to that.
Like you, I have approached the oud from my experiences with the lute and other stringed instruments of the 16th and 17th C Western societies. As there may be more historical information surviving about the construction of the lute than there is about the oud as the two instruments are historically connected and - structurally - are essentially the same, I figure that oud makers might find something of interest and relevance among this information.
Here, for example is part of a text relating to lute barring taken from Harmonie Universelle by Marin Mersenne, Paris 1636 a massive and important theoretical work of the time. It is taken from Book 2, Proposition 2, pages 49-50. It is in 17th C French with phonetic spelling so I will follow this posting with my translation into English.

jdowning - 3-14-2007 at 07:49 AM

.... and here is my translation of the relevant bits of the text for information. I am not fluent in French so my translation may not be perfect but should, I trust, reasonably convey what Mersenne was saying.

SamirCanada - 3-14-2007 at 08:18 AM

Did you ever concider this... beuty is the ears of the beholder.
Wich makes me say that the basis on wich we determine a good sounding oud cant just be scientific.
Because If you can produce the loudest possible oud doesnt mean its the best sounding oud.

Personaly I cant stand if the trebles on my oud are ringy and really loud. I like a more closed sound to them so that they sound rounder and softer.
You can get pretty close to that result with mathematical calculations Iam sure (if your aiming for that). But not if your aiming to maximize every aspect of the sound. So in the objective of trying to build the perfect oud that responds to all kinds of frequencies we might be going against the true sound an oud should have. Thats why you will hear many people complain that there oud has too many overtones. And thats why that before one gets into trying to maximize the sound of a oud they should know how a good oud sounds. If its made to sound like a lute or a guitar then its not a oud anymore, It looses its soul I find. The master makers before us left us there beutifull ouds and it is an evolution of many many years to get here. And they are a few aprentices left to carry on the art.
And on a side note many oud makers will make 2 identical ouds and they will sound differently it must depend on a ton of reasons that cannot be controled by the human hand. my guess
This is based solely on my opinions and some of my fellow oud players.
What do you think about this?

Dr. Oud - 3-14-2007 at 08:39 AM

I don't know about the number and/or positions of bars/braces in French lutes as the translation mentions only 6, while the ouds I've observed (more than a few) have 7 or eight. re: mass centroid analysis - It would take only a few experiments with different pieces of wood to discover that the stiffness of each pice varies even between pieces of exactly the same dimensions and at the thin dimensions of an oud, varies quite dramatially. It is the stiffness that determines the physical response to the string vibrations, not the mass. Stiffness is not necessarily related to the mass, but rather depends on grain density, ring count, runout and varies even within each specific species of wood. I have discussed the dynamic analysis of soundboards with simulation engineers at my company and they feel that as the components are stressed at the boundary of their limits, the analysis cannot be conclusive. The complexity of the structure also defies computer simulation, much less a simple mathmatecal analysis. I still believe it's art, not science.

jdowning - 3-14-2007 at 08:53 AM

I agree with you Samir. Certainly I would doubt if an oud that only had two bars - for example - situated, say, above and below the main soundhole, would have much of the anticipated sound characteristics that defines an oud. These structural features for oud, lute, guitar or what have you have all been basically figured out through long tradition but it is still up to the individual luthiers, through experience, to get the best response and finest sound from the instrument that they are building given the particular characteristics of the materials they have on hand.

Ahmad, to return again to the question that you opened your topic with - i.e. are the ends of the bars glued to the ribs - here is another reference that may be of interest - taken from Musick's Monument by Thomas Mace, London 1676 in which he is giving instruction on how to repair and maintain your own lute. He confirms the importance of the bar ends being firmly glued to the ribs.
"Because the belly being so very thin, and only supported with six or seven small weak bars, and by the constant stretch of the strings, (which is a great strength) the belly will commonly sink upon the first bar next above the bridge, but sometimes upon any other, and so cause it to let go its hold at the ends of the bar, and then your lute will jarr and grow unpleasant.
And if it not timely amended, worse inconveniences will follow, which may endanger the spoiling of the belly. Therefore, whenever a bar is loose, let it be quickly amended or presently set your lute down to a lower pitch, or untwist your strings, and lay it by, till you can get it mended."
Sound advice especially in the wet English climate of Mace's time when "Misty Vapours and Dampness" played havoc with the lute and its tuning.
I may post the full text of Mace's instructions later as a separate thread as they are just as relevant to the oud as they are to the lute.

jdowning - 3-14-2007 at 09:13 AM

Doc - the number of bars in Mersenne's description is only the basic minimum number.There might be more (and there usually are in a lute) - as Mersenne states - dependant upon the experience of the individual luthier and the type of lute being built, number of strings etc.

setarmaker - 3-14-2007 at 12:35 PM

wow! Thanks guys. I deeply appreciate mike for putting this beautiful forum site together. It is a wonderful tool so we all can chat and educate each other. I truly loved the translation of lute construction. thanks a ton. I thinks my English is not good enough to have a better explanation for my thoughts. I believe most part of your discussion. As I said before, I don't play Electric guitar! I play Iranian Tar and Setar. By ear not by note. I trust my ear and tune piano by my ear so confidently. So, I'm deeply involved with spiritual part of the deal. you are all right. It's not all about math. Or we may even say that, it has nothing to do with math.

fair enough. As I said, we do not need to know any math to follow up on my primitive theory. One can simply do it by a trial and error. I have seen so many amazing lute instruments in my life which proof what you all say. lot's of them did not even meet the minimum qualifications of a wonderful instrument but, they sounded brilliant! I've seen soundboard which, you won't believe their sound quality with an ugly grain pattern. Todays luthiers may use such wood for their fire place only. It seems like the old masters new how to talk to those woods and new how to ask them to become a part of their life by sounding well. I know all of these and I appreciate your reminder. What I posted, came to my brain as a matter of second as soon as I saw an oud's soundboard for the first time. I may change my mind on the second or third time! who Knows?

My friend suggested that, I better to do my project by myself and try to experience it first. Certainly may friend. I did not mean to have others working on any project of mine.
Please don't get me wrong. Come on guys, I didn't ever mean that, by using this method, your oud will sound better, lauder, warmer or more brilliant at all. you can always use a microphone and an amplifier to make a 10,000 Watt of sound. That's not my point and far away from my subject.
My point is that, lets see thing from all directions. for instance, If I go to a physician for a particular disease, the Dr. would not necessarily need to have self experience on that type of disease to help me out. In todays world, People work for each other without knowing each other. so there might be some luthiers out there, who may want to go a head and give it a shot.

I'm going to take your advise and start exercising on the matter. I will definitely post some pictures and the out come of it in a near future. The only problem is that my workshop is not prepared for oud work and most of my tools are designed for making tar and setar. But that's not a big deal. I think that Khalaf's site may help me with some needs. Thanks again for all of the info available on this site. So my project will start with turning down a damaged oud. I'll try to document everything by taking photos so, interested luthiers may help me with different steps.
at the end I was wondering if it would be o.k. for me to print a copy of the translation for my records. It was extremely interesting. Please let me know if I would have the permission.

Thank you all, Ahmad

jdowning - 3-14-2007 at 01:37 PM

Keep up the good work Ahmad. I for one look forward to your future discoveries that I hope you will post on this forum to the benefit of everyone.
Here is some good news. Interested in your theory, I tested the belly of an old oud that I am currently restoring and find that the centroid (the 'centre of gravity' of the area of the belly) coincides quite closely with the centre of gravity of the belly itself - the balance point in this case being just below the bar under the the large rose. I did this fairly quickly and roughly - dividing a scale drawing of the area of the belly into squares and estimating the area that way - deducting the areas of the roses etc. and of course the balance point is affected by the non vibrating part of the belly glued over the neck block which should be deducted so this is not very precise. This oud has symmetrical barring so I am not too surprised at this result but at least it is one possible data point for you to work with. I also have a couple of (reproduction) lute bellies that I will also check out in the same manner to see what I can find. It is unlikely that there are very many oud bellies lying around - detached from their bodies - but if there are, others may want to carry out the same test or at least check the position of the balance point and send you an image of the belly so that you can determine the centroid.
Not sure how useful all of this will be to luthiers in the creation of fine instruments but interesting to explore further if you have the time.

John

ALAMI - 3-14-2007 at 02:44 PM

Most interesting,

It is still an idea and needs to be tested, but as we are still at the theory stage, I will just expose the outlines:

John is right not much oud bellies lying around, and no one is going to rip off his Nahat neck to test a theory....well ...we still can do it.
I am thinking of a computer simulation system using dynamics rules. The oud can be built as a 3D model which is not very complicated, I agree that we can assume the material isotropic due to its low thickness ratio.
If we have the exact measures of a soundboard and its braces (luckily measured during a repair) and the density of the woods used for the board and the braces (or at least the relative density, ie the braces wood is 1.2 times more dense than the soundboard wood) the model will be reliable.
Same for the belly.
I received few days ago Dr oud book, It has all the dimensions so theoretically, we can build a 3D digital Nahat model (still need info on density- density would change over age but I don't know if the variance is signicative).
Don't get me wrong, the simulation will only be reliable on the static state physical and geometrical side, it can determine the relation between the center of geometry and the center of gravity but not reliable in anyway to simulate the vibrations (the sound of the oud).
The 3D model can keep its construction history so it could be adapted to different models made by different makers,
The only problem is that this experiment needs time.

setarmaker - 3-15-2007 at 05:49 AM

Wonderful. I don't know how to thank both of you gentleman. obviously, without the help of professionals and experts, I will get to no where. When I see there are some people who are interested in researching, I get more motivated and it really doesn't matter if we don't get a good result. Any conclusion would be useful as far as continuing it or stopping it right there. Back to the old damaged oud that I have received, ( this is the one that made me wonder about all these).
This oud belongs to Master Abd-al-wahab Shahidi in Denver Colorado. He is a wise old singer and oud player. still sings well and plays a good oud in age of 86-87. I explained the situation to him and I asked him if I'm authorized to start restoring his oud with nope for the acceptance. At the moment, I felt like, nothing would make him as happy as my offer and then he returned, off course, Ahmad. I trust you and I believe that, this instrument won't become worse than what it is know.

Seems like I had a good luck in my first move. The rose is missing so the sound hole is wide open. The soundboard in an awful condition. Not only crack are developed in over 10 areas but also the surface is not smooth anymore and where the cracks are, you can see sinking or swelling on either side of cracks. I'll try my best to detach the belly with no damage to the ribs edges. I'll be posting some pictures of the progress. I have a simple steam machine which was designed and built by myself. it's neat. It's got a simple temperature control gage and I use it for bending purposes. I can strongly say that Not more than 10% of failing was recorded in my bending process so far. I have a little note book where I saved my bending experience notes. It's a good tool to show steaming specifications for different type of wood as far as their species, thickness, original MC, thus, the steaming time which should be somewhat optimum. Pictures will be posted next time. I use a smaller hose in order to get a better and more concentrated steam out of the machine for detaching purposes and in order to keep ribs away from steam. Since it has not been used for a long time, I don't want to spend time on it and I can just use a kettle with a modified output to get the same result. Hope this will go well.

Take care, Ahmad

Dr. Oud - 3-15-2007 at 07:32 AM

Quote:
Originally posted by ALAMI
....
I am thinking of a computer simulation system using dynamics rules. The oud can be built as a 3D model which is not very complicated, I agree that we can assume the material isotropic due to its low thickness ratio....I received few days ago Dr oud book, It has all the dimensions so theoretically, we can build a 3D digital Nahat model ....
If you examine the front illustration and the dimensional drawings you will realize that I have modeled oud in 3D already. If you like I can send you a iges or step file of the model. I did this once before to generate a cam program to machine a plug to build a mold to fabricate a carbonfiber oud. The engineer doing the molding didn't solve the acoutic problems with the CF soundboard and I haven't heard from him since. send me an address if you'd like the files.

ALAMI - 3-15-2007 at 12:05 PM

Great news Doc, IGES is a fine format, It is a great time economy.
Just sent you an email.

I just had half an hour today to make a rough test on 3D bowl virtually attached to a vertical wire and tested gravity on it.
Then I deformed the geometry a bit making the bowl assymetrical and run gravity again.
Test attached as Quicktime movie (mp4)

jdowning - 3-15-2007 at 03:00 PM

Interesting discussion Ahmed, ALAMI and Doc. Here is some more data that I can contribute to this project.
Today I looked again, in more detail, at the oud belly that I have and also of a scrapped belly from a copy that I made, some years ago, of a well known late 16th C lute by Giovanni Hieber (Venice/ Padua - Italy). This lute is one of the few surviving instruments that has remained essentially unmodified over the centuries. It has a string length of about 60 cm and has seven courses - so not too different geometrically from an 'average' oud.
I built this instrument from measurements and drawings of the original provided by professional luthiers that I knew at that time. The attached images show the original lute and my copy of the belly. As can be seen, the barring is asymmetrical (as is usual with lutes) - ignore the inked in 'fan bracing' at the lower end of the belly , below the bridge, it does not exist in the original (I used this scrap belly to illustrate a talk that I gave about lute construction some years ago). The inked in area at the top of the soundboard is the area glued to the neck block that does not form part of the vibrating area of the belly but does affect the location of the centre of gravity (C.G.) of the belly when it is detached from the body of the instrument. This section ,when discounted, lowers the measured centre of gravity a little towards the bridge - a correction that must be made when comparing the location of the centroid and centre of gravity. In my calculations, I made this correction simply by taking moments (of area) about the measured CG point - balancing the moment of the area on the neck block against the balancing moment of the total area of the belly. This gave me the corrected shift of the C.G. towards the bridge. Here are my results in summary:-

Egyptian Oud (see my postings Restoration of Egyptian Oud - part 1 or Traditional Oud Barring). After correction of the Centre of Gravity balance point as described above, the centroid position and C.G. coincided approximately within 0.25 cm.

Hieber Lute copy. Again the centroid and C.G. coincided approximately within 0.25 cm.

Given the fairly crude methodology that I used to determine the C.G. of the belly (balanced upon my finger) and estimation of areas (division into squares) as well as using approximate moment of area calculations to correct for the redundant area of the soundboard over the neck block, these results (if I have done the math. correctly) indicate a possible correlation between coincident C.G. and centroid as proposed by Ahmad. However, these are only two data points and more will be required to confirm these findings.

Melbourne - 3-16-2007 at 06:02 AM

I have only briefly read bits and pieces of this most facinating thread, as I can hardly keep my eyes open anymore! But I would love to hear what Jameel has to say about these points. Jameel you about?? :D

jdowning - 3-16-2007 at 11:33 AM

Note that in my analysis of the Egyptian oud and Hieber lute bellies I have taken the geometrical centre of each soundboard to be the 'centroid' - that is the areas above and below a line drawn through this point across the belly are equal(and deducting the full area of the soundholes).
This, of course, may give a different point from the true centroid which is determined by integrating area moments - requiring the application of calculus for an irregular shape like an oud belly - a calculation that I don't feel inclined to attempt manually. Perhaps the computer program currently under consideration could handle this?

I should also mention that the balance point or centre of gravity for these two detached bellies occurs at the same point - that is close to the bar at the bottom of the rose on the bridge side. Curious about this, I checked out two other bellies from instruments that I am currently building - a copy of a 17th C lute (liuto attiorbato) and a reconstruction of a 16th C guitar - and found the balance point to be in the same location below the soundhole. Note that these two bellies are unfinished so have yet to be trimmed to size which might affect the C.G. position but not by much I imagine.
Interesting, given the wide variation in geomentry of the bellies.
The attached images show the two bellies in question - for information. The liuto attiorbato is asymmetrically barred with bars 2 and 3 above the bridge position canted to be higher on the bass side than on the treble. Note also the below bridge 'fan bracing' alluded to by Mersenne.

setarmaker - 3-16-2007 at 05:08 PM

Perfect. I cannot thank you enough for all hard work you have done so far. The interesting point that you brought up, does make sense. I did not realize the neck block as none resonating part comparing to the rest of the soundboard. Your analysis for the moment of area based on the negligible area of neck block part is in fact helping us to understand more about the relationship between C.G and centroid, specifically in our case. Nice job cutting that. To me finding exact location of the centroid by mathematical ways is well possible, but we may not need to be that accurate. For you and I, the concept is clear and just takes time to do it. What really is important is that, I doubt if luthiers and old masters have ever gone through the complex calculations. I’m now working on the traditional ways of doing this so our friends who have no idea about complexity can follow through the procedure. Your result of 0.25 cm is just close enough for me to continue coinciding concept. You’ve had a few experiences in a short time and I bit we will find more and more down the road.
I’d like to emphasize that the centroid for all similar shapes are the same no matter what size they are e.g. H/3 for any size of triangle or intersection of diagonals for all rectangles, squares, hexagonal, octagonal and so on. So we could expect that, most of the centroid would fall in to somewhat the same vicinity you just mentioned. Because, most of the ouds have almost the same shape of the soundboard, But of course not the same size. So if Ky (K =as coefficient & y= distance from end block up to the neck block just like K=1/3 in case of all triangles) is the location of the centroid for this particular shape, the K could be a constant and Y would differ case by case. That’s may be why in any oud you can see the centroid somewhere close to the big sound whole.
O.K. back to my project, Please see attached pictures and I will explain every step in the next posting. Detachment came out nice and fully safe.
It seems like someone screwed this up before. It makes you Cry. Must be around 70 years old. We’ll discover. It used to be a wonderful piece of artwork, when It was original. I’ll try to post the sound of it played by Mr. Shahidi.

pictures

setarmaker - 3-16-2007 at 05:38 PM

steaming

picture3

setarmaker - 3-16-2007 at 05:41 PM

detaching from neck block

picture4

setarmaker - 3-16-2007 at 05:47 PM

detaching from neck block

setarmaker - 3-16-2007 at 05:49 PM

Look at who questionable is the missing Bracing on this one.

Faladel - 3-17-2007 at 04:44 AM

Please ¿can you post the lable??
Thank you
;)

jdowning - 3-17-2007 at 05:40 AM

Thanks for the update on moments of area, centroid etc. Ahmad. It is amazing how much can be forgotten after 50 years - which was the last time that I recall having carried such calculations!
If the centre of area calculation that I did for the Egyptian oud and Hieber lute soundboards is a good enough approximation - given the relatively standard belly profiles of ouds - then this is something that can easily be determined by someone who does not have the knowledge, time or inclination to get into integral calculus - or perhaps some easy to use software could be developed (if it doesn't already exist) to do these calculations - one that has graphics capability to read the profile geometry of a specific soundboard etc.
My low tech. calculation just requires dividing the surface area of the soundboard into squares and doing a bit of reiterative adding and subtracting to find the point where the areas above and below the "centroid" are equal. A bit tedious and prone to error but something that could easily have been done by luthiers in centuries past.
As a lute maker, I have always been interested in the bar below the bridge that seems to be a common feature of ouds. It is not found in any surviving lutes to my knowledge but I wondered if it did at one time - being replaced by the below bridge 'fan bracing' in lutes of a later period. I did instal such a bar as an experiment on one of my lutes and (to my ear) it improved the resonance of that instrument - but these things are very difficult to judge objectively.

Peyman - 3-17-2007 at 05:53 AM

Ahmad, it's interesting that there is a picture of Mr. Shahidi inside. Also, it's good to hear he is doing well.
Keep up the good work.

the bowl and label

setarmaker - 3-19-2007 at 05:48 AM

All right. I’ve been so busy over the weekend and no production at all, except couple of things. Just look at the label. It says, (Amal= Made by) ( Ouste= Master) (Luthier’s name=Ali Al-Avad) (dated either 1967/3/1 or 1917/3/1) the second number of the year is under question. ( made in Baghdad) ( Al-dokkan= store # There is a dot right on top of #1 which could make it to look like a six in Arabic numbers. So its at least 40 years old. Up to you guys to figure that out.

I’m positive that from the ugly looking Titebond glue. This soundboard has been replaced by an unprofessional worker who did not even care about the label and did not cover it during restoration. We don’t do that, do we?

The neck had three quite big dents filled with wood filler! The neck looks to be intact but the owner says that it would bounce up and down by tuning high or low respectively. So, when the soundboard was detached, I noticed a small gap between the neck and neck block. I do not want to touch the neck, Since I believe that it could be adjusted and fixed by shimming. Also the neck block is split over the top of the wood pin all the way horizontally. Poor instrument has been Beat up or poorly repair? Either way, it’s sad and I’m in optimistic in putting it back together with your help and advise and give the health back to this old good sounding instrument. Please reply for any suggestions and advise. Ahmad

I think that AJAMI was just a nick name for the builder , since his generations were Iranian Arabs from Southwest part of Iran, or something els, I'm not sure! I just know the used to call him master Ali Al-Ajami.

OUD means Just one OUD but AL- AVAD may mean numbers of OUDs as a plural for OUD. Please correct me if I'm wrong. I just understand a little Arabic.

setarmaker - 3-19-2007 at 06:17 AM

I cannot say how beautiful this bowl has been made. Unbelievable workmanship, very simple, and the wood is outstanding.

setarmaker - 3-19-2007 at 06:41 AM

Now, It took me few minutes to balance this soundboard on a just needle type vertex. It actually took me 5 minutes to do this. When it was perfectly in equilibrium, with a little air movement it would start bouncing without collapsing. very much accurate to the tenth
of mm. Next using a self standing square, I marked the bottom edge of the soundboard. then measured from the mark to the tip of the needle. It was exactly 21.5 cm. Now make no mistake this is showing nothing yet. means we'll find out if the bracings where designed in write location which I doubt . because this instrument had completely lost it's original sound after the poor repair.

jdowning - 3-19-2007 at 08:54 AM

Ahmad - did you balance the soundboard without the detached (or missing) brace?
I am interested in how the ribs inside the bowl have been finished. It looks from the images that the interior has been left in a rough finished state and from the label image it looks as though a toothing plane has been used - to judge from the grooved marks on the ribs? This might be the same type of finish that I have found on the oud that I am restoring (see Restoration of an Egyptian Oud - part 1 on this forum) an unusual feature by all accounts. If so, could you post a close up image of the interior surface with a centimeter rule alongside for scaling purposes? Thanks

setarmaker - 3-19-2007 at 08:43 PM

J,

Yes, I did balance it the way it was detached. I already joined my 5years old sitka which was bought from LMI. It's AAA and I'm sure its a pretty good quality wood. Pictures will be posted accordingly.

Due to layer of dust inside the bowl, I did not notice the roughness, but as soon as you requested it I did clean up and the answer is positive. Attached pictures show a type of fork edge plane work on the ribs. When I make setar or tar, I never smooth out inside of the bowl. It really effect the sound of high frequency strings. Sometimes when the sound sustains, you'll hear an undesired wolf tone. I'm not sure if the maker has been trying to eliminate the same effect. I believe that, grooves are about one (1) mm apart. The top ribs are both smooth.

I'll have bunch of good stuff in the next posting.

setarmaker - 3-19-2007 at 08:46 PM

this one is from the side ribs and the previous one was from the bottom center and left and right of it.

jdowning - 3-20-2007 at 08:55 AM

Thanks for the images Ahmad. Looks as though the interior finish of the ribs is almost identical with that of the ribs on my old oud.
I forgot to ask but is the underside of the belly also finished in the same way? It is on mine.
All very interesting.

Mazin - 3-20-2007 at 10:28 AM

Dear Ahmad

The lable of Usta Al avad Oud.

El Awo'waad means the man who plays Oud. I t is very common in Iraq to adopt a sure name from a man's career, but also can mean' oud making'. The year is 1967 because Iraqis were using different form of dating (the Hijri) at the beginning of the century. I do not think that 'Al Ajmi' related to this master.

Thanks for sharing.

Abdul

setarmaker - 3-21-2007 at 09:15 AM

Hey all,
Abdul, thanks a lot for your time and corrections you made on the label and I’ll pass it to the owner. The owner was telling me t hat, he’s had the oud even before that date (1967) and he remembers when and how he had received it. I noticed that a little black spot has been dropped right next to the real number and caused the problem. It’s not ink at all, but the color is amazingly resembling. But It really doesn’t matter. I don’t want to touch it or remove it. Keeping it like this is more professional to me.

Yes J. that rib is also roughed by the same plane. I planed and sanded the new soundboard down 2mm. Do I need to go more? Then I cut the rose and did the inlay. Persian walnut Pick guard is installed. (It’s my own design and I hope you all like it) and This is the time for me to fined the centroid and mark it before putting any bracings on. As soon as this point is marked, the next wood be calculating the Height bracings. This will be a desired height for just one of the braces based on experience and the height for all other ones will be transferred by proportional spanning. For example, If I start my first Brace below the bridge with height of 15 mm and span of 260mm then my factor would be 15/260= 0.05769 for all other spans. So where I have span=370mm then the brace height would be 0.05769 x 360= 20.7 mm and so on. Now what I need from you is to tell me what should I try for my first height by your experience? I’d rather to use NAHAT’s bracing method of height. The date for this oud is as follows:

top plate's length= 500 mm widest span= 360mm So, just please give me a good height of either one of bracings, so I can start this research and to see where is going to be the location of the last bridge by keeping the centroid untouched.

I’ll post some pictures for your info. Thanks a lot for sharing your skill and knowledge.
Ahmad

setarmaker - 3-21-2007 at 10:52 AM

Let me know if you need explanation on the photos. The most important message in the up right image is that how we really can keep the centroid in the same place and how wisely we can determine what we want. This can be a quick answer to Dr. oud's question regarding asymmetrically distributed bracing. It cannot be worst that this situation. Isn't it right?

setarmaker - 3-21-2007 at 11:02 AM

I just post some pictures of my decorating although they are not our subject and just so you don't get bored too much.
The bottom right photo shows where I had my needle point standing also where the Yellow self standing square was marked as the edge of the soundboard. the top right picture may confuse you! don't worry, the fingerboard is not glued yet. I'll do that when the soundboard is adjusted perfectly home.

setarmaker - 3-23-2007 at 08:20 AM

All right, I have finalized the soundboard’s thickness based on NAHAT’s samples and Dr.'s verifications e.g. 1.8-1.85mm in the center tapering to 1.00mm to 1.3mm to the sides edge. Obviously, my soundboard wouldn’t be balanced on the same exact centroid that I had before. Because the first trail C.G was with the same thickness all around the shape. When tapering to1.3mm towards treble strings against 1. mm for the bass strings, your actual centroid shifts toward the heavier half of the soundboard. See the red dot on the soundboard. But, no problem at all because when I’m going to locate my last and key bracing, I will set the soundboard on the real original centroid and so, this will guarantee remaining of the centroid. Now, buy doing this, your last bracing may not necessarily stay parallel to the others which makes sense, since we play with it to meet the balance on the centroid. I’ll post bunch of pictures from the process and I’ll try to send a video clip on the web as well , but this may take a while for me.

One thing that I need to emphasize again is that the process is to be utilized when you have good quality wood and skills in hand. I’m just trying to research on the physical and geometrical part of this huge deal and to find some realistic relationship between them.

I’m still waiting for someone who could tell me what would be the best average height of bracing to consider. I’d appreciate it.
Ahmad

jdowning - 3-24-2007 at 06:31 AM

Ahmad - I didn't understand your earlier response to my question about the surface finish underneath the old soundboard. Has the underside been finished with a tooth bladed plane - like the ribs of the bowl have?
I do not have any historical data on ouds but I do for lutes. If it is of interest, I have checked my files for measured belly thicknesses from surviving lutes of the late 16th to 17th centuries. For 12 instruments belly thickesses - regardless of the size of a lute (from soprano to bass) are all generally well below 2mm in thickness and often closer to 1 mm with quite a wide variation in thickness on individual instruments. Note, however, that the surviving bellies are somewhat worn after 400 years or so and most of these measurements were taken close to the edge. Nevertheless, these tend to confirm that average belly thicknesses for lutes are similar to those of the oud - to judge from the data provided by Richard.
As for choosing the precise thickness for a new belly that would depend upon the material of the soundboard and its particular physical characteristics to some extent. Likewise for the barring. These are often found to be assymetrical in both placement on the belly and in shape - which, as I understand it, is often the case with ouds as well.

setarmaker - 3-24-2007 at 11:38 PM

J,
I must have misunderstood you first time. sorry about that. The answer to your question is negative. The back of the soundboard is just in a regular shape. It's smooth but not perfectly glazed. Thanks for the info you posted about soundboard thicknesses. It seems like you are the only one who is wiling to help. I certainly appreciate it. Well, I did not want to wait too much, so just went ahead and chose a 22mm for the widest braces' height and finished my bracing. Now I'm going to start the most important part of the journey and work on the last bracing by the method which is under investigation. I'm receiving a pretty good tap tone so far, but taping on the part where the last bracing is not installed yet is somewhat flat and it tells me that I need some more bracings around the tapping area. after the last bracing is installed, I'll check the tap tone by my experience. If I'm not getting the desired sound, I can always sand and reshape the bracings to approach that, But I still need to keep the centroid stationary in my modifications until, I have the desired tone and centroid.

I hope that I'll be able to wrap up the project in couple of days so I can post all necessary pictures and descriptions. thanks again, Ahmad

setarmaker - 3-26-2007 at 11:45 AM

I'm back happy! I think the project has a very satisfying result. Everything is put on now. Before even gluing, I installed one string of each size and checked the sound. As I was guessing, it sounds very nice. the tap tone on free hanging soundboard was really good as well. I did the whole process as I explained and as you can see from the pictures. Pictures are from finalizing the soundboard's thickness and so on. I used a perfect size and elastic bonji cord and installed it around the edge. This was My simple Idea to help th bowl pushing towards end of bracings from both side. Please see pictures and notice wood blocks I have used to ensure the compression. The centroid is perfectly untouched and coincide with C.G. as you see in the photos. Next I made my custom band for my inlay since I did not have any in hand. To do this I just utilized the old soundboard as a mold.

I'll get back to you when the instrument is ready to play.
Ahmad

setarmaker - 3-26-2007 at 11:51 AM

see the progress. What I did was to just tape the bracings on the back and then balanced the soundboard on the centroid by using my last bracing on the top of soundboard. This allowed me to move the last bracing and try to make the equilibrium. As soon as that hapend, I marked the location and transfered it on the back.

I did this one more time after I actually glued all other bracing and did get the same result for the position of the last bracing. Man this sounds beautiful.

setarmaker - 3-26-2007 at 11:58 AM

more pictures.

setarmaker - 3-27-2007 at 05:04 AM

work in progress.

setarmaker - 3-27-2007 at 05:09 AM

bracing ends were perfectly attached to the top rib with no failing. I think the bonji cord did a wonderful job. Bonji cords may not do the job if the ribs are too thick or extremely solid. if the top rib is 2.4mm to 2.7mm you'll get the best result.

setarmaker - 3-27-2007 at 05:13 AM

I'll post the finished job's pictures in couple of days.

setarmaker - 3-28-2007 at 07:46 PM

here is the finished product. Thanks to my wife for making the rose for me. It looks beautiful to me. I didn't have oud strings so I just tried the old ones which were rusted due to several years of being seated on the oud. They still sound pretty good and I just cannot wait until I receive the new Pyramid set.

setarmaker - 3-28-2007 at 07:57 PM

Look at the soundboards figure. Its an outstanding wood that I've had it for five years in my stock. it might be around 10years old from the cutting time.
Just for your info, I used a beautiful piece of Rosewood for the fingerboard plus a very dens South American cherry for the edges. I made my own inlay band from same rosewood and maple with a finished width of 4mm. The soundboard is French polished twice only and sanded with 0000 finest steel wool very gently.

setarmaker - 3-30-2007 at 06:35 AM

Just a brief sketch from inside of soundboard to show what caused the balance.

Jameel - 3-30-2007 at 07:13 AM

Nice work. I look forward to looking closer at your technical drawing. Please post a sound file when you have one.

setarmaker - 3-30-2007 at 08:35 AM

Thanks Jameel,

I will definitely do so. I reviewed and tried your suggestions on fitting bracings against th ribs. Since I'm not as skilled as you are I had some difficulties at the beginning but I got used to it. Your method of using sand paper to do this just reminds me the way we adjust violin bridge legs on the soundboard or the way I fix bridges on my setars. It's neat and thanks for mentioning it. Believe me or not, The centroid method helped me to think about a %99 accurate way to do this without any nightmare. I think when you are free to position all bracings and keep the last one for modifying the centroid, this will allow you to have a nice template for ever and use that for all ouds with the same design. this is what I did and got almost %100 of positive result.

What I did was that I got a piece of plywood 3/4" thick. I cut it exactly same as inside curve line of the edge ribs ( the edge where my bracings need to be fit). then marked the centerlines of all bracings the way I had marked them on the bowl as you thought us. When I transfered them on the plywood, I cut grooves about 5mm wide and 250mm deep so my bracings wood freely located inside of grooves. Since the perimeter of the plywood is exactly similar to real ouds shape, I started filing the bracing ends against the edge of the groove. Doing this took me one minutes for each bracing an 10 minutes all together. You may not believe me but, They all fit perfectly.

Now I can save this template for all other ouds with this particular shape. Although, in the next oud I will have different soundboard and center of mass, but remember my bracings can be located just the same way I did here on this oud and the last bracing will play the role again thus, I can still use the template any time. I want you to try this method for yourself and correct me if I'm wrong. I'll post some sketches when I get a chance.

Thanks again.

setarmaker - 10-12-2007 at 05:44 AM

Hello my great friend. I'm back a gain. Sorry for being late. I have two good news. First that we had our first baby and she is four month old now plus cute and healthy. I hope this will explain my delay. second, is that I have two grate things to show you. I worked hard on the concept that I brought up. yes, the centroid and the center of mass. To proof how the waves effect or do not effect the centroid of a mass, I'm going to ask you to refer to the links below and just follow what is explained below. Not need to start all over since, you know what we have been after. You can play with the attached applets with different frequencies, modes and combination and find out exactly what is going on. you can go up ti infinity on combinations but, in an instrument you just need some of those combinations. When you get on the link do this:

First link http://www.falstad.com/circosc/ is about a circular plate. Please turn on "show frequencies", "color" and sound. you can click on each single square at the bottom knowing that the top left is the first mode of a frequency. next you can add another square to combine it with the first one or unplug the first and try the second mode in single. Now you have to see these modes with different views. in the menus choose solid, wireframe and play with 3D or 2D options.

Now, what you find is that the center of mass and the centroid which are coincide are not playing a big of a role. Just bouncing on the same axis no matter what combination if taking effect. I'm sure that an oud belly is more complicated, but the centroid and the C.G. will behave similar if and if they are coincide. You may notice from the wireframe option that the more you get far from that center the bouncing has bigger range till you get to the edges and this is where the wave reverses it's direction (up and down). The beauty of these plates is that they are fixed on the edges just like the oud belly is.

Now try the link http://www.falstad.com/membrane/ to see it does not matter what shape we are working on. The centroid and C.G. are still the same way.

try to work with the applet for at least 30 minutes to feel it. and give me your input. By the way, Mr. Shahidi is not available and would not like to be broadcast, so I could not upload him to show you the result of the oud that I repaired for him, but he Okayed just to put the sound of it for you. As I mentioned I'm not an oud player but I know what a good sound is. The sound of it is lovely except a little weak on the first string when played in high frequencies near the belly. I believe this is just a lack of my skills. I bet you will get much more result since you are more skilled.

Thank you all, Ahmad

jdowning - 10-13-2007 at 12:37 PM

Thanks for the link to falstad.com Ahmad. That is an interesting site to explore.
Congratulations on your new family member. Enjoy!