Roger Hill

An additional opinion of the whole sound perception phenomenon is that most of the "break-in" of audio components that people hear is in fact, brain break-in. Let me give you an example of this. Silver conductors have a reputation among audiophiles as adding a certain shrillness to the sound. I once bought some teflon insulated coax which had silver-plated copper strands for both conductors. I made a pair of new interconnects, put them in the system, and thought WOW! that is shrill. A few days later they sounded great. Break-in? Well, I then made a second pair and when I installed them, I didn't hear them "break-in", the system sounded about the same. I think my brain had adjusted. The break-in I believe in is measured as a shift in resonant frequency after a new loudspeaker driver has been driven by a signal generator for a while (an hour or so for a woofer is plenty). I think what you ask about has to occur and would (perhaps) be measurable in shifts of some of the fundamental modes after playing-in of a new violin, maybe even after a new steup of a violin. The stiffness constants would be reduced and modal frequencies would decrease if micro-fractures have anything to do with this. I think Don should be able to tell us what happens with his new violin as it plays in. I've also encountered some who recommend a good twisting and squeezing of a new violin as being helpful. Not sure I'm brave enough to try that. I think the limitation on this would be whether such a loosening-up, if it occurred, would produce a large enough frequency shift that any measured difference was greater than the inherent inaccuracy of the measurement system.

Fracture may be too strong a word. "Delaminate" from places where it doesn't belong is another way to think of what I am referring to. Glue holds well between two structural pieces, but not well where is contacts only one structural piece. You find some relatively brittle stuff on the lead wires to voice coils, etc.

Hi John: In my experience, which includes magneplanars, the break-in of a loudspeaker is largely one on fracturing the excess glue used in its manufacture. With the magneplanars, the mylar has to be bonded to itself in places, to the edges of the interior steel plate, etc. Those bonds are not terribly precise and the excess glue, which inhibits vibration, has to be fractured. I don't for a minute believe it takes 100 hours but that is a different issue.

There are numbers of us here who do not subscribe to Strad. Could someone provide a summary of the Denis arching approach so that we have some idea what you are discussing? Thanks,

My head is ........spinning

But I've just finished my first violin, god knows there is not 300 years of quality thought in it. I'm happy not to repeat the mistake of too wide a thin area around the island in the back. That was something I had in mind to experiment with. I'll pick a different mistake for my next one Thanks, Don. Good info.

Hi Don: By "standard" graduation patterns, I assume something along the lines of Sacconi. Would you mind telling us what the deviations were, and what you will do differently next time? A brief description of the arching would be helpful also. What I am hoping for is something along the lines of the Zygmuntowicz request of Strad that he should have left us a little book telling us "if you thin a bit more here, the sound changes to ___________" i.e., would you mind filling in the blanks a little more completely? Thanks,

Most important is to have the proper steel. Best is blue spring steel from McMaster.com. I like the .015" thickness best. A lifetime supply will cost you about $25. Another good steel (but which will not hold an edge as long) is a stainless steel ruler from Home Depot. Under no circumstances should you try to use the various cabinet scrapers from Woodcraft, etc. The steel is simply too hard to get a good edge on it. I wasted sever months trying to figure this out. To shape your scrapers use a bench gringer to get the outline you want. First, you have to get the steel into workable size pieces. To do this hold it firmly in bench vise and with eye protection, work gloves and a hammer, work it back and forth, bend it as far as you can and then break it with the hammer. Shape with the grinder. Now, clamp it to a piece of hardwood and file the edge to 45 degrees all around the perimeter, getting it as sharp as you can with files. You are really trying to get a knife edge, but only on one side. Now take the rough edged scraper and sharpen the filed side on your "Scary Sharp" papers all the way down to 2500 grit. I then polish with tripoli stick (Woodcraft, about $5.00) and the felt wheel on the bench grinder. For a burnisher, use the back side of a gouge or preferably, the round shank of a phillips screw driver. Pull the (sharpened side) edge of the scraper along the burnisher with the plane of the scraper about 30 deg to the axis of the burnisher. The pressure against the burnisher should be very light. Try on maple, then spruce. Some spruce will give only very small curls. (See the Strad3d dvd of Sam Z scraping a spruce top to see that spruce will not produce big even curls as will maple). You may have to touch up some areas of the scraper with the file and then resharpen the whole thing. Also, experiment with some small area on the scraper as to how much pressure to use when you burnish. You can usually resharpen by wiping the back only on the wet/dry paper and reburnishing. If not, back to the front side and all the way through the Scary Sharp. Edit: get that ruler at Office Depot, not Home Depot

Hi Don: How do you balance the 1 kg weight on top of the indicator shaft without affecting the reading? Thanks

Hi Don: Tell us what the horizontal axis of your graph measures. (or did I miss it?) Thanks,

My pleasure Craig. You might be interested to know that one of my scientist friends was the college room mate of a kid named Moog, who did quite well fror himself with that synthesizer idea. Don't remember the Mellatron but were I aware of it I am sure that at one time or another it too would have siezed my attention.

Thanks, Melvin. Simply stunning work. I can't imagine how you could ever improve on this violin

Beautiful work, Melvin. Joe himself would have been proud of such work. Tell us a little about the arches and what you use for varnish. Thanks,

Thanks, Craig. Not poor form at all, and yes, it is exciting. Age 68. background includes a PhD in physics from long ago ('67). Career took me away from science to become a business man (real estate and finance). Have lost the physics analytical skills I once had, but still have good physical intuition for classical physics (mechanics and E&M). Very good machinist with old-time methods (not CNC) I can also do things with a file that are really impressive, but that takes one heck of a long time so I can't screw up very fast. Thorough understanding of and well equipped for good measurements of metal work and have figured out how to convert to wood. Monarch lathe and Bridgeport mill in the garage. Good at design and building of tube-type stereo equipment and measurment and design of loud speaker systems. Ears are definitely not as good as they used to be. Worlds most anal fly fisherman

Thanks, David. Good information.

Hi Craig: If I seemed short with you that wasn't my intent. Just hoping to get you experts to stop and think of graduations you started and experimented with, what changes you've made to them, why you've made them and what the results were. Obviously, anyone with persistense can start with a plan and do trial and error forever. On the other hand, if I don't have to repeat common mistakes I would prefer not to. In pursuit of that, I have read everything I can find on the internet on arching and graduation. I am not ignorant of the literature. I have noticed some things that I have never seen experts comment on. For example, I have found no references anywhere that notes that if one measures the outside width between the bridge feet, and then measures the distance between the inside f-hole notches, the two bridge feet are located at the symmetric centers of percussion of a beam of length equal to the width between the notches. Now, I can't be the first guy to notice that, but it would certainly be nice if some of the experts would come right out and say what they know. There have to be things you and others have discovered about graduation that can be passed on. If there are things you want to protect as proprietary, ok. I certainly respect David Burgess when he says exactly that. All I am asking from you experts is that when you comment on something in a manner that indicates that you have some specific approach to what you do, one that is amenable to some methodical, repeatable process, then tell us what it is, why you thinks it works, etc. I am all ears. I am not trying to avoid paying my dues, only to minimize the amout of those dues. BTW, my first violin is in the white, I am currently getting organized to build a quick and simple light box. I am going to do two violins side by side for my next two.

BUT, BUT, BUT there is some starting point, some inkling of response, some hint of experience, some gut feeling that tells him where to start and what that starting point will produce. Otherwise, we are saying to a newbie maker "well, make 200 violins and then you will have a feel for wood, arching, graduation and varnish" I think you experts are more eloquent, more able to explain what you know, more communicative of what those who taught you than to say, "well, you really need the experience of making 200 violins before I can talk to you."

Tell me more: how are the grads "properly proportioned" to the model, arching, wood stiffness,etc? Thanks

Hi Jim: Would you expand on this a bit? Need a more graphic description of what you are saying. Don't need to spare the gory physics. Thanks

Oded: good observation. There is a very obvious similarity between the vibration pattern Schleske shows and the graduation pattern.

Nope, I'm relying on all of you experts to tell me about the grads and their history. But, since the Kreisler has been described as the finest solo violin ever (I think by Wurlitzer?) and I am assuming that both of the Strads sound good, whoever did this had something in mind that appears (to me) to be the same for all three. Wood properties probably changes around the violin plate. It would be a surprise if the elastic constants are exactly the same all around. But for modeling purposes it is a simple and practical approximation. Anders, that is why I am wondering if the observed patterns might be an attempt to homogenize the elastic constants.

Well, Andres, their carving skills greatly exceeded mine of today, and with only a graduation punch I think I can carve a top that is uniform to within a few tenths of a mm, rather than a 2+ mm variation over the top. Are you saying that because of 1200 violins between them, it is only a random occurrence that these three tops appear to be graduated with the same malice of forethought? Maybe so, but I don't buy it. It looks to me as if they were trying to accomplish something very specific with the thinning in the various areas, in that particular pattern. I wish I could tell what. Has anyone ever looked at the variation of (vertical) stiffness at various areas of a top plate? Dean, tell us more about anchor points

David: The Harrison poster says the measurements were done by Loen and King. The Strad pictures are accompanied by one paragraph which doesn't identify how the data were obtained. The del Gesu article says that Loen, et al have measured 46 del Gesu's, but does not state whether these eight were among them. Think we would need to have Loen's book for a more definitive answer to your question. I don't think my abilities in analysis and certainly not construction will ever justify purchasing either this book or the Biddulph book.

OK, Anders, tell me more of what you are thinking. The Hacklinger gauge used by Loen is at least used by the same individual, who has also used it a lot, so I think errors due to inconsistency in technique are minimized. I have been under the impression that the gauge itself has an inherent accuracy of at best a few tents of mm. The accuracy of the ct measurement is stated to be between .5 to 1mm depending on the operator. Would you expect a practiced ct operator still to be unable to match the accuracy of a Hacklinger map, despite being able to map a much greater number of points and produce a more accurate contour? (I certainly don't believe that the top of the Betts is uniformly 1mm)

Thanks, David. There is a massive amount of information here. Interestingly, the Betts Strad is one of the violins. The top thickness plot conveys a totally different sense of the graduation than the Betts measurement of Jeffrey Loen. My instinct is that the CT scan method is capable of greater accuracy. Perhaps the plotting program used with the CT scan was not asked to do a fine enough resolution of the output to distinguish the character of the graduation.