The important part isn't the alloy. I mean, yeah, it gives them a leg up - but there's nobody willing to put in the time and money to figure out what's in their cymbals, because it's just more worth it to make your own.

The Zildjian name is more valuable than any of their cymbals. If you don't buy Zildjian cymbals, it's because they're too expensive, or it's because you don't like the sound. It's a better idea for a company to go and make a cymbal that fills one of the other niches in the market than to try to replace Zildjian.

It's sort of like Coca-Cola's secret recipe. If you wanted to put the time and effort into it, you could figure out what's in it. But why bother? Coca-Cola's success is due to the brand, not to the taste.

It may not be just the composition (which could easily be figured out with a mass spec or the like). How metal alloys are formed and treated can have vast effects on their properties.

For instance, https://en.wikipedia.org/wiki/Wootz_steel

There are machines [1] designed for this sort of thing, my father used to "own" a JEOL one that he used for research on all kinds of stuff from Iron Age artifacts to gas turbine blades.

[1] https://en.wikipedia.org/wiki/Electron_microprobe

This family has apparently been doing a lot of things right from a business and operations management perspective. But their alloys are not "secret". It's actually never been a secret :) Production of cymbals is one of the lesser-known deep-rooted traditions in Istanbul.

http://www.aa.com.tr/en/culture-and-art/468267--holy-metal-i...

FWIW "Zildjian" is a mix of Turkish and Armenian. "Zilciyan" would be the Turkish spelling where "Zil" is cymbal or bell in Turkish, the -ci suffix adds a "seller-of" or "manufacturer-of" meaning and -yan == -ian, "descendant-of" in Armenian.

Sarcastic comments aside, it is not just the "quantitative" or "stoichiometric" (if I may) side that counts.

In alloys, there is what is called the Phase Diagram.

Depending on how you mix stuff (proportions), the temperature you mix them at, the pressure you mix them at, and the rate at which these change during the operation, you will have completely different characteristics.

There's also what is called the Time-Temperature-Transformation diagram.

It's not enough to set parameters, how they change matters, too.

The difference between hard steel bayonnette and a flexible shaving blade comes to mind to illustrate the influence of the rate of change of temperature, for example.

And here I only talked about a binary phase diagram. In the case of the article, it's a tin-silver-copper (as far as we know). The result is a solid. In a phase diagram, if you have it easy, it is finding a point in a continuum if you want to reverse engineer the alloy 100%.

[0]http://chemwiki.ucdavis.edu/Textbook_Maps/Physical_Chemistry...

[1]http://www.sv.vt.edu/classes/MSE2094_NoteBook/96ClassProj/ex...

All those things are true, but they just make it harder to duplicate. Not impossible or even especially difficult for people who know this stuff.

Plus there are only a limited number of ways you can mix them, a few trials and checking which come closest, then refining from there would get you there as well.

You can't centrifuge molten metal to separate the constituent elements in the same way you can't centrifuge salt water to get the salt out.

However, you could think of a mass spectrometer as working slightly along those lines: send charged ions though a curved magnet, and vary the magnetic field to steer the mass (i.e. the element) you are interested in to the detector.

The easiest way to determine the composition of a cymbal would be to use a handheld XRF analyser-that is a convenient method scrap metal yards use to identify alloys. If you are willing to trade time for precision & accuracy, you could use ICP-MS or ICP-OES or laboratory XRF.

If you centrifuge salt water you'll get a concentration gradient and if it's saturated at the end you'll get salt crystals. Alloys aren't solutions, so it won't work, but you'll still get a gradient. You could iteratively refine by removing the unwanted half from two spins and combining the wanted halves for a third spin. I'm not sure this is practical...

Do people actually do this? Can the accellerations you reach with laboratory centrifuges induce sufficient concentration gradients in salt solutions? Surely the salt begins dissolving as you stop accellerating?

No, people don't do this in practice, and I haven't seen it done, but it could be done if you had the right equipment and a salty enough solution. Just decant the water at the end, salt doesn't dissolve immediately. I don't know how fast you'd have to go.

Sure, in the same way you can look at the ingredients of an M&M's bag and replicate the chocolate candy.

Or I'll bet the accountant knows the formula by looking at the order books.