Swedging the keywork - pros and cons.

There's a lot of voodoo and mystique surrounding saxes - but when you strip all that away all you're left with is the fact that the sax is a machine, and a relatively crude and simple one at that (at least in engineering terms). And like any other machine it's subject to wear and tear.
Your brand new horn starts out life with an action that's tight and slick - or maybe not so tight and slick if you bought a cheap one. From the moment you pick it up and start playing, it starts to wear - and it does this through the process of friction. Each key is mounted on a pivot of some kind, and for most woodwind instruments it'll either be a rod screw, or it'll be a pair of point screws. If you're unfamiliar with either of these terms, check out the article on point screws. As for rod screws, well, read on...

A rod screw (or hinge screw in some parts of the world) is pretty much as it sounds...it's a rod, with a thread on one end of it. A typical modern sax will have around a dozen or so such screws. There'll be one on the crook key, three for the palm keys and another one for the front top F link, one for the octave mech, a pair for the side Bb/C keys, one for the G# cup key - and depending on the design of the sax maybe one for the low C/Eb key and one for the low C#. And then there are the two long rods screws for the upper and lower stack (the left and right hand main key groups). The rest of the keys will be mounted on point screws.
Dealing with wear on keys mounted on point screws is generally pretty easy - but worn keys mounted on rod screws require a rather more complicated approach.

Let's have a look at the problem in detail.
The upper example in the graphic below shows a shiny new sax key sitting on its rod screw - rather exaggerated for clarity. It's a very simple mechanism; the key barrel (or hinge tube...in those other parts of the world again) is just a brass tube and the screw is a solid rod of steel. It's usually silver steel, but it could be plain mild steel, stainless steel or, less commonly, some kind of alloy such as nickel silver. Most barrels will have one or more attachments, such a key cup, a touchpiece or some sort of connecting arm.

In order that your shiny new key doesn't wobble about, the tolerance between the diameter of the rod screw and that of the hole through the barrel (known as the bore) will be reasonably close. I say reasonably because while there are set standards for clearances and fits in engineering, woodwind keywork just isn't built that accurately. And with good reason.
For a start, the keys themselves are quite soft and flexible - and the keywork as a whole has no provision for preventing dust and grit getting into the key barrel. It has to work over a wide range of temperatures and each key must be free to move with only the lightest touch.
So what sort of tolerance are we talking about? Ideally it's around 0.01mm, which is just under half a thou in old money. And that's across the diameter - so the actual clearance between one side of the rod screw and the opposing face of the bore will be half of that.
To put that into a real-world perspective you can take a rod screw with a diameter of 2.84mm and slide it nicely into a stack key from a Conn 10M. The key will move freely with no sign of wobble. Pop a 2.85mm rod in and you'll notice that while it will just about go into the barrel, the key will no longer be able to move freely on it. 2.84 was a sliding fit, 2.85mm is an interference fit - which gives you a very rough guide as to the clearance required.
This ideal clearance is denoted by the thin red lines on the upper example of the graphic.

Now, I said ideally because keys are rarely made with such accuracy. If I took all the keys off a Conn 10M (and we're assuming it's factory fresh) I could bet you a pie and a pint I'd be able to find at least one key in which the 2.85mm rod would be a perfect sliding fit. And I could probably find a good few where the rod would slide part of the way in and then begin to bind up - and this will be due to inaccurate boring of the barrel or a slight bend in it.
And all of this assumes that the rod screw itself is dimensionally accurate over its length...which isn't always the case.

Anyway, this is our starting point with a new key; a rod screw of a known diameter and a barrel that's bored out to be 0.01 larger. Everything is nice and tight. But as time goes by the key will wear - and when it does so the wear will accumulate on both the bore of the barrel and the rod screw (though to a lesser degree), as shown in the lower example. The key now has at least twice the clearance it had from new.
What this does to the action is quite subtle. For the smaller keys and those that stand alone (i.e. not connected to any other key) it hopefully won't make that much of a difference. The keys will operate normally and unless the wear is extreme it's unlikely to result in a leak on a key with a pad on it. However, it will probably be rather noisy and advanced players may notice that the key feels imprecise in operation and the horn doesn't quite speak as quickly and responsively as it once did.
Here's a perfect example of a wobbly key, and shows you precisely how you can test your own keys for wear. Just grab one end of the key and try to wobble it from side to side. In this instance there are more problems than excessive clearance in the key barrel because you can actually see the barrel sliding back and forth between the pillars. It was made like this.

For keys that are linked to other keys - such as on the main stacks - the effect of this wear is rather more insidious.
The principle of linked keys is pretty simple; you press one key down and another one goes down at the same time. Press your low F key down and you'll see that the key cup above it (the Auxiliary F) also comes down. Such keys are linked together by a bar coming off one key that sits over the foot of the other key. Inbetween this connection there'll be a thin buffer - either a piece of cork or felt - and this is the means by which we 'regulate' the movement of both keys so that they close together at exactly the same time.
But there are a lot of factors working against this regulation. There's the flex in the keywork; you'll be hard put to see it, but it's there. There's also the force of the springs trying to push the keys back up. And then there's the potential compression of the regulation buffer and, finally, that of the pads. It's all conspiring against you to lift that linked key up and create a nasty little leak.
So you can imagine that it makes a great deal of sense to minimise these factors. This can be done by fitting thin and firm regulation buffers (though this can induce key noise) or by using firm pads (again, can be noisy - and sometimes uncomfortable under the fingers for some players) - but as for the flex in the keywork, there's really not a lot you can do about that.
But now chuck in some wear in the key barrels. You press the F key down, the Aux.F key comes down with it...the springs push back, the regulation buffer compresses a bit, the pads compress, the key flexes...and all this opposing force looks around for somewhere to go. And whaddya know...it finds that the keys are worn and can move on their pivots. So the Aux.F rises. Oh, it won't be by much - but bear in mind that a gap of just one thou between the pad and the tonehole constitutes a leak. If it's just the one key that's worn it probably won't be that much of a deal - you'll usually be able to blow right through it. But when one key has worn it usually means that others have too - so by the time you get down to the low notes you could be looking at a number of leaks due to worn keywork. That's rather harder to blow through.

I mentioned firm pads, and at this point it's worth saying that they can be very unforgiving. They're going through a bit of a trend these days and I see quite a lot of horns that have been fitted with them that have subsequent problems. Part of this is down to being badly fitted and set, and part of it's down to insufficient preparation. With so little give in these pads a perfectly flat tonehole and a level key cup is a must. But it should also be noted that a tight action is necessary. You can do all the tonehole and cup levelling you like, and spend an hour setting each pad - but if there's even a slight bit of wear in the keywork a hard pad will punish this oversight mercilessly.

It's hopefully clear, then, what effects a worn action can have - but it gets a bit worse, because wear tends to beget wear.
Here's our worn key at rest. Because there's a spring acting against it (to hold it open), the force of the spring takes up the wear and causes the key to tilt along its axis. You can see that the key barrel is now only touching the rod screw at each end.
Each time you press and release the key the contact point shifts up and down the barrel - but the area that will see the most contact will be that at each end of the barrel. So this'll wear more because this is where most of the friction will be concentrated.

Over a period of time the situation gets worse - and then another kind of wear creeps in.
Because the key barrel is now canted, it no longer sits square to the face of the pillar. This means that friction is concentrated in one area rather than spread over the entire end of the barrel. You can see that the rear portion of the barrel has some clearance - so it's not touching the pillar - but the front portion is. Little by little, bit by bit, this'll wear down...and at a faster rate than had the pillar been making full contact with the end of the barrel.
And now it gets even worse - because with that increased gap on one side of the barrel there's more scope for larger particles of grit and dust to find their way inside the barrel. How it does this is quite fascinating. In my article on Oiling the action - for geeks I noted that the process of pressing and releasing a key causes the key barrel to flex. It's only very slightly, of course, but in so doing it acts like a pump - pushing oil out of the barrel as the key goes down and sucking it back in again as it's released. The oil that gets pushed out gathers where the end of the key barrel meets the pillar. Because it's such a small amount of oil, surface tension prevents most of it from dribbling down the pillar - but while it's hanging around it has the potential to pick up any grit that may have adhered to the oil left behind from previous pumping cycles. As the key rises and sucks the oil back into the barrel, it carries this grit with it.
The grit now mixes with the oil and acts like a very, very fine abrasive - and this can only mean one thing...more wear. Sure, it won't happen overnight - but each time you play the horn it gets just that tiny little bit worse. Leave it long enough and the key will end up like that top D key - being able to wobble on its rod screw and being able to rattle about between the pillars.

The above diagrams are somewhat exaggerated - but also simplified. In practice the wear patterns can be rather more complicated. Take a look at this worn lower stack rod screw.
It's off a mid '60s Selmer MkVI tenor - so it's had around 50 years to get into this state, but judging by the overall condition of the body work it's a horn that hasn't seen a very great deal of use.

You can see a polished portion in the centre of the action where the low D key barrel sits, and this area is where current wear is happening. You can see that it doesn't go all around the circumference of the rod, and in fact it lines up with just the lower rear section of the key barrel.
As this section wears down, those polished point will move up and down the rod - so this is just a snapshot of where the wear is right now.

You can see that the pillar section is a slightly darker colour, and we'll talk about that a little later in the article.
But look just to the left of the pillar section, where it meet the D key section. See that distinct ridge between them? If you slide a fingernail up the rod you can actually feel it catch on this ridge. And just like the shiny spot lower down the rod, this ridge doesn't go all the way around the rod.
Again, we'll come back to this point a little later in the article...

As for those thread-like lines around that area, they were caused by a foreign object when the rod screw was either fitted or removed at some point. It may have been a burr on the key barrel when the horn was assembled in the factory, or a particle of grit that later found its way into the barrel...and left its mark when someone dismantled the lower stack.

So what can be done about all this doom and gloom? Simple - have the keywork tightened up. Yep, take it along to your local friendly repairer and have them swedge the keys. Job done, right?
Well, maybe not. Let's have a look at what swedging is, and talk about the pros and cons. First off, though, it's pronounced 'swayjing'. Don't ask me why, that's just the way it is.
Swedging is, to all intents and purposes, a form of crimping. What the process does is deform the key barrel by compressing it. This does two things; it reduces the diameter of the barrel (internally and externally) and it stretches the barrel. In this way it can deal with wear within the barrel and wear between the pillars. Very handy.

As you might imagine, there are special tools for this job - which, in the main, tend to be swedging pliers. There are two ways to acquire them, the first being that you buy a pair or two from a repair supply store. The second way is that after the ones you've bought have broken, or failed to give you the results you desire, you make your own...just like the ones you see here.
Every repairer has their own idea about what makes for a good pair of swedging pliers - so designs and sizes will vary - but they all have one thing in common; the business end of the pliers should be highly polished. This helps to prevent marking the keys.
I say 'helps' because it can often be quite a brutal process; you may need to apply a considerable amount of pressure to the key barrel - and even with the smoothest jaws and perhaps a protective grease, wax or film, there's always a chance that you'll leave what we call 'witness marks' behind. Sometimes its inevitable. If a key has poor quality or flaking plating or lacquer, just the application of pressure alone will damage it.

There's also the chance that you might leave ripples behind. Some keys may have been harshly buffed at some point, which leaves the barrels out of round (some metal has effectively been polished away) - and this will affect the neatness of the job. If there's any grit between the tool and the barrel it will leave score marks - as may any imperfections in the tool's jaws.

The tool at the bottom is a variation on a theme, being what we call a collet swedger. It works on the same principle except that it uses a collet (a sort of chuck) to apply the pressure. It can produce a neater job sometimes, and can deal with very short barrels that would otherwise be difficult to handle with pliers...but it won't exert as much pressure.

Swedging is particularly hard on the hands - especially if you're dealing with nickel silver keywork (much tougher than brass), such as you might find on the King Super 20. After a day's swedging on one of these beasts it can take you a good half an hour before you can actually open your hand again. To overcome this problem you can buy/make a larger tool...or, if you're a cheapskate like me, you can pop a couple of steel tubes over the handles to act as an extension...thus significantly increasing the amount of pressure you can exert for less effort.

Here's how the tool works.
You start by cleaning everything. The tool's jaws, the exterior of the key barrel, the interior and the rod screw. You fit the rod into the barrel, apply your favourite barrier to the jaws then clamp them around the part of the barrel you want to tighten, and then squeeze the pliers hard.
Various techniques can be used. Some people favour clamping down and rotating the key or the pliers, moving slowly back and forth along the length of the barrel. Others prefer to clamp down in one spot, release the pliers, rotate them or the key then clamp down again. Or perhaps a combination of techniques depending on the key in question. The collet swedger is used in a similar fashion, although it doesn't take too well to the second method.
As with many repair techniques there's a bit of a knack to it. If you go steaming in with a swedger you might find that all your key barrels end up slightly bent. This is because swedging induces torsion in the key barrel - and unless you're careful to balance this out, you'll run into problems. They won't be especially difficult to correct, but it will take time. On a key like the one on the right, which has two separate barrels, you will nearly always have to realign them after swedging. But by far the biggest 'gotcha!' is attempting to swedge a barrel onto a rod screw that's worn in the middle. I'll leave you to work out why that rookie mistake is a bit of a doozy...

It's at this point I'd like to have a moan about one of the swedging techniques I see on a regular basis - which I call 'end-swedging'.
In the above paragraph I mentioned that you move the pliers along the length of the barrel. The reason you (or at least I) do this is because wear is present along the entire length of the barrel. If someone asked me to remove the wear, I remove it all - or at least as much as is possible - and not just a bit of it.
The thing is, end-swedging does the job. If you take your worn out horn to a repairer and they end-swedge they keys and return it to you, you'll probably be very happy with an apparently now-tight action. For a while, at least.

Here's what end-swedging looks like.
You can see that the barrel has excess clearance along its entire length, except for the ends of the barrel where the swedging tool has done its work. Because the wear has been taken out at each end, the rod screw can't flop about inside the barrel. But now there's only a limited amount of contact between the rod and the barrel, which means that all future wear will be concentrated at those points...and when you concentrate friction, you increase the rate at which wear happens.
In a relatively short space of time the barrel ends will wear and will need swedging again.
It also leads to a rather unpleasant look - because the diameter of the barrel ends will now be slightly smaller than the rest of the barrel - and if the process is repeated you almost always end up with a rather ugly taper.

Why do people do this? I don't really know. I suppose it saves time, but then doing half a job always saves someone time...and costs someone else more money in the end.
I've been told that some people say that it allows for the key to maintain a reservoir of oil in the barrel - and so it does, for about as long as it takes for someone to press the key down. The resultant pumping action will push out the excess oil, but the increased clearance further inside the barrel will not allow all that oil to be sucked back in again. You just end up with oily pillars, and your action wears that much faster. In any event, the normal amount of clearance is entirely sufficient to maintain a film of lubrication between the rod screw and the key barrel.
So I consider it to be nothing more than a bodge - a temporary repair that gets you out of trouble for a while when you're a bit short on funds. Or a quick way of making a worn out horn appear to be in better shape than it really is prior to selling it. Fortunately it's often quite easy to spot if a horn's action has been end-swedged - just look out for witness marks or tapered/reduced barrel ends.

There are also some limitations as to what swedging can do and how much wear it can take up, and this is largely down to the ability (or otherwise) to get the tooling onto the key.
So far I've shown key barrels with even wear right along the length - but in practice a barrel doesn't wear in this fashion. Here's what really happens. There's a certain amount of wear that extends along the entire length of the barrel, but as the key becomes more and more able to tilt along its axis, the wear begins to increase at each end.
Once this happens it leads to another problem, because the key can now tilt enough to allow a fulcrum point to develop further along the barrel - and this will typically gravitate towards the spot where the tip of the key's spring sits (which may or may not incorporate a key arm). In this instance (on the right) the tip of the spring sits in the right hand key arm - and the tilting of the key combined with whatever flex there is in the key has led to a 'pocket' of wear forming.

You can probably see the problem that arises from this; namely that there are a couple of areas where you're not going to be able to get at with your swedging tools - and so even given your very best efforts you're simply not going to be able to remove all the wear. However, you'll be able to deal with most of the wear in the middle of the key - and you have enough 'meat' at each end of the barrel to ensure that whatever swedging you're able to manage, it will at least prevent the key from being able to tilt for quite some considerable time.

But here's another example. This time the key arms are positioned at the ends of the key barrel - and just as in the above graphic we have two point of wear that can't be addressed. You can make the rest of the barrel as tight as you like, but you really can't do much about the wear beneath the key arms. You could 'peen' (essentially squash) the ends of the barrel, but that leaves you with an awfully small bearing surface that simply isn't going to stand up to a great deal of wear.
In this instance you'll have to rely on the centre portion of the barrel to do most of the work - as it largely will for a good long while. But such keys are more prone to rapid wear than the example just above.

On a similar theme there are keys where the key arms are so close together that not even the thinnest pair of pliers can get into the gap between them - as well as keys that don't really have much of a barrel at all, such as the lower end of this Auxiliary B key from an old alto.
It's just about got enough of a barrel to get a tool on, but the job would still leave around 50% of the barrel still with wear in it. On many older horns you'll find such keys with no barrel at all - there's just a hole drilled through the key arm. In such instances peening the arm is about all you can do...unless you want to do a proper job and fit a bush. There are also keys where the design precludes the stretching of the key barrel - such as the low C on later Martins; and it's not all about saxes, there are keys on flutes and clarinets that can't be swedged.
Which leads me onto another moan. I can understand why many older horns had such foibles, but it really is disappointing to come across a modern horn where no thought has been given as to how the action might be tightened up after a good few years of use. Granted, it's thankfully uncommon on saxes these days - but it still crops up from time to time...and it really ticks me off. There's no excuse for it - and I think the world would be a much happier place if manufacturers of all kinds of gadgets were required by law to throw their products at a repairer before being allowed to place them on the market. I feel better for having got that off my chest....

If we look even closer at what happens when a key is swedged we can spot another issue.
The whole process is somewhat 'organic' by nature. I mean, you're trying to compress a brass tube of possibly varying diameters against a rod screw of potentially equally variable ones.
One way to mitigate some of that is to replace the rod with a new one. By swedging a key and (very) carefully cutting it in half - then running an accurately ground bar smeared with a little marking blue along the bore - it's possible to see that the bore of the barrel isn't compressed evenly along its length. There are ripples, peaks and troughs which leave wear patterns like these shown on the right. OK, so this is being extremely picky - but it nonetheless shows that while a great deal of a wear has been taken out, it's by no means all of it...and you have to add that to the areas that you simply can't get your tooling on.
It's essentially all about percentages. A new, well-built horn comes with an action that has 100% contact between the rod screws and the key barrels. Now, I say "100% contact" but in fact it isn't really that at all. An accurately bored barrel and a close-fitting rod screw will have space for a film of oil between them that keeps them apart - at least while the key is under no stress. If you were to hold a cleaned and degrease key barrel upright and drop an equally clean rod screw into it, it would pass straight through the barrel....then hit the floor and ping off somewhere you'll never find it. At least not until you've made a new one.
The moment you apply stress you'll force some of that clearance closed and rubbing will occur. The lubrication will go some way to limiting the effect of that, but it won't prevent some wear occurring. What that 100% really means is that there is potential for full contact - and that any rubbing due to the key being stressed will be distributed over as wide an area as possible. When you lose a percentage of that potential contact it shifts the rubbing action into an ever-decreasing area, which then wears faster.
A repairer's job is to take a worn action which might be running at, say, 70% contact and bring it back to as near as possible to 100%. And as we've seen from all of the above, that can be a tough call.

So what's the alternative?
Well, as already mentioned, changing out the rod screws will help. This allows you to swedge down to a known and accurate diameter. Another little trick is to overswedge. You can either do this by really bearing down on your swedging pliers (at the risk of damaging the lacquer or plating, and your hand) or you can use a set of very slightly (and only very slightly) undersized rods - and thereafter lap the original or replacement rods in with a suitable compound. It's worth bearing in mind that lapping will reduce the diameter of the rods very slightly, so its often worth using the old rods for most of the work and then replacing them with new ones...which won't usually require a great deal of lapping, if any at all. But even if you don't damage the finish, it's likely to leave the barrels looking a bit uneven.

All well and good, but there's another spectre looming on the horizon.
Y'see it's not just the key barrels and rod screws that wear - the pillars wear too. "How can that be?" I hear you ask, "Nothing moves in them...they just sit there!". And so they do, but every time you press and release a key it imparts a knock to the rod screw, which in turn imparts a knock to the pillar(s) through which it runs. It's a bit like the proverbial bird pecking away at a mountain - give it enough time and the damage will show.
When this happens it means that you can tighten up the keywork all you like but the keys will still be able to move laterally. And what's worse is there's sometimes more free play in the pillars than there is in the keywork - probably because pillar wear is something that gets overlooked during servicing.
Here's a good example of it... and you can really see how much movement there is on the rod screw.

How that translates into action movement can be quite spectacular.
There's a bit of free play in the key barrels on this soprano, but by far the bigger problem is that the entire stack is able to move back and forth. It's not down to wear on this occasion, rather it's down to poor manufacturing - but the end result is exactly the same.
This has grave implications for the internal regulation of the key stack, and means that whatever combination of keys you press down - no matter how well-set the pads are - there will always be a leak on the stack. That's a big problem on any horn - - and I can tell you that it's the most common cause of vagueness in the low notes, but it's especially problematic on something as small as a soprano.

Fortunately there are a couple of ways to deal with this pesky problem, the first of which is (to my way of thinking) a bit of a bodge.
You can peen the pillars. This requires the use of a peening tool, which is nothing more than very hefty pair of pliers that has a steel dome or ball on one of the jaws and a flat on the other. You position the dome in one of the holes, place the other jaw on the other side of the pillar....and squeeze like hell. As the ball gets forced into the pillar it displaces the metal around it...some (and only some) of which ends up reducing the diameter of the hole. You then repeat the process for the other side of the pillar.
It sort of works, but it doesn't do a lot for the face of the pillar, and even less for the side of the pillar that houses the head of the screw. And because it's a bit of a bodge it's not really a repair that will last a very great deal of time. Fine if you're on a budget or just want to flog the damn horn, but not what you'd call the 'gold standard'.

A more advanced method is to bush the pillars. This is rather complicated. You must first ream out the pillars oversize. Then you turn down a brass rod to fit the hole you just reamed, and you drill an accurate pilot hole through it. You then insert the rod into the pillar and soft solder it in place. You trim up the ends to match the width of the pillar head and then you ream the bush out to the required size...taking care that the hole you're reaming aligns with any others along the path of the rod screw.
It's a fair bit of work, and it carries with it the risk that the lacquer might get scorched in the soldering process - but thereafter it restores the pillar to its original specification. And it won't wear out again for a very long time.

And now we come back to that earlier photo of a worn lower stack rod screw.
It can be clearly seen that the wear is irregular - and from that you can surmise that the rod screw is no longer truly round. And neither is the bore of the key barrel. If you were able to fully swedge a worn key barrel down onto a worn rod screw, it would bind up because various parts of the rod screw and key barrel would be more oval than round. The best that you could do is aim for a sort of 'working average'. And as for that section with the ridge worn into it - I really don't know what you could do about it, save for replacing the rod - but that still leaves you with an out of round barrel to deal with

But what if I told you there was another method of tightening the action that restores it to at least as good as new spec (often better), that doesn't mark the key barrels, that deals with both wear in the barrel and on the barrel ends at the same time...and sorts out the wear in the pillars too?
Welcome to the wonderful world of key reaming.
Simply put, you don't 'reduce' the free play - you cut it out altogether - and then you make up the difference by fitting an oversized rod. How much oversized? Well, that depends on how far you want or need to go. Let's go back to that poor old Conn 10M I referred to at the top end of this article. It typically has a rod screw diameter of 2.84mm. By the time the horn's done a few tours and been through a few owners (none of whom have looked after it terribly well) the action is likely to be - as the saying goes - worn to bloody buggery. If it's that badly worn then you'll probably have to step the rod screw size up to 2.92mm - which, if you've done your maths, is 0.08mm larger. It's not much, is it?
It might even be the case that the larger rod will fit straight through the keys and the barrels and thus be almost a straight swap - though this would put the action back to state similar to it having been merely swedged well. For many players this is a sensible option. It restores most of the accuracy of the keywork, doesn't damage anything and works out reasonably cheap - given that all the repairer has to do is cut a thread on the rod and cut it to length. And perhaps ream out the odd key/pillar or two. However, it doesn't do much for stretching the key barrels - but then with a tighter fit on the new rod screw you don't have to go at the swedging pliers with a gorilla grip.

If the 2.92 rod fits through most of the keys you might want to consider going up a step. This would typically be 3.00mm - which is still only 0.16mm larger than the original rod. This would probably require you to ream all of the keys and pillars - but in so doing you clean up the bores of the barrels and take out most, if not all, of those wear pockets. And because you've had to remove a slightly larger amount of material from the bore of the barrels, the process of doing so will stretch them - usually enough to to take up the axial (end to end) wear.
Of course, the rod screw sizes vary from horn to horn. Selmers, for example, use a 2.78mm rod for the main stacks and larger keys - so the most likely step-up for a well-worn MkVI would be 2.84mm.
So what are the cons? After all, it's clear that swedging the keys has a few - surely reaming the keywork has some?
From a mechanical perspective, none that I can think of. Yes, it's true that the process of reaming the keys removes material (which can't easily be replaced), but it's not quite as clear-cut as that. For example, wear removes material - so if you're doing a straight swap to a slightly oversized rod there's a good chance that you won't be removing any material at all. If you step up a size and have to break out the reamers, you're still not removing very much material at all. And remember what I said earlier - that wear begets wears. By making a half-arsed job of removing such wear you're just pushing the problem further down the line. It's much better to go in with all guns blazing and sort it out once and for all...or at least for the next half-century.

As to the reamers themselves, typically these will be shop-made tools. They're simple to make, cheaper and just as effective (if not more so) than shop-bought tools...assuming you can even find them in the appropriate sizes and lengths. The replacement rod screw steel is typically silver steel (also known as drill rod), which means it can be heat treated to harden it - so not only do you have a ready supply of potential rod screws, you also have exactly the right material from which to fashion a reamer. At a pinch you can even make a reamer (for brass) out of plain steel. You'll get one or two cuts out of it before you need to grind a new cutting edge on it - but you'd really only want to go down this route for a one-off job or an obscure size of rod.
Little geeky tech-tip here; when hardening these reamers I've found I get better results by not tempering the blade of the reamer after hardening.

There are a variety of designs to choose from, but I've found that the basic D-bit reamer (as shown above left) is all I need most of the time - but I also use a tri-point style of reamer (shown above right). Takes a bit longer to make but tends to work better on harder metals, such as nickel silver. And as with all tools there's a bit of technique involved. I tend to start by degreasing everything and use the reamers dry on brass. Nickel silver sometimes benefits from a little cutting oil, but for the most part you have to rely on feel. You can certainly make life a lot easier on yourself by taking one or two undersized cuts before going for the final size. It's not that a well-honed reamer won't tackle a reasonably large step up, it's just that the risk of the tool binding in the barrel increases. When this happens it places you in a very precarious position, in which you have to decide whether to push on through or try to back the reamer out. If you make the wrong call there's a chance that the reamer will lock itself tight in the barrel and snap in two - and then you're in a whole heap of trouble. Unfortunately there are few hard and fast rules about the process, especially since what works well for the keys on one horn might not work at all well on the keys from a different one depending on the grade and quality of brass used. In the case of Chinese Ultra-Cheap horns - which benefit tremendously from oversized rods from new - pushing through with gusto is the way to go.
I also find the length of the cutting blade makes a difference to the finished job. An acute angle on the cutting face will make for a faster cut, and one that can remove a great deal of material - but the surface finish of the barrel's bore won't be so smooth, and so will usually require quite a bit of lapping. A less acute angle makes for a longer and more gently tapered cutting edge, which gives a nicer and more progressive cut - but it comes at the cost of speed and some weakness in the tool.

I guess that some people will be horrified at the prospect of cutting chunks of brass out of a horn, and some might suggest that changing the specification of the rod screw affects the horn's originality or that reaming the barrels makes them thinner.
My response to that is that swedging the keywork changes the specification of the key barrels and also makes them thinner - so it's six of one and half a dozen of the other. Reaming, at the very least, does a much better job of restoring the functionality and accuracy of the action. Yes, it will probably make the wall of the barrels slightly thinner - but not really by very much at all...and with a slightly larger rod screw fitted you've actually increased the bearing surfaces (if only fractionally) which results in less wear over a given period of time...and as mentioned earlier, wear makes the barrel wall thinner anyway.
As for cutting chunks of brass out of a horn - well yes, it does do that...but so does tightening the action by swedging. You're still going to have to square up the ends of the key barrels and face off the pillars in order to ensure a precise key fit. It's all about choosing a method of dealing with a problem that does the least amount of damage and gives the best results in the long term.

I'm going to finish up this article by showing a truly ghastly example of what can go wrong when you either don't know what you're doing and/or don't have the right tools.
This is a fine old Conn 10M. It's survived for around 90 years with little more than a few minor dings and a spot of fading lacquer here and there. In other words a very nice example of a well-treated vintage beauty. Or at least it was up until the point that someone decided to tighten the action.
This, my friends, is one of the worst examples of swedging I've ever seen in my entire life. From looking at the serrations on the right hand side of the key it's clear that a pair of linesman's pliers were used - like these ones on the right. It would still be quite bad if they'd merely crimped the barrels, but you can see on the other end of the key that they've rotated it too. Oh, the humanity.
And as much as I turn my nose up at the practice of end-swedging, on this occasion I'm thoroughly relieved that they adhered to this make-do method. Can you imagine what would have happened had they run the pliers all the way along the key barrel...?