This comes up fairly regularly so I thought I'd just put together a comprehensive post, outlining some of what I've learned through my experiences and experiments with different foundries and magnets.
Rough cast guitar magnets and “polished” guitar magnets are made the same way (at a given foundry). The AlNiCo alloy is mixed, heated in a crucible until molten, then poured into green sand molds. Once filled, the molds are then subject to a calculated and complex series of heating and cooling cycles at different temperatures for different times, and usually exposure to strong magnetic fields during some of these cycles. All of this depends greatly on the particular type and properties of the magnet being made and the particulars of a given foundry. Variation in any of these can result in a different sounding magnet, at the end of the day.
The difference in manufacturing process comes after this. The "polished" magnets, having been cast a bit larger, are then run through a surface grinding process which grinds off the rough surfaces, bringing them to the final size on all sides, while being cooled in a solution to prevent heat from physically changing the material. That’s why I keep putting “polished” in quotes. They aren’t really polished at all and usually show tooling marks from the surface grinding. This same process is done to rough cast magnets, but only on the North and South polar faces, leaving the other four surfaces rough. Many rough cast magnets also end up being ground flatter on the other four surfaces, as well, if they are so rough and unusual in shape that they are outside the range of tolerance to fit the order. That’s why some rough magnets may appear to be smooth on only part of a surface.
Any differences heard between rough cast and polished magnets are likely due to the magnets coming from different foundries, or being made in different batches with loose tolerances or sloppy practices, rather than the actual surface texture of the magnet. Especially regarding inexpensive modern magnets.
Three “polished” magnets from three different foundries will likely measure and sound a bit different from each other.
Similarly, three rough cast magnets from three different foundries, will also likely meter and sound a little different from each other.
From the same foundry and time, a rough and “polished” magnet of the same type are likely to sound the same. Especially if they are from a precise and automated foundry with high standards and tight tolerances (but most aren't).
There are some rare times that a rough cast magnet can change the sound of a pickup due to the rough surface. It's not common, because generally both the polar faces of the magnet, which contact magnetic surfaces in the pickup (keeper bar and slugs), are surface ground flat, even on rough cast magnets. An example is some of the A3 magnets that Gibson used in P-90s in the early 1950's. They were made with only the South face of the magnet polished, with the North face left rough like the other four surfaces. That’s because in P-90s, the North polar surface faced the outside of the pickup, not making contact with anything, and only the South sides faced inward, contacting the keeper bar. In that case, when some of these magnets made their way into PAFs in 1957 (and, although very rare, I've seen them in pickups as late as the mid 1960's) this rough surface on the North polar face of the magnet which contacts the slugs could be so rough that it creates an air gap by having very poor contact with the slugs. Small air gaps cause magnets to do strange things in physics. This can change the sound, in rare cases.
So, generally, when a difference is heard between rough or "polished" humbucker bar magnets, it's because they came from different foundries or at least different runs of AlNiCo production. Some foundries, especially the lowest-bid type, can be extremely inconsistent. Not just the alloy used, but the heating/cooling cycles temperatures and times and field exposures all change the resulting properties of the magnet. I've seen how some of these foundries are run and experienced the results. The processes are not automated and terribly inconsistent. On the other hand, some magnets are cast here in the US with military precision and produce extremely consistent results. ...and one will pay for those. They cost quite a great deal more than the offshore alternatives. Further creating differences in magnets, some manufacturers go so far as having foundries alter the alloy or process to achieve somewhat different results for a given magnet type, intentionally. Some foundries follow these instructions well, some don’t. Others likely just ignore them entirely and ship what they ship. It can be a real mess depending who one deals with in manufacturing.
There is another type of polished-looking AlNiCo magnet, which is not sand cast at all. It's called a "sintered" magnet. Instead of having a molten iron-like alloy poured into green sand molds like the two described above, sintered magnets are made by pressing a powdered AlNiCo dust into a die, then compressing it under huge amounts of pressure and heat until it solidifies into a permanent shape. These magnets are magnetically weaker but cheaper to make.
I've not seen typical guitar magnets which are made with a sintering process yet, though it's possible they are out there. If you see flaws/exposed cavities in polished magnets, you know they are cast, then surface ground. Those flaws are bubbles and the sintering process doesn't allow for those bubbles/cavities.
At the end of the day, not all magnets of a given type achieve the same result in a guitar pickup. There are a great deal of variables in materials, process and tolerance that result in different chemical, physical and ultimately sonic properties of a pickup magnet. Control and consistency is how a manufacturer can use that to their advantage, rather than experience it as a problem. Personally, as I've tried to simplify and cut down on the types of magnets I use, over the years, I've only found valid reason to increase them. For example, I now stock five different A5 types, to achieve significantly different voicings in pickups. They really do sound noticeably different and measurements do reflect that they are different, as well. Magnets are relatively cheap, even for American ones. That's one upside. Pickup magnets are a rather inexpensive thing for musicians to play around with and experience the differences for themselves, in their own gear, compared to most of the equipment in the music industry. Happy swapping!
