Tool Selection for Digital Artists

On Ye Olde Computer Forum, a discussion sprang up recently about whether or not the Mac Pro would continue to be a product.

It took me a few posts to really nail my thought process with this one on the head, but I think I've got it.

The discussion (on the forum and on the Internet at large) centers around whether or not the Mac Pro is a viable product for Apple, and whether or not it would be reasonable for them to update it (i.e. would it make them any kind of money at all, or would it be a loss leader like the xserve was) and whether or not a scaled down version of the Mac Pro we've got today (i.e. a single socket system, or something with fewer PCI/e slots, or fewer disk bays) can really be capable of filling the same spot that the Mac Pro does.

The arguments I see on the forum are essentially "yes, the Mac Pro is an essential part of Apple's line-up, both for people who need the computing horsepower for straight-science activities as well as the people who need the computing horsepower or the massive number of interfaces for specific artistic equipmen.

The folks in the first group one, who are using Macs for science, would originally have joined the Mac fold in the early 2000s when products made by NeXT were being integrated into Apple, and the biggest one was the OS itself. Apple advertised pretty heavily at the time that its computers (both desktops and portables) were UNIX computers, and in an era where Silicon Graphics and Sun looked like they were floundering as badly as Apple had previously been, this must have been very attractive. Additionally, by the time this advertising took full effect (in 2003 and 2004), Microsoft, Adobe and Macromedia had versions of their products up and running on Mac OS X. Apple's value proposition was that you'd have one computer that could do your scientific research using classic UNIX apps, your custom code, and your desktop productivity. And if you wanted, this one computer could be a laptop. These folks probably never required much out of the Mac, hardware-wise, except that it be reasonably fast and that in some cases, it attach to instrumentation. (These days, science instrumentation is prooobably USB, but that's somebody else's problem.)

The folks in the second group admittedly feel a little bit more entitled to say what Apple should and should not build, because they may have been with the Mac from the start, in 1984, building page layouts on the original Mac and PageMaker. I'm going to talk about the second group, artists, a lot in this article.

The primary justification for the Mac Pro's existence as it stands today is two-fold: The first is that it has a lot of computing oomph. As I write this, you can drop in two six-core processors running just shy of 3GHz, and that includes hyperthreading, so you can run twelve concurrent threads. The second justification is that it is very flexible hardware. Even if you only require only "modest" amounts of computing horsepower, you can insert that special PCI Express card for video capture or for Photoshop acceleration (if we'll presume that such a card exists for the Mac, and that you'd want one instead of more ram and more cores, but whatever) or for sound input/output and digital signal processing.

Back to Ye Olde Computer Forum: One of the justifications I heard about keeping the Mac Pro around was just laughable at best. This argument revolves around folks doing audio processing either professionally or as a hobby. As the argument goes, in a lot of cases, you just can't substitue one effect or tool for another. A reverb was given as an example. It may not be possible to make the reverb functionality on a new tool (Say, Logic or GarageBand) do what you were able to make it do in the way you wanted to on your old tool (say, a physical card from Digidesign or a ProTools plugin.)

This person continued by saying that it's part of why he will always keep a Mac Quadra around.

whoa. wait. stop. The Mac Quadra family was introduced in 1992 or so and discontinued in 1994. The most powerful member of this family (which I've got in a storage unit somewhere) happens to have had a 40MHz Motorola 68040 processor and some DSPs in it. This is not a high end computer we're talking about.

So here's what struck me: He and I were probably talking about people at different phases of their careers. Somebody just starting out in audio production is most definitely not going to pull a digidesign card from 1993 as their primary tool for making music. It's just the same as how somebody starting out in digital photography is unlikely to pull a Sony Mavica off the trash heap and try to interface it with their brand new computer. It just does not make sense.

Somebody who has been established in that field for, oh, say, twenty years and has a digidesign setup from 1993 in the closet may pull it out for a specific project, but unless they were unwilling to uprade through the whole 1990s, it's unlikely that the Quadra in question still plays a significant or active role in their studio. Reverb A may not be possible anymore, but that's probably at least as much due to the fact that Reverb A is no longer stylish as due to the fact that Digidesign had a very strong grip on the entire ecosystem for creating Reverb A, and then failed to update it to run on newer computers.

The one thing I'll say about keeping old equipment around is that it's possible the task at hand is super specific. Our older audio artist might make the stylistic decision to continue using Reverb A, for example, or our older photographer may be in the specific business of black and white digital landscape photography, with a camera such as the Kodak DCSProfessional 760m.

The other part of my thought process about the Mac Pro is that it's just against the entire concept of how Apple wants you to treat the Mac ecosystem. People who intentionally purchase Macintosh computers because they are Macs don't really want to think about the specifics of how and what their computer is doing, and what parts make this possible. This is a big part of why Macs cost more, and are less configurable from the factory. A typical Mac user doesn't care what wifi chip or graphics chip their computer has -- they care that it works well, and few Mac users are likely to recognize a specific component that has an upgrade or maintenance need when the machine is far into its lifecycle.

To put it another way: the users see the Mac as a closed system, because that's the way they want to see it. The Mac Pro is the opposite of this because it forces its user (even if only very briefly) to consider what is and isn't needed for their particular use case. For example, while purchasing a Mac Pro, you must decide how many processing cores your task needs, what type of graphics processing your task needs, and what memory and storage requirements your task has. When purchasing an iMac, you simply decide how big you want the display, and whether you want the "good" computer, the "better" computer or the "best" computer. It's the same when buying a Mac laptop -- you decide how big you want it, and you decide from a list of components you can customize therein. (Mainly processors with marginal performance differences, and of course storage/memory, which users are more likely to understand than different CPU grades, different wifi chips, raid configurations and different graphics options.)

So, is the Mac Pro a reasonable product? Ultimately, for pure computational horsepower, it's just not there. IBM, Dell, HP, and Sun/Oracle all have quad-socket servers, and IBM, HP and Sun/Oracle have eight socket servers. If you're running a high-performance computing application, you had probably moved away from the Mac (and Apple's 2005 attempt at a "workgroup cluster for bio-informatics") long ago.

As a flexible member of the existing Mac family, it's sort of reasonable. Of course, it's against the whole concept of the Mac as Apple wants it to exist today, and most specialized cards (such as the Black Magic Design Intensity, or a FibreChannel adapter for a SAN) are already available for USB, Firewire or the new Thunderbolt interface.

Most modern art-forms that are centered around the computer are as much about the tools as they are about the processes, expression, and creation. It's a fairly simple fact that these tools have been changing and are continuing to do so. It's arguable as to whether or not these changes are positive or negative, but they're happening, and they're probably making the creation process cheaper, faster and more available.