Plugins for Audiophile Music Players
Betraying the analog roots of audio, this time I’m diving into plug-ins, those fun and versatile bits of code that can enhance your digital audio listening experience. The first time I heard about plug-ins was when Adobe added that feature to Photoshop version 2. In the audio world, Digidesign took what was at the time a rather nasty sounding application named Pro Tools and drove it to the number one Digital Audio Workstation on earth, in no small part thanks to their early inclusion of a plug-in SDK.
A plug-in, according to Wikipedia, is a “piece of software which enhances another software application and usually cannot be run independently.” Correct! Plug-ins are relatively small and inexpensive packages of code; relative to their host application that is. They are created through an SDK or Software Development Kit, a set of rules and tweaky tools that simplify many of the gory details of interoperating with the host application.
There are a half dozen or more plug-in formats, though for our purposes, we’ll stick with two. The VST or Virtual Studio Technology format was created by Steinberg as their in-house answer to Digi’s own proprietary plug-in formats. VST is a hellish tangle of poorly documented methods, designed to challenge even an experienced programmer. That hasn’t stopped gazillions of kids from brewing up VST “virtual instruments” designed to act as music synthesizer “voices.” Bust out a MIDI keyboard, instantiate a VST instrument, and start laying down some tasty virtual djembé rhythms.
Here, we aren’t so much concerned about virtual instruments as we are with VST “effects,” that umbrella label for all things DSP. Digital Signal Processing effects encompass EQ, dynamics, pitch shift/SRC (sample rate conversion), reverb synthesis and delay, and more esoteric functions like recovering HDCD data embedded in a track.
There are plenty of VST effects out there, some of them free for the downloading. Since the VST format is cross-platform, you will need to make sure to download the version that matches your operating system. The one fly in the ointment regarding VST on Windows is that JRiver, the audiophile player for Windows, doesn’t support the format. There are plenty of VST hosts on Windows, but the big kahuna isn’t one of them. foobar2000 doesn’t support VST natively either, but at least you can get a native VST wrapper, a sort of software adapter that bolts onto foobar. A similar situation holds true for Mac OS: if you’re looking for an audiophile player on the Mac that supports VST, there isn’t one. Fidelia used to but that feature was dropped in version 1.2.1.
There is another plug-in format, Audio Units, that’s particularly suited to the audiophile host of choice… In 2007, Apple Computer released Audio Units (AU), a plug-in format for their four year old release of Unix. Back in those days, Apple still cared about their developers, and their SDK was well documented and relatively easy to use because it leveraged Core Audio, the sophisticated abstraction layer for all things audio in the operating system.
So much for the history… what’s really of interest to us audio phreaks is that not one but two Mac audiophile players support the AU format; Fidelia and Pure Music. (*) Fidelia has three slots for instantiating plug-ins, while Pure Music, with its More Has To Be Better approach, offers 14 plug-in slots. In general, Pure Music does a much better job of plug-in support than Fidelia, with all installed plug-ins showing up in the menu.
(*) As of June 3rd 2013, just prior to publishing this article, Audirvana+ 1.5 supports AU plugins also – Ed.
Or, if you’d like a bit of a challenge, download the cross-platform Audiomulch… the trial version will keep you busy for a while and it hosts both VST and AU plugs. Do a Google search for “free au plugins” or “free vst plugins,” and a treasure chest of riches will present itself.
Wait a minute! What good are plug-ins? For listening-only, not production, I value shuffling, measurement and EQ. Let’s start with shufflers. Strictly for headphone heads, shufflers are signal processors that simulate the interaural crosstalk one experiences when listening over loudspeakers. As you can imagine, a single speaker produces sound that both ears hear. The ear closest to the speaker hears a direct version, while the other ear hears a delayed and EQ’d version, due to the acoustical effects of your head, torso and pinnœ. A shuffler circuit, either in hardware or software, provides a simulacrum of loudspeaker playback when listening over headphones. (See the PDF of Stereo Shuffling: New Approach – Old Technique.)
I admit to a decided lack of enthusiasm for most shufflers, the free ones anyway – but since they’re free you should give them a listen if you have a host application that will run them. One favorite software shuffler I own is an accessory to Fidelia, a proprietary plug-in if you will, called FHX. It does a darn good job and is reasonably priced. The best software shuffler I know of, at twice the price, is the Ircam HEar by Flux. It is remarkably good, providing a good sense of the width and depth of soundstage you’d expect, but also supporting high sample rates and boasting what is basically a one knob interface:
On to measurement, a category where plug-ins really shine… You see, a measurement plug-in, by definition, doesn’t modify your bit stream. So, you can luxuriate in your cosseted bit perfection and still eat your eye candy, so to speak. Measurement plug-ins can visualize many aspects of your audio. For example, Medla’s MAnalyzer shows the power spectrum of your music:
Since plug-ins are placed in the signal path in series, you can do things like instantiate a shuffler first and a goniometer after that and actually see the difference that the shuffler is making in your stereo image.
Sidebar: The basics of EQ
Let’s look at some terms and concepts related to equalization. Here is a screenshot of MIO Strip, a “channel strip” processor from Metric Halo that includes dynamic range controls and equalization:
A kitchen sink of processing, as it were, it’s meant to be applied to an individual microphone or instrument. What we’re interested in is the highlighted stuff along the bottom. At left is a 6-band parametric equalizer. Parametric EQ are so called because you have control over all the parameters, unlike a graphic or fixed frequency EQ where you’re confined to predefined “center” frequencies. At lower right is a graph of the selected EQ’s response. Box #1 contains the labels for the columns of controls. From left to right, these are positive or negative gain in decibels, resonant frequency in Hertz, and bandwidth in octaves. The gain is easy to grasp: more or less amplitude at certain frequencies. The resonant frequency is the frequency at which the settings produce the maximum effect. Finally the bandwidth is (roughly speaking) the range of frequencies over which the EQ has an effect.
Box #2 highlights one filter section, a 1st order parametric filter with a gain of +15 dB, a center frequency of 1 kHz, and a bandwidth spanning 2 octaves.
The last highlighted area in the MIO Strip screenshot above is a results graph, with frequency on the horizontal axis, and ± gain on the vertical axis. The arrow points to the center frequency, which is 15 decibels above unity gain. Notice how the audible effects range from 100 Hz all the way to past 8 kHz.
Some sample EQ types are: high pass, passing through the high frequencies and cutting lower frequencies; high shelf, a super handy type that gives you constant boost or cut over a range of frequencies; bandpass, which chops off frequencies symmetrically above and below the resonant or center frequency.
Filters can of course be combined in arbitrary ways. One particular interesting combination arises when you combine a high and low shelf, both with a fixed amount of gain – you have what your Dad’s hifi called a “loudness” filter:
One measurement plug-in that is near and dear to my heart is the Pleasurize Music Foundation’s (PMF) TT Dynamic Range meter. The TT DR meter generates an integer value representing an approximation of the dynamic range of your material. It’s a back door way of addressing the “loudness wars” that have plagued popular music. By objectively seeing the dynamic range, or lack thereof, of your music, you may get your hackles up and do something about it. Visit the PMF (Pleasurize Music Foundation) website to find out more…
EQ, or equalization, the long-hand name that describes its original use, is a blanket term covering the quotidian tone controls of yore. EQ provides frequency-dependent gain, and is a fundamental part of every form of electronic entertainment and communications, from open reel analog decks and the AAC tracks on your portable music player, to telephone calls, hand wired tweakophile crossovers and WiFi connections. Let me rattle your cage a wee bit; There is nothing wrong with EQ! Without EQ, modern life would literally be impossible.
All the material you listen to – and I do mean all – has had EQ applied to it at least twice in its life, during recording and in preparation for distribution, and usually many times more. Both analog and digital recording relies on EQ. Without it, the process simply would not work. In addition, engineers often reach for EQ to either correct a perceived flaw or modify what he or she is hearing so, why can’t you? After all, our little hobby is about musical enjoyment, not adherence to occult and unspoken laws…Humm, now that I think about it…
If there is one killer app for plug-ins, it has to be EQ. Yeah, I know…you personally wouldn’t touch the stuff. I’m not sure how EQ or equalization has come to be a dirty word in audiophile circles. There are some truly great, quality hardware choices out there from GML, NBS and others. My pockets, however, aren’t that deep. I can afford a hundred bucks or two for a stellar software implementation. I like low coloration in my gear so I start with something close to the master recording. That said, I have no qualms about then processing the audio if I feel it will subjectively improve the content, my mood or playback system. A good tool for said improvement is EQ or, should I say, good EQ.
As with everything else in audio, there are simplistic implementations that sound rather nasty, and there are more thoughtful designs that sound simply loverly, to quote Lerner and Loewe. Plug-ins, as with any piece of audio gear, are only as good as the wisdom and ability of their designer. Steven Massey, the principal propeller head of a well respected plug-in boutique, had this to say in his blog: “…designing plugins requires a understanding of electrical engineering. It is, in fact, electrical engineering in virtualized form. The program source code is simply the schematic. And, the software compiler is the soldering iron that gets the circuit board (executable code) built.” A textbook implementation of a DSP function gets you the basic functionality but, it’s the subtleties that make for truly great processing.
Here’s a good EQ example: take a pair of in-ear phones, and EQ out the canal resonance that results from the trapped air column between ear drum and transducer. With a bit of experimentation, you can usually improve the subjective “sound.” I routinely listen to my Etymotics with a personalized combination of Siegfried Linkwitz’s recommended resonance reduction and Laurie Fincham slight LF shelving boost. The result: taking a great sounding pair of cans, and making them much better, all with a little EQ.
Some free, high quality EQ plug-ins are SPL Free Ranger by Sound Performance Lab and Elysia’s niveau filter. Some not-free EQ plugins I like are Flux’s Epure and PSPaudioware’s Neon HR, both spendy (for a plug-in) but beautiful. These puppies require licensing hardware, a USB dongle. For less money, NUGEN’s SEQ1 is a solid choice. Also, it employs file-based licensing so, no dongle needed.
When evaluating an EQ, the quickest test is to dial in a 1 kHz center frequency, then see how much boost is required for you to notice it. The less boost it takes, the nastier the filter. High quality EQ will typically require twice the amount of gain change to be as audible as an el-cheapo version. Use the free Blue Cat EQ plug-in as an edgy example, and compare the same settings to Melda’s linear phase EQ. I’m sure you’ll hear what I mean… Another factoid of interest is that better quality EQ plug-ins “honor” the sample rate of the host. So, if you’re listening to a 88.2 file, the EQ should operate all the way past 40 kHz. Try adding some high frequency shelving boost at 22 to 30 kHz and listen carefully… This is where “air” lives, and a small amount brings some delightful openness to dark systems.
I hope I’ve whetted your DSP appetite, there really is a whole world of listening enjoyment out there to be explored. Plug-ins are fun, often free, and demo versions of everything under the sun are available for downloading. Before you drift off to sleep tonight, remember to close the closet door, and keep reminding yourself “… EQ is nothing to be afraid of…”
- Don’t Crack Freeware Plugins
- ApulSoft Ports of Free Plugins
- KVR Audio
- Dubstep Productions: Ultimate Free Plugin List