top of page
  • Michael Fay

A Case for Lossy Music Formats

Updated: May 9, 2020


Downloaded Lossy Files are Viable for Loudspeaker & System Evaluation

Presuming you’re in the audio business on some level, and are an “active” listener, you’ve probably started a collection of reference recordings you think are exceptional. You’ve also probably found it helpful, if not necessary, to make this collection highly portable so it’s available, on a moment’s notice, where and when needed.

As a pro, or an aspiring one, you just never know when you may be presented an opportunity to evaluate a new or existing loudspeaker or system. Think trade shows, rep demos, client system evaluation and troubleshooting, factory tour demos, and shoot-outs. Having a highly portable, widely compatible file storage system makes it easy to say, “Yes, I have a few tracks I’d like to listen to.”

Background

When I first started thinking carefully about this topic, CDs were at their peak as the ubiquitous digital audio storage and distribution format. While they are still a viable option, cell phones, tablets, and computers are now the storage devices of choice. Each of these is just as convenient, if not more so, than carrying around one or more discs in a wallet, and obviously the storage capacity is much greater. Further, when we store our music as files on a small device there’s no need to carry or ask for a CD deck.

This brings us to the subject of file formats that are widely available and compatible with your favorite device. MP3, FLAC, WMA, WAV, ACC, OGG and AIFF, come immediately to mind. Some of these are lossless formats, others are lossy. But, when you only want to buy one track at a time, not the whole album, a lossy download is often your only option. When that’s the situation I think there’s one clear choice, MP3. Why, because of availability and compatibility.

According to the MakeUseOf.com website, https://www.makeuseof.com/tag/audio-file-format-right-needs/ “Nearly every digital device in the world with audio playback can read and play MP3 files, whether we’re talking about PCs, Macs, Androids, iPhones, Smart TVs, or whatever else. When you need universal, MP3 will never let you down.”

So, is promoting MP3 and other lossy formats to the pro audio community blasphemy? I think not!

The Never Ending Search

Over the years, I’ve spent a lot of time looking for really good examples of music, production and recording technique. If you’ve seen my article, The Mental Side of Mixing, [Insert LSI reference link here] you know I’ve searched through tens of thousands of recordings (via Rhapsody, Zune, Spotify, etc.) to find my collection, which currently numbers about 45 tracks of varying musical styles. It only takes me a few seconds to tell if a track has something sonically-special to consider.

Why so many favorites you might ask? That’s an excellent question. I strongly believe the music you use must have something in common with a client’s requirements for sound reinforcement. If your reference tracks are tools, it’s about having the right tools for the job.

There’s an important psychological/emotional bridge that needs to be crossed when selecting songs to use for showcasing a new system to a client. They must be something the client can easily relate to. Can you imagine using an AC/DC or Metallica track to tune and showcase a new Presbyterian church system? No, I don’t think so. How insensitive. I wouldn’t use an Ed Sheeran ballad to showcase a new dance club system either.

Here is a short list of my favorite tracks for system testing, tuning and voicing. There are several different musical styles, but all have something very nice to offer. My full collection has about a 60/40 mixture of high bit rate MP3 (60%) and WAV files that I use without reservation.

Reference Tracks

Reference Track Sampler

Night Winds – Randy Dorman and Rick Harper

Calling Elvis - Mark Knopfler and Dire Straits Some People - LeAnn Rimes

Can't Find the Words - Kim Richey

Pop Culture – Thomas Dolby I Will Remember - Toto Poquito Mas – Infected Mushroom

Capricorn - 30 Seconds To Mars

You can find the full list of my favorite tracks here: www.gracenoteds.com/more, look for the MF SysTune link.

I keep most of these tracks handy on various storage media and on one small, portable device. I have a tiny thumb drive on my key ring and a number of compilation CDs, but my favorite is a FiiO X1, high-resolution, lossless music player. This device is compatible with the lossless and lossy file types listed below, and unlike phones, tablets and most computers that use headphone outputs, it has a dedicated line-level output with excellent specs.

There are one or more specific things I'm keyed into on every track. Examples are often kick, snare, bass and acoustic guitar. Sometimes I’ll use a specific track just to listen to the high-hat, or another for the harmonics of an orchestral triangle.

If I only get one track to use to evaluate the voicing of a single speaker or system, and I’m not looking for brute-force power handling, it’s a track called Night Winds on Taylor Guitar’s Sounds of Wood & Steel III sampler. It’s an amazing recording, and a must-have track for any reference collection.

Exceptional vocal recordings are hard to find. The LeAnn Rimes vocal is exceptional. Want to stress test your system? Use the 30 Seconds to Mars or Infected Mushroom tracks. Need to focus on the subwoofer section, bring up the Thomas Dolby track. The Bill Payne recording has the best jazz acoustic piano recording I've been able to find, plus a lot of percussive ear candy. The Kim Richey track will rip your head off if your system is not right in the upper mid-range. And last, but not least, the Mark Knopfler track is my favorite overall track for mainstream rock and blues.

Too Often MP3 Gets Blamed for Poor Sound?

As for the debate over the sonic quality of the various lossless and lossy file formats; I contend that with most 21st century mass-market music, it’s the original production that often sounds bad, regardless of the final file format or compression scheme used. In other words, don’t automatically blame the lossy format if a track or whole album sounds like it was recorded and mixed in a metal shed full of Styrofoam and gravel, and has about as much warmth, depth, space, clarity and punch as cold, leftover turnips.

If poor sonic quality is what the artist demands, fine, but don’t automatically blame the compression algorithm if you don’t like the sound of the recording.

Acceptable Bit Rates

A few years ago, while working at Sound Image, I performed some blind listening tests with a few very talented and experienced professional mix engineers, and one very well known loudspeaker designer. The conclusion reached was that no one could consistently pick out a 256 kbps bit rate MP3 track vs. a ripped WAV track (or even the original disk).

Experimental data shows that for average listeners, 128 kbps is the point where MP3 becomes consistently worse-sounding than higher bitrates and uncompressed audio. But, when higher bitrates (196 kbps and above) are compared to uncompressed files, the differences are less and less audible.

The distinctions between those are so small that they become statistically insignificant, and in many double blind studies (here's just one), uncompressed original files rank somewhere below high bitrate compressed formats.

My takeaway is: if you want or need to use lossy files, do so without any embarrassment.

Take Your Own Blind Listening Test

To start, find a track that you really like and that you think has exceptional audio quality. Buy the original disk, then also download the same track as a WAV or other lossless file, and at a bit rate of 256 kbps or higher from an on-line music site. Now days I usually use Amazon for individual lossy tracks, or one of the HD download sites listed below, if they have the track(s) I want.

You might also find it interesting to download the same track at 192 kbps, 128 kbps or even lower, so you can find the point at which the sound becomes unacceptably degraded. Next, rip the original track as a WAV or other lossless file so you can burn everything back to a new CD. Scramble the sequence in some random fashion when building the new test disc.

Now you are ready for playback. Use the best listening equipment you can get access to and ask a friend to play the tracks in random order. Ask yourself if you can hear any difference between the various downloaded bit rates, the ripped lossless file, and the original disk? Also, try asking some friends, with good ears, to pick out which is which. I think you will be surprised. I was.

For me, the magic number is 256k. Nothing more is mandatory, nothing less will do if I have a choice. Another quick way to test your ability to recognize different bit rates can be found here:

https://www.npr.org/sections/therecord/2015/06/02/411473508/how-well-can-you-hear-audio-quality

Lossless and Lossy Download Sites

Here are a few sites that offer both lossless and lossy music downloads:

HD Tracks - http://www.hdtracks.com – Offers multiple lossless formats for each artist

7 Digital - https://www.7digital.com – Offers lossless and 320 kbps lossy downloads

Bandcamp - https://bandcamp.com/ - Offers lossless and lossy downloads

Here are three of the most popular lossy music download sites:

Amazon offers 256 kbps MP3 for the most part

iTunes offers 256 kbps ACC for the most part

Spotify Premium offers, with a paid subscription, up to 320 kbps Ogg Vorbis

Summary

So, unless you’re downloading music to play through your audiophile stereo system, in your acoustically-refined listening room, and you’ve spent thousands if not tens of thousands of dollars on loudspeakers, tube preamps, cables, Class A monoblock amplifiers, etc., there’s no reason to avoid using a high bit rate, lossy files for system evaluation and tuning.

© Copyright Michael Fay 2013 - 2018


56 views

Recent Posts

See All

Going Way Beyond

Complex automixing using a virtual digital mixing console I was so tempted to give this piece a geeky marketing title or sub-header suggesting VR (virtual reality), AR (augmented reality), or AI (arti

Acoustic Shadowing

Unpacking Wave Numbers, and Introducing the Wave Ratio Thesis Ok I get it. This topic may be too esoteric for many, but if you’re truly interested in learning all you can about sound propagation and a

bottom of page