Emil Torick On Why The MP3 Was Good Enough

I just came across this short, 2-minute+ video sponsored by the Audio Engineering Society (AES) from 2000 and found it very interesting and relevant, especially in view of all the anti-Pono talk of late. Watch the video and see if you agree.

Here's my favorite quote:

"We had the MPEG-1 Layer II which seemed to be almost transparent. Ah, and then along came Layer III and it was pretty clear to most of us that it wasn't of, ah, adequate quality. But the fact is that it was good enough and today MP3 is the dominant encoding system.

We need to remind ourselves that there's something that is equally important than the absolute perfect quality and that is that the marketplace will accept something which is good enough if it has other desirable attributes."

Since we know that different people hear differently, I'm of the opinion that "good enough" is not a universal value and Mr. Torick's comments make it clear that the MP3 was not of "adequate quality" for everyone involved. But it turned out to be good enough—"MP3, good enough for people who don't really listen." (I made that up).

From a Facebook post by John Krivit written on the occasion of Mr. Torick's passing in 2010:

Emil's career in Audio Engineering began in 1958, when he joined the professional staff of CBS Laboratories. It was the beginning of a twenty-eight year association, which resulted in more than sixty technical publications and sixteen U. S. patents, a number of which were awarded for his inventions of audio signal processing devices for broadcast and recording applications. As the CBS Vice President for Audio Technology, Emil led all advanced audio and acoustics research for the CBS Radio and Television and for CBS Records. He also served as Director of R & D for the CBS Musical Instruments Division which included such companies as Steinway Pianos and Fender Guitars. Emil Torick’s most lasting accomplishment came not from his great scientific mind but from his heart. He started in 1984, and ran for 25 years, the AES foundation which has awarded thousands of dollars to almost 200 students pursuing graduate degrees in the field of audio. This work lives on as scholarship funds that can be applied for each year.

deckeda's picture

What's interesting is that "this is what they're willing to buy" is interpreted as "this is what they want" -- a classic and oft-repeated mistake, because it doesn't actually offer an alternative, let alone attempt to sell one.

Lossless downloads have never been attempted in the mass market. They've always been the exception. There's always been this overriding perspective of what's good enough for the customer is good enough for the customer, and don't bother asking us about metadata, or tech details.

Azteca X's picture

I am a huge advocate for FLAC, whether redbook or 24/88, 24/96 and occasionally 24/192. I like lossless, period.
However, I think it's silly to dismiss lossy codecs, particularly the MP3, as not being useful. Sure, many of us have connections that allow for more than 1MB/s download speed. But what about the 56k days? Early, crappy broadband? Current crappy connection speeds the world over? Spotty 3G cell towers? I think it's short-sighted to imply that if you can't listen to lossless audio you just shouldn't bother. How many people have fallen in love with music or encountered life-changing art through cheap radios and boomboxes, iPod earbuds, car stereos, etc? Here, in the US, among middle-class folks, we have a good shot at excellent audio for reasonable prices. But let's not be so short-sighted as to think everyone is in the same place.
I have a Spotify subscription and think their "higher streaming quality" option sounds pretty excellent though they sometimes don't have my preferred masterings of older albums. It's let me discover tons. And it's not FLAC - it's ogg. For my modest work desk setup, it's great, and I've learned and enjoyed lots.
Mastered for iTunes ain't bad, either.
I look forward to the continued improvement of lossy codecs, such as the new-ish Opus codec. If not for hi-fi listening, for better VoIP calls and streaming. I look forward to the day in the near-future where I can check "lossless, please" on my Spotify account, but while infrastructure and financial access to broadband is limited, many people will rely on lossy codecs - and I think they should be able to access as much as the rest of us.

Michael Lavorgna's picture
I agree with you regarding the relevant uses of lossy encoding. My main issue with the MP3 is that is has become synonymous with "download" and there are still people who suggest it is "good enough", indistinguishable from CD-quality and higher resolutions which is just not the case. While different lossy codecs do offer more transparent results, I think its important that we strive to make CD-quality equate to "download" making the distinction between lossy formats, CD-quality, and HD more obvious.
jim tavegia's picture

HOw about "space-saving" music delivery. Isn't that what MP3s really do? I just have a hard time with some music services telling me that 2448 is high rez. To me that has to start with 2496.

Remember all those "space-saving" appliances that weren't really better.

Archimago's picture

I don't see how one would judge 24/48 *not* being high-resolution.

jim tavegia's picture

I, personally, have not been able to discern an appreciable difference from redbook to 48 khz. I can hear the increased improvement going to 2496 even in my own amateur recordings. I read somewhere that most commercial music recorded on ProTools is done at 2448 which makes no sense for me given affordable HD space and with archiving in mind. It just seems to me that 2496 should be the base recording format in 2014 and that 24192 should be considered for long term masters. Great quality converters are so affordable these days. It just seems to me that going to 2448 is just one step behind in quality which the industry always seems to want to do. I would agree that it is a big leap for ITunes. If the masses will make the jump to 2448 that would be a good thing, but I think the file sizes may impede that acceptance and limit their portable library. For them it has always been about convenience. Maybe things will change.

jim tavegia's picture

Sounds more than acceptable to me not even knowing the bit rate. Certainly listenable while working at the computer. That darn convenience thing.

jim tavegia's picture

Appears to be 320kbps.

jim tavegia's picture

This will only change if a significant number of current MP3 downloader demand higher resolution and if the likes of ITunes see a decrease in revenue. For me it still had much to do with the native remastering from either a good quality master tape or from a 2496, or better, master file. I don't believe that many could pick a 2448 file from a redbook one, but that would be an interesting test. If they chose the 2448 file then it would beg the question why you didn't want many of the high quality redbook titles that have been out for decades?

I do think that Pono will help move this along, but the shape still bothers me and not very pocket friendly. I can still put my Tascam DR-2d in my shirt pocket and carry a number of DSHC cards for various music choices, Jazz, Classical, Vocalists etc . My larger cards are 32 gig, but 64 gig cards are avaiable and I would think that 128 will not be far behind. These new Sanyo Eneloop rechargable batteries for $10 for 4 keep my battery costs down. 2510 MAH . Duracells are 2100 MAH.

Michael Lavorgna's picture
...is rumored to be moving to 24-bit downloads in June. Whether this happens next month or next year, I believe its inevitable. It will be interesting to see how Apple decides to differentiate between a lossy download and a 24-bit download but it will certainly change the download landscape dramatically.
Johnny2Bad's picture

We have the established standard of mp3-ish bit rates (mp3.128 up to AAC 320) because that was the reasonable file size to fit the affordable Hard Drives (64 and 128 GB) and the reasonably available data rate for downloading files when Apple was negotiating the original legal music criteria with the labels who, let's remember, came into the negotiations screaming and kicking in the early parts of the 21st century.

If 1 Terrabyte Hard Drives could be had for under $100 in 2002, we'd be using probably Redbook bitrates for our online music ecosystem. But they weren't.

Digital "standards" are always a day late and a dollar short ... something is introduced and a decade later becomes equivalent to what it purports to replace. A sub-2 megapixel camera was also the basic affordable option at the same time iTunes first came upon the scene. 5 MP was a $4-figure option. That's just the way it goes, and for some reason, people have short memories when it comes to digital performance.

People seem to assume you could always have affordable download speeds simply because they can't remember what it was like back in the day. But I remember waiting hours, and I don't mean just two hours ... to download Linux installation disks in 2002 or so that were a fraction of what size today's typical standard download ... OS updates come on DVD's and in some cases are not available on media at all but only via download. And I have always had the fastest internet available, starting with my first "High Speed" connection in 1995, which was early in terms of what most people had available to them in North America.

Analog has always enjoyed what amounts to infinite data rates, but that's not the point.

Digital Cameras had their best sales during a time when they could not equal an analog film equivalent, and it's somewhat ironic that now when you can actually buy a camera that rivals film in resolution, the digital camera is an industry that is struggling to sell product; the not-as-good cellphone is the dominant device to create still and video images.

The whole history of digital is and always has been about "not as good" as what it replaces, often with enthusiasm by consumers. I am amazed when people seem to think Audio is some special case; it's not. "Good Enough" even when it wasn't good enough is the norm, not the exception.