Compression vs Art (Radio Programme)

[Audio_ELF]

This programme was on Radio this morning, not yet had a chance to listen but thought it might be of interest to some here.

 

Not sure if this link will work for those outside the UK BBC Radio 4 - Compression versus Art If it works; the programme should be available until around 6th February 2016.

 

Trevor Cox asks whether compression can detract from our enjoyment of recorded music - does it matter that what we hear may not be the same as what the musicians heard in the studio? How important is high quality reproduction? He looks at attempts to make music recordings sound louder and louder (the so-called Loudness War) and asks whether anything is lost in the process. And he considers whether making audio file sizes smaller, so that they take up less space on portable devices, means that some of the musical detail is lost. He talks to record producer Steve Levine (who produced Culture Club among many others) mastering engineer Ian Shepherd, the musician Steven Wilson, members of the BBC Philharmonic, and Dr Bruno Fazenda, Senior Lecturer in Audio Technology.

 

If anyone wants to listen and the link doesn't work ... let me know will see what I can do.

[orgel]

Works fine on my iPad, which is currently located in the US.

 

Seems interesting, I'll try to listen in the car on my way to work.

 

--David

[Audio_ELF]

Works fine on my iPad, which is currently located in the US.

 

Seems interesting, I'll try to listen in the car on my way to work.

 

--David

Thanks

[orgel]

Good show! I don't think there's much in there that'll shock the CA crowd, but it's well done nonetheless and a good overview of the subject.

 

The one thing that set me back on my heels a bit was that there's research showing that some people (wasn't clear what percentage) prefer a soupçon of loudness-wars-style compression when compared with an uncompressed original. The researcher's speculation about the reason for that is on the mark, I think.

 

--David

[crenca]

Good show! I don't think there's much in there that'll shock the CA crowd, but it's well done nonetheless and a good overview of the subject.

 

The one thing that set me back on my heels a bit was that there's research showing that some people (wasn't clear what percentage) prefer a soupçon of loudness-wars-style compression when compared with an uncompressed original. The researcher's speculation about the reason for that is on the mark, I think.

 

--David

 

The reason he gave was simply that since most of the music folks have ever listened to is compressed they are are simply used to it, so they expect it. I want to say that there might be more to it, but that is only an "intuition".

 

This was an excellent program and really covered this subject quite well.

 

Edit: I will add that one interviewee (third one I believe) asserts blandly and without any qualification that "people" can not tell the difference between 320 and 16/44 "according to studies". This of course is problematic and not even true without serious qualifications...

[gmgraves]

The reason he gave was simply that since most of the music folks have ever listened to is compressed they are are simply used to it, so they expect it. I want to say that there might be more to it, but that is only an "intuition".

 

SQ wise, FM radio hasn't been worth listening to for about 50 years! And not just in the USA/Canada either It's the same in Europe and the UK too. Here's the reason:

 

In the 1950's, with the advent of the LP, magnetic tape, and high-fidelity condenser microphones from post-war Germany and Austria, FM could finally reach the performance potential that Major Edwin H. Armstrong had designed into the format when he invented it in the late 1930's. Before that, the program material had consisted of noisy, frequency response limited 78's, lacquer transcriptions, and ribbon microphones; mostly 44 BX and 77 DX models from RCA with frequency responses that fell off like a rock at just a hair over 10 KHz (remember, these mikes were designed for AM radio with it's 7KHz top - a few so-called "clear channel" AM stations went to about 10KHz, but they were rare). Of the two, the 77 DX was the better design, for although it peaked at 10 KHz, it was only 6 dB down at 20 KHz - still not exactly hi-fi!

 

By 1950, with everything in place, FM was sounding glorious. Even in high-population markets such as NYC, Boston, Washington DC, Chicago, etc. FM stations were few - a real niche market. The FCC spaced FM stations fairly far apart on the dial and didn't really care too much is they over modulated regularly. They weren't legally supposed to do it (the rules gave them +/- 200KHz of bandwidth) but there was nobody within MegaHertz of any assigned FM frequency, so, in reality, nobody cared if FM stations broke the rules. The result was uncompressed and unlimited full frequency response (20-20 KHz) low distortion radio (in metro areas, anyway). This state of FM development lasted until Jan, 1962 here in the USA. Since the introduction of the stereo LP in 1957/58, people had been clamoring for stereo radio. A number of competing formats were considered (Crosby, Halstead, EMI, Zenith, and General Electric among them) and were quickly whittled down to two. Both used the FCC mandated Subsidiary Communications Authorization (SCA) subcarrier rules to carry the second audio channel, but did so in a very different way. The first was the Crosby system. It was the simplest, and audio-wise, the best. It used the main FM carrier for the right channel and the subcarrier for the left. Both channels were full frequency response (20-20K) and full quieting. The Crosby system had one very serious drawback, however - and that should be apparent to you gentle readers already. If you had a mono FM receiver, you could only listen to one of the two channels! Zenith and GE had come-up with systems that were so similar (they differed only in small details) that the FCC asked them combine their efforts and thus the GE-Zenith system was born. It used a variation of the matrix stereo LP system proposed in the middle '50's by Paul Weathers of turntable and pickup fame. What it brought to the party was that it was mono/stereo compatible. The main channel carried the summed left + right channels, and the subcarrier used a difference signal derived mathematically from left minus right! The stereo demodulator would then provide the math to derive separate left and right channels. It had tow characteristics that pretty much guaranteed it's acceptance as the stereo FM standard, first in the USA, and then all over the world. This characteristic was, of course, that the GE-Zenith system was monaural compatible! It iced the cake by also not using up so much station bandwidth that a station had to decide between stereo-casting and offering a subscription service (Muzak) which, of course, brings in extra income to the station. Ergo, with the GE-Zenith system, a station could offer both stereo and SCA!. Unfortunately, as with everything, there is no free ride and everything comes with trade-offs. The GE-Zenith system had the following problems: reduced audio bandwidth, (50-15 KHz as opposed to 20-20 KHz of mono FM or the Crosby system), poorer signal-to-noise specs (especially in suburban areas which were further from the transmitter than was the downtown area), and increased susceptibility to multi-path distortion. However, the rush to stereo brought about another innovation which further reduced FM sound quality, an innovation which was to, eventually completely ruin FM as a source of high-fidelity music.

 

When SCA was made legal by the FCC in 1955, it became necessary for those participating stations to regulate their modulation. It wouldn't do for the main FM carrier, to spill over into the subcarrier and mess-up the revenue stream that came from the paid background music service sold to stores and offices. So, for the first time, modulation limiting was needed. The early limiters (mostly made by Fairchild in the 1950's and early sixties) were simply terrible. If a radio station exceeded 100% modulation, the limiter simply clipped the over modulation; i.e. turned it into a square wave (just like a clipping amplifier) with nasty-sounding results. Luckily, only a few stations in each market broadcasted SCA, and they were mostly MOR (Middle-Of-the-Road) stations and could afford to cut-back on their overall modulation level, because audio quality wasn't their main concern. With the advent of GE-Zenith stereo broadcasting, it suddenly became necessary for every stereo station to employ hard limiting to avoid distorting the subcarrier. Then, by about 1965, FM caught-on with the college students and within months, new stations starting popping-up all over the dial, adding types of music that hadn't been on FM before: rock-n'-roll, folk, country, international, etc. Suddenly the FM dial, once a niche market, scarcely populated, became crowded, and the FCC insisted that it's modulation rules be strictly adhered to. Suddenly everybody had to back-off on their over modulation and listeners found that they now had to increase the volume on their radios to get the same volume as before. Broadcasters started to clamor for equipment that would restore their former program levels without allowing them to over modulate. To the "rescue", came designer Ben Bauer of CBS labs with his Audimax compressor, and his Volumax limiter. Together, these two work horses of FM broadcasting promised complete control of radio station modulation while maintaining a maximum level of broadcast volume. Together, these two devices became both the answer of the needs of broadcasters, and the final nail in the coffin of FM quality. Soon, competing FM stations were turning up the compression on their signal more and more to be the loudest station on the dial, and the "loudness wars" began. In the 1970's an Engineer named Bob Orban designed a single unit to do both limiting and compression called the Orban Optimod. Bob (who was a friend of mine) tried to take audio quality into account and certainly his device was much better sounding than the CBS units (I took part in a good number of his listening tests at the time and provided feedback to him), especially when used judiciously. But even he couldn't guarantee how a radio station would use the thing, and when abused, the Optimod could sound every bit as nasty as the earlier designs. I've seen a number of articles condemning the now universally used Optimod technology, but I can tell you that when used correctly, the Optimod (now in it's umpteenth iteration) is a very benign sounding device. But FM broadcasters just want louder and louder. Anyway, by the time the Optimod was ready for production, it was too late. The promise of FM as a quality music delivery system was already irreparably broken. I'm sure Edwin Armstrong was spinning in his grave by 1970!