Hi-fi and high-end: what is the difference and what is just marketing

In short: Hi-fi once meant that a device met an official standard; today it is just a word with no obligation behind it. High-end was never a standard. It is a way a group of enthusiasts looks at music, and in shops it is mainly a price category. The research is clear: the speakers and the room have the biggest effect on sound. With well-designed electronics and cables, the differences are small or zero when you compare honestly. Expensive gear can be beautiful, well made and a joy to own. But that it will sound better is not guaranteed.

Hi-fi: a word that once guaranteed something

Hi-fi is short for high fidelity. It means fidelity to the original sound: what comes out of the speaker should be as close as possible to what was recorded.

In 1966 Germany introduced the standard DIN 45500, which set what a device had to meet to be allowed to call itself hi-fi. It was used across much of Europe, and old hands still remember it. For example:

Device	What it had to meet (simplified)
Amplifier	range from 40 to 16,000 Hz, distortion at most 1 %, at least 2 × 6 W
Loudspeaker	range from 50 to 12,500 Hz, distortion at most 3 % in the midrange and 1 % in the treble, 10 W power
Turntable	speed fluctuation at most 0.2 %, cartridge range 40 to 12,500 Hz
Tape recorder	range 40 to 12,500 Hz, signal-to-noise ratio at least 50 dB

For its time this was useful. A buyer knew they would get at least decent quality. Today almost any cheap device beats these values. The human ear hears roughly from 20 to 20,000 Hz, and a speaker that only reaches 12,500 Hz would impress nobody today.

In 1996 the standard was replaced by the European EN 61305. But that one only says how devices should be measured, not what results they must reach. Since then the word hi-fi is not protected. Anyone can print it on a box.

Today's stickers are similar. A maker buys the Hi-Res Audio logo from a Japanese association based on its own data. It only says that the device plays high-resolution files and reaches 40 kHz, not whether it sounds good.

High-end: a movement that started with listening

High-end is not a standard or a technical term. It arose in America in the 1960s and 70s as a reaction to how equipment was judged back then.

• J. Gordon Holt founded Stereophile magazine in 1962. He wanted equipment judged by how it sounds, not just by spec sheets. He coined many words still used to describe sound today, such as "warm" or "bright" sound.
• Harry Pearson founded The Absolute Sound in 1973. He made live music the yardstick: the sound of real acoustic instruments in a real space. He judged equipment by how close it came to that, and he did not trust measurements much.

Why it made sense back then: in the 70s, amplifiers were sold with excellent numbers in one measurement that still sounded bad, because the wrong thing was being measured. Listening revealed problems that the tables hid. This is high-end's real achievement.

Over time, though, "listen, don't just measure" turned into "measurements mean nothing". Part of the high-end press still rejects blind tests, that is, comparing without the listener knowing what they are hearing. Interestingly, Holt himself wrote in 2007 that high-end lost its credibility in the 80s precisely because it rejected blind tests.

In shops today, high-end mostly means price. An amplifier for €300 is "hi-fi", one for €5,000 is "high-end". The High End show, held for years in Munich, is the world's largest exhibition of such equipment.

How to find out what you can really hear

Imagine a wine tasting. When you see the label of an expensive wine, it tastes better to you. Not because you are lying, but because your brain mixes what your tongue senses with what you know. Sound works the same way. That is why science compares blind: the listener does not know which device is playing.

Equal volume matters too. If one system plays even half a decibel louder, most people call it better, more detailed and livelier, even though they don't notice the difference as loudness. Many "audible differences" in shops are just this effect.

What research has shown

• Toole and Olive (1994): 40 people rated four speakers, once blind and once with the speakers in view. When they could see them, they rated the big, expensive ones higher. Blind, a small cheap speaker did slightly better than a $3,600 rival. Experienced listeners were no more immune to what their eyes told them than anyone else.
• Olive (2003): 268 people, including salespeople, students and reviewers, rated speakers blind. Almost all of them liked the same speakers. Trained listeners were just more precise and faster. So people have similar taste when price and brand don't get in the way.
• Olive (2004): speaker measurements predict fairly reliably how much people will like them. Across 70 speakers, the prediction matched the actual ratings closely (correlation 0.86). What you hear most is a flat response, even dispersion to the sides and good bass. Floyd Toole added that price is not a reliable indicator of sound quality.
• Meyer and Moran (2007): they inserted a conversion to CD quality and back into expensive systems. In 554 trials, listeners spotted it 49.8 % of the time, which is chance. A later review of many studies (Reiss, 2016) found a small difference for high resolution: trained listeners got it right about 60 % of the time. Not zero, but not "night and day" either.
• Cables (Greenhill, 1983): a very thin cable with 1.8 Ω resistance was audible. Ordinary two-core cable with a cross-section of about 1.3 mm² could not be told apart from an expensive brand-name cable in a blind test.
• Amplifiers: in comparisons with matched volume and no overload, people usually cannot tell good amplifiers apart. Richard Clark's well-known challenge, with a $10,000 prize for telling two amplifiers apart blind, by his account never ended in a win.

Where the differences are big and where they are small

Part of the system	Typical differences between products	Audible?
Room	peaks and dips of 10 to 20 dB in the bass	always, very clearly
Speakers	response deviations of several dB, different dispersion	yes, clearly
Speaker placement	bass changes by several dB	yes
Tube amplifier with high output impedance	response changes by about 1 dB depending on the speaker	often yes
Good transistor amplifier, DAC	fractions of a decibel, distortion below 0.01 %	at equal volume, usually not
Speaker cable of adequate thickness	hundredths of a decibel	no
Power cable, cable "direction"	nothing measurable	no

This gives you a simple rule. Money and time have the most effect where the differences are biggest: in the speakers, their placement and the room acoustics. You can read more about the room in the article The room plays along.

Price and quality: where the link breaks

Up to a certain price, more expensive is better. You won't get a good speaker for €50. But then come diminishing returns: each extra euro brings less. With electronics this goes to the extreme where an expensive device measures worse than a cheap one.

• According to Audio Science Review measurements, a $6,000 converter (DAC) had distortion and noise at roughly −75 dB. $200 converters reach −110 to −117 dB, so they are a hundred times cleaner. Even −75 dB may be inaudible in music, but here price does not mean quality.
• A $15,000 tube amplifier delivered only 1.8 W at 1 % distortion in Stereophile's measurements. The reviewer liked it anyway.
• The other way round: some expensive devices measure excellently, such as the Benchmark AHB2 amplifier or Genelec studio monitors. So expensive doesn't have to be bad. Price just guarantees nothing.

Is high-end just marketing?

Not all of it. It makes sense to sort it into three groups.

What is real

• Good speakers. A flat response, even dispersion, low distortion at high volume and deep, clean bass cost money and you can hear them.
• Matching to the room, acoustic treatment and correct placement.
• Craftsmanship, design and durability. Solid veneer, proper connectors, service years later, the joy of a beautiful object. That is real value, just not in the sound.

What is taste

• Tubes and turntables. They sound different because they measure differently: they add subtle distortion and change the response. Many people like that. It is legitimate, it just isn't higher fidelity but a pleasant colouration.

What has no support in the evidence

• Cables for hundreds or thousands of euros, when a cheaper cable has a sufficient cross-section.
• Cable directionality, the claim that a cable sounds better in one direction. Manufacturers provide no measurements for it.
• Audiophile power cables. Measurements found no difference even compared with an ordinary extension cord.
• "Burn-in" of cables and electronics. With drivers, the suspension really does loosen a little in the first hours; with cables, nobody has proven it.

In 2007, when a maker of $7,250 cables agreed to a one-million-dollar blind test, it withdrew before the test, so there was never a result.

How not to get fooled

1. Compare at equal volume. Without that, every comparison is worthless.
2. Compare blind. A friend who switches and doesn't tell you what is playing is enough. Or use the blind ABX test on Loudsmith, which matches the volume and scores the result for you. If you spot the difference at least 9 times out of 10, it is real.
3. Ask for measurements. For speakers, mainly the on-axis and off-axis response. If the maker has none, that is an answer too.
4. Put your money where the differences are big: speakers, room, placement. Buy the rest decent and sensible.
5. Buy for joy too, if you want. Just know what you are paying for.

What this means for building speakers

For the home builder this is good news. The speaker has the biggest effect on sound, and that is exactly what you can design and build yourself. Excellent drivers cost a fraction of the price of a finished high-end speaker; most of the shop price is the cabinet, labour, distribution and brand. With a good design, an honest crossover and measuring, you get a speaker that holds its own in a blind test even against far more expensive ones.

If you want to go further, an active crossover with DSP removes most of the compromises of a passive crossover. See the article Hypex FusionAmp.

Small glossary

• Frequency range: from the lowest to the highest notes a device plays. The ear hears about 20 to 20,000 Hz.
• Decibel (dB): a unit of loudness. 1 dB is the smallest difference a person notices; 10 dB sounds about twice as loud.
• Response: a graph showing whether a device plays all notes equally loud. A flat line is faithful.
• Distortion: sound that the device adds and that wasn't in the recording. Given in % or in dB.
• Blind test: a comparison where the listener does not know what they are hearing.
• Diminishing returns: each extra euro brings less improvement than the one before.

Sources: DIN 45500 standard (via hifi-wiki.de and German Wikipedia), F. Toole and S. Olive (AES 1994, JAES 2003, AES 2004), F. Toole (Acoustics Today 2013), E. B. Meyer and D. R. Moran (JAES 2007), J. Reiss (JAES 2016), Stereo Review 1983, measurements by Audio Science Review and Stereophile.