The first thing that you should keep in mind before you read the following discussion on how the frequency response reflects on what a product actually sounds like is that listening for yourself is worth a thousand words. There is no better way of getting a feel for how the frequency response translates to sound than comparing what you hear to a graph side-by-side.

So we strongly recommend that you compare your own products to the graphs provided in our reviews if it has been reviewed - if not, we would recommend that you check out a product that has been reviewed the next time you visit your local dealer.

Once you have compared three or four products to their corresponding graphs, you should develop a better feel for a ball park estimate of what a product sounds like just off its frequency response - again, there is no substitute for listening for yourself.

Important note

The ideas outlined in this article are about trends in how the frequency response generally translates to sound - they are in no way foolproof or without exceptions.

The wisdom that there is no substitute for listening for yourself also applies when you're choosing to purchase a product, but the graphs could still help if your local retailer isn't willing to open a fresh pair of Grado PS1000s for a test drive. The areas of evaluation here will mirror those that are discussed most frequently in our reviews.

This area pertains to how much presence the bass and treble have relative to the midrange (around 1kHz). All frequency graphs on Golden Ears have already been adjusted to reflect the perceived sound, so a raised bass relative to the midrange is indicative of significant bass presence, and similarly a raised treble reflects an abundance of treble.

The personal preferences on what is 'neutral' (or more frequently, 'desirable') are often different from a completely flat line, so you should have your own personal standard in mind as you look at these graphs.

Curve for calibrating the Bruel & Kjaer Type 4128 HATS (Head and Torso Simulator)

Loudspeaker Room-Equalized Response
Five Recommendations 

Relative responses from five different recording studios

(Courtesy of Etymotic Research)

The calibration for measurements made by Golden Ears is based on data provided by B&K, the manufacturer of the measuring equipment used, and data on the Room EQ released by Etymotic Research - these data are based on the average sound perceived around the world and are well-received audio industry standards.

Of course, sound doesn't have colors (unless you're a synesthete, that is) - but many people rely on visual descriptions when they are trying to convey what they hear, especially when they have to make metaphors to someone unfamiliar with audio. This is usually done in terms that are also applicable to photographs, such as bright, warm, fuzzy, out of focus, clear, tinted, dark and so forth, in relative terms (to a flat response in audio, and an unaltered photograph in the analogy).

One thing that we should try to avoid is taking the word 'coloration' literally and saying that 'product A sounds red' or 'product B sounds blue' - this is not in relation to a prototypical standard, and is completely irrelevant when trying to objectively express the perceived sound.

So without further ado, let's try Photoshopping photographs to see what we mean by coloration:

A Photoshop Curve Window

The window above shows a Curve Window from Photoshop - the horizontal and vertical axes both represent brightness with the lower left corner being dark and the upper right corner being bright.

For example, the upper right side being moved up (red arrow) would make the bright areas even brighter, and the lower left side being moved down (blue arrow) would make darker areas even darker. And in the audio analogy, this is how the same transition would appear in an equalizer.

Equalizer

The difference between the Photoshop curve and an equalizer is that whereas in an equalizer an abundance of bass (characterized as 'dark') is obtained by pushing the low frequencies up, the Photoshop curve requires you to pull the curve down in the darker areas for the same effect.

If that's all made sense, let's look at some real pictures:

Bright ares are made even brighter, and dark areas are made even darker

In audio terms, the bass and treble are both emphasized

Original

Adjusted using the curve

In audio terms, the bass and treble are both emphasized

(if used in moderation, this could bring the sound alive) 

Bright ares are made even brighter, and dark areas are also made brighter

In audio terms, the bass is lacking with a treble boost

Original

Adjusted using the curve

In audio terms, the bass is lacking with a treble boost

(in the photograph, the darker areas are lacking contrast - in audio, the deep bass is not expressed clearly)  

Bright areas are made darker, and dark areas are made even darker

In audio terms, the bass is boosted with a lacking treble

Original

Adjusted using the curve

In audio terms, the bass is boosted with a lacking treble

(the sound becomes dense and dark)

Bright areas are made darker, and dark areas are made brighter

In audio terms, the bass and treble are both lacking

Original

Adjusted using the curve

In audio terms, the bass and treble are both lacking

(the midrange is emphasized in comparison, which has the effect of reducing contrast)

Hopefully, this provides some understanding of how the coloration relates to sound - the equalizer is analogous to the Brightness Curve.

This area pertains to how the treble sounds - as a rule of thumb, more treble in the frequency response makes the sound more defined, clear and resolved. On the other hand, more bass leads to masking, which reduces definition and clarity.

The resolution is closely related to the treble range - having a wider treble response is like having sharper pencils to draw with.

Resolution and treble range go hand in hand - the resolution tends to increase as the treble range widens. Resolution is defined as how finely the sound is expressed, and a wider treble range helps the sound draw a more precise picture.

Imagine drawing on a canvas with a crayon and a 0.5mm mechanical pencil: the pencil sketch with finer lines will be able to better depict the object - this is also the case with a wider treble range.

This area pertains both to the amount of bass and the treble range. The amount of bass is directly proportional to the frequency response, but the bass quality depends on the deep bass range.

Wide bass and treble ranges are both necessary for a high-quality bass.

With deeper bass range comes a sense of energy - even though your ears cannot hear sounds below 50Hz too well, your body can feel it. This is why bass is much better felt in speakers than in earphones or headphones.

The response time of the bass is also related to the treble range: if the treble range is not wide enough, the bass cannot express sudden changes in sound, which hampers the response. Thus, a product that has a wide treble response tends to have a responsive and quick bass sound, whereas the bass of a product lacking in treble sounds slow in comparison. This is not a steadfast rule, but soft bass sounds are associated with a narrow treble range, while firm, powerful, responsive bass sounds are found on products with a wide bass range and a wide treble range.

Distortion refers to when the sound is altered significantly from what it should normally sound like - distortions rarely happen on earphones and headphones, but are commonly found on cheap speakers where turning the volume up causes the amplifier (for active speakers) to clip or when a resonant standing wave is set up inside the enclosure due to poor design.

This area doesn't have too much in relation to the frequency response - it pertains to the efficiency of the product, which is a measure of how intense a sound is produced when input a given signal. Generally, more efficient units are able to express finer sounds, and will benefit even more (in fineness) from a response peak around 7kHz.