We can put a man on the moon and bring him back alive, but...Beowulf wrote:The science of sound perception (psychoacoustics) has been attempting to identify and quantify the qualitative aspects of sound for some time. Music waveforms can be recorded and analyzed to determine which characteristics are associated with listener's perceptions. Qualities such as sustain, attack, harmonic content, etc., can easily be measured. However, even with quantified values, individual listeners will differ in auditory sensitivity to various characteristics and in the end the values will have to be compared to subjective perceptions for validation. Audiophiles have been trying to work this out for decades and J. Gordon Holt (a man with "Golden Ears") devised a comprehensive descriptive dictionary of audio qualities in an attempt to standardize the use of "silly subjective labels," and provide a clearer definition of sonic characteristics.
An example of the difficulty of assigning numbers to audio qualities: attack can be measured as the risetime of an audio waveform from the initial impulse to maximum sound level. In addition this is a characteristic which is related to the player's technique, nails, etc., and thus to develop a measurement scale those variables would have to be taken into account. Subsequently, the time for the sound waveform to decay (decrease in level) to inaudibility (which will differ for each listener) can also be measured. However, listeners differ in their perception of and preference for "attack" and "sustain". So, who sets the standard? Someone with "Golden Ears" or do we use the average of 1000 listeners? In the end, this will very likely not be in agreement with my ears.
Colour may be defined as the presence of a particular balance and intensity of various harmonics...some people like a predominance of 2nd and 3rd harmonic content (which will tend to be perceived subjectively as "warmer" and "sweeter") and others prefer a predominance of higher order harmonics (which will tend to be perceived as "sharper", "clearer" and "brighter"). It also appears that "projection" and "sustain" of a guitar's sound is greater when the fundamental tone is not accompanied by a large proportion of harmonics.
It's not that I think this is impossible to do...just that it would involve an enormous amount of research and development to create a measurement tool that could be calibrated to an individual's preferences.
You might find this article which contains the entirety of Holt's Audio Glossary useful in clarifying the meaning and description of terms such as those you note above: http://www.stereophile.com/reference/50 ... tEepPod.97James Lister wrote:Yes, you can measure most of the properties that make up the tonal characteristics of a guitar. As has been said, the problem lies in relating these to what people hear, but that doesn't mean it's not a worthwhile thing to do. Even if we continue to use "woolly" descriptions like "warm", "sweet", "fragrant", these can still be useful if we can find a way to demonstrate what these terms mean to most people.
I actually tried to make a start at doing this on the forum about 6 years ago here: viewtopic.php?f=11&t=52898
The idea was to try to identify what the majority of listeners meant by some of the more common terms used to describe the tone of a guitar. I started with this list:
..and attempted initially to reproduce what I thought these terms might mean by taking a recording of a single guitar, and playing about with a graphic equaliser (and a few simple effects).
Unfortunately not enough members responded to get any useful data together. I'd be happy to consider trying it again if there was enough interest.
Yes, it has as I noted in my post above: specifically the section referring to harmonic content as related to sound character and waveform risetime/decay as related to attack. Another example: playing without nails will produce a warmer sound with less "attack", i.e., fewer higher order harmonics and a slower risetime in the waveform...as in less percussive sound. To measure this, simply record a note and then use a computerized programme to analyze the spectral content...or use a spectrum analyzer app:Rick Beauregard wrote:Thanks everyone for your attempts at answering my question and clueing me into the similar threads. I am still frustrated though. I am not asking the question "how do you describe the color purple." I am asking what specific combinations of red-green-blue are in this shade of purple. This is defined quantitatively for colors of every hue. I may prefer baby blue to burnt orange, that's my personal preference. But I should be able to specifically define each color that is repeatable. Finding verbal descriptors for whether red is raspy or warm is not the point. For example, to say a guitar has balanced tone says nothing to me. To say striking this note creates multiple peaks in these frequencies and suppresses these others seems like a more precise way to describe subtle differences in sound quality.
I'm sure we have some physicists and sound engineers who can solve this and present a repeatable, if somewhat simplified model analogous to the RGB scale for color. It has probably already been done.
Something you and me and Christopher Parkening have in common! That's where the comparisons end for me.Beowulf wrote:Thanks...now it's time to get out my Silver Doctor and go fly fishin' for same salmon.Rick Beauregard wrote:Great answer by the way.