dave stojan wrote on Thu, 01 March 2007 20:57 |
If you search the LAB archives you'll find the Yamaha Handbook has been pointed out incorrect on this particular subject. Think about it - if 3db is barely noticeable how would 10db sound twice as loud?
|
Then just about every other respected audio and acoustics text and reference must also be wrong.
You have to keep in mind that what is being addressed is perceived changes in loudness due to changes in sound levels, not signal levels. Human hearing is not linear in frequency or level nor in the perception of several other factors, thus perceived loudness becomes very complex where absolute level, frequency content, tonal components and duration among other factors can affect the perceived loudness. So assessing changes in loudness is quite different than addressing absolute voltage or power level changes. I think that trying to consider a big picture that can encompass voltage levels, electrical and sound power levels, sound pressure levels and loudness levels quite understandably confuses many people.
For one thing dB numbers, such as dB Sound Pressure Levels or dBV and dBu, are ratios derived by referencing one number to another reference number, the term "dB" by itself is actually meaningless without some reference value. Conversely, loudness is actually measured in absolute units of phons and sones rather than in dB units. But more importantly, what is actually being related to the effect on loudness is not a change in the voltage or electrical signal level but rather a change in acoustical Sound Pressure Levels (SPL). Trying to relate a specific change in a fader level directly to a change in perceived loudness is not really a valid relationship as there are numerous factors that can keep this from being a direct, much less linear, relationship.
What is generally stated is that all else being the same, a change in Sound Pressure Level of 3dBSPL is perceived as a readily noticeable change in loudness, while a 10dBSPL change in level is perceived as a halving or doubling in loudness. I think that this generalization is usually applied in the context of trying to create the same signal, only louder or softer, and in that sense it is valid. If there are differences in the frequency content, duration, etc. fo the acoustical signal then this generalization certainly may not be true, but I think that is somewhat beyond the issue at hand.