The fretting about baffle step issues results from the error of "hearing" with your eyes.
When you look at the theoretical damage done by a step or edge on a freq response trace, it looks terrible. But when you are listening to music (or white noise) two things are different.
First, there is no step to hear; nothing is "happening" except some undetectable small deviation in loudness in a narrow band. Harmonic distortion is something happening and you might hear it.
Second, there is no baseline from which to hear the deviation; with your eyes, you want to see a flat line but with your ears, there is no such line, just tone colour.
Think of somebody playing a guitar. Each note has partials and other bits of grace all over the place. By eye, that would be different, but by ear just how would you know if the player was wearing a sweater while playing that day? There's no baseline.
I think things are a bit different with reflections and their timing. A thread for another day.
B.
When you look at the theoretical damage done by a step or edge on a freq response trace, it looks terrible. But when you are listening to music (or white noise) two things are different.
First, there is no step to hear; nothing is "happening" except some undetectable small deviation in loudness in a narrow band. Harmonic distortion is something happening and you might hear it.
Second, there is no baseline from which to hear the deviation; with your eyes, you want to see a flat line but with your ears, there is no such line, just tone colour.
Think of somebody playing a guitar. Each note has partials and other bits of grace all over the place. By eye, that would be different, but by ear just how would you know if the player was wearing a sweater while playing that day? There's no baseline.
I think things are a bit different with reflections and their timing. A thread for another day.
B.
It also depends of the ear of who are listening. For example, I am not unable to distinguish La 440 from La 432, but I can recognize small levels of distortions.
I'd only want to see a flat line if I was floating in space...
When you say 'small deviation in a narrow band' and it could in reality be everything under 300hz down by 6db I disagree that it is 'undetectable'
Rob.
When you say 'small deviation in a narrow band' and it could in reality be everything under 300hz down by 6db I disagree that it is 'undetectable'
Rob.
Well, the deviation in loudness in a certain frequency band (not always that narrow) actually is one (or the biggest) of the effects. It does not add distortion but it can change the room impression and dispersion pattern, depending on the frequency and the baffle dimensions.
Well, a non-linear frequency response is a colouring of the sound. The theoretical dramatic change often does not happen that extremely in praxis. That's because of 3 things:
Firstly, the sound source isn't a point, even at fullrange or coax speakers, it's always an effective surface of a certain size, depending on the geometry of the membrane and partitial movement and frequency - membranes do not move uniform as a perfect cylinder (or whatever the shape of the membrane is) over the whole frequency range (there are a few exceptions like plasma, ESL and some foils).
Secondly, if you don't use the worst case (a round baffle with centered driver), the actually effective membrane surface got a different distance to the edge. Both 'smears' the baffle step a good way over a certain frequency range. With an exact positioned measurement microphone you can still measure it but the tonality of the sound also consists of the indirect/reflected sound in the room which can partly smooth it out a bit. That has an effect on the room impression or how some details of a voice or instrument is perceived though.
Generously rounded edges or bevels at the edge of the baffle also reduces the effect of the baffle step, sometimes drastically. Absorbent foam or fabric layer on the baffle can reduce it too (especally at high frequencies).
Thirdly, it depends on the directivity of the driver. A fullrange driver, a horn or waveguide probably already starts to beam at the frequency of the baffle step and the reflections of the sound waves are already on a lower loudness and therefore not adding much up to hardly anything.
- Status
- Not open for further replies.