Melbourne music spotlight: Stewart Kohinga
Ever noticed how much mediocre music lives on the radio, regardless of the channel? Well, in the interests of promoting musis that is actually worth hearing, that you might not hear on the radio, let me introduce some local talent. Stewart Kohinga is Melbourne singer and songwriter who recently performed live as part of the Live at Baker street series.
I would describe his music in poetic terms, but you can hear it for yourself on youtube >
Stewart Kohinga online:
Facebook page
Website
Live at Baker Street series >
I would describe his music in poetic terms, but you can hear it for yourself on youtube >
Stewart Kohinga online:
Facebook page
Website
Live at Baker Street series >
Melbourne event : live music to CD
This is a rare treat. A studio in Melbourne this Saturday are offering a live music experience where you then get to hear the recording on a $150k system.
Read more on StereoNET >
There is also a session on Thursday 30th.
Read more on StereoNET >
There is also a session on Thursday 30th.
RSA HE2 almost there
HE2 progress - the box is almost complete.
This first version will be my demo speaker. Since various aspects of this build proved to be very labour intensive, I will be making a simpler version based on CNC machined panels. This one includes bitumin rubber damping, which is very labour intensive to apply, and quite expensive. It will be an upgrade option added to the standard model.
Coming up:
- clear finish
- directivity measurements
- passive crossover
Loudness perception and bass
We've all seen the equal loudness contour:
It tells us quite a few things, many of them missed in casual observation. The first thing we might notice, is that our hearing is most sensitive to midrange from about 200 Hz - 14 kHz. Bass is subjectively "turned down" indicating that it's actually not very important in the functional sense. Bass doesn't tend to occur in nature in isolation of much higher frequency content. We also note that our ears are especially sensitive around 3.5 kHz.
This was made clear to me when as a teenager, I measured my first system. I found a huge peak around 3.5k and playing test tones and measuring with an SPL meter outside the science lab in high school, those 107 dB peaks were just plain nasty! No doubt that was a big part of what made little 15W speakers sound loud to my teenage ears.
This 3.5k peak in sensitivity does have implications. Many midrange drivers have cone breakup around this region, and many crossovers are not far removed from this point. A very large part of our internal hearing mechanism is dedicated to 1 - 7 kHz primarily, this region is critical.
Notice how things change with level. At a typical quiet background noise level of about 26 dB, we need over 90 dB at 20 Hz before we perceive the bass as simply matching in loudness. Many subwoofers fail to even register as a quiet background noise level! Now if we listen at a moderate level of 60 dB, we need 110 dB @ 20 Hz or about 94 dB @ 40 Hz. We probably need about 25 - 30 dB increase in bass level to get it to sound about the same in level across the bass spectrum.
If we boost the level to 80 dB, we need about 20 dB increase in bass level. The bass is starting to sound a bit more solid. If we increase the level to 100 dB, we are now in the very loud range, and the bass only needs about 12 dB boost to match the level.
So why don't we all boost our bass by these crazy amounts? These effects are partially allowed for in the recording process, so you don't want to invert these curves and use them for EQ. But these curves reveal that the tonal balance changes with level and in this regard, it is the bass that changes the most.
I had previously thought that the equal loudness contours explained the harshness of some speakers when turned up loud. The contours don't support that view. Perception of treble stays about the same at high levels. In reality this might be explained in terms of system stress where distortion causes the treble to sound louder, especially where the amp is clipping.
This was made clear to me when as a teenager, I measured my first system. I found a huge peak around 3.5k and playing test tones and measuring with an SPL meter outside the science lab in high school, those 107 dB peaks were just plain nasty! No doubt that was a big part of what made little 15W speakers sound loud to my teenage ears.
This 3.5k peak in sensitivity does have implications. Many midrange drivers have cone breakup around this region, and many crossovers are not far removed from this point. A very large part of our internal hearing mechanism is dedicated to 1 - 7 kHz primarily, this region is critical.
SPL and tonal balance
Notice how things change with level. At a typical quiet background noise level of about 26 dB, we need over 90 dB at 20 Hz before we perceive the bass as simply matching in loudness. Many subwoofers fail to even register as a quiet background noise level! Now if we listen at a moderate level of 60 dB, we need 110 dB @ 20 Hz or about 94 dB @ 40 Hz. We probably need about 25 - 30 dB increase in bass level to get it to sound about the same in level across the bass spectrum.
If we boost the level to 80 dB, we need about 20 dB increase in bass level. The bass is starting to sound a bit more solid. If we increase the level to 100 dB, we are now in the very loud range, and the bass only needs about 12 dB boost to match the level.
So why don't we all boost our bass by these crazy amounts? These effects are partially allowed for in the recording process, so you don't want to invert these curves and use them for EQ. But these curves reveal that the tonal balance changes with level and in this regard, it is the bass that changes the most.
I had previously thought that the equal loudness contours explained the harshness of some speakers when turned up loud. The contours don't support that view. Perception of treble stays about the same at high levels. In reality this might be explained in terms of system stress where distortion causes the treble to sound louder, especially where the amp is clipping.
What do we learn?
- 10 dB in the midrange sounds twice as loud
- 6 dB in the bass range sounds twice as loud - this number decreases at high SPL levels
- Maximum sensitivity - 200 - 7k
- Peak sensitivity 3.5k
- A dip in sensitivity occurs around 10k
Waveguide damping
This is how I add damping to plastic waveguides to improve their performance. I use bitumin rubber which is mainly used for waterproofing applications. It has mass, adhesion and flexibility which make it work here also.
Raw plastic waveguide:
To ensure good adhesion, the plastic surface is scratched up with coarse sand paper:
Then the bitumin rubber is diluted and applied as a primer:
When it dries, the surface is slightly rough and has the kind of texture that easily bonds to successive layers. Masking tape is applied.
After a few hours, a thick first layer is applied with a metal paint scraper and a brush.
Then a second layer includes fine screenings, which add greater mass. The screenings are cheaper than the bitumin rubber, ensuring that it goes further, but they also improve the performance.
A third layer seals up the mix of screenings and bitumin rubber. This locks them in the middle. Without a layer underneath, they can tend to lack adhestion. Without a layer on top, they can tend to come off.
Langganan:
Komentar (Atom)