|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
207.242.206.146
In Reply to: RE: How to reproduce the full dynamics of a scissor cutting paper without clipping posted by tomservo on April 18, 2017 at 09:55:36
I have a pretty crappy scissors here - they don't make a "snip" so instead I popped a Coke can open 4 inches from the microphone.
In one second the levels went from a minimum of 14.7dBA minimum (FAST RMS) to a maximum of 87.5dBA FAST RMS. These FAST time weighted values will correlate (somewhat) to the loudness of the signal. The PEAK pressure on the microphone reached 114.3dB!That means within less than a 1 second interval an amplifier would have to cover over a 100dB swing in level without clipping. The difference between the MAX level(during the peak) and the Max Peak was 27dB. That 30dB difference requires 1000:1 amplifier headroom! It is a rare recording that shows and preserves that difference.
The Gray "Cursor" and values for that second shown to the right.
The recording engineers place limiters and compressors on the recording or mastering process that prevent that almost 30dB crest factor (Peak to RMS Ratio) of something real! The do this because human hearing doesn't or won't hear the instantaneous peak anyway. And if they recorded the signal preserving the dynamics on the CD, you would be turning your volume knob full clockwise to hear the signal - and then your amplifier would be sure to clip!
The Sound Meter uses TWO 24bit A/D converters overlapped 30dB and sampled at 48kHz to cover the measurement range up to 144dB or so.
Note: the Crappy scissors cutting paper had a Peak level 77.9dB and Max RMS level of 56dB - still a 21dB crest factor - typical recordings are only 13-20dB!
"The hardest thing of all is to find a black cat in a dark room, especially if there is no cat" - Confucius
Edits: 04/20/17Follow Ups:
Hi Jim
I am glad you posted here, you should have a good selection of toys.
I had not heard of the scissors test but I did find the peak levels indicated by the 2204 slm on impulse hold, were a lot louder than the subjective loudness would have suggested.
A problem is our sense of loudness is both frequency and time weighted and not strictly sound pressure like a microphone but if one wants to produce "the whole signal", one has to produce all of it.
Not knowing what / if any integration was in the meter on impulse hold, the cumbersome / failsafe way to judge instantaneous peak pressure is to capture the mic signal on a storage scope (or a repetitive signal is easy to capture on a regular scope) and use the mic sensitivity to calculate the peak pressure.
Lastly on the subjective audibility of very short things;
About 15 years ago, a well known amplifier company was going around with an abx box and their new amplifiers.
They came to the shop and from home I brought in my trusty Threshold stasis intended to be in the mix.
I set up a system I was familiar with and had a CD of wave files (of to me were revealing and difficult recordings). We had Crown's, a Chevin, the QSC's and a few others. Fast forward 6 hrs and the amps generally had two slightly different sounds, it was all very subtle behavior on the decay side of transients and only audible on a few tracks but repeatable. AND then there was the Threshold.
It sounded exactly the same as the better category amps but at a decent but not that loud level (on sensitive speakers) it sometimes was less dynamic.
Enough different that I grabbed an oscilloscope and looked at the amp outputs (as the front indicators said we were at peaks about -16dB) and saw that on the biggest peak, there was occasional instantaneous clipping.
It was these very short "limiting" events which were it, not at all audible as traditional "clipping" but instead were only detectable when you heard with and without as I was here, being able to switch to a much larger amplifier and back.
Since then I have urged people if they have a scope, to look at the amp signal (perhaps not with class D) and see if there is any clipping at listening levels.
Best
Tom
And if you do the same thing with the microphone at a normal distance away (say 2 feet)?
I think Inmates point is that sure, at 5 inches the scissors will make 150 db peak but at 2 feet away it will be a LOT less.
You sort of have this backwards. Sure at two feet the level will be less. but you loudspeakers are 6-7-8 feet away so compared to two feet away they have to be putting out 6dB or more output than the Coke can at 1 or two feet. 114dB adjusts down to about 104dB in a typical room - then add 6dB for the loudspeaker distance distance and adding two loudspeakers (assuming a mom placement gets you back up to 110dB peak capability.
While it could be the level that you are interested in, there is also the interest in preserving dynamics. Sure in a Public address world - the loudspeakers would have to be putting out a lot of SPL to make a coke cane sound realistic at 100 ft. So high efficiency is paramount in that application. But I feel that if we really had recordings the preserved the dynamic range of a piano or drum kit or even the bite of a reed, many audiophile stereo systems would be driven into clipping.
Another think the measurements bring out is that it is important to listen in a quiet environment to preserve dynamics. The best and most expensive listening rooms I have been in are designed for 10dBA or lower background noise levels - with the HVAC running! The standard microphone on the sound meter can't go that low but (as you see in the pic) we do make a mic that can go that low...
"The hardest thing of all is to find a black cat in a dark room, especially if there is no cat" - Confucius
"You sort of have this backwards."Not really. His point is simply that the sound of scissors cutting paper, at some "normal" distance, say from your hands to your ears, is a very small sound, and that, as you move the microphone (or your ears) closer, it's louder. At a normal distance, it ain't anywhere near 140 dB peak.
If you want to reproduce that sound, and hear it from 8 feet away AS IF you were 2 inches away, that's a different ballgame.
BTW, regarding studio rooms... When I was in Studio D at Fantasy in Beserkeley, it was like being in heaven! Quiet, spacious, nice visual asthetics, and Alison was easy on the eyes, too. Forget recording - I could hang out in that room just to unwind or have lunch!
Anyway, regarding sound levels... Did you know that the SPL of a trumpet can be up to 200 dB at the bell of the horn?! Holy crap! Somebody told me this years and years ago. Ever since, I've wondered if most trumpet players know this, 'cause especially these days, lots of them use "clip-on" microphones on their horns. I remember recording the Buddy Rich band, and the trumpet soloist was practically eating the microphone, and it sounded like a f-ing kazoo. On the other hand, Doc Severinsen uses a Sennheiser 441, and, doesn't eat it.
:)
Edits: 04/21/17
I heard 150db at the bell not 200db. What was your source for that number?
I don't recall the figure but it is quite high.
This high speed video captures what they called a shock wave produced by a trumpet however a traditional shock wave is visible on the rarefied side.
https://www.youtube.com/watch?v=Cr4ro2lftZU
The second video is a system i developed the transducers and control electronics for which here is operating in the mid 160dB's at about 21Khz and at least up to 175dB, i never saw any kind of shock wave even at those pressures (although at 175dB one could light a cigarette with acoustic friction).
https://www.youtube.com/watch?v=0K8zs-KSitc
I don't remember, it was a long time ago. Anyway, I play the trumpet, and it can be VERY loud at the bell. Heck, it can be very loud 15 feet away. I went to a clinic/master class given by Doc Severinsen about 3-4 years ago. He picked up a student's horn, pointed it right at me, and darn near parted my hair with a scale going up to high C.
:)
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: