[Scott[_6_]]

On Sep 5, 10:27*am, Sebastian Kaliszewski
wrote:
Scott wrote:
On Sep 4, 10:39=A0am, "Arny Krueger" wrote:
"Scott" wrote in message
...

[quoted text deleted -- deb]

The problem with all of this is considering the room tone to be a
noise floor the same way you have noise floors in the equipment
itself. The noise floor in real life is 0 in that all the "noise" in
real life is actually signal not noise. I want to hear the sound of
the room. that isn't noise that is the sound that transports me as a
listener to that space. The true dynamic range of real life is about
120 dB depending on one's thresholds of pain.
Back in the real world, the primary sources of room tone in most
contemporary live recordings is HVAC noise and/or other forms of atmospheric
or structure-borne noise from the environment.

Every time I do a spectral analysis of room tone from one of my recordings I
see the LF spikes from the HVAC air movers and hiss from the turbulent air
in and around the ducts and vents.

Everybody who wants to suffer the economic slings and arrows of building a
120 dB dynamic range recording system in order to produce 60 dB dynamic
range recordings of HVAC and traffic noise can be my guest! As things
stand, I'm usually producing recordings of them with 30 or so dB dynamic
range, and seems to produce little concern on the part of the paying
customers. A nicely done fade in and out at the beginning and end of the
song, and all seems well.

I believe the question posed was what is the dynamic range of the real
world for us as listeners. If you don't like the sound of the rooms
you are recording in I suggest finding better rooms. But that sound,
whether or not you approve of it is part of the real world and is not
noise in the same sense as you have noise in the gear itself. so if
one is actually interested in capturing everything one can hear one
does need a dynamic range of at least 120 dB. At least.

Nope. Try to play 120dB SPL signal into your ears (not for long as permanent
damage will occur promptly) and then check your hearing threshold. It won't be
even close to 0dB for considerable time (measured in minutes). Even 60dB SPL
talk will sound badly attenuated (to the border of being intelligible).

It's called accomodiation. It occurs all the time to us. For example during
silent night ticking of wirswatch sometimes disturb me falling asleep. Yeat
during the day in the very same room I'm not even able to hear thet very same
wirs****ch. That's in a city. In my family's summer cottage, far from any
traffic, during a night if a weather is calm I could hear bark beetles feasting,
hear my own body noises, etc. Simply noise floor at my summercottage at calm
night is porbably around 20dB SPL, in my city apartment it rather does not fall
below 30dB (maybe 28dB if all windows are shut hardly and neighbours are asleep
or left for vacation) and 40dB during a day. We could hear 10 maybe 20dB below
wide band noise floor but not much more.

But what you describe is masking. Not accommodation. And certainly
listening to anything at 120 dB will cause temporary desensitization
if it is constant. But that affliction aside we still do hear at both
ends of that dynamic spectrum.


Populated concert hall is not going to have 20dB SPL ever. It's rather ~35dB if
audience is behaving (i.e. to coughing, no mobile phones in 'meeting mode') and
seats are in good technical order. In case of clubs, rock concerts, etc, noise
floor is (often many) tens dB higher.

True but that 35 dB is "signal." You are hearing it because it is
there. and if we make a recording of it we would need all of that 35
dB of signal to capture what we hear in that original acoustic event.



Besides, real life orchestra does not play at 120dB when listened from audience
perspective (even first row). And were talking about real life performances not
all nonmusical stuff like jest engine at 20m distance. So 105dB of properly
dithered CD quality signal is plenty enough

who said anything about an orchestra? The question was what is the
dynamic range of our hearing. I can hear plenty of things that are
louder than an orchestra.


If one is
interested in getting it without gross distortion one needs
substantially more headroom.

Please support that statement with actual technical arguments.

That's easy. If you have 1 bit of signal what is the distortion as
expressed in percentages? If you have 2 bits of of signal what is the
distortion? You need headroom, plenty of it if you are worried about
getting low level signals with as little distortion as possible.