From: Vicki Powys <>
> Walt and all,
>
> I checked out the Sound Out possibilities. On my iMac running OS9, in Pe=
ak
> LE, I can choose ASIO or Built In. Built In seems better than ASIO and i=
s
> the default. With Lang's insect sounds, Built In (listening via Peak) an=
d
> Quicktime (used separate from Peak) were identical in listening results.
OS9, and the Peak that runs in it are virtually entirely different than
OSX and the Peak that runs in it. They appear the same, but all the
underlying programming is quite different. ASIO appears to have gone
away in OSX to be replaced by Core Audio and two versions of HAL (and
others depending on your hardware). There is still a built in option, in
the system sound options, and it could be using Core Audio/HAL or
whatever. Much more complex.
> I cannot try Walt's Magic trick as Peak LE does not support sample rates
> above 65 kHz and I don't think I can change the bit rate either.
The particular rate I gave was the highest rate my OSX G4 running a OSX
native Peak will allow. I did try the various steps below that one, and
the higher I went the better the results. Even the 64 you can do helps a
bit.
Peak won't change the bit rate within a open file. You have to do a save
as... and set a new bit rate in the options for the aiff file. Then open
that new file. I'm not sure how much of all that is built into Peak LE,
been a very long time since I used it.
> The sample files at 14 secs and 1.2 MB are too large for me to download a=
ny
> more. 5 second samples would be plenty.
I know all about that, my connection is barely faster than yours.
Particularly at uploading the files to the server. And they are eating a
bunch of server space, so will not be up all that long.
> My assessment of the sounds is same as before, and Walt's magic files
> sounded poor, compared with the original 48kHz. Interesting that none of=
us
> agree on this!=20
There are three people here, all have listened to the samples, and we
have three opinions right here. Though not far apart.
With the variations in equipment, the variations in our ears and so on,
it's a wonder if we all hear the same thing in anything.
My threshold of hearing at frequencies from 10khz to 20khz is nearly
even, and at about 40-60dB louder than my threshold of hearing at 1khz.
I can hear these samples, but have to turn the gain on the headphone amp
up a lot. At that gain the errors in some samples in the lower
frequencies become very audible even though they are a lot quieter. I
can even hear Rich Peet's Bat, with lots of gain cranked in. My son, in
comparison does not have to crank in much extra gain to hear the high
frequencies in the sample. To make it even more fun for me, my left ear
is less sensitive than my right, something I think is due to all the
driving I've done in my lifetime with the window down.
In these samples we have a wide high frequency band that is by sonogram
the loudest thing in the recording. We also have a much fainter narrow
band at 5khz and at about the same level sound below 160hz. In the
original Peak and Soundhack samples sonograms show additional noise
introduced between about 2khz & 5khz, a conversion error. This extra
noise is at about the same level as the other faint parts.
Now think about two simplified types of hearing. One like mine will hear
the high band as the same volume as the low band only if it's 40-60dB
louder. It is nearly that much louder, but not quite. So, when I amplify
the high band to hear it, the low frequencies also get amplified up into
my audible range at nearly the same time. I hear all the various
frequencies at similar apparent volumes. Now think about a person who
has little high frequency loss, my son, no extra amplification is needed
for the high frequencies and he is fairly evenly sensitive to all
frequencies. Since the high frequency is so much louder than the others
my son will hear it before the lower stuff reaches his threshold of
hearing as he turns the gain up. The high frequency band will definitely
dominate the sound picture, or even be all heard. He will judge the
quality of the samples mostly on the high frequency band, hardly hearing
the others. And since most of the differences Lang and I hear are in the
lower stuff, may come to different conclusions.
On the basis of your statements that you believe yourself to have very
good high frequency hearing, I would think you probably are judging
these samples on the basis of the high frequency band. If you have a cut
filter available, try filtering out everything above 10khz or so and
listen to the lower stuff. That was something else I did with these
samples was to cut out various bands so as to listen to just each part.
That all, of course, leads into a discussion on the perils of
equalization, but that's a separate subject.
Walt
________________________________________________________________________
________________________________________________________________________
|