- 8MM Camera Noise -Removal Steps
- Posted by Parvardigar on May 18th, 2008
I was hoping to get some help on this audio track.
It’s located here: http://meherbabaaudioarchives.blogspot.com/
I’ve attempted noise reduction and fail. This material was shot with
a
8mm handheld camera. The camera machine noise is horrific.
If I can get a few suggestions that would definitely contribute to
understanding how better to use
The Adobe Audio tools. Thanks for any assistance. John M
- Posted by Paul on May 19th, 2008
Parvardigar wrote:
I had a listen, and yes, that is one horrific noise. It is
like a fan/motor noise, has two dominant frequencies, but
occupies a wide band. The AGC of the camera, causes the noise
component to drop, when the human speakers are speaking. It will
take a fancy tool to get rid of that. Only the staff on
"Miami CSI" will be able to fix it :-(
Good luck,
Paul
- Posted by Parvardigar on May 20th, 2008
On May 19, 10:44*am, Paul <nos...@needed.com> wrote:
Thanks. I needed to hear that. At least I know I have one real world
solution. . . CSI Miami Crime Lab.
- Posted by webpa on May 20th, 2008
On May 18, 3:23 pm, Parvardigar <parvardi...@yahoo.com> wrote:
If you're in Windows: try GoldWave.com. They have an evaluation
version that is (IIRC) completely functional. There is a noise
removal function that is quite effective under some circumstances.
May help you.
- Posted by Parvardigar on May 20th, 2008
On May 20, 9:41*am, webpa <we...@aol.com> wrote:
Thank you. I'll try it out, and let's hope for the best.
- Posted by Richard Crowley on May 21st, 2008
"Parvardigar" wrote ...
You missed the emoticon. CSI was mentioned specifically
because it is NOT anything remotely resembling any sort
of "real world" capabilities. Their techniques exist only in
the fertile imaginations of their non-technical writing staff.
Their magical feats in the lab are generally laughable to
people who know what is possible in the real world.
I've reduced (never *removed*) similar noise using the
noise reduction feature of Adobe Audition. You can let
is "sample" the noise by defining a snipped where there
is only noise. Then you can try applying varying amounts
of reduction of that noise profile. Note, however that the
stronger you make the noise reduction, the worse it makes
your desired audio, so you have to make the (frequently
difficult) tradeoff decision.
- Posted by Paul on May 21st, 2008
Richard Crowley wrote:
Just for fun, I tried a couple things in Audacity, the free
audio recording/editing tool. It isn't really set up for
anything too fancy.
There is a noise removal tool, that is supposed to take a
sample of user provided background noise. What the tool is
supposed to do, is run an FFT and spot the noise frequencies
and that provides a profile of what to filter on the real
sound recording. When I tried the tool, using a snippet of
sound from the recording, where no human voices were present,
the tool crashed !
The second thing I tried (knowing in advance it was pointless),
was using the notch filter. You can download a separate filter
effect (not in the main program) and install it. Using this,
I filtered more than 30 different discrete frequencies. When I was
finished, the human voices were no more audible than before.
http://audacity.sourceforge.net/nyquist/notch.ny
The thing is, when the notch filter is used, on so many
frequencies (because the background noise is wide band),
it just attenuates everything. I knew it would be pointless,
but I had to try it anyway.
To do the filtering, would require a tool with better
heuristics. The thing is, the noise is hopping around in
frequency, with time. And the hopping around is random looking,
and not necessarily predictable. (I know this, because
I used the Sonogram program, to analyse the sample.)
And this observation is spot on :-)
"NOT anything remotely resembling any sort of "real world" capabilities"
The CSI TV programs are filled with impossible things,
wrong things, for the purposes of dramatic effect. For
example, if there is an explosion depicted in the program,
would you expect them to give correct chemical details
on how it was done ? Of necessity, some things in that
program, will be glaringly incorrect for good reasons.
Part of the unintentional humor in that series, is
finding all the nonsensical "science" in it.
Paul
- Posted by Parvardigar on May 25th, 2008
On May 20, 10:29*pm, Paul <nos...@needed.com> wrote:
"You missed the emoticon." No I did get it...I was just teasing.
and Paul...thanks for working over the sound clip. Maybe I'll have to
remove the sound track - and get someone to lip synche in the vocals!
seeing that there is no rememdy with this extreme audio matrix of
machines and vocals mingling together. Thanks for the input, and
insight.