[[email protected]]

On occasions i have someone coming to the studio to rescue an old
recording in which the vocal is buried in traffic noise, also speeches
where the noise clouds what is being said. Sometimes I can get pretty
decent results but others it's impossible to achieve due to the nature
of the recording.
I generally use iZotope RX to remove noise/unwanted frecuencies though
it depends on what is happening on the recording.
I was wandering what kind of things/tools do you try when attempting
to do such a thing, it obviously depends on the nature of the noise
but I wanted to know if someone does something different from what I
do that I could try.

thanks!

[Johann Spischak]

schrieb im Newsbeitrag ...
On occasions i have someone coming to the studio to rescue an old
recording in which the vocal is buried in traffic noise, also speeches
where the noise clouds what is being said. Sometimes I can get pretty
decent results but others it's impossible to achieve due to the nature
of the recording.
I generally use iZotope RX to remove noise/unwanted frecuencies though
it depends on what is happening on the recording.
I was wandering what kind of things/tools do you try when attempting
to do such a thing, it obviously depends on the nature of the noise
but I wanted to know if someone does something different from what I
do that I could try.

thanks!



By rescue old recordings there are two ways:One is, with software by "surgery" cleaning from unwanted noises. There are however another aspect, the distortion too. That goes on special custom build hardware basic, what we have. After we make a digitalizing and/or resampling, can be judged how much was noise and how much distortion from that sound. Only after that can come surgery if it is still necessary at all. In the most cases not!

Regards
Johann Spischak

[[email protected]]

Sometimes I use band pass filters, which are not available on RX but
they are on a few plug-ins I own, though if the recording is someone
speaking I haven't tried one, i'll give it a try.
What I do often is look at the spectrogram and try to remove what
bothers me from there, but sometimes the speech is so buried in the
noise i can't make the noise go away without taking away a major part
of the sound i want to recover, and at that times is when i find
myself not being able to achieve a good result.
I think you mean something like that when you say surgery? And what
kinds of things should you try to do to remove distortion? There is
one declipping in RX though by no means it's a magic bullet, and Waves
Decrackle sometimes help my, others not.

[[email protected]]

Sometimes I use band pass filters, which are not available on RX but
they are on a few plug-ins I own, though if the recording is someone
speaking I haven't tried one, i'll give it a try.
What I do often is look at the spectrogram and try to remove what
bothers me from there, but sometimes the speech is so buried in the
noise i can't make the noise go away without taking away a major part
of the sound i want to recover, and at that times is when i find
myself not being able to achieve a good result.
I think you mean something like that when you say surgery? And what
kinds of things should you try to do to remove distortion? There is
one declipping in RX though by no means it's a magic bullet, and Waves
Decrackle sometimes help my, others not.

[Johann Spischak]

schrieb im Newsbeitrag ...
Sometimes I use band pass filters, which are not available on RX but
they are on a few plug-ins I own, though if the recording is someone
speaking I haven't tried one, i'll give it a try.
What I do often is look at the spectrogram and try to remove what
bothers me from there, but sometimes the speech is so buried in the
noise i can't make the noise go away without taking away a major part
of the sound i want to recover, and at that times is when i find
myself not being able to achieve a good result.
I think you mean something like that when you say surgery? And what
kinds of things should you try to do to remove distortion? There is
one declipping in RX though by no means it's a magic bullet, and Waves
Decrackle sometimes help my, others not.



Under surgery I mean eliminating (cutting out, suppressing, adding de-phase noise and so on) after very carefully analysis. That all on software basis, like with Cube Tech Forensics.

By hardware way what we make, first we must recognise from that sound what distortion and what noise is. By playing the sound on different hardware you can hear what moves with the sound together and what is changed depending the hardware. In our epuipment (my own build-soldered) is very easy, since it has a so high S/N ratio, that the question own noise - own distortion became question of chicken-egg question. The most distortion disappears already by the first resampling and then we need either to repeat it, or to cut the exactily defineable dirt away.

Regards, Johann

[[email protected]]

Thanks for the advice guys!
Johann, what do you do when you resample? Upsample, downsample? How
does that take the distortion away?

Juanma

[Johann Spischak]

schrieb im Newsbeitrag ...
Thanks for the advice guys!
Johann, what do you do when you resample? Upsample, downsample? How
does that take the distortion away?

Juanma


Neither up nor down. Since my technology is hardware based I do not need to change the file to run it trough my mastering computer. I need only open the audiofile, do something, then cancel it, then save. Ready. Another version, when I convert the audiofile to DDP, then back to wav. Ready. It is astounding how much from original sound is hidden in an audiofile. The ground is the accumulation of quantisation noise-distortion because of digital signal deformation through glitches. The sound goes in the most cases down, appearing more body, deepness, natural timbre, even weight-volume together with better plasticity, the original room wall to wall. At the same time you will hear what belongs to the original record and what is for example artificial echo, given to it by post production (it floats only so in the air and has nothing to with the room, since the original reverberation is back, sounds very tasteless) But again, it is special hardware wich is extremly noiseless (-220 dB) full of technics, which are not existing on the market because of their prices or complexity.

If your problems with that record looks unsolvable, you can send me the original to take a look at it.

Regards, Johann

[[email protected]]

Sounds cool man, I'd like to get my ears on stuff like that someday.
And if I ever get a record so valuable that is worth bothering you,
i'll definitely send it.
Do you use the same hardware when it comes to cleaning a signal that
has been recorded in a noisy environment and has traffic noise or wind
noise that is periodic and not easy to profile with a denoiser?

Juanma

[Mark1]

But again, it is special hardware wich is extremly noiseless (-220
dB) full of technics, which are not existing on the market because of
their prices or complexity.



Ha Ha that's a good one...

Mark

[Mark1]

Under surgery I mean eliminating (cutting out, suppressing, adding de-phase noise and so on) after very carefully analysis.

What exactly does "adding de-phase noise" mean?

this should be good...

Mark

[Richard Crowley]

"Mark1" wrote ...
Under surgery I mean eliminating (cutting out, suppressing, adding
de-phase noise and so on) after very carefully analysis.

What exactly does "adding de-phase noise" mean?

this should be good...

Perhaps it means he's imbibing the Christmas Schnapps a bit early.

[Ben Bradley]

On Fri, 12 Dec 2008 09:17:22 -0800, "Richard Crowley"
wrote:

"Mark1" wrote ...
Under surgery I mean eliminating (cutting out, suppressing, adding
de-phase noise and so on) after very carefully analysis.

What exactly does "adding de-phase noise" mean?

this should be good...

Perhaps it means he's imbibing the Christmas Schnapps a bit early.


I was trying to give the benfit of the doubt, thinking this could
be a language barrier thing, that the poster doesn't know enough
English to correctly describe his 'technics.' The website for the
poster's email domain, which appears to describe his services, looks
to be in German.
But then there was the -220dB thing, and that means the same in
every language.

No doubt he uses the Denon AK-DL1 for all his interconnects.

[Johann Spischak]

schrieb im Newsbeitrag ...
Sounds cool man, I'd like to get my ears on stuff like that someday.
And if I ever get a record so valuable that is worth bothering you,
i'll definitely send it.
Do you use the same hardware when it comes to cleaning a signal that
has been recorded in a noisy environment and has traffic noise or wind
noise that is periodic and not easy to profile with a denoiser?

Juanma


Well I am using only this one for all purposes, because the difference is too big already by first listening the original. I make mastering - remastering mainly and by some old originals I have many of this sort of small detail questions. In the most cases the customer must listen the sound without its dostortions - but also without any destroying changes - and decide what to do. The charakter of the wholerecord becomes so much different, softer, smoother, that in most cases a "surgery" would damage more than rescue.

The example with the wind is good: It sounds like FFFF meanwhile undistorted like HHHH. This way it is more a part of the scene than a disturbing noise. First when it is distorted can cover other voices, which are important to hear, however these will be also cleaner,so this is an interaction between them.

I was asked to put some sound onto our website for demonstration for our customers, so it wll come in the next few days together with some other languages.

I whish you a happy weekend
Regards
Johann Spischak

[Johann Spischak]

"Ben Bradley" schrieb im Newsbeitrag ...
On Fri, 12 Dec 2008 09:17:22 -0800, "Richard Crowley"
wrote:

I was trying to give the benfit of the doubt, thinking this could
be a language barrier thing, that the poster doesn't know enough
English to correctly describe his 'technics.' The website for the
poster's email domain, which appears to describe his services, looks
to be in German.
But then there was the -220dB thing, and that means the same in
every language.

No doubt he uses the Denon AK-DL1 for all his interconnects.


No doubt you don't speak german at all, however you passed some picture too which - I am afraid - would rather irritate you. :-)))

The foto shows an Audio-CD from tempered glass, made with Direct-to-Glass-Mastering, no stamper, no pressing, this is the end product. Of course it is up to you, is it for your pocket or not. Technically and in sound unbeateable. (Psst! There are some downloadable ppt presentations too, but in french, again nothing for you, I am afraid, so you will stay closed in your small world)

I whish you a happy weekend
Regards
Johann Spischak

[Scott Dorsey]

In article ,
wrote:
On occasions i have someone coming to the studio to rescue an old
recording in which the vocal is buried in traffic noise, also speeches
where the noise clouds what is being said. Sometimes I can get pretty
decent results but others it's impossible to achieve due to the nature
of the recording.

If you want to learn about this stuff, the AES Forensics Committee has
some panels and classes that are very much worth checking out.

A good introduction to the basic theory can be found in Tom Owen's 1987
paper "Forensic Audio-Video Theory and Applications" which is probably
available as an AES convention preprint.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."