What is the best clipped peak restoration tool?

[Industrial One]

What is the best clipped peak restoration tool out there to date?

[Federico]

> What is the best clipped peak restoration tool out there to date?

Haredware? Software?
F.

[Industrial One]

On Sep 22, 12:49 am, "Federico" > wrote:
> > What is the best clipped peak restoration tool out there to date?
>
> Haredware? Software?
> F.

Software.

[gareth magennis]

"Industrial One" > wrote in message
...
> On Sep 22, 12:49 am, "Federico" > wrote:
>> > What is the best clipped peak restoration tool out there to date?
>>
>> Haredware? Software?
>> F.
>
> Software.




Well, well, well, Mr Industrial one, you have gone 2 posts now without
mentioning penises penetrating male anus'. Clearly your meds are working.

Well done.


Gareth.

[[email protected]]

On Sep 22, 6:17 pm, "Gareth Magennis" >
wrote:
>
> Well, well, well, Mr Industrial one, you have gone
> 2 posts now without mentioning p*n*ses penetrating
> male *n*s'. Clearly your meds are working.

And here is Mr. Magennis conducting the tired,
worn-out exercise called "kicking the skunk,"
the results of which are utterly predictable.

Clearly Mr. Magennis' is off his meds..

[gareth magennis]

> wrote in message
...
> On Sep 22, 6:17 pm, "Gareth Magennis" >
> wrote:
>>
>> Well, well, well, Mr Industrial one, you have gone
>> 2 posts now without mentioning p*n*ses penetrating
>> male *n*s'. Clearly your meds are working.
>
> And here is Mr. Magennis conducting the tired,
> worn-out exercise called "kicking the skunk,"
> the results of which are utterly predictable.
>
> Clearly Mr. Magennis' is off his meds..



I totally disagree. This poster very rarely contributes anything but a
diatribe of juvenile homophobic nonesense. Yet he seems to be quite an
intelligent guy. But seemingly not aware enough to realise how
innappropriate and ridiculous his actions are.

He is not a skunk, he is a someone presently stuck in a skunk like place,
but most probably in a position to remove himself from this hole and go
somewhere better.

If I feedback to him, albeit in a mocking tone, that he rarely writes
anything without some homophobic content, then maybe this will have some
effect for the better. It certainly is not the best policy for gaining the
information he is looking for in this newsgroup, and I believe his interest
here is genuine.


He just needs to grow up a lot and to realise his current methods do not and
will never work.



Gareth.

[Richard Crowley]

"Gareth Magennis" wrote ...
> I totally disagree. [Industrial One] very rarely contributes anything but
> a diatribe of juvenile homophobic nonesense.

Which is why many of us have just plonked him outright.

> Yet he seems to be quite an intelligent guy.

Have I missed anything substantive?
I didn't think so.

[Industrial One]

Ok, since DJ Pierce is one of the credible, smart posters around here
that has been useful before, I'll try to make my post more
intelligent.

I need a tool that will significantly reduce clipping if it can't
completely remove it. I'm aware that a lost signal can't be accurately
restored, but if we can estimate the amount of overdriven decibels
from the width of the clip, perhaps an algorithm could extrapolate the
lost signal from the existing samples before the beginning of the
clipped samples? Something to turn http://i38.tinypic.com/5juvt1.png
into http://i35.tinypic.com/2agowg5.png instead of http://i34.tinypic.com/33epyjk.png.

[[email protected]]

Clip restoration tool in CoolEdit is what I've been using.

[Arny Krueger]

> wrote in message


> Clip restoration tool in CoolEdit is what I've been using.

As have I. It's not the solution of anybody's dreams, but if you work with
its parameters, it does about as good job of a reasonable-appearing cleanup
as I've seen.

The problem is that if you clip a wave a little, no cleanup at all is
usually required. The ear is fooled or maybe hears a slight brightening or
hardening.

If you clip a wave a lot, so much evidence about what the wave should be has
been lost, that any restoration has to be based on guesswork.

Based on past experience, it appears that attempting to restore the waveform
may not be the best approach.

Rather, doing something like approximating what overdriven magnetic tape
might be more effective. Basically, if you overdrive mag tape you get a
fast-acting combination of soft clipping and bandwidth reduction.

[[email protected]]

Arny Krueger > wrote:
: If you clip a wave a lot, so much evidence about what the wave should be
: has been lost, that any restoration has to be based on guesswork.

Yes, if you clip a lot, it is very likely the original wave is
considerably more complex than just a parametric interpolation.
Despite that, I have noticed that it can sometimes slightly
improve the sound.

[Geoff]

Arny Krueger wrote:
> > wrote in message
>
>
>> Clip restoration tool in CoolEdit is what I've been using.
>
> As have I. It's not the solution of anybody's dreams, but if you work
> with its parameters, it does about as good job of a
> reasonable-appearing cleanup as I've seen.


Sound Forge has a Clipped Peak Restoration tool. Seems to work in a fairly
basic way.

geoff

[Industrial One]

I've tried Sound Forge. The declipped waveform *looks* reconstructed
and close to the original, but still sounds distorted nonetheless. The
ClipFix plugin for Audacity is somewhat better, the one for Nero Wave
Editor is highly regarded by some but the application is unstable, the
instructions suck horse cock and the declipper didn't do any better
than Sound Forge and ADDED distortion. The CoolEdit tool looks
promising. What parameters would you guys recommend? Should I amplify
by -15 dB before I run the restoration tool or no? What is the best
choice? I don't care if the operation will take forever to do, I need
the restoration to be maximal.

[Tony[_11_]]

Any lossy encoding of lightly clipped material (eg MP3) will improve things slightly, if
the clipping artifacts are judged less important than the rest. Of course that also
introduces other problems, and can lead to much worse clipping on playback. But the latter
can at least be countered with MP3Gain or any manual pre-DAC digital gain control (eg drop
the gain in WinAmp's EQ), or by scaling back the clipped waveform (eg -6dB) before
encoding (better). I would never suggest that MP3 encoding should be able to rival a
dedicated clip removal tool, but it's better than one might expect (when the gain is
reduced before encoding to avoid playback clipping), and Cool Edit's dedicated tool is
worse than one might expect.
Tony

[Mr.T]

"geoff" > wrote in message
...
> Sound Forge has a Clipped Peak Restoration tool. Seems to work in a fairly
> basic way.

Yep, that's been my assessment of it too, and rarely much, if any,
improvement.

I agree with what others have already said, slight clipping is not so much
of a problem, in fact most modern CD's use it to a greater or lesser extent
in the mastering process. (not that I think that is a good thing however)
Major clipping can rarely be improved much beyond some high frequency
filtering.

Given the 100dB+ dynamic range of even some cheap sound cards these days,
clipping is *easy* to avoid in the first place, so no tools should be
necessary anyway.

MrT.