Marky
posting's fun
Joined: 30/06/04
Posts: 560
Loc: Boston, MA
|
Reversing compression for sampled music
#44623 - 03/11/04 04:57 AM
|
|
|
|
Anyone any tips for un-doing compression when sampling existing tracks ?
- Mark
-------------------- "Nothing in the world can take the place of persistence. Talent will not; nothing is more common than unsuccessful people with talent."
|
giles
new member
Joined: 26/10/03
Posts: 555
Loc: belgium
|
Re: Reversing compression for sampled music
[Re: Marky]
#44625 - 03/11/04 05:31 AM
|
|
|
|
waves C4 using the uncompress preset or too much limiting preset. G
|
adrian_k
Joined: 30/01/03
Posts: 1741
Loc: Gloucestershire
|
Re: Reversing compression for sampled music
[Re: giles]
#44648 - 03/11/04 08:26 AM
|
|
|
how on earth does it do that? once a transient has been flattened how do you get it back?
|
Ultimate Fish
posting's fun
Joined: 06/12/02
Posts: 1910
Loc: York, UK
|
Re: Reversing compression for sampled music
[Re: adrian_k]
#44679 - 03/11/04 09:26 AM
|
|
|
|
You don't. But using expansion can help restore some dynamic range.
-------------------- need to get rid of this stupid sig...
|
adrian_k
Joined: 30/01/03
Posts: 1741
Loc: Gloucestershire
|
Re: Reversing compression for sampled music
[Re: Ultimate Fish]
#44731 - 03/11/04 10:50 AM
|
|
|
|
and then I guess you could use something like transient designer to shape the sample..
|
Hugh Robjohns
SOS Technical Editor
Joined: 25/07/03
Posts: 18539
Loc: Worcestershire
|
Re: Reversing compression for sampled music
[Re: Marky]
#44756 - 03/11/04 11:35 AM
|
|
|
|
Tools like the transient designer are designed to work on single sources, not complete
mixes -- there won't be enough clean 'transient' for the device to control the envelope
properly in a mix.
Expansion can sometimes be used to increase the dynamic
range of previously compressed material, but it is rarely effective -- especially if
complex multi-band compression has been used.
DBX works basically with simple
2:1 compression on the record pass and a complementary 1:2 expansion on replay... and
think how rarely that works properly and hjow easily it is messed up!
Hugh
-------------------- Technical Editor, Sound On Sound
|
adrian_k
Joined: 30/01/03
Posts: 1741
Loc: Gloucestershire
|
Re: Reversing compression for sampled music
[Re: Hugh Robjohns]
#44761 - 03/11/04 11:39 AM
|
|
|
|
ahhh.. wrong sample, I was thinking single note/hit samples.
yes this doesn't
sound like a goer at all.
|
Urthlupe
member
Joined: 20/09/02
Posts: 380
Loc: West Midlands, UK
|
Re: Reversing compression for sampled music
[Re: Marky]
#44766 - 03/11/04 11:55 AM
|
|
|
|
Hi Mark
Apologies - you are the unfortunate recipient of my 'Large Post of the
Year'.
There are a couple of techniques involved in 'uncompressing' whose
succes will depend on the material, your ear and the amount of faff you can take.
I do a bit of restoration in my diddly studio and sometimes it can be something of a
vicious circle. Usually it is the most damaged waveform which requires the most
remediation, but the more one applies then the more that remedial process can be heard
operating. Therefore often the material most in need of processing will suffer most from
it. It's a bit like spelling banana - you have to know when to stop. So beware.
First up - upward compression. From its name you might think this is what you want but
in fact this is best viewed as a better (but long winded) way of compressing in the first
place. Simply put - pass your material through a compressor and add this in small
proportion back to the source. This technique uses the downward action of a standard
compressor to raise low-level passages leaving high levels untouched, but to my ear also
allows a degree of dynamic lift not present in standard compression - it has been called
'parallel compression' for obvious reasons. The amount of lift is controlled by the amount
of compressed signal added back in. Use a very low threshold to force the compressor into
almost permanent gain reduction and a fast attack to preserve transients. Be careful to
delay the source by the time it takes the signal to travel the compression circuit. See -
long-winded. Bob Katz explains it well.
Second - upward expansion. This is
easier and what you want for your restoration of dynamics. A standard compressor fools the
ear into perceiving material as louder where in fact, above the threshold it has been
reduced. If, rather than reducing signal above the threshold you increase it, then you
have achieved upward expansion. You need a bit of kit capable of doing this. The two I use
are, in hardware the DBX Quantum and in software Waves Linear Phase Multiband Compressor
(Masters Bundle). I think that Waves C1 and C4 are also capable. This technique is
particularly useful in multiband compression (in fact I can't remember using it broadband)
where the expansion can be focused within a specific frequency range.
Last bit
- I know some folks think 'Ooh Waves - overpriced', but take my tip - do not buy from a
dealer. Buy direct from the Waves site via download, prices are 30% cheaper (I jest not)
and you are not limited to bundles - you can purchase the individual plug-ins which you
need.
You can wake up now guys - Im done.
Loopy
|
*gregster*
Ceilidh godfather
Joined: 07/01/03
Posts: 598
Loc: Glass Cow
|
Re: Reversing compression for sampled music
[Re: Marky]
#44804 - 03/11/04 12:50 PM
|
|
|
That' not a large post, m8 don't worry. Can I ask a question in a similar
vein... Anyone know if it's possible to reduce the effect of an over-zealous
tremolo effect on a recorded part? I've got a guitarist using a tremolo pedal which sets
the fluctuation rate completely at odds with the tempo of the song, and it just sounds a
bit poor. g
-------------------- Ceilidh Band Scotland
Wedding Music Scotland
|
Urthlupe
member
Joined: 20/09/02
Posts: 380
Loc: West Midlands, UK
|
Re: Reversing compression for sampled music
[Re: Marky]
#45048 - 03/11/04 08:12 PM
|
|
|
Hi Greg That is an interesting question... I guess in theory you
could attempt compression with extremely fast attack and release, depending on the length
of the tremelo cycle that might work. The possibilities for problems however would be
legion. Maybe a look-ahead compressor might help. The other more 'mechanical'
approach that comes to mind would be to duplicate the track and move one forward in time
by half of the tremelo cycle. You may also unintentionally achieve some interesting
filtering effects that way! Thank you for the kind comments about my size by
the way  Loopy
|