Dynamic Impact Restorer Plug-in needed -Sound Forge - PC

Subject:Dynamic Impact Restorer Plug-in needed Posted by: plasmavideo Date:6/24/2003 12:10:26 PM Hi all, I'm looking for a DirectXplug-in for Sound Forge (or a way to do it with the existing SF plug-ins) to simulate the old analogue DBX boxes that provided "punch" for overly compressed and processed recordings to restore some dynamic realism. The DBX boxes improved transient response and did some fancy expansion and peak-unlimiting. I've tried to setup the standard dynamics settings in SF to provide some peak expansion but have not been able to come close to what I'm looking for. Any suggestions would be appreciated. Thanks Tom
Subject:RE: Dynamic Impact Restorer Plug-in needed Reply by: Geoff_Wood Date:6/24/2003 11:23:43 PM There are 'expander' DX plugins available. Try Googling"DX Plugin Expander' and you'll get heaps. geoff
Subject:RE: Dynamic Impact Restorer Plug-in needed Reply by: kilroy Date:6/25/2003 12:16:12 AM Sigh We get alot of these kind of files for mastering, mostly from home studios with amatuer engineers at the helm who thought it would be cool to smash the cahonies out of every last track in the mix in order to get a loud mix. It goes something like this... Client: "Can you make our record sound better?" Mastering Engineer: "Uh...in what way?" Client: "Well...it sounds kind of flat" ME: "Uh huh...sorta like a pancake, does it?" Client: "Huh...?" ME: "Kinda like there's no dynamics?" Client: "Erm...no whats?" ME: "No dynamics...you know, feel...vibe...that sort of thing." Client: "Oh, I dunno...maybe...it's like there's no punch or nothing. We recorded it with really great stuff but our performances didn't seem to translate the same as we play live." ME: "Yeah, you really gotta hate it when that happens...I know I do." Anyway, all fun aside...honestly, the only way we have had any "success" "saving" these kind of files is using multiband expansion techniques, probably in the area of four bands. I personally have not used any more than this because I have not been able to get useful results using any more than four, it just gets too weird. Well...a record with one whole db of dynamic range for the whole duration is arguably even wierder... As far as settings go there is virtually no way to determine what these might be without actually listening to the file, because you will be basing this all on the spectral balance and instrumental makeup/performance of each song. You should try to use very high quality phase linear processes for this kind of thing, other wise you will be very likely introducing significant phase shift artifacts, something that will show up far more easily in this particular case. Two good examples that we have used and like are from Waves and Samplitude, the latter only available in the full app. If it is me doing the work I usually end up going with the Samplitude algorithm. Hardware, even really good stuff, usually doesn't cut it for these kinds of tasks. Not surgical enough and you have no phase linearity so stuff gets really whacked out when you get extreme with it...and extreme is the watch word for the majority of these kind of jobs. I mean...if you only have a wee, wee bit of dynamics and you want even modest improvements you are going to have to kick booty and leave subtlety for good tracks. Sadly, in the end it's one of the crappiest jobs a mastering engineer could ever imagine having to do. My feeling is that once the punch is gone...it's gone. I've never once "rescued" songs by "uncompressing" to my satisfaction, and depending on the project I may even refuse it because it does not leave you with any way to properly represent what good mastering can do. My first response is always to explain what's gone wrong and see if they can possibly go back and rebuild the mix. It's just plain scary how many times that is not an option with some folks. Anyway, best of luck.
Subject:RE: Dynamic Impact Restorer Plug-in needed Reply by: plasmavideo Date:6/25/2003 1:12:18 PM Kilroy, Yep! I'm a former radio station chief engineer/dj who was always asked by the PD to "turn up the processing 'til yer ears bleed" and then was asked to "now make it sound good, there's no presence". Interestingly, the current project involves some old radio airchecks that were dubs of dubs of dubs with eq and compression apparently added at each dub! You can guess how they sound. Thanks - I'll continue playing and searching. Maybe I'll have to add DBX boxes on the analog in to the computer from the tape deck as a last resort. Tom

*Sigh* We get alot of these kind of files for mastering, mostly from home studios with amatuer engineers at the helm who thought it would be cool to smash the cahonies out of every last track in the mix in order to get a loud mix. It goes something like this...

Client: "Can you make our record sound better?"
Mastering Engineer: "Uh...in what way?"
Client: "Well...it sounds kind of flat"
ME: "Uh huh...sorta like a pancake, does it?"
Client: "Huh...?"
ME: "Kinda like there's no dynamics?"
Client: "Erm...no whats?"
ME: "No dynamics...you know, feel...vibe...that sort of thing."
Client: "Oh, I dunno...maybe...it's like there's no punch or nothing. We recorded it with really great stuff but our performances didn't seem to translate the same as we play live."
ME: "Yeah, you really gotta hate it when that happens...I know I do."

Anyway, all fun aside...honestly, the only way we have had any "success" "saving" these kind of files is using multiband expansion techniques, probably in the area of four bands. I personally have not used any more than this because I have not been able to get useful results using any more than four, it just gets too weird. Well...a record with one whole db of dynamic range for the whole duration is arguably even wierder... As far as settings go there is virtually no way to determine what these might be without actually listening to the file, because you will be basing this all on the spectral balance and instrumental makeup/performance of each song.

You should try to use very high quality phase linear processes for this kind of thing, other wise you will be very likely introducing significant phase shift artifacts, something that will show up far more easily in this particular case. Two good examples that we have used and like are from Waves and Samplitude, the latter only available in the full app. If it is me doing the work I usually end up going with the Samplitude algorithm.

Hardware, even really good stuff, usually doesn't cut it for these kinds of tasks. Not surgical enough and you have no phase linearity so stuff gets really whacked out when you get extreme with it...and extreme is the watch word for the majority of these kind of jobs. I mean...if you only have a wee, wee bit of dynamics and you want even modest improvements you are going to have to kick booty and leave subtlety for good tracks.

Sadly, in the end it's one of the crappiest jobs a mastering engineer could ever imagine having to do. My feeling is that once the punch is gone...it's gone. I've never once "rescued" songs by "uncompressing" to my satisfaction, and depending on the project I may even refuse it because it does not leave you with any way to properly represent what good mastering can do. My first response is always to explain what's gone wrong and see if they can possibly go back and rebuild the mix. It's just plain scary how many times that is not an option with some folks.

Anyway, best of luck.