Gear / Technical Help > Post-Processing, Computer / Streaming / Internet Devices & Related Activity

fixing levels over zero in soundforge

(1/1)

jerryfreak:
lets say you had a hot recording without clipping then you EQ'd it and it pushed the levels over zero before you rendered the final version

something like the left channel in this pic:



is there a simple limiter that can be applied to fix these without normalizing all data in the file?

nassau73:
I'm in no way familiar with Soundforge. However, if I'm understanding your issue correctly, would it not be possible to just reduce amplification before anything else?

jerryfreak:

--- Quote from: nassau73 on April 20, 2019, 10:54:06 PM ---I'm in no way familiar with Soundforge. However, if I'm understanding your issue correctly, would it not be possible to just reduce amplification before anything else?

--- End quote ---

compression/limiting would reduce those peaks locally - but can affect dynamics. im not super well versed in compression and avoid it where possible

normalizing to lower volume of all data can be done... but unneccesarily affects all data

i ended up starting from scratch andnormalizing a few dB before equing

next time i will keep levels at -6dB or below if i plan on EQing later

inoutoffocus:
give izotpe rx a shot. the declip (digital signal) process can work wonders. take a few minutes and your done.

checht:

--- Quote from: jerryfreak on April 21, 2019, 02:13:04 AM ---
--- Quote from: nassau73 on April 20, 2019, 10:54:06 PM ---I'm in no way familiar with Soundforge. However, if I'm understanding your issue correctly, would it not be possible to just reduce amplification before anything else?

--- End quote ---
[snip] normalizing to lower volume of all data can be done... but unnecessarily affects all data

--- End quote ---
Not super expert, but my understanding from folks on TS who are expert is that post-processing in a modern DAW has an insignificant impact, especially in a 32 bit space. I believe that normalizing down a bit and then eq would work well and the impact on all data not an issue.

Navigation

[0] Message Index

Go to full version