How to improve the quality of certain sound files added by mods

[bajs11]

There are quite a few very good mods that also added custom voice files.

The problem is that some of them sound like they were recorded by low quality microphones so I am wondering if there are ways to improve them.

If it's possible then which software should I use.

[Crosstieger]

You can't just magically improve an audio file if the source itself is of bad quality. Your PC can do a lot but not magic.

 

Either the author did a terrible job at compressing the audio or just has a terrible mic. Only the author can fix this.

[montky]

while I acknowledge Crosstieger's sentiments,

it often is much quicker and easier to re-record if that is an option,

I recommend speaking to folks from the Audio thread here at NMM,

and at ThisWeekInTech, AVClub fora and elsewhere,

as well as SoundCloud and AskFM.

there are many sound engineers and sound-enthusiasts to be sought.

 

 

 

bottom line, you can use some pluggins with Audacity,

though Cubase has some very nice isolation inverse-wave math pluggins,

and AGI-GAH wave-form stuff.

You can then add reverb, SC4, hard-limiter and other option filters,

to isolate some sound.

 

you can then 'sketch-layering' to fix it or improve the listenability of the audio.

you can reverb-cue for the node in-game too,

so you can achieve neuromorphic, anagogic and cacophonous-dischordant diverging stuff,

triaxlating, and fade from white-noise/random etc.

 

Restoring sound and up-resolutioning the audio from various live-performances,

that is a hobby of mine for a while now.

specifically, sound-from-hologram, sound-from-qDotler, and sound from image.

I was intrigued by it after people pointed to things such as "can you hear the shape of a drum?"

and explained how to inverse-isolate different sounds within a sound.

----

some variants and performances are exceedingly rare,

and often only one or two recordings exist, period.

so, don't let folks say its not a hobby or a thing hehe.

 

However, if you are recording things like instruments/orchestras, and it's "red-lined"

or reverberation from the microphone casing itself,

that is very difficult to isolate and inverse-cancel.

 

 

[ehtyeci]

There are a few things you can do, but they all depend on the source files, the nature of the issues and how proficient you are using audio tools and a DAW like Cubase, PT, Ableton Live or Audacity (free).

Badly normalized audio with high floor-to-ceiling dynamics can be tamed using a compressor (no, not Winrar). All professionally recorded voice audio use some degree of signal compression anywhere between 1:1 to 2:1 with fast releases for volume consistency without perceivable pumps or dips. The actual values are arbitrary and depends on both the source and target genre.

All mics record with a signature boost or scoop somewhere in the frequency spectrum. Laptop mics severely lack in the lows and sound very high-shelved with a boost in 5kHz to 8kHz area. You can normalize the spectrum using a parametric EQ to take out peaks that are obvious and boost low-end using a low-shelf. Always apply low-cut and high-cut filters on top of any EQ shelving.

If the source audio has background noise between each line, you can use a noise gate that automatically tries to lower the noise. However it can't lower the noise while the person speaks. This is because the gate works on signal threshold, and at a certain threshold it closes the throughput, and only brings it back once the signal reaches a given value. You can also use the noise gate to take out acoustics from an audio caused by untreated rooms, but again, it only works after the person has spoken.

If you are proficient with a DAW, you may use these tools and automate them to produce different values for different times to further tailor changes. This is very time consuming and requires the technical know-how. Actually, all of these tools require some degree of audio engineer's ear and fundamental understanding of the tools involved.

TLDR;

You can add compressors to increase loudness and tame peaks of the wavefile.

You can add EQ to shape the spectrum.

You can add noise gates, compressors or inverted expanders to control the noise floor or acoustics.

All of these require experience and knowledge to use for profitable results.