So far what I do is

Take the sound effect I recorded
Drop it in pro tools Cut up The audio so I have pieces of sounds I will keep Then I name the sound
Fade the sound on the back end
Add a compressor to the track Do gating and EQing if needed
Then I put a limiter on the main output track and thats it.

Is this the same process for you? Am I missing any thing? Is that the process you use to get the sound ready for your library?

  • Thanks everyone. This helps. So leave it raw until it gets used in a project. I need to learn this Metadata thing.
    – ShaunKelly
    Dec 5, 2010 at 17:42

7 Answers 7


I don't use compression, EQ or denoising as I feel that is presupposing the final use for the sound...

You also don't mention metadata editing & embedding

  • @Tim Prebble I got a fuzzy sound in this recording of a child talking. It was recorded at -22db because it was picking up that fuzz even at that level. So the audio track is normalized to -2db. You can here the fuzzy sound in the background. Would you gate that out before saving it to your library? here is the track soundcloud.com/shaunkelly/child-talking-03-norm-02-04
    – ShaunKelly
    Dec 5, 2010 at 0:38
  • I would never use gating... if a sound had a high background it would go into the library as is, and when it was used in context I would make a call on how to deal with the background. This would likely involve editing and ramping bg in/out, but again the issue of bg depends on the use eg if its a quiet part of a composite sound the bg may not even be noticeable... Re your example, 90% of production dialogue has far higher bg than that file
    – user49
    Dec 5, 2010 at 18:49

No, I don't do any of that. No fades, compression, gating or anything. FX go into my library untreated, as recorded, unless it needs a blatantly obvious notch or something. I agree with Michael, I think processing like that should be carried out in context rather than in isolation. With the quality of current restoration plugins such as RX2 being as good as it is, it's also tempting to go crazy cleaning up your recordings but you have to remind yourself that these plugins will be even better in years to come, so why commit your FX to today's technology?


Just a note from my experience with FX libraries: i think fading the head and tail of the file is a waste. I mean, i do my own fades on it (who doesn't?) when i bring it into my tracklay. And if it's an ambience that i need to loop, the parts of the track affected by the fade have to be chopped off anyway.

Also, i'd avoid any kind of dynamics processing (limiting/gating/compressing), as you can always do that in your tracklay if necessary. I'd only eq out any really ugly frequencies that negatively affect the nature of the effect.

I guess i'm mainly talking about ambiences here. For spot FX, i'm sure you could give it some eq or light dynamics to make it something you can instantly use; just make sure you keep an untreated version too. Not that i'm an expert; but these are my opinions so far.

  • i think the only reason to add a short fade in/fade out is to avoid the glitch of non-zero crossing, eg when auditioning sea tracks I really dont want to hear a glitch every time I audition every file....
    – user49
    Dec 5, 2010 at 19:02
  • @tim i was actually just thinking about that. Maybe editing the effect to make sure it begins at zero power would be a good idea. Of course, length is no big deal if you record your own. I just get frustrated by SI or HE ambiences that are 1 minute long! Dec 5, 2010 at 20:21

All I do is cut, fade and metadata...and I don't fade out tails of effects if I can help it. I prefer to leave the natural decay, if it's at all possible. If I have a noisy take, I'll usually toss it from the "cut" list (but I always keep the raws). I might do some minor levelling, but no compression or limiting. I'm of the same opinion as Tim. I don't know how I or someone else may use it in the future, so I'd rather leave it as is for greater flexibility.


I usually cut it into pieces, try to remove noise and EQ it. I sometimes use compression, but in 90 percent of all cases I dont compress it. I don`t have use any limiter on the putput but try to have peaks at -2dbFs.

As it comes to layer several (already edited) sounds into one sound, I use compressors, limiters,etc.


I usually don't use compressors unless I go for over the top style on my effect. Your chain seems pretty nice, a quick tip is if you like to mess more with the compressor you can try inserting the eq before the compressor and that way by messing with the eq you are also making the compressor play different. I guess you know that but thought to share the tip anyway.

I also use a nice feature to mix many sounds together called "free item positioning" in Reaper, search for it, its a time saver when it comes to many layers of sounds like different power levels of magic and stuff like that, when you want to bounce your regions in stems and that way do one bounce and get everything exported. Probably I can't explain it well, do a little search about the feature and I'm sure you gonna get it.


I've been known to eq if there are problems with the recording, but never affect dynamics. I also run everything through Izotope and will spectral repair out things like an intermittent squeak or bird chirp - but I'm pretty hard core about leaving blemishes in if the alternative leaves artifacts.

also, as mentioned above fades are pretty conterproductive.

When I'm building stylized effects I go nuts with layering, plugs ins and processing, but that's a different thing from prepping recordings made in the field or in the studio.

things not mentioned in the OP:

  • metadata incl mic, mic setup, literal descriptions, texture descriptions, codewords, project, etc
  • tagging with photos
  • folder structure

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.