Hi everyone,

My recordings are all noisy to some extent, but I want to say it's only preamp noise. Should I process my audio for NR just now and get rid of the raw file or should I leave that NR to appreciation for each and every case I'll need this particular effect in?

With the recording gear I have (Tascam DR-100, AT8015) I need to crank up the gain quite a bit to record at a reasonable level and unfortunately it's noissssssssssssy. Haven't had the chance to try with another mic yet but I'll do that asap!

But back to the NR. Should I keep that for later and let my sounds live with their noise?

  • 1
    Do not batch process Noise Reduction. If you do, at least make sure you are doing it to a copy of your library, and not the original. About your levels, your recorders mic preamp has an impedance of 1.2k and your Microphone has an impedance of 300 ohms. Jul 18, 2010 at 20:25
  • 1
    Let's talk about impedance then! I take it the combination I have is not ideal. What should I be looking for? Jul 18, 2010 at 22:47
  • 2
    If anyone is interested I'm reading microphone-data.com/pdfs/Mic%20impedance.pdf at the moment and I'm learning some more stuff! Jul 18, 2010 at 23:00
  • The rule of thumb seems to be a ratio of 1:10 for mic to mic-pre impedances. Jul 18, 2010 at 23:15

4 Answers 4


My recommendation is to keep the original files and also the mastered sounds in separated folders. I have a folder of raw recordings and also the respective folders on my personal library.

When I copy my recordings from the recorder's memory, I first rename all the files, then make a backup and work over the originals. Normally I first mix the channels if it's needed, then remove unwanted sounds, cut/add silences, fades, etc. All the commmon editing stuff. Then I use some EQ if it's needed, or noise reduction, depending on the sound. EQ can do a lot of corrective things if you know how to use it correctly.

You have to be careful with Noise Reduction. From my experience, it's better to start using some EQ, notch filtering, etc. If there's more specific noise, it's good to try with some noise reduction (mostly RX and Z-Noise in my case).

And the levels... For me it depends on the sound that is going to be mastered. The first recommendation I would give is to record with the final levels in mind, so you don't have to worry too much about it when mastering your recordings. Obviously it depends of the circumstances and the sounds recorded. Some sounds are fine on it's recording level, with other kind I try to get -0.5dB peaks, etc.

In this final step of mastering, expanders/limiters such as L2 or multi-band tools such as Waves C4 or McDSP ML400 are really great. That kind of tools are pretty versatile, both for noise reduction and expanding/limiting tasks. Of course it depends on whatever fits better for you and what exactly do you need. That's because is useful to keep the raw files, in case you want to go back and re-edit the files.

For me the main goal of mastering sfx is to get clean sounds. Just that, everything else such as adding puch, warmth, etc it's more for an editing/design process.

  • I didn't intend to use NR before EQ, it's just that in my case of white-ish noise EQ is not an option :( A lot of great advice though, thanks! Jul 19, 2010 at 9:02
  • Thanks, guys! And @Justin, yes. When you have some experience mastering your recordings, you know what to use from the beginning, so you start with the right tool. It's faster and you get the best results :-) Jul 19, 2010 at 15:04

Always keep the raw files. Hard drives are cheap, and you can go back and do it better if you learn better techniques in the future.

  • 1
    or the technology improves... Aug 25, 2010 at 18:29

I'd agree with the suggestions to keep the raw files, even if they're just backed up somewhere and the drive collects dust for years. I agree with Ryan. Typically, I'll hold off on noise reduction until the mix stage of production. In context, I may get some helpful masking going on (yes, masking can be helpful), which also allows me to keep the full spectral range of the sound in its natural state.

If you need to go for noise reduction at that stage, at least you know it's worth your effort. And for anyone who hasn't learned this yet, don't forget the golden rules - filters first, interpolation second, then broadband noise reduction.

[And just in case someone didn't know those rules, the second part is: filters (i.e. EQs, de-essers) get used primarily on harmonics (i.e. 50/60 cycle hum), interpolators (i.e. declickers) on clicks and pops, and broadband noise reduction is used on noise with wide spectral content (think white noise)]


@Justin I use to clean my audio files one by one, never use batch process without knowing what I'm doing in detail, so yes, de-noise them with a tool like iZotope RX or something like that, and of course keep the original files somewhere. Also try always to eliminate noise while you record by using different mic angles and other similar techniques.

  • @Panozk, obviously miking technique always comes first :) Apr 26, 2011 at 18:06

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.