Taming excessive peaks in an old recording

I am attempting to clean up a copy of a recording made back in 1966 of my family sitting around the kitchen table talking to the neighbors from across the street. The original recording was made on a reel to reel my father used to listen to talking books on ( my father was blind). The original tape is long gone but I do remember borrowing a deck from our sound guy and transferring the tape to cassette some twenty years ago.

I transferred the cassette recording into Sonar and have begun cleaning up the recording to leave as a legacy for my kids. There is no tape hiss to speak of, but there are many extreme peaks and mic overloads, especially when us kids got ahold of the mic.

This is my first foray into actually editing something badly recorded and I'm not sure what is the preferred tools to tame the loud peaks. I don't think there is much to do about the distortion in some spots.

I have read about normalization, but don't think that would apply here. I read an article that suggested using a limiter to clamp down the overloads. I have just started exploring that option with a brick wall type limiter that has a peak suppression preset.

Am I heading down the right road? Or can someone suggest better tools or methods?

Thanks,

Ardjunc

ardjunc
I am attempting to clean up a copy of a recording made back in 1966 of my family sitting around the kitchen table talking to the neighbors from across the street. The original recording was made on a reel to reel my father used to listen to talking books on ( my father was blind). The original tape is long gone but I do remember borrowing a deck from our sound guy and transferring the tape to cassette some twenty years ago.

I transferred the cassette recording into Sonar and have begun cleaning up the recording to leave as a legacy for my kids. There is no tape hiss to speak of, but there are many extreme peaks and mic overloads, especially when us kids got ahold of the mic.

This is my first foray into actually editing something badly recorded and I'm not sure what is the preferred tools to tame the loud peaks. I don't think there is much to do about the distortion in some spots.

I have read about normalization, but don't think that would apply here. I read an article that suggested using a limiter to clamp down the overloads. I have just started exploring that option with a brick wall type limiter that has a peak suppression preset.

Am I heading down the right road? Or can someone suggest better tools or methods?

Thanks,

Ardjunc

I'd download and use izotopes RX Pack (I think demo mode is good for 10 days) 
 
The clip tool will help make overloads sound much better if setup properly.   You may want to use a compressor AFTER that tool is applied to level out the mix, but you need to restore the audio first.  Using the limiter/compressor without this will only make the sections more uniform, not actually make the clipping sound any better. 
 
You mentioned not much noise to speak of, but you might also run the Dialog De-Noise plugin on it, it really works on dialog recordings. 

Other than cleaning up the occasional pops and clicks as the recorder was being turned off and on and perhaps some overall eq I am not convinced there is much to be done with the quality of the recording. Picture a cheap mic sitting on a table with several people talking then the kids come in and start competing for attention. The recording is fairly listenable as is, my goal is to level things out which is why I gravitated to compression first then apply any enhancements later.

I'll experiment with different types of compressors and settings. I have always recorded the best possible take in the first place so not much had to be done in the mix. This project will certainly provide invaluable experience in how to "fix it in the mix".

It never hurts to tackle the worst to learn the quickest.

Ardjunc

First a Wave editor is easier to work with for stereo files. Sonar does not allow for quick and easy drag mouse rendering. You have to jump through a few menus each edit so it's way slow. 
 
In a wave editor you can easily see the peaks. So you find them, zero in, and apply a negative gain to bring them closer to the rest of the material. you also can apply gain to the quiet parts. Your goal is to have the whole track "look" smoother. 
It's as easy as highlighting the sections and applying the tool needed. Sometimes, like for a POPPED P,  EQ can be applied instead of gain reduction. 
 
Once the track is under control you then apply normalization to bring it up close to 0. Then you check the RMS peak level and if it is say at -17 db you then apply a peak limiter or leveler with about 4 db to get the level around -13db or so. 
I use Wave Lab elements but there are many other wave editors, some free. I like Gold Wave. Sony Sound Forge is also real good and comes free with a Sony USB turntable these days! 

^^^ Almost correct. Anonymungus! is 100% correct insofar as you do want to get the levels as consistent as possible via automation, then apply compression for the final touch.
 
The problem is that volume envelopes are post-fx, so wildly-varying levels are just going to confuse a compressor and potentially make things worse.
 
What you need to automate is not volume, but gain. Unfortunately, SONAR does not support gain automation. However, you can easily achieve this using a gain plugin. Blue Cat Audio offers a free one in its Gain Suite. You'll most likely want to use the mono version, as I assume the recording is not stereo. 
 
If there are sections where the audio is actually clipped from people speaking too loudly into the microphone, then some sophisticated restoration software will be needed, and even that will likely only be partially successful.

John
I would think this is ideal for the CA 2A Leveling Amp.

Agreed - And maybe an expander/gate to clean up the low-level stuff.  

ardjunc
Other than cleaning up the occasional pops and clicks as the recorder was being turned off and on and perhaps some overall eq I am not convinced there is much to be done with the quality of the recording. Picture a cheap mic sitting on a table with several people talking then the kids come in and start competing for attention. The recording is fairly listenable as is, my goal is to level things out which is why I gravitated to compression first then apply any enhancements later.

I'll experiment with different types of compressors and settings. I have always recorded the best possible take in the first place so not much had to be done in the mix. This project will certainly provide invaluable experience in how to "fix it in the mix".

It never hurts to tackle the worst to learn the quickest.

Ardjunc

There were plenty of memorable recordings done with not much more.  The recordings of Robert Johnson are terrible by today's' standards, but inspired generations of blues and rock players.