Being a noob to Audacity and using the program simply to tweak and restore previously recorded audio, I realize I’m not worthy to participate on a forum with true audio specialists, but I hope someone can share their expertise and provide me with some direction in cleaning up a file.
I was recently invited to a private acoustic performance of my favorite recording artist through a radio station contest, however the guests were not allowed to bring in any recording equipment, video or audio. The only item I had in my possession to capture some of the performance was my iPhone. I used the “Voice Memo” function on my iPhone to record the show, which I realize was not the optimal device to capture the audio, but it was the only available option. I was left with a Mono 8000Hz .m4a file of the recording that I can import into Audacity 1.3 Beta, but after searching the forum for “concert,” “distortion,” and several other key words, I am still confused with how to proceed with the clean up.
The problem I’m having is that clearly, the measly mic on the iPhone was not capable of recording such a performance and I definitely overloaded the mic, giving the file reasonable distortion and clipping, especially during high vocal notes, low guitar chords, and squeals of the over-enthused crowd. I understand I cannot completely repair the problems, nor do I intend to; I simply would like the best result possible for my own enjoyment, but I am left perplexed by which effect I should use to resolve the difficulties and in which order I should apply them. I apologize if this subject was previously mentioned in a post, but I promise I tried to find an appropriate solution.
I always thought these things has a certain auto volume control, but this one didn’t.
The overload is permanent. There are tools to help overload and clipping, but not when a third of the performance is distortion. The tools will just give you different distortion.
Audacity 1.3 has an effect called Clip Fix. It’s intended to repair one high note every fifteen minutes.
The telephone sounding fidelity is caused by you recording it…on a telephone. I’m betting that’s permanent, too. Audacity 1.3 > Effect > Equalization. I don’t know that there’s a tutorial on that and it’s difficult to use. High sounds are on the right and low sounds are on the left. Click and push the blue line up to get louder down to get softer. You can add multiple click points.
Here’s an equalizer I designed to get rid of truck rumble in a recording. I picked each of those little white dots on the curve
Thanks for the suggestions Koz. I appreciate you taking the time to analyze it. I know the quality is poor and I’m realistic about my expectations; I’m not expecting any miracles, just hoping for “audibly tolerable.”
If that wasn’t depressing enough, nobody in the audience has ever walked away from a live performance with usable show and that includes hundreds of camcorders.
“My video sound is cracking and snapping at each one of the bass notes. How do I filter that out?”
The grownups and news people have special microphones and they provide additional circuits between the microphone and the camcorder to avoid overload. They, at least, will be missing the cracking and snapping. Their camcorders are also recording in much higher sound standards, so they can filter and edit later.
I think you’re in the position of being able to make it different, but not better.
I just wanted to let you know that the best solution I found to remove the distortion and most of the crackling on high notes and audience screams was the comb filter. I kept up my search long after reading your posts and continued to try different effects and plug-ins, and though it does give it a bit of a hollow sound, it subtracts most of the awful crackling from the overloaded mic.
Thanks again for taking the time to analyze my issue!
It was the David R. Sky Comb Filter that I used on it’s default settings. I also tried the Steve Harris Comb Filter from the LADSPA plug-in pack, but it gives the recording a weird echo/reverb effect that doesn’t help. Like I had initially wanted, the comb filter made the recording “tolerable”. While it took away about 95% of the crackling, the overall recording was left with a hollow, airy sound and in the areas that had clipped originally, most of the song and voice dynamics remained, but a “windy” noise was introduced. I did, in fact, try the Clip Fix effect you had mentioned also, but results were dismal. It actually added more of the crackling I was hoping to remove. Previous to the comb filter, I had equalized by adjusting the low end of the am radio curve to match your suggestion. To remove the hollowness, I equalized again, then finished with Chris’ Compressor. Not a perfect fix by any means, but considering the condition the original recording was in, I ended up with a somewhat decent result, at least to me.
You had mentioned the “telephone sounding fidelity” in one of the earlier posts. The lower frequencies don’t seem to be much of a problem, but there’s not much of anything beyond 3500Hz. Most recordings have frequencies that extend to approximately 20000Hz. I was wondering if there is a plug-in, similar to Steve Daulton’s Vocal Exciter that could introduce “harmonically related distortion to high frequencies to give more presence”, available for a 8000Hz, 16-bit PCM recording. Or would I have to change the project rate to 44100Hz or 48000Hz, export, then import again to use it? I’d like to give the recording a more “balanced” sound, if possible.
The maximum frequency that is possible in a recording is a bit less than half of the sample rate - so for a recording made at 8000Hz sample rate there will be no audio frequencies above 4000Hz - this is physics (Nyquist Theorem) and there is nothing you can do about that. In practice, the maximum frequency is somewhat lower than half the sample frequency, so as you have noticed, the high frequencies drop off rapidly above 3500Hz.
The vocal exciter plug-in can never put back the true high frequencies that were in the original sound, but it attempts to “make up” some high frequencies that are roughly similar to what may have been there. In order to do this, the sample rate needs to be high enough to allow those extra frequencies, and as we have already seen, 8000Hz sample rate just can not do that. So yes, before using a vocal exciter you will need to resample the track to a higher sample rate (for example 44100Hz) as described by Trebor.
“Resample”: boy, I feel dumb. Trebor, you saved me from several extra steps. Thanks.
Wow. First Koz, now Steve; I just have to say I am humbled to have my post reviewed by such brilliant minds. Before I initially posted, I examined many of both of your reviews and suggestions so I must say that you’ve been a tremendous help to so many people, especially me. Your combined understanding of all things audio have shed light on so many issues, so I really appreciate you taking the time to review mine.
Again, I’ll just reiterate my disclaimer that I am new at Audacity, so please forgive any exceedingly dumb questions I may have. Steve, thanks for the explanation of sample rates and frequencies; It all makes a lot more sense to me now. As far as the Wet/Dry settings on the Vocal Exciter, would I be correct in assuming that by adjusting the slider toward “Wet”, the levels of the effect are increased, in essence “saturating” the recording with more of the harmonic distortion? Also, as the creator of the plug-in, do you have any suggestions for an ideal setting for the slider that should be used in my situation or an estimation of how much distortion could or should be achieved?
I think the biggest challenge I’ve encountered with the plug-ins for Audacity is getting acclimated to the terminology. As a audio amateur, I would truly appreciate a dictionary of sorts to decipher some of the audio language used so I don’t feel as ridiculous when I have to come to the forum to ask the pros seemingly dumb questions. Sorry guys.
Your questions have been clear and well formed so please don’t feel embarrassed about being new to working with digital audio. Familiarity with the terminology will come as you gain experience and learn more (we are all learning - audio is one of those wonderful things where the more you know, the more you realise how much you don’t know).
Wet and Dry:
With many effects you will usually want a mix of the “effect” and the original unprocessed sound. The Wet/Dry mix is simply the proportion of processed sound (the effect) compared with the amount of unprocessed (original) sound. “Wet” refers to the processed sound and “Dry” refers to the original unprocessed sound. Making the mix more “Wet” means that there will be a greater proportion of the effect. Making the mix more “Dry” means less effect, more of the original unprocessed sound.
To be honest, I can’t even remember exactly what that effect does
I don’t appear to have a copy of it on this machine - if you could post the plug-in on the forum or post a link to where you got it from I can take a look. (to post the plug-in as an attachment onto the forum you may need to “zip” the file - alternatively you can copy and paste the code into a message). It’s likely to just be a matter of playing with it to see (hear) what sounds best.
Unfortunately this is not going to work. That effect takes existing high frequencies and creates additional (higher) harmonics from them. It assumes that the input sound already has some high frequency harmonics which it then “enhances”. In the case of a low sample rate recording, even if it has been resampled, there are no high frequency harmonics for the effect to work with.
The effect could be modified to operate on lower frequencies from the input sound, but that sounds horrible - like it has been recorded too loud and distorted.
