Improving sound - words for software

  • Thread starter Thread starter sebo106
  • Start date Start date
S

sebo106

New member
Hello,

This is my second post on this forum. In first one I found out what should I do to record a voice over in a decent quality. I already made some recordings with ZOOM H4, 96kHz, WAW. The recording consists of separate words, which will be cropped and stored as separate samples to use in a vocabulary teaching software.

I attach a sample of the recording which wasn’t processed anyhow or changed – apart from saving it to MP3 because of the large file size of WAWs.

here is the sample: View attachment sample Rec1.mp3

I am new to sound editing and I wouldn’t really want to know everything about it as I will only need it for this specific task – to make the words sound clear for the learner. I know that the most important to do is normalization and noice reduction but also the recording needs to be passed by some effects to sound properly.



I use Adobe Audition software

What I would ask you is advice for:

First of all which processes should I do since the beginning and what should be the sequence of them. What is most important - which filters should I use to make the higher tones up that it would sound more clearly, warmly and impressively than now. Generally – how to improve the recording. What would you suggest?

Secondly – the P’s explosion. As you can hear in the sample the word DROP has the explotion ( it happens when P is at the ends). I tried some advices from forum and used FFT filter  Kill the Mic Rumble but that doesn’t entirely remove the unpleasant effect. I wonder if there is a way to get rid of them or I would need to rerecord those words (it’d be nice to avoid rerecording ;-),

I will be very very grateful for any of your advice, suggestion or opinion. I am new to audio editing, and just want to learn how to improve recording in this specific task.
Hope you can help me with this issue,

Thanks in advance. Sebastian
 
Use of a pop filter in front of the mics will help with the splosives. Also try to angle slightly away from the mic. It is the air hitting the mic that is causing this. They are difficult to remove after recorded.
 
sebo106 said:
I know that the most important to do is normalization and noice reduction but also the recording needs to be passed by some effects to sound properly.
Click to expand...

Hi,
I never do either of those things.
I suppose normalising may be useful for what you're doing, but noise reduction shouldn't be necessary unless there's an underlying problem.
If there is, I'd address that first.

Jimmy's right there. A pop filter will help a lot with plosives.

You may want to use an equaliser to boost the high end a bit and make your samples clearer/brighter.
That's a matter of taste but an equaliser would be the right tool.
 
96k is overkill for your application. 44.1 or 48k would be perfectly adequate.

If you record well you shouldn't need any kind of noise reduction.

A pop filter will prevent plosives, but you can also edit them by applying a high-pass filter to just the spot with the boomy sound. Prevention is the better option.
 
I used pop filter for the recording but maybe the mic was too close to the reader. On next recording session I will be more aware of it and rise the distance.

In the sample you can see that there is some noise which gets high when I set the high tones up somehow. So I guess I HAVE to use a noce reduction for this recording. I also guess noise reduction decreses the quality of the whole recording, isn't it?
 
If it's noise introduced by the recording gear then fair enough.
If it's introduced acoustically then you'd be best to try and fix it at the source.

I haven't scrutinised your recording, but I'm thinking of things like computer fans, lights, etc.
 
Steenamaroo said:
If it's noise introduced by the recording gear then fair enough.
If it's introduced acoustically then you'd be best to try and fix it at the source.

I haven't scrutinised your recording, but I'm thinking of things like computer fans, lights, etc.
Click to expand...

ZOOM H4 is a portable device so the computer was off. there were no other electric sources of noise. The room was treated properly I suppose. You said about light - near the mic is a lamp which is supposed to light the text. Is it possible that it couses noise on the recording?
 
Hello again.
I would be really grateful if someone with some audio experience could check my sample and give some advice over it, like:

- removing the noice best technique with low quality loss,
- bringing up the high tones and improving the sound by giving it clear, bright sounding
- Also I'd need to reduce the frequency from 96 khz to 44,1. I tried to do it in Adobe Audition simply by saving it in WAW file with the reduced frequency and - I hear no differance at all! I was reading that the reduction can couse some unpleasant effect as 96 isn't multiplication od 44,1 and I expected some troubles. Has anyone got experience in such cases? Is it possible that there are no unwanted effects or am I neary deaf ? ;-)
- doeas anyone know how to split the file for separate words? Is there an way to make it automatic or should I make it manually. I don't see any automatic option in Adobe Audition, I tried many free dedicated small softwares but they aren't good for this task. Maybe someone would know how to make it faster? So far I select one word and save the selection as a new file,

I will be really grateful for your help.
 
1. Noise: if you don't have access to a fancy noise remover, try an expander with low ratio 1:1.5 or lower just to push the hiss down a little. Don't use a noise gate or it will only sound worse with the noise switching in and out.
2. Tone: EQ used carefully.
3. Sample rate reduction: It wouldn't surprise me if you got good results. I've had good results with an ancient Korg D1600 multitracker bringing 48 kHz down to 44.1 kHz sample rate and 96 kHz has a bunch more samples to work from so it should be better.
4. Splitting: For separate files for each word you'll probably have to do it manually. You'll need to make sure you leave out the mouth noises.

You will need to do some de-popping either with a side-chained compressor or a multiband limiter set to only affect the pops.

For this kind of work I don't think you need super fidelity. I've heard plenty of language tapes and podcasts that aren't as good as yours and they sold for real money. It's only the pops that bother me. For future recordings make sure the mic isn't directly in front of the vocalist so the pop blasts will miss the mic. The EQ could do with a bit of tweaking but don't make it too bright. Actually a gentle roll off in the low end would be the first thing I'd try.
 
It's been a while since I used it, but I recall that Cool Edit Pro (the predecessor to Audition) could do nondestructive editing in the multitrack view. Otherwise selecting and pasting to a new audio file may be the simplest approach.

The downside of downsampling is overstated.

My favorite de-popping process is to select the offending spot and apply a high pass filter at around 500Hz-2kHz.
 
to igi 616: thanks for your comprehensive reply. I'll check them in practice nxt week and share the result.

I wish a good weekend for you and all members!
 
You must log in or register to reply here.
Back
Top