• Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.

should I extract the soundfile first for noise reduction or clean it while editing?

W

whytte

I'm not happy with the sound of the short film (interviewing people) because of:

1) I have used only the built in mike on the compact digital camera
2) my compact camera was not so close to the sound source (the person)
3) variety of unwanted background noises were present at the scene

It is somehow difficult to catch the words of the person I have interviewed at some scenes

Should I extract the sound file from the movie ( with something like Pazera Free Audio Extractor or any other programs you could name please? )

and use a noise reduction filter with Audacity (or what other programs ?)

Or run the noise reduction filters in the movie editors ? (like Sony Vegas? or any other you could name please?)

I have about 1-3 minute each x 8-10 interviews
 
Sort by date Sort by votes
I usually extract the audio and run it through a specialized program like Reaper before throwing it back into Premiere (or whichever program you use).

This can be hard to put back into the clip if you don't have a clap or sound spike in the audio to reference by...
If you DO extract it from the video, try not to cut the track 'coz you will have a hell of a time trying to make their voices match.
 
Upvote 0 Downvote
In a proper audio post the production the sound is always separate from the visuals. At the big-budget level the noise reduction takes place during mix down. At the low/mid budget indie level it's generally done during the dialog editing or during pre-mixes.

BNR (Broadband Noise Reduction) and EQ will not get rid of background sounds; it's geared towards hums, buzzes and rumbles. Programs like iZotope RX, the Waves NR package and the Cedar system - all probably beyond your budget - can sometimes do a creditable job, but nothing is a cure-all. You may want to try reducing the noise between spoken lines with volume automation.

I use a combination of plug-ins; BNR, iZotope RX, Soundsoap Pro, EQ and volume automation. It can take a long time even for me to get only marginal results on badly recorded production sound.

As the old computer dictum goes, "Garbage in, garbage out."
 
Upvote 0 Downvote
josh-ozirigpro said:
I usually extract the audio and run it through a specialized program like Reaper before throwing it back into Premiere (or whichever program you use).

This can be hard to put back into the clip if you don't have a clap or sound spike in the audio to reference by...
If you DO extract it from the video, try not to cut the track 'coz you will have a hell of a time trying to make their voices match.
Click to expand...

Thank you for sharing your experiences.

I have tested the Adobe Audition today and am kind of enthusiastic about it for the moment whereas I was hopelessly down when I heard that NR doesn't do much kind of comments

Thank you for pre warning about not to cut the track for matching with the video later on

I haven't heard the Reaper yet. I'll do a search on that too

all the best
 
Upvote 0 Downvote
Alcove Audio said:
BNR (Broadband Noise Reduction) and EQ will not get rid of background sounds; it's geared towards hums, buzzes and rumbles. Programs like iZotope RX, the Waves NR package and the Cedar system - all probably beyond your budget - can sometimes do a creditable job, but nothing is a cure-all. You may want to try reducing the noise between spoken lines with volume automation.

I use a combination of plug-ins; BNR, iZotope RX, Soundsoap Pro, EQ and volume automation. It can take a long time even for me to get only marginal results on badly recorded production sound.

As the old computer dictum goes, "Garbage in, garbage out."
Click to expand...

Thank you Alcova,
I totally agree with you on the "Garbage in, garbage out."

I'm also beginner amateur in the field so I guess I'll search for those plug ins and the programs you mentioned, for the moment

all the best
 
Last edited:
Upvote 0 Downvote
RX will help a lot but I'm afraid you don't have enough voice to work with on the tape.

I suggest re-shooting the interviews - it will save you time and thus money in the long-run to re-shoot with proper sound gear and getting it right - not to mention a better product.

No-one likes Space Monkeys (bi-product little noises behind the voice once you've de-noised it)
 
Upvote 0 Downvote
I always do my audio processing after the locked edit... that way, I'm only taking the time to scrub the audio you'll actually be using and not wasting CPU cycles (time) on the stuff that the audience will never hear anyway.
 
Upvote 0 Downvote
I feel much better today :) partly due to your very valuable insights into sound field and partly my positive trial with the Adobe Audition (which did not require much knowledge to do a basic experiment) turn out to be somehow promising results at few attempts, whereas I was very panicky and :( down earlier

Alcove Audio: I appreciate the links very much. I'll use them and print them to pdf for for my learningpedia collection

ROC: I may not be able to record some of them again, (I am also trying to avoid the voice-overs, if I could get away with.) On the other hand, this should be a lesson for me for future endeavors

knightly: that's an important pre warning I should keep in my mind. That'll not only save a lot of time, but will not take my initial enthusiasm away when I go along with your advice



Thank you to you, all, I really appreciate all your prompt and timely advice, I couldn't manage without
 
Upvote 0 Downvote
Alcove Audio said:
In a proper audio post the production the sound is always separate from the visuals. At the big-budget level the noise reduction takes place during mix down. At the low/mid budget indie level it's generally done during the dialog editing or during pre-mixes.

BNR (Broadband Noise Reduction) and EQ will not get rid of background sounds; it's geared towards hums, buzzes and rumbles. Programs like iZotope RX, the Waves NR package and the Cedar system - all probably beyond your budget - can sometimes do a creditable job, but nothing is a cure-all. You may want to try reducing the noise between spoken lines with volume automation.

I use a combination of plug-ins; BNR, iZotope RX, Soundsoap Pro, EQ and volume automation. It can take a long time even for me to get only marginal results on badly recorded production sound.

As the old computer dictum goes, "Garbage in, garbage out."
Click to expand...

I see there you are using iZotope RX. I haven't tried it myself, but have heard good things generally spoken about it and wondered for a while whether I should get it. I mainly use the Waves NR plugs. How do they compare in your opinion?
 
Upvote 0 Downvote
Alcove Audio said:
I haven't used the Waves, so I can't compare. But I really like the RX Advanced; far superior to SoundSoap Pro.
Click to expand...

OK thanks. I find the Waves stuff works well if you are careful to set the parameters oneself, rather than relying on presets. I have made my own presets as starting points. I have read on other forums where some have claimed that the iZotope RX has less artifacts than the Waves stuff. Not sure if it's true or just personal opinions though, which is why I was interested in your experiences.
 
Upvote 0 Downvote
I have made my own presets as starting points.
Click to expand...

As do we all.

A lot of the peeps I know who do audio post are very pleased with RX Pro, but when you reach their level it's just another tool in the arsenal. They'll have RX and Waves and SoundSoap - even duplicating on or two of them - all on one track. Each one does some things really well and others only fair.

Every one is drooling to get their hands on the "budget" version of Cedar.
 
Upvote 0 Downvote
You must log in or register to reply here.
Back
Top