Hi, the audio I want to cleanup sounds like there is interference in the microphone when the person talks… it only happens when the person talks so I can’t capture a noise print at all in Audition. Can I have someone listen to a clip and make suggestions? I have tried to figure this out on my own and just don’t know what to do to make it better and I can 't re-record it.
Thank you
Hi ristenK,
I didn’t reply to this post earlier because I don’t use Audition to clean/repair audio files, I use RX. You might find this helpful and my apologies if not…
Now it would appear that the audio in this example is exhibiting some of the signs of a lossy codec. I think the biggest issue is ‘Transient Smearing’, but then I haven’t listened to the complete audio file. It’s not a microphone issue but still not good news as you have already figured out.
However most of it can be ‘cleaned’ (to a degree) but be warned, in this case you will need to do almost all of the work by hand, that makes this very time consuming and not cost effective if you are getting paid for your services at a fixed/set price.
Some advice…
First of all make sure that the original file is safe and work only on a copy - this goes without saying but you’ll be surprised at how many people forget to make a copy!
If you are editing as well, I suggest you do that as your first step as there is no point spending time ‘fixing’ audio that is going to be thrown away later. But only edit approximately a single minutes worth in total together for now. Why will become clear later on.
One of the better strategies I’ve found with this type of problem is to just turn down the offending items to the point that they are just audible in the background and not remove them completely. Removing them leaves ‘holes’ in your audio which seems distract the listener easily. Fortunately the ear is fairly forgiving when these ‘artefacts’ are at low volumes, the brain focusses more on what it wants to hear and treats the items that have been turned down as background noise - ideal in this case.
Unfortunately you will need to identify all of the areas that are in need of repairs by ear although you should start to recognise some problems visually and then confirm the problem area by ear to be repaired.
When I’ve been handed a file like this I usually just edit the first minute together quickly, think ‘rough-cut’. Then time myself repairing that rough-cut minute to work out how long the whole audio file should take assuming that there are no other surprises. If there are any other issues you will again need to assess if you can repair the issues and how long they will take/cost to fix. Then I send the edited/repaired minute and if required, any other examples of issues/surprises to be dealt with back to the client for them to hear/approve and also warn them just how much it will cost to fix the whole file. Make sure you ask for a bit more on top to make it worth your while because repairing this type of thing gets very tedious very very quickly. The bit on top also gives you a little maneuvering room with the price if it’s too costly for the client.
Hope you get it sorted,
IanT.
Thank you Ian… I’m not sure how you are singling-out and turning the transient smearing down… do you just take the end of each small audio area that has the smear and turn the volume down?
Hi ristenK,
The answer to your question is yes.
If you want to do a better job (and if you have time) is to use a frequency and time based selection. You gain more control especially when transient smearing occurs in the middle or at the beginning of a word.
Be careful with the amount of processing applied because the audio will sound ‘chopped’ if it is overdone.
Remember, this isn’t a great solution and unfortunately doesn’t work in all cases. I just hope you are getting paid well for this…
IanT.