I have a cd of an audio file recorded by a mini digital cassette recorder. I have rippd the file to a .wav on my hard drive and the quality of the audio is terrible, full of noise and warbling throughout the file. The warbling is so bad that you can barely make out what is being said, sometimes not at all. I am running CS5.5 Production Studio so I have access to Audition and Premiere, but I have not had any luck with either one yet. It's starting to look like there is nothing I can do to fix it since it was recorded in such poor quality. Any help is much appreciated.
Well it's pretty distorted and you've got modulated noise in it. You stand some sort of chance of reducing the background noise a bit, but none with the distortion, I'm afraid. Will reducing the noise improve the legibility? Perhaps a little, but I don't hold out much hope for this at all.
Nope, no luck improving the quality of the recording. I got ahold of the original source, a microcassette, and it was no better. I was holding out some hope that someone improperly digitized the recordings, but the problem was in the intitial recording. Thanks for the feedback.
I've been in this situation before and I've got about 4 minuets while my tracks mixdown in CS6 trial so here it goes...
You may not be able to improve the 'quality' but you may be able to improve the intelligability of those voices. You could start with noise reduction in CS6 by selecting a section of noise and then opening the 'Noise Reduction' effect. Set the 'Noise Print' options to the highest settings and then capture a noise print by clicking the 'Capture Noise Print' button. Then click the 'Select Entire File' button. Then it's just a matter of sliding the noise reduction % slider around to find a nice spot where the noise is tolerable around the voices. In this case I would say around 35%.
Then you can use hiss reduction on light to remove some of the nasty hiss. The next step would be a bit of equalization. I like to use the 'Mastering' effect for this because I can also use the 'Loudness Maximizer' to fix the amplitude of the signal. The EQ section is usually boosting the Low Shelf and the High Shelf by about 6dB. The Low frequency at about 260Hz and the High frequency at about 2400Hz. The mid or 'Peaking' I use to cut a bit of the center frequencies to eliminate some harshness... Usually the mid is about -3dB at about 400Hz.
That usually clears it up a bit but the settings usually need a bit of adjusting to get right. I suggest setting up the mastering rack after the initial noise reduction. It's a shame that the noise reduction plug-in can't be inserted into the mastering rack for this though. It runs so well on this computer even with the highest settings.
Hope that helps. Have to go now...
PS. Great job to the developers on CS6. It's amazing and blazing fast here. Mixdowns/Exports seem to be in real-time up until the last 30 seconds though.
Following on from what RonNovy said, I had a quick play and, indeed, was able to make it a bit more understandable.
I did the same on the NR as Ron suggested, then played some with EQ etc.
However, I got the best results in spectral view. First I selected and deleted all the upper frequencies in one fell swoop...but, after than, I got painstaking and "painted" around the obvious voice bits and deleted rubbish bit by bit.
Don't expect it to be in any way good quality--but, if it's important, you might at least understand more of the words.
Europe, Middle East and Africa