Strange artifacts after Noise Removal

[SaintNick]

I have cassette with a 15 minute recording of voice + guitar that I need to digitize.
It's quite possible this is a copy of a copy. I don't know.

After listening, I noticed after the end there were two layers of noise. The first noisy "silence" after the end of the performance, the second much quieter hiss of what must be the tape itself.

So I decided it would be clever to remove the noise under the recording in two steps: first remove the tape hiss, and then in a second step remove the remaining performance recording hiss. At each step analyzing the best settings for particular noise profile.

I studied and struggled with the four controls, which I do grasp when each explained, but I am not experienced enough to apply empirically.

The result is, whatever I do (unless using far too aggressive slider settings), I end up with with a strange kind of quiet but high pitched artificial singing background noise, in the place of the noise and hiss.

Does this ring any bells for anyone? It sort of does for me, but it's been so many years since I heard it.
Could this be the result of Dolby-C on the recording? Or due to the age of the tape, magnetic leakage from each loop of tape up or down to the next?

[Trebor]

I decided it would be clever to remove the noise under the recording in two steps: first remove the tape hiss, and then in a second step remove the remaining performance recording hiss.

One application of noise-reduction is sufficient : the "silence" after the performance is the noise-profile to use.
If the "much quieter" hiss of blank tape is present on the recording it will be included in the "silence".

The result is, whatever I do (unless using far too aggressive slider settings), I end up with with a strange kind of quiet but high pitched artificial singing background noise, in the place of the noise and hiss.

Each application of noise-reduction inevitably creates some digital processing artefacts.
The "high pitched artificial singing" you are describing could be the consequence of noise-reduction.
You can attach audio to posts in his forum, e.g. below ...

Dubya, before-after too-much noise-reduction (and not enough smoothing).mp3

(463.67 KiB) Downloaded 191 times

[sparksnsmoke]

You may have used TOO MUCH noise reduction,or your sample had other frequencies included that were not just tape hiss.

First, get a sample of only tape hiss.Sample an area and try it. You can get it from ANY area of the tape itself where absolutely no program material has been recorded.It would be wise to play that tape as loud as possible to hear the hiss.Remember to digitize the program material using the proper Dolby B,C etc AS YOU transfer to the PC.You are also adding more noise from the transfer via the Signal/Noise ratio of your particular sound card and tape deck.

For a free program,Audacity works reasonably well.There are others but have a much steeper learning curve to get approximately the same results.