filters produce or reveal new noise on tape

[citguy]

I have been processing "concert hall" tapes which have various degrees of hiss or hum on them. When applying filters, both hum and hiss, I am either "creating" or revealing low level noises (in spaces with no music content) that sound like burbbling water or "chipmunk chatter". I suspect the cause of this is the inablility to get a really "clean" profile from a tape that has ever present "room ambiance". In other words, the concert hall has a "sound of its own" especially with 3000 people present. Hardly a studio setting. I have had to resort to just using treble cut on some of these tapes. Sometimes I can "get away" with using "hum and treble" but the combination of "hum and hiss filter" is definitely not possible. I have considered the possiblity of noise being "generated" by the tape at the playback heads. I have demagnetized and cleaned without significant improvement although I have discovered that magnetized heads can generate "new noise" at playback.
Stan

[Derek]

Stan

The characteristics you describe are typical of a file that has been over-filtered. This happens, as you intimate, when the noise profile includes other sounds along with the tape hiss. You really need to search for a section of the tape where there is no extraneous sound apart from the hiss – and then to capture this as a new noise profile. It should then be possible to hiss filter without the unwanted burbling.

You can rest assured that there is nothing wrong with your tape player. The hiss filter works by selectively removing frequency components of the signal. If too much is removed, it has the adverse effect on the sound quality that you have described.

In any event, from your description of this recording, I would advise only using the hiss filter on its minimum setting (8dB reduction).

[citguy]

Thanks Derek. I have found that the results vary with the "engineer" who made the recording and the "mike placement" either high in the hall or close to the orchestra. I have also found that taking an extremely small profile (.1 sec) and expanding it, then cleaning any "bumps" out of it with "cut and paste" can sometimes give me a cleaner profile. Does the "profile" average all the content or just create a mask based on everything it finds? I believe "hall ambience" is not a static signal but a background that has it's own frequency albeit a very low one.
(I posted this reply to the wrong messeage before but deleted it)
Stan

[citguy]

Hello all. I have continued to struggle with the hiss filter as it applies to different tapes. I think I have correctly come to the conclusion that what I have been calling hiss (background hiss from a blank tape) is really hiss plus a certain amount of "open mike room ambiance" which varies from one recording to another. Derek warns about having "no recorded material present" when capturing a profile but I, maybe falsely, assumed dead space was the same as "no recorded material". The perplexing aspect of this is, on some tapes, "open mike dead space" can be used as a profile but on other tapes it produces an unuseable result. Anyone else experiencing the same?
Stan

[Glenn]

Hi Stan;
You wrote-

on some tapes, "open mike dead space" can be used as a profile but on other tapes it produces an unuseable result.

My experience extends to music recorded in a studio. Reverb or hall ambience must not be present in the sample as it 'tints' the recorded music. Any attempt to remove ambience will also affect the music. I tried extremely short samples before with some limited success. It leaves fewer artefacts but has less effect, particularly with long wavelengths. You can try repeated passes but a new sample needs to be made before each pass.
Glenn

[ramiller]

Hi Stan,

I have also noticed this burbling noise with the hiss filter. Its not unique to wave corrector--I have two other programs that also do noise sampling and filtering with about the same results with the exception that I can do less than -8db (which sometimes helps). I mostly convert lp's to cd, but have done a few open-reels and cassette's and have found this to help....

Get a noise profile from a blank, never recorded on tape. (an unbiased tape is usually quieter than an erased tape, the erase head, even on a blank tape can add 2 to 3 db noise... Sometimes a blank is actually noisier (rare) if it has been exposed to a magnetic field so be aware).

On a cassette use a high quality blank tape like tdk or maxell and record noise samples at 120 microseconds eq (normal) then 70 microseconds (chrome and metal) with dolby off, then repeat with dolbyB on. If you have an open-reel same applies but playback eq is 50 microseconds at 15 and 7.5ips and 90 at 3.75. Some late open-reels had EE settings (chrome tape) and eq is 35 at 7.5 and 50 at 3.75. The higher playback eq's give more hiss.

This will give a good noise profile for your tapedeck that will be added to anything you play on it. This works best if you always use same output level when you record. Also make sure everything is as clean as possible to get smooth tape travel. Different brands of tape will vary slightly, The quietest cassette tape is BASF true cromium dioxide according to a cross reference I once owned. Tdk and Maxell are cobalt treated ferric tape and about 2 db noisier (but can take about 3 db more above 0db therefore have a better sn ratio).

Append an appropiate profile to your wav file and use it for the filter profile. Usually this gives a more steady-uniform hiss that has less of the burbling noise, but I admit this doesn't always work--Tape scape on old worn tapes (especially open-reels) tends to be masked by the hiss, and brought out by the filter.

Randy

Randy