Stereo Tool

Hi,

It seems that the azimuth display is capable of showing between plus and minus 100 µm.

Is this the displacement in time at 1.875 in/s (4.76 cm/s) ? It makes sense, but I wanted to confirm.

Also, is this algorithm only dealing with time displacement or is it trying to reconstruct the original signal and remove the comb-filtering within the individual tracks?

Thanks,

Richard

Hell Richard,

If you use the version in Tape Restore Live it will also restore lost highs (that's the comb filter you're talking about, if it doesn't get TOO extreme anyway).

The version in Stereo Tool does not and cannot assume that shifts are caused by the tape hear for casette tapes or reel tapes, so it cannot restore anything. But, if you convert the result to mono no comb filter issues should occur because the channels are aligned.

The shift in um corresponds to the shift that a cassette tape head would have to cause the amount of offset that's measured. Cassette tapes run at 4.76 cm/s indeed so that's also part of the assumption!

Hi, Hans,

Thanks! But I am confuse. I see a graphic (octave) equalizer in Tape Restore Live. I do not see a special filter that adjusts for the in-channel comb filtering that occurs.

Comb filtering occurs either within one channel or between two channels. In the case of a cassette tape, each stereo channel is 0.6 mm wide and the guard band is 0.3 mm wide.

On an NAB two-track tape, the tracks are roughly 2 mm and the guard band is roughly 2 mm. On a DIN stereo tape, the tracks are roughly 2.75 mm and the guard band about 0.75 mm.

So, in the case of a cassette, the lowest null frequency of both channels of a stereo head is about 40 % of the null frequency of a single channel. In an NAB head, it's about 33 % and in a DIN head about 44 %.

The details of this are posted on Jay McKnights site in a Robert Morrison paper:
http://home.comcast.net/~mrltapes/morri ... -tapes.pdf

I was not looking for a manual filter, but it MIGHT be possible once one dials in the track configuration and the speed, to calculate a reverse comb filter from the time displacement. One would not want it to be too strong because the frequency notches are deep and one would not want to bring up the noise too much, but the notches are also narrow.

The Cedar paper shows a bit of that in Figure 6:
http://www.cedaraudio.com/intro/az_intro.html

My website shows the relative track configurations:
http://richardhess.com/notes/formats/ma ... cassettes/
http://richardhess.com/notes/formats/ma ... reel-tape/

(just providing this for reference--I'm sure you know it, but if you're trying to do the calcs, it might be useful).

THANKS!

See the 'Fix bias' setting below AZIMUTH, it contains two sliders that are set up correctly for cassette tapes, for other types of tapes you'll need to change the value. Unfortunately it's years ago that I made this thing and I don't remember the maths well enough to be able to tell you the correct settings...

Thanks! I guess that feature is not going to be added to StereoTool any time soon.

It doesn't really make sense in a (broadcast) audio processor because if you play music from different sources you would have to change the settings for each track... That's why it's not there in the first place.

I agree with you, but one can no longer assume cassette source material, either.

Since this tool has use outside of broadcast--and you may have one of the better azimuth correctors available -- certainly one of the most cost-effective -- I was hoping there might be a market for the higher-functionality azimuth corrector for speeds from 1.88 in/s to 30 in/s. It doesn't have to be free, just cost-effective and a VST plug-in.

In fact, it would be good if the azimuth and the compressor/equalizer/limiter/loudness corrector were a separate VST, I think. The loudness corrector plus the multi-band compressor really improves the audibility of oral history and other poorly recorded spoken word recordings.

Thanks for your consideration.

Richard