I probably should clarify...

Think of the azimuth error situation as a right triangle.

The "perfect" azimuth is one leg that is vertical. The actual recorded 
azimuth is the hypotenuse. The azimuth error is the leg which is horizontal.

The length of the horizontal leg of the triangle can be translated into 
the time delay and the phase shift between the two channels.

When the two channels are summed, even small errors which CAN be 
equalized out, grow into huge errors and cannot be recovered, so the 
re-timing between stereo channels prior to summing will produce far 
better results than not using it and summing the two channels will 
reduce the noise taken from either channel.

For high-quality music work, using a full-track head is superior, but 
for oral history production work (especially from cassettes) there is 
merit to summing the two tracks, especially because mono cassette 
machines are hard to find and few (if any) were made with the quality of 
the top stereo machines.

Here is an example using a cassette:

15 arc minutes (1/4 of a degree) azimuth error.

Single channel  -3 dB point ~  9 000 Hz
Summed channels -3 dB point ~  3 400 Hz
Single channel  -6 dB point ~ 12 000 Hz
Summed channels -6 dB point ~  4 600 Hz

Sliding the second channel in time to be in alignment with the first 
provides a huge improvement in the high-frequency response.

I contend that, if desired, losses of up to approximately 6 dB can be 
equalized out as long as there is no additional boost and a rolloff 
after the - 6 dB point (to avoid increasing noise).

Cheers,

Richard



On 2014-09-10 10:25 AM, Richard L. Hess wrote:
> Hi, Andrew and Tom,
>
> I am (very) slowly working towards a paper on azimuth.
>
> The full-track mono tape is, of course, the hardest to play from an
> azimuth perspective (of the common formats)--I hate to think about the
> 1-inch two-track!
>
> Transport guidance is a huge issue and I have found that the Studer A80
> is magical in that regard--and the Studer A810, not so much. The Sony
> APR-5000 is between the two.
>
> For those of us who have taught ourselves how to adjust azimuth by ear
> (sort of the same thing as focusing a camera lens in many ways), we do
> not need crutches, but I have come to realize that using a two-track
> head for "factory workers" might be beneficial.
>
> For one project I proposed (but we chose not to implement) a wide/narrow
> head for full track mono. We batted around a few different
> configurations--including a long discussion with Greg Orton. I was
> thinking of something like 0.120 and 0.04. The nice thing is even if
> things go south, you still have a good percentage of the highs on the
> 0.04 track, albeit noisy.
>
> With that said, for oral history cassettes, I use, in addition to manual
> azimuth adjustment, the azimuth compensation feature of
> www.stereotool.com. This allows excellent channel summing for improved
> noise, assuming both channels were recorded.
>
> There is a similar feature in iZotope RX Advanced.
>
> Cheers,
>
> Richard
>

-- 
Richard L. Hess                   email: [log in to unmask]
Aurora, Ontario, Canada                             647 479 2800
http://www.richardhess.com/tape/contact.htm
Quality tape transfers -- even from hard-to-play tapes.