Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site clyde.UUCP
Path: utzoo!watmath!clyde!saf
From: saf@clyde.UUCP (Steve Falco)
Newsgroups: net.audio
Subject: Re: Cassette Head Alignment
Message-ID: <825@clyde.UUCP>
Date: Thu, 28-Mar-85 09:02:52 EST
Article-I.D.: clyde.825
Posted: Thu Mar 28 09:02:52 1985
Date-Received: Fri, 29-Mar-85 00:13:48 EST
References: <161@ucbcad.UUCP>
Distribution: net
Organization: AT&T Bell Laboratories, Whippany NJ
Lines: 57

> After all this, what is the problem?  Is the Nak correctly aligned
> and does that mean that auto-reverse decks are no good?  Or does
> it mean that this particular Nak deck is very poorly aligned?
> Or something else?

Some comments:  first, the tracks on a cassette are not interleaved they
way they are on reel-to-reel, so your pictures should have been:

/////////////////////            /////////////////////
/////////////////////            /////////////////////
\\\\\\\\\\\\\\\\\\\\\            /////////////////////
\\\\\\\\\\\\\\\\\\\\\            /////////////////////

Next, I think the thing that NAK worries about is that tape in a cassette deck
isn't guided very well.  In fact, there are a few tiny tabs of metal on
the heads for all the guide functions (shown as > and <):

  -------      -------
 >|     |      |     |<
  |     |      |     |
 >|     |      |     |<
  -------      -------

The tape can weave around a bunch so turning a tape over doesn't
guarantee it is still perpendicular.  Don't forget, tape isn't perfect,
the width varies a few thousanths yet the guides must be wide enough for
the widest sample of tape so it can skew around differently in one
direction than another.  Also, tape is allowed to be slightly warped
rather than straight:
                                               ------------------
-------------------------         -------------                  ----------


                                               ------------------
-------------------------         -------------                  -----------

	perfect				warped (exaggerated)

Altogether, you can see where the medium is not 100% transportable to
different machines.  Note, this gets much worse as the tape gets wider
because a given error angle has more width to diverge in:

narrow         wide (more divergence)

|   /              |          /
|  /               |         /
| /                |        /
|/                 |       /
		   |      /
		   |     /
		   |    /
		   |   /
		   |  /
		   | /
		   |/

	Steve Falco