When you say dream, was there any tick pop or massive tracking error or was it simply transparent?
I was basing what I wrote on this 1993 Scott Dorsey DATHeads post:
From: Scott Dorsey <oldaddress[AT]blah.blah.blah.nasa.gov>
Subject: DAT splicing
Date: Thu, 22 Apr 93 09:10:37 EDT
Yes, I have made a DAT spliceblock, by milling a 4mm groove down a
conventional spliceblock. Any good precision machine shop should be
able to do such a thing for you, and shouldn't charge more than a
couple of bucks. The problem is getting the splicing tape.
I found the easiest way was to apply 1/2" tape perpendicular to the
tape itself, then trim the edges. If anyone out there remembers the
old Gibson Girl editors, this is pretty much the same idea, although
it's a bit difficult to get it into the splice block. You'll get
some trash at the edit, though.
The trash is the result of two factors, the first being the thickness
of conventional splicing tape. It's going to slow down the tape as
it passes through the machine and really produce a lot of errors. The
other factor is the fact that you are making a 45' cut through a tape
with a much more acute helical scan. It's possible to make a splicing
block whose angle matches the tape scan (does anyone out there remember
editing 2" quad videotape by hand?), but it's not easy to make, not easy
to use, and not really needed for emergency work anyway, which is all
I would recommend using the thing for.
--scott