[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: VMs: Re: Re: Inks and retouching

----- Original Message ----- 
From: "Nick Pelling" <nickpelling@xxxxxxxxxxxxxxx>
To: <vms-list@xxxxxxxxxxx>
Sent: Wednesday, July 28, 2004 4:08 PM
Subject: Re: VMs: Re: Re: Inks and retouching

> Hi GC,
> At 12:54 28/07/2004 -0600, GC wrote:
> >I'm at least happy you're using terms like "consistency" and
> >because that's exactly what we're seeing. The 8 in f3v.14.1 is
> >with your counter-observation above, as are the e's in f3v.8.4 and
> >One may identify specific features consistent with a specific subset, but
> >single subset of observation creates the entire image.

> Well: no, not really. As per f3v.14.1 / f3v.12.2 (the word-initial <Sho>)
> f3v.10.1 (the <oShos>, the original ink appears to be affected by the
colour of the vellum underneath (including bleedthrough), relatively
independently of the darkness of the ink.

I will grant this observation to a degree.  The question is how much is
pigment, and how much is perception - needs some study.

> Perhaps you're not seeing what I'm seeing with this glyph. Look again at
> the shape of the end of the word-final <a> in f3v.1.4 - as well as the
> stroke shapes, the stroke ends are downward curved, arched round. Correct
> me if I'm wrong, but this appears to have been written with a more rounded
> nib, whereas the rest of the letters on the page have been written with a
> flatter nib. Same author?

We're crossed up here.  I was directing my attention to f3v.1.4, and you are
speaking of f3v.1.3, ending in {o9}.  I've never recorded the {9} as an {a},
and wondering why, I went back over the British library images, the copyflo
and microfilm images, and the faint trace of a tail is evident in the black
and whites.  this is the normal shape a {9} takes, and if you look at
f3v.2.1, you also have a similar {9} missing just about all of its tail.

There's actually a long 'tail' to tell here, about the difficulties in
transcription and how I've fought these over the years.  In the absence of
better images that reveal the traces of these faded tails, I came up with an
a temporary glyph representation to identify the "curved" {a}, (maintaining
the detail to the best of my ability), and have kept these on the back
burner until such a time as they can be properly identified.  I remember
when the MrSids files first came out recently, that I remarked "not all o's
are o's, and not all a's are a's".  I was specifically referring to this one
feature, and the discovery through the MrSids files that many of these
"curved a's" still have no tails, even in the best imagery available.  This
one however, does have a tail, and is therefore a {9}.

> Anyway, perhaps we're simply missing the simply point about why the first
> part of the stroke making up EVA "n" is often heavier - that the writer
> pushed his quill upwards (down-right to up-left), before completing his
> stroke with a "plume". There are a good number of asymmetrical <Ch> pairs
> on this same page where the left <C-> half is heavy and the right <-h>
> is light - there's even a two-halved <o> (f3v.12.2). Perhaps for these
> pages, the scribe pushed his quill upwards in this way a lot of times, but
> wasn't yet experienced enough in writing Voynichese to control the flow of
> ink completely (it is only the start of the first quire, after all).

Curious suggestion.  I had rather thought that the scribe moved the pen in
the usual direction, but wanted a wider line outlay, so either repositioned
his quill or applied more pressure.  Possibly both.  Either way, we're
getting at the same thing - These are a function of the writing instrument,
and to that we may add the act of writing itself.

> >No word-final a's, is this a Voynich rule?  In the contiguous herbal
> >there are seven, and though that's not a large number, I am reminded that
> >for every Voynich "rule", there is something to be found that does not
> >follow the rules.  These examples are located at:
> >
> >f3v.1.4
> >f3v.8.5
> >f11v.4.8
> >f14r.2.5
> >f32v.3.9
> >f44v.8.7
> >f45v.9.6
> I'll go after these another day (it's late, it's hot...)

Is England hot this time of year?  And damp as well?  I wonder how this
would affect the writing of a manuscript - I mean I've often wondered if
someone could look at a page and get an idea of the temperature and weather,
and indication of the time of year.  Naw..... :-0

> >IMHO whoever wrote this 'a' wrote the others.  The real
> >question is - is there any evidence of retouching on this a?  The simple
> >fact that it's darker than surrounding text does not immediately imply
> >retouching, yet you've already jumped to this conclusion.
> Errm... so the wispy tail underneath it doesn't count as evidence, then?

See above, we were accidentally comparing different words.  I don't think
the faint tail counts as retouching.  You have a few on this page alone, and
hundreds throughout the manuscript.  I can give you a list at the push of a
button if you're really interested, but all you have to do is choose any
page and you'll find similar - if not completely matching  - examples.

> >The effects you describe from the bleed-through is not the case at all.
> >the dark background were the cause of the ink appearing darker, EVERY
> >inside the dark background would appear darker than those outside.  Only
> >heavier inked portions of the glyphs appear to be affected, not the
> >sections.

> No, plenty of lighter portions are affected as well. AFAICS, the ink isn't
> well-behaved enough to conclusively prove or disprove either of our
> theories on this page (a familiar story,*sigh*).

Well, I was filtering out the *apparent* darkness and trying to focus only
on only heavy ink outlays.

> >   This effect is relational to the pigmentation portion of the ink,
> >not to the sections where the ink is too thin to cover the bleed-through.
> >Where the pigment is too light or non-existent, this effect does not play
> >through.
> We're getting into the realms of psychophysics and perceived brightness
> here (back to retinex algorithms, again). Suffice to say that you have to
> be extremely careful when comparing brightnesses over a changing
> - the eye uses relative mechanisms for doing this, not absolute ones, and
> so what we see can be quite wrong, no matter how careful you are.

I'm glad you brought up this phenomenon, and I was somewhat wondering if
this holds true for black and white and greyscale imagery as well.  some of
us were weaned on copyflo, and I still maintain a scan of it on my computer.
I also have virtually every image that has ever been released of the
Voynich, as well as a couple that haven't.  F3v was probably one of the
first images I looked at and noticed this 'dark/light' phenomenon, and it
was probably one of the first images that cleared up most of the problem for
me, but it was in black and white.

Takahashi's scans of the microfilm are almost identical to copyflo images,
the copyflo being made from this microfilm.  In the microfilm image we see
the leaf background of the bleedthrough, and in many instances the glyphs go
dark at the border of the shadow.  In one case, a darker curving leaf makes
everything within it darker.  Compare this to the effects we see in the
MrSids images.  Here we can also see more clearly the 'grain' of the vellum,
and see how many of the darker glyphs sit within the heavy grain.  What we
don't see in this microfilm image is the faint outline of the red ink in the
bleed-through image, and how the moisture from its application affected the
text on this folio.  But there's good news for you, Nick - this does
demonstrate conclusively that the images were painted *after* the outlines
were drawn and the text was written.  :-)

> >Are there any "retouched" glyphs you see on f3v besides the 'a'?
> As I mentioned before, I'm suspicious about both word-final <-a>'s on line
> 1, and the <-o-> preceding the first one: but while I'm really not seeing
> retouching everywhere, I'm definitely seeing it in places.

One folio down, and only a couple hundred to go! :-)

> Incidentally, has anyone tried colour deconvolution on this page,
> specifically on the end of line 2? This finishes "okai", which kind of
> suggests that the rest of the word might be the "-n" finishing the next
> line (which would be out of place otherwise), or obscured the daubed blue
> paint.

Not likely if my understanding is correct, but I'd like to see the results,
even if they're negative.

> >I said that features can be categorized in subsets, and when this can be
> >done, the effect is systemic in nature, not systematic as retouching
> >be.  I also implied if not directly stated that there are several subsets
> >here, the above discussion being on only one of these.
> If you think that subsets can overlap (ie you can have a reinking splodge
> being affected by bleedthrough, etc), then I'd agree.

Then we're in agreement.  Some folios have little or no bleed-through, and
these are were the definitive subsets may be drawn.  Many effects on folios
like f3v are indeed overlapping, yes.

> >I look forward to the retouching evidence.  So far we have Jorge's 'm'.

> I think f1r has more than you admit to, f3v's top line has some, the ink
> stain on f93r has quite a few, and the "okeos" in f73v Jorge mentioned
> seems to be in a different hand. This is only the start of a list, but
> probably not the whole list... we shall see. :-)

I have personally disqualified f1r from the competition.  This is something
I did very early, before this discussion even began, and for more reasons
than you may know at present. There's writing all over this page, most of it
barely visible by now.  It's been chemically enhanced/degraded, and simply
closing the book on the page has allowed some of the chemical to interact in
vapour with the rest of the script.  It's written in four sections at four
separate times, and I would not doubt for a moment that there has been
alteration of this page by others, even if to doodle a retrace in deep
thought.  No one has touched f1v however, and very few pages exhibit signs
of external tampering.  It's almost as if f1r became a 'blog' where someone
wrote a comment, someone else cued in.  I'm just happy they restricted their
activities to this page.  The one major exception - there are early two
hands at work in the text on f116v apart from the text at the top - but I've
covered that already.

F3v's top line - we're not in agreement on this.  You found these
suspicious, but I didn't know that suspicion qualifies as proof.  I don't
see signs of retouching here.  If this is one of your candidates, you need
an image like Jorge's {m} that clearly delineates between original and
retouched text.

F73v - There is without a doubt two separate inks involved in this folio,
not just the <okeos>.  The second and fourth inward circles of the rotunda
are separately inked, and the breasts were clearly drawn (or retouched) in
the process.  But who can tell if this was done an hour apart or a century
apart?  Was this done by the author or someone else?  Again, pull the glyphs
from their positions, compare and examine the loops and significant features
of the writing against the rest. The probability is extremely high that this
was done by the author at a later time.  What you should get from this is
that these 'labels' were of some importance to the author, since he went to
the trouble to go back and add them.  A few 'breast strokes' in the process,
well.. as I said, it's kinda Freudian.

Ink stain on f93r - Nick, my laughing is waking the neighborhood, but I'm
trying to laugh loud enough for you to hear me all the way to England.  You
of all people should know how much time I have invested in this folio.  I
can recite both the script and the plaintext for this folio in my sleep!  I
know the four errors made on this page by the author, and I'd ask you to
correctly identify them.  You know I only release certain details of f78r
because I "own" f93r and protect that knowledge.  Bringing this folio into a
serious discussion on "retouching" shouldn't be a laughing matter however,
so plese parden my lak of edikette, and let me get serious about this folio.
Retouching is evident within the ink stain - and the logic would be, that
the ink stain somehow obliterated the glyphs so they were in need of
renewing?  NOT, NAY, never happened.

You remember my diatribe about how moisture causes the pigment to adhere in
places?  Virtually every glyph within the boundary of this 'stain' suggests
that this observation is true.  The ink throughout this folio has remained
rather consistent with age, so we get a clear picture.  The glyphs that
change dark-to-light and vice versa as they pass the borders of the stain
are too numerous to mention, but if you can't gain an overall sense by
looking at the image, we'll go through them one by one.

Your idea that the darker background enhances the brightness/darkness of the
glyph in question, and therefore 'skews' the visual results, has to be
checked against the final {m} in f93r.7.4, as example, which apparently
lacked the pigment outlay necessary for full adherance.  Examine the
single-letter glyph at f93v.29.5 for an exemplary indication of the
correctness of my assumption about the pigmentation and its adherance
qualities as relates to moisture.  This folio is full of text-book examples
of exactly what I've been describing.  Void the color and return to the
microfilm, and you get a sense of the 'grain' of the vellum and its specific
effects on pigment adherance in areas unaffected by the 'stain'. Applying my
theory, we can deduce that this 'stain' was administered after the writing
and before most of the pigment had worn off by use.  Remember - the pigment
is particulate, not part of the base, as I see it - and the organic base did
not react well with the particulate pigment.  This suggests that the pigment
was not organically based, or if it was, not well matched to the ink base.
In any event, the particulate reacts to moisture, and is therefore
water-soluble, while the base does not appear to react to water or 'stains'
such as we see on f93r.

And if I might be so impolite as to suggest that the term "ink stain" may be
inaccurate?  Anyone knows that the only way to remove certain inks is by the
application of an identical ink, and then a rubbing or application of a
different chemical while it's wet.  What?  You've never laundered a check
your life?  Get REAL! :-)  If this stain were caused by a similar ink, it
would have eroded the base, which has not occurred in this instance.  In
fact, this stain demonstrates identical qualities to those of water color
bleed-throughs, suggesting that this stain was water-based, not ink-based.
English tea, anyone? :-)


> ______________________________________________________________________
> To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
> unsubscribe vms-list

To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list