The Voynich Ninja

Must admit I didn't explain my way to find the "M" amongst VMS letters:
as I am convinced it is just simplified Kyrillic, I took the freedom of simplifying the most obvious kyrillic letter by myself:



Cutting off the half of M and adding a loop lead to the write-in-one-strike feature nearly all Voynichese characters have and resulted in the well-known letter k




Thereby, okeos  would read by my method ( You are not allowed to view links. Register or Login to view.p , 1st comment ) as (е)моет / (e)moet, which translates to моет = "washing" in russian and quite similar word stems in ukrainian and belarussian with the same meanings (мити, мыць).

"Washing" is not quite satisfying since the debate about soap, so I took care of the other word here: okoroeey , which may read
as "Emeseuya" or as well as "Emesiua" (емесеыя / емесиыа), as the use of o still is not fixed to "e" or "i" (not really necessary) and y may also be used as the ending instance of a. All "foreign" word findings and names within VMS - until now - had a slightly strange, but useful ending in letters, so this was enough to find "Emesis" as best translation.
You are not allowed to view links. Register or Login to view.

Emesis was considered as one of the basic health treatments during the middleages across whole europe:
to invoke this you will soon find the suggestion of Phytolacca acinosa ( You are not allowed to view links. Register or Login to view. ), 



If this darkblue, comb-like thing of VMS is part of a plant at all, it seems hard to recognize it in images of blossoms; the root has an interesting form, but also a clear pink apperarance -- as always, the plant and herbal drafts´comparisons alone are not very convincing.


But among the suggested emesis treatments very soon this wonderful chrystal is also named:



It is copper(II)sulfate, indeed also known as blue vitriol.
This was medically used for exact the "Emesis" purpose in China already ~3,000BC(!) and was known also in europe's ancient times for that function and main material in medieval era.
This stuff could be found, mined, grown or just traded by someone who could.

Thereby, the "washing" description doesn`t have anything to do with soap use, but means the `cleaning´ of upper digestive tract with rude methods which I assume are summarized here in the blue category. This would explain best why a single blue "block" appears very uniquely in the herbs´ section at all.

That allows 2 conclusions:
- the short 1-word comments at most herbal parts are generally not their names, but a description of their functions or similar attributes
- these listings appear to be inventories for the (mostly different) cans they are aligned to.

I´ve suspected for a longer time that the cans are shown with packing lists and each´s content in different categories: 
trained people could find the right treatment by 1 word only, and if this content is limited to the few shown herbs, it would not be necessary to give too detailed drawings as some "blue chrystal" or "blueish plant part" can easily be distinguished from some other things in the same can. 
Use may be in a pharmacy or, well-packed, during travels.

And even if both translations turn out to be completely wrong, coppersulfate is still good alternative to explain this detail here in VMS.

Yes it is vitriol, as I previously stated. 
But the word has nothing to do with "моет".
br: Vessy

Let's keep this on topic.

(for context You are not allowed to view links. Register or Login to view. )

I've added an MSI viewer that can be used to examine split channel views of all publicly available MSI images. The whole thing was "vibe coded" using Anthropic Claude, my job was mostly cursing, testing and uploading to Cloudflare. 

You can click on a point of interest of the main image to zoom in to the split channel view and then you can click on channels to view them individually. Loading channel images might take a while. I only included narrow-band channels, normally they are the most interesting. I couldn't include all the channels, since Cloudflare has a hard limit on the number of files in a project.

UV is upper left (and usually the most interesting one when looking for faint writing), IR is lower right.

If you want to reference a particular piece, you can create a permanent link via "Copy link" button. This should work for split channel images and for individual channel images.

f1r: You are not allowed to view links. Register or Login to view.
f8r: You are not allowed to view links. Register or Login to view.
f17r: You are not allowed to view links. Register or Login to view.
f26r: You are not allowed to view links. Register or Login to view.
f47r: You are not allowed to view links. Register or Login to view.
f70v1: You are not allowed to view links. Register or Login to view.
f71r: You are not allowed to view links. Register or Login to view.
f93r: You are not allowed to view links. Register or Login to view.
f102v1: You are not allowed to view links. Register or Login to view.
f116v: You are not allowed to view links. Register or Login to view.

---

(BTW, I never noticed the weird way 3 is written in 93 from 93r: You are not allowed to view links. Register or Login to view.

Was this a common way of writing 3, with an extra section at the top?)

I added some basic navigation to the single channel view. It's a bit glitchy (the navigation won't initially work if the view is accessed via a permanent link), but it looks like this level of complexity is already challenging for Claude, so I'll leave it as it is.

(07-06-2025, 10:59 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.I've added an MSI viewer that can be used to examine split channel views of all publicly available MSI images.

Thanks a lot for this very valuable tool!

The image names seem to be "MB", then the (dominant?) wavelength in nanometers, then a two-letter mnemonic for the channel, then a digit that indicates the direction of the light.  So all images that end with the same digit could be compared pixel by pixel, whereas those with different digits like MB450RB0 and MB450RB1 are affected by the relief and bending of the vellum. 

Some channel mnemonics like "UV", "IR", "GN", "RD" are obvious.  The "RB" and "LB" are hues of blue; perhaps "royal blue" and "light blue"?   "CN" could be "cyan".  No idea of what "AM" means, but it is some shade of yellow.  Which is "amarillo" in Spanish, but...

I am looking at those images, but I am not sure yet about what it is that I am seeing...

All the best, --jorge

Thank you, I only understood the wavelengths. Instead of 0, 1, 2 at the end the original images have some two-digit numbers, different for different sets. I had no idea what they meant, but I had to standardize the names to simplify the task, so I just replaced them with sequential numbers. I think you are right and it has something to do with the light source, it never occurred to me before. As far as I understand, for a rough surface changing the direction and the apparent size of the light source would highlight different details, this can be very useful.

The original TIFFs are 16bits per channel, but only something like 11 bits contain useful information (the range of values is from 0 to 2^11 - 1), if I remember it correctly. The processing code takes the cropped part, sets the darkest pixel to black and the brightest to white, applies gamma of 2.0 and then linearly squeezes the result into 8 bits. This normalization works fine for most images, but for the edges of the folios when parts of the crop contain bright or black objects from the area around the book, the page itself could become low contrast.

Also I had some success with producing enhanced images using a linear combination of TIFFs with weights obtained via linear regression from various points of interest.

In this thread there are some results for f1r:
You are not allowed to view links. Register or Login to view.

In this thread some information about what I believe could be an almost invisible signature like feature on f116v:
You are not allowed to view links. Register or Login to view.

(08-06-2025, 11:09 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.  
No idea of what "AM" means, but it is some shade of yellow.  Which is "amarillo" in Spanish, but...

Oops, "AM" must be "amber", meaning  between yellow and orange.  It was one of the standard colors of early LEDs, and is the official color or the "yellow" traffic lights.

I suppose the images with "MB" prefix use light of "Medium Bandwitdh", while the "WB" ones
use light with "Wide Band" spectrum.  The "TX" prefix apparently means transmitted light.

It should be fun to explore the latter.  

One weird thing I already noticed on TX870IR_045_F and TX625RD_042_F for page You are not allowed to view links. Register or Login to view. (Aries Light) is that some stars that are filled with the light yellow paint are more transparent to those bands (red and infrared) than the blank vellum!  For instance, the star in the outer band at 08:30.

The only explanation I can think of is that the binder used in that yellow paint, which must be transparent at those wavelengths, fills the cavities and pores of the vellum so that it scatters less and hence transmits more.  Like paper becomes translucent when stained with oil or wax.

All the best, --jorge

(10-06-2025, 08:17 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.One weird thing I already noticed on TX870IR_045_F and TX625RD_042_F for page You are not allowed to view links. Register or Login to view. (Aries Light) is that some stars that are filled with the light yellow paint are more transparent to those bands (red and infrared) than the blank vellum!  For instance, the star in the outer band at 08:30.

The only explanation I can think of is that the binder used in that yellow paint, which must be transparent at those wavelengths, fills the cavities and pores of the vellum so that it scatters less and hence transmits more.  Like paper becomes translucent when stained with oil or wax.

The folios can become translucent. There has been an interesting example recently, when it turned out that a faint label that even made it into a transliteration file was just the mirror image of a perfectly normal label showing though from the back of the page: You are not allowed to view links. Register or Login to view.