QTR-Quadtone RIP

Thanks for the Keywizard tip - sadly it's 300 quid in the UK and Gretag 
can't say for sure it will work...

So this is the problem, still:

It's a ED 100 densitometer made by EFI that I'm trying to get to work.

So far I've used ToolCrib from X-rite (free!) to test that the instrument is 
working and it does seem that it is.  Setting Toolcrib to look for 
"generic"  in the drop-down box and setting serial protocol to COM 1, 
9600 baud, pressing the green "READ" button on the densi returns a 
value to tool crib.

Toolcrib will auto-paste these values into Excel....  (looked like I was 
getting somewhere) like this:

MGI RD-200 000255 00095B 00077A 00301A
MGI RD-200 000256 0007DA 000C24 003A47
MGI RD-200 000257 0007DD 000C27 003A64
MGI RD-200 000258 0007DF 000C26 003A5B
MGI RD-200 000259 0007DE 000C25 003A5E
MGI RD-200 00025A 0007E0 000C49 003C5C
MGI RD-200 00025B 0007E0 000C30 003BC5
MGI RD-200 00025C 0007E4 000C3E 003C3B
MGI RD-200 00025D 0007DF 000C2A 003AA9

The first two blocks are always the same - device type?

The next block increments with each reading. Reading number I assume.

The next three blocks seem to be hex values - RGB ?
The do change when measurements are taken on different reflective 
surfaces and stay constant if successive readings are taken on the 
same surface.

Looks like it works ok as a generic serial device.  Here's the question 
finally -

Is there a prog or excel macro which will convert these raw readings to 
something useful for reading density?

Is there a (free or cheap) prog which will talk to this instrument as a 
generic serial densi and produce the same info?

Lots of info there, sorry, but it's probably best to set it all out rather than 
have to be asked.

Thanks
Andy

OK I know I'm answering my own posts!

> Toolcrib will auto-paste these values into Excel....  (looked like I was 
> getting somewhere) like this:
> 
> MGI RD-200 000255 

The useful part of this string are the values:
 
00095B          00077A              00301A

They are R-G-B as HEX values.

I can get all the measurements neatly and automatically into Excel, 
(each click on the densi pastes another reading into the sheet) then 
converted to decimal.  The lower the number the more reflective the 
surface.


How do I convert these  decimal RGB values into something that QTR 
will enjoy?

RGB to L ?  How to do that?

 I Have an IT8 target (Wolf Faust) which I might be able to use to check 
calibration...

Andy

What you need is the single Vis (density) reading rather than the RGB 
readings. On the X-rite 810 there is a button to toggle between the two 
modes. You can convert density readings to Lab L figures (although qtr 
understands density figures as well).

If there is no vis mode an average of the RGB readings _may_ suffice.

In the files section of this group there is a spreadsheet which contains 
formulas to convert between Lab L & density.

Howard

andywebsterus wrote:
> OK I know I'm answering my own posts!
> 
>> Toolcrib will auto-paste these values into Excel....  (looked like I was 
>> getting somewhere) like this:
>> 
>> MGI RD-200 000255 
> 
> The useful part of this string are the values:
>  
> 00095B          00077A              00301A
> 
> They are R-G-B as HEX values.
> 
> I can get all the measurements neatly and automatically into Excel, 
> (each click on the densi pastes another reading into the sheet) then 
> converted to decimal.  The lower the number the more reflective the 
> surface.
> 
> 
> How do I convert these  decimal RGB values into something that QTR 
> will enjoy?
> 
> RGB to L ?  How to do that?
> 
>  I Have an IT8 target (Wolf Faust) which I might be able to use to check 
> calibration...
> 
> Andy
> 
> 
> 
> 
> 
>  
> Yahoo! Groups Links
> 
> 
> 
> 

-- 
Howard Shaw
glassman@...
www.howardshaw.org

Andy

You can use MS Excel to convert hex string values to decimal numbers. Use
the function HEX2DEC. For example I converted the values below and got:

HEX: 00095B	00077A	00301A
Dec: 2395	1914	12314

Obviously you'll have to convert a few known patches (such as the Wolf
target) before you can figure out the units and scaling. If you end up using
excel, xrite has a free utility called xkey that will import keystrokes from
a serial port directly into an application such as Excel.

Whatever the units are, I imagine you'd find tools and formula at Bruce
Lindbloom's site:

http://www.brucelindbloom.com/index.html

Keep digging.

Tom Moore

> -----Original Message-----
> From: QuadtoneRIP@yahoogroups.com [mailto:QuadtoneRIP@yahoogroups.com] On
> Behalf Of andywebsterus
> Sent: Monday, January 22, 2007 3:34 PM
> To: QuadtoneRIP@yahoogroups.com
> Subject: [QuadtoneRIP] Re: EFI ED-100 - again
> 
> OK I know I'm answering my own posts!
> 
> > Toolcrib will auto-paste these values into Excel....  (looked like I was
> > getting somewhere) like this:
> >
> > MGI RD-200 000255
> 
> The useful part of this string are the values:
> 
> 00095B          00077A              00301A
> 
> They are R-G-B as HEX values.
> 
> I can get all the measurements neatly and automatically into Excel,
> (each click on the densi pastes another reading into the sheet) then
> converted to decimal.  The lower the number the more reflective the
> surface.
> 
> 
> How do I convert these  decimal RGB values into something that QTR
> will enjoy?
> 
> RGB to L ?  How to do that?
> 
>  I Have an IT8 target (Wolf Faust) which I might be able to use to check
> calibration...
> 
> Andy
> 
> 
> 
> 
> 
> 
> Yahoo! Groups Links
> 
> 
>

It looks like the device you have is reading RGB density. The x-rite
810 (that I have) also does this. In the case of the 810, the density
readings are performed according to a standard called Status-A density
readings. The 810 also will take Visual density readings. These are
the usual density values we refer to in BW printing. I have not been
able to find a way to mathematically relate RGB density to Visual
density. I tried simple averaging and various more complicated
correlation techniques. Nothing was satisfactory. I quit looking once
I realized that the Status-A RGB values bear no relationship to the
RGB values used by Photoshop, for example.

Your best bet would be to determine if there is a setting on your
densitometer for Visual density.

Tom Moore

--- In QuadtoneRIP@yahoogroups.com, "andywebsterus" <andy@...> wrote:
>
> OK I know I'm answering my own posts!
> 
> > Toolcrib will auto-paste these values into Excel....  (looked like
I was 
> > getting somewhere) like this:
> > 
> > MGI RD-200 000255 
> 
> The useful part of this string are the values:
>  
> 00095B          00077A              00301A
> 
> They are R-G-B as HEX values.
> 
> I can get all the measurements neatly and automatically into Excel, 
> (each click on the densi pastes another reading into the sheet) then 
> converted to decimal.  The lower the number the more reflective the 
> surface.
> 
> 
> How do I convert these  decimal RGB values into something that QTR 
> will enjoy?
> 
> RGB to L ?  How to do that?
> 
>  I Have an IT8 target (Wolf Faust) which I might be able to use to
check 

> calibration...
> 
> Andy
>

You can brute force calibrate the ED100 by using Wolf's IT8.
His data that comes with the target gives all the info you need.
The basic approach is this:

    1--Read the 23 step grayscale and note all three values for each
       step.
    2--Convert to decimal.
    3--Using the known RGB/CMY values from Wolf and the Y cie value
       along with the readings,use cubic spline interpolations to
       get the appropriate density numbers. You would use the Y cie     
       values and interpolation from the Green readings to get the
       proper values to first convert to Lab L* and then to d'log
       visible if that is what you wish.You can get the formulas
       for the Lab and d'log transforms online.

Note that if you plot the relationships of readings to known values 
you will see what looks somewhat like a geometric curve so the
instrument is not linear. Hence the cubic splines. To extrapolate each
extreme beyond the IT8 data you can use the slopes of the last and
first two points.This will offer reasonable accuracy in those areas.
The instrument is quite precise (with some deficits).

Now for the bad news. The densitometer is extremely sensitive to UV
induced fluorescence in papers so surrounding light can and will
influence readings at both extremes and can be severe. The only way I
found to overcome this is to cover the sample and instrument with a
dark cloth while making readings. It does have a very good dynamic
range so if you have the standards you can get very high density
readings with it. It also is very repeatable.

I have 2 of the ED100s and use them for B&W with QTR. They have proved
to be more accurate than the process densitometer I paid 100s of $ for
 some years ago.I would be very pleased if I could find a way to get
them to store readings for later transfer as they apparently are
capable of doing.

Hope this is of some help to you.

Regards
Duane






--- In QuadtoneRIP@yahoogroups.com, "Tom Moore" <r.t.moore@...> wrote:

>
> It looks like the device you have is reading RGB density. The x-rite
> 810 (that I have) also does this. In the case of the 810, the density
> readings are performed according to a standard called Status-A density
> readings. The 810 also will take Visual density readings. These are
> the usual density values we refer to in BW printing. I have not been
> able to find a way to mathematically relate RGB density to Visual
> density. I tried simple averaging and various more complicated
> correlation techniques. Nothing was satisfactory. I quit looking once
> I realized that the Status-A RGB values bear no relationship to the
> RGB values used by Photoshop, for example.
> 
> Your best bet would be to determine if there is a setting on your
> densitometer for Visual density.
> 
> Tom Moore
> 
> --- In QuadtoneRIP@yahoogroups.com, "andywebsterus" <andy@> wrote:
> >
> > OK I know I'm answering my own posts!
> > 
> > > Toolcrib will auto-paste these values into Excel....  (looked like
> I was 
> > > getting somewhere) like this:
> > > 
> > > MGI RD-200 000255 
> > 
> > The useful part of this string are the values:
> >  
> > 00095B          00077A              00301A
> > 
> > They are R-G-B as HEX values.
> > 
> > I can get all the measurements neatly and automatically into Excel, 
> > (each click on the densi pastes another reading into the sheet) then 
> > converted to decimal.  The lower the number the more reflective the 
> > surface.
> > 
> > 
> > How do I convert these  decimal RGB values into something that QTR 
> > will enjoy?
> > 
> > RGB to L ?  How to do that?
> > 
> >  I Have an IT8 target (Wolf Faust) which I might be able to use to
> check 
> > calibration...
> > 
> > Andy
> >
>