HomeRoast Digest

Topic: Agtron (can of worms) > RE: +1st crack and city roast (9 msgs / 277 lines)
1) From: miKe mcKoffee
Ok Ok. To truly discuss roasts levels on an even scientific playing field
there is an industry standard method and at it's best requires an Agtron
spectrophotometer. For quite a bit "cheaper" the SCAA came out with a set of
8 color tiles used to compare ground samples for only abour $300.
Ok, who's gonna suggest using paint sample swags again:-)
Kona Konnaisseur miKe mcKoffee
URL to Rosto mods, FrankenFormer, some recipes etc:http://mdmint.home.comcast.net/coffee/Rosto_mod.htmUltimately the quest for Koffee Nirvana is a solitary path. To know I must
first not know. And in knowing know I know not. Each Personal enlightenment
found exploring the many divergent foot steps of Those who have gone before.
	From: homeroast-admin
[mailto:homeroast-admin] On Behalf Of Scott Marquardt
	Sent: Thursday, January 18, 2007 12:27 PM
	I think that's more of a piece of practical advice for home roasters
learning about the process (who also may not be likely to measure temps
reliably) than it is a relegation of temperature to an irrelevant status for
roasting. And the other reason is probably 'cause he knows his temps are
off.   ;-) 
	But to obviate argument, "off" is a matter of degree. I'm pretty
cocky about my measurements, but I know they could be better. So I'm humble
about my cockiness.    ;-D

2) From: raymanowen
On 1/18/07, miKe mcKoffee  wrote:
"When the theme hits the bass, I dance the Jig!" - -Virgil Fox at the
Mighty Wichita (ex- NYC Paramount) WurliTzer- 1976

3) From: Justin Marquez
On 1/18/07, miKe mcKoffee  wrote:
I have the paint chip set. It is interesting, but unless you have
really good control over the roaster and can make exact-same roasts
back to back, the tiles only tell you what you "GOT, after cooling and
NOW".  Of course, if you have a fair amount of a coffee and can
control the roasts exactly each time, then the tiles will help you to
home in.  So can roasting by time and temp and taste tests as well.
If it was real easy, we wouldn't be having this discussion now would we?
Maybe that's unfair.  It IS easy to get good results and GREAT coffee.
Sometimes it is difficult to make two roasts exactly the same.  (Maybe
the Gene Cafe makes that more likely now without spending a gazillion
bucks on a sample roaster.)
Safe Journeys and Sweet Music
Justin Marquez (CYPRESS, TX)

4) From: raymanowen
A quick thought about the Agtron Spectrophotometer Farce.
A lab spectrophotometer will give a readout of the relative
photometric intensities in the visible spectrum, extending to the
invisible near and far IR and UV.
A printout that should resemble the profile of the Rocky Mountains
from 39N, 105W is SHEER FOLLY.
If some magician could actually use it to divine exact Visible Colors
(They Cannot), Hand him a mirror and tell him it's a printout of an
If he says, "This is a very light Cinnamon roast," I wouldn't spoil
his Charade. I just cleaned the rabbit's cage.
The surface color won't tell the profile, the rest of the story is
beneath the surface, not visible at all. The 0.9999 Coffee Bean that
extracts coffee can't even be seen until you grind it up. Then you
shouldn't spend time looking, but Brewing!
If you must have numbers, save your money and check eBay for a Cosar
SOS-40 color densitometer. MacBeth made a similar instrument- they're
called Reflection Densitometers.
You can pick one up for <<100$. You'll be First on your block to be
able to blow Agtron's doors off.
The "cones" in your eyes' retinas can resolve Red, Green and Blue, the colo=
rs of color TV. The "rods" are much more sensitive to Luminance but do not =
resolve colors. 
The densitometers resolve intensities of the four printers' color inks- Cya=
n, Magenta, Yellow and Black. Why waste money on a Spectrophotometer when i=
ts data is almost totally useless?
Cheers -RayO, aka Opa!
To Grind or Not to Grind- That is the question.
On 1/18/07, raymanowen  wrote:
"When the theme hits the bass, I dance the Jig!" - -Virgil Fox at the
Mighty Wichita (ex- NYC Paramount) WurliTzer- 1976

5) From: miKe mcKoffee
Fascinating "Agtron Farce" diatribe Ray, but measuring color does not =
to apply to the Agtron spectrophotometers. If you're gonna bash 'em read =
on 'em!
Excerpt from Agtron manual from the cheaper of the two Agtron
spectrophotometers designed for coffee roast evaluation (list $9500.00):
"How Does the AGTRON M-BASIC II Evaluate Roast Development?
The M-BASIC II does not look at the color of the coffee or the lightness =
darkness of a sample. Instead, it measures the amount of near-infrared
energy at specific wavelengths
reflected from the surface of a sample. The wavelengths selected relate =
that directly correlate to the development of soluble organics. The =
flavor of coffee is related to soluble organics. By ignoring color, the =
school method, and
concentrating on the development of organic compounds, the M-BASIC II is
capable of
identifying subtle changes introduced by various roasting strategies. It
also provides a
much more accurate and repeatable method for controlling roast =
End excerpt.
Obvioulsy the relatively inexpensive SCAA tile sets "based on" Agtron
eqivalent roast stage colors do rely on color... For further Agtron =
amusement another excerpt:
Undeveloped 100 75.4 -NO TILE
Extremely Light 95 71.7 #95
 90 68.0 -NO TILE
Very Light 85 64.3 #85
80 60.6 -NO TILE
Light 75 56.9 #75
70 53.1 - NO TILE
Medium Light 65 49.4 #65
60 45.7 -NO TILE
Medium 55 42.0 #55
50 38.3 -NO TILE
Medium Dark 45 34.6 #45
40 30.8 -NO TILE
Dark 35 27.1 #35
 30 23.4 -NO TILE
Very Dark 25 19.7 #25
Extremely Dark 20 16.0 -NO TILE
Organic Matter Reduced To Carbon 0 0.0 -N/A
 Agtron 90 Score similar to Cinnamon Roast
 Agtron 30 Score represents the nominal development for Italian / =
Copyright Agtron Incorporated 1986 / Revised 01/30/04

6) From: Jeremy DeFranco
     You are partially correct when you state that the suface color (or
Agtron wavelength measurment) will not tell the profile of the bean. As has
been mentioned so many times on this list, and what I hope every list member
realizes by now, is that final bean temp (AKA ~ color) will NOT determine
overall taste profile as much as the temperature profile throughout the
roast does. HOWEVER, it is a great method to standardize roast degree when
speaking with fellow "collegues". AND it can also be used as an excellent
method to determine the eveness of heating of your coffee beans with a given
profile. For example, you can take Agtron readings of the outside of the
beans, and then compare this reading to ground sample of same beans. A
higher delta (difference in Agtron readings) means more uneven roast, while
a lower delta translates into more even roasts. A very high delta (> ~12-15=
possibility of burnt outside, raw inside). A very low delta (< ~ 5-7 =
possibility of baked beans). So, SOME information can be garnered by proper
use of Agtron.
     As far as what the Agtron actually measures, I beleive it is the amount
of electromagnetic radiation emitted by Trigenelline (don't know off the top
of my head what wavelength this absorbs at). Trigenelline concentration
decreases with longer roasts. So the lower the Agtron value, the lower the
concentration of Trigenneline and hence, the Darker the roast. There's my
$0.02 on the Agtron. Maybe in 20 years I will be able to afford one, but in
the meantime, I'll be waiting for your lower cost "Opa" model to hit the
homeroast market! Cheers!
---The surface color won't tell the profile, the rest of the story is
beneath the surface, not visible at all. The 0.9999 Coffee Bean that
 extracts coffee can't even be seen until you grind it up. Then you
 shouldn't spend time looking, but Brewing!

7) From: Scott Marquardt
On 1/18/07, Jeremy DeFranco  wrote:
It depends on what range of temperatures you mean by "the roast." If
you mean "anything after 340 or so" I'd be less inclined to mildly
dissent.   ;-)
Unless things have changed, it's impossible to use whole beans either
(a) in an agtron or (b) with the tiles, "properly" -- since the grind
and the actual lighting for viewing (that for tiles) is just so.
However, this delta talk is spot-on.
Grind just the outside of the bean!   ;-)
I want an Agtron control system, myself. Something for the popper, y'know.  ;-)

8) From: Scott Marquardt
On 1/18/07, miKe mcKoffee  wrote:
And that's the fascinating thing -- the darkness of roast is not a
correlate to the proportion of trigonelline remaining. In a real
sense, the expensive little tiles give unwarranted grounds for
confidence. Nothing but the analyzer itself will do.

9) From: miKe mcKoffee
In fact the Agtron spectrophotometers are designed to test both whole bean
and ground. Check the spec's! :-)

HomeRoast Digest