DIY Color Meter - Page 80

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
Brewzologist
Supporter ♡
Posts: 1179
Joined: 7 years ago

#791: Post by Brewzologist »

pcofftenyo wrote: Now I've got figure out calibration, like what/where/how to open the roast_meter.ino file and edit for offset but I might be able to do that in what, a text editor?
Yes, you have to open the .INO file in the Arduino IDE, edit it, and recompile the sketch. Pretty standard stuff. Read the section on calibration in the users guide for where to make the change in the INO file: https://github.com/juztins-lab/roast-me ... alibration
pcofftenyo wrote: Some blue light keeps blinking on my main board and only one of the red lights glows on the sensor board, but I guess neither is a big deal.
Neither is a big deal for this project. If you would eventually like to know what those LEDs do, see here: https://learn.sparkfun.com/tutorials/ho ... e-overview

User avatar
yakster
Supporter ♡
Posts: 7340
Joined: 15 years ago

#792: Post by yakster »

You can also try the Bluetooth sketch (roast_meter_ble) and change the calibration factor on your phone using a web interface.
-Chris

LMWDP # 272

pcofftenyo
Supporter ♡
Posts: 354
Joined: 7 years ago

#793: Post by pcofftenyo »

The user's guide mentioned a reading of 230 for the baking soda.

Is this the number?

pcofftenyo
Supporter ♡
Posts: 354
Joined: 7 years ago

#794: Post by pcofftenyo »

yakster wrote:You can also try the Bluetooth sketch (roast_meter_ble) and change the calibration factor on your phone using a web interface.

I've still got the training wheels and crash helmet on.

boren
Posts: 1115
Joined: 14 years ago

#795: Post by boren »

Can an inexpensive colormeter such as Linshang LS170 (about $60 at AliExpress) be used instead of a DIY hardware sensor? It supports many different color spaces - CIE Lab, Luv, LCh, Yxy, CMYK, RGB, Hex. Is any of these straightforward to convert into Agtron?

User avatar
Brewzologist
Supporter ♡
Posts: 1179
Joined: 7 years ago

#796: Post by Brewzologist »

boren wrote:Can an inexpensive colormeter such as Linshang LS170 (about $60 at AliExpress) be used instead of a DIY hardware sensor? It supports many different color spaces - CIE Lab, Luv, LCh, Yxy, CMYK, RGB, Hex. Is any of these straightforward to convert into Agtron?
Most coffee colorimeters measure in the near infrared range around 880nm. If this meter can do that it might work. Converting to Agtron is yet another issue not easily solved. At a minimum you'd need a reference meter you could compare to this meter to try and derive a conversion formula.

jpender (original poster)
Posts: 3910
Joined: 12 years ago

#797: Post by jpender (original poster) »

boren wrote:Can an inexpensive colormeter such as Linshang LS170 (about $60 at AliExpress) be used instead of a DIY hardware sensor? It supports many different color spaces - CIE Lab, Luv, LCh, Yxy, CMYK, RGB, Hex. Is any of these straightforward to convert into Agtron?
At first glance it doesn't look like a great fit. But the DIY meter is based on a pulse oximeter chip so who knows?

boren
Posts: 1115
Joined: 14 years ago

#798: Post by boren »

Brewzologist wrote:Most coffee colorimeters measure in the near infrared range around 880nm. If this meter can do that it might work. Converting to Agtron is yet another issue not easily solved. At a minimum you'd need a reference meter you could compare to this meter to try and derive a conversion formula.
Do you know why is measuring at near infrared important when differences in how dark the roast is are so clear in the humanly visible range? At least that's how it seems when looking at reference cards that claim to be "Professionally color-calibrated to original Agtron/SCA Eight Color Tiles. The color accuracy of less than 1.0 ΔE00 is achieved under reference condition", such as these:

https://www.amazon.com/Coffee-Color-Cha ... B0C2XF454L

The card from Sweet Maria's doesn't come with similar claims about accuracy, but the various Agtron level on its back are similarly easy to distinguish with a naked eye:

https://www.sweetmarias.com/roasted-cof ... -card.html

jpender (original poster)
Posts: 3910
Joined: 12 years ago

#799: Post by jpender (original poster) »

boren wrote:Do you know why is measuring at near infrared important when differences in how dark the roast is are so clear in the humanly visible range?
I believe it is because coffee reflectivity is greater at those wavelengths. This allows for greater precision. If the goal is merely to match the discrimination ability of a human eye then why use a color meter at all? Even the DIY meter can't match the Agtron machines. That's why the Roast Vision scale is coarser than Agtron. Using blue light will just make it worse.

Or maybe you can make it work. But why start with a disadvantage like that?

renatoa
Posts: 770
Joined: 7 years ago

#800: Post by renatoa »

Brewzologist wrote:Most coffee colorimeters measure in the near infrared range around 880nm. If this meter can do that it might work. Converting to Agtron is yet another issue not easily solved. At a minimum you'd need a reference meter you could compare to this meter to try and derive a conversion formula.
Somewhere in this big thread there is buried a .pdf document with a serious enough study showing the differences between IR and two visible light wavelengths. They are minimal, as expected. A standard deviation less than 1 Agtron unit, if I remember well...
If you are patient, search for it.

btw, we measure the beans or grounds when they are cold, not hot, right ? IR is a property of HOT objects.
IR spontaneous emission does not exist in nature at ambient temperature, unless it's an emitting electronic component.

On HRO usage of display colorimeters for this purpose, i.e. Agtron measurement, was been discussed some time before starting of this thread, you can look there also.