Roast Vision - a new and affordable roast level sensor for $299 - Page 7

Read about current and future offerings from the site sponsors.
espressovision (original poster)

#61: Post by espressovision (original poster) »

Agtron Gourmet Conversion update (12/15/2020):

After using a US 20 standard sieve to ensure that 70-75% of the grinds pass through (SCA cupping protocol), we replaced and continued testing our previous samples to ensure that the Agtron Gourmet measurements are accurate.

There was a slight shift (approx. -3 in Agtron Gourmet) in measurements across all of the samples, indicating that our original grind size was slightly off for testing. We have updated all scale images on our website to reflect the new conversion, so no need to update anything on your end unless you are using a temporary copy of the first scale (screenshot, handwritten notes, etc.)

The difference can be seen here:

The new conversion scale is here:


#62: Post by Hotep »

Ordered. Bravo Zach. Bravo. Can't wait to start using this thing.

Team HB

#63: Post by Jeff »

I just ordered one of these, to help satisfy my curiosity about where the beans I buy fall between "medium" and "we waved them over the flame for a bit". I see that there was a slight shift in the conversion due to grind size. I can't really justify buying sieves too. How much does grind impact the readings?

espressovision (original poster)

#64: Post by espressovision (original poster) »

Thanks Alan and thanks Jeff! I've already sent Alan's out, but Jeff I'll get yours boxed up this evening and ship it out tomorrow morning.

The target grind size is Turkish, but as long as you use a very fine espresso grind, it will be sufficient for accurate readings.

And no need for sieves at all - we have a good demonstration here showing the effect that an increasing grind size has on the final reading.

We used the sieves to validate that we were using the proper grind size (SCA cupping protocol) for the Lighttells CM100+, which we used to take Agtron Gourmet measurements, and correlate those measurements with our own Roast Vision readings of the same coffee (ground for Turkish).

We've had a few customers recently reach out to ask about whole bean measurements, so I'm actually running some tests this week to see if a very coarse grind measured in the Roast Vision can be used as a proxy for whole bean readings (taken on the CM100+).

My intuition says that it might work, but based on the principles it shouldn't. I'll post the results regardless, and explain further once I have some data to show.
★ Helpful

espressovision (original poster)

#65: Post by espressovision (original poster) »

Ran the tests, and I would not recommend coarse ground measurements at all. Far too much variability between readings of the same ground sample, and the effect of fines is tremendous, and would certainly lead to large discrepancies based on the use of different grinders.

This is not unexpected, as the normal requirement for finely ground coffee in a flat layer on the acrylic, is almost the exact opposite as an irregular surface of coarsely ground coffee with a varying amount of fines.

Here is a quick screenshot of the results for reference, and even though there might be some correlation between coarse readings and whole bean measurements, unfortunately the Roast Vision design is not suited for exterior bean measurements (or coarsely ground as a proxy).

User avatar
Supporter ❤

#66: Post by Peppersass »

I'm skeptical about whole-bean measurements anyway because they don't tell you about the interior roast level/color, which is crucial to taste. If you use a high charge temperature and drop early, you can have a bean exterior that measures medium or dark, but is less developed on the inside and measures medium-light or light when ground.

I suppose whole-bean measurement can be a quick way for commercial roasters to determine if the exterior color of a batch of beans matches the exterior color of a previously successful batch of the same bean. In that case, chances are good that the target profile was properly repeated. But I'm not so sure that's as useful for the home roaster who switches beans often, is in the process of learning, etc.

espressovision (original poster)

#67: Post by espressovision (original poster) »

Agreed - whole bean alone is certainly limited for taste/flavor.

The way I've seen people use them is to compare the external and internal measurements to see how "even" the roast was throughout the bean, where a "hot and fast" roast will likely have a much darker exterior than interior, compared to a slower and more gradual roast.

I think most would say that the more even the roast, the better, although I wouldn't be surprised if a certain amount of variation between external and internal was necessary to have some complexity in flavor...

Supporter ❤

#68: Post by kbell »

Zach, how tightly is the lens sealed to the housing? Eventually the lens will need to be cleaned with some type of cleaner to remove oils and smudges. What do you recommend, Q-tip with a bit of Simple Green or alcohol?

I received mine a couple weeks ago, and don't roast that often, but this little tool is very informative in helping be better understand my degrees of roast. Thank-you.


espressovision (original poster)

#69: Post by espressovision (original poster) »

Thanks Kevin!

The lens is sealed from coffee particles but we do not currently validate that the seal is water tight, so we don't recommend any liquid cleaners. Additionally, the seal is made with silicone so please be careful if you still choose to use a liquid cleaner as they might cause the silicone to swell and loosen the seal.

The acrylic lens can be wiped clean with a microfiber cloth like you would use for eyeglasses, and works very well to remove the coffee oils and any residue with light pressure.

We're exploring ways to improve the lens and silicone seal design for future versions, hopefully eliminating the need for exposed silicone and making it easier to keep clean.

User avatar
Supporter ♡

#70: Post by hankua »

I just got one on loan and started comparing it to my V1 Tonino in the can. It's pretty cool it can get a general conversion to Agtron readings for people who are buying Artisan coffee. And maybe the coffee I thought was more medium is medium light on the scales. I work with someone who used an Agtron on a daily basis, he could guess the readings after awhile. Maybe that's a useful aspect of the Roast Vision?