Sovda Technical Brand Ambassador and fellow coffee lover, Nicholas Flatoff, spent an afternoon playing with our Pearl Mini optical sorter and photo fluorescent beans to create a surprisingly delightful cup of coffee.
Sovda Photo Fluorescence
Nicholas Flatoff: Welcome to the Sovda podcast. This is Nicholas Flatoff, technical brand ambassador.
[00:00:06] Joining you from our Portland showroom for an interesting and slightly different type of podcast today. We normally do interviews, and this is going to be a monologue, but I still think it's going on some interesting things that we are doing with the equipment that we have available to us in Portland in order to better understand coffee and better understand the capabilities of our own equipment, and really just in general, to do a little bit of experimenting and see what we can't learn. So I'm pretty excited about this. The topic today is photo fluorescence in green beans. And when combined with optical sorting, what does that do to the flavor and the cup? So this one has a little bit of an interesting backstory.
[00:00:59] Katie, who is the head roaster at Torch Coffee, which roasts out of the showroom here using the equipment, and we have them roasting on a Loring S15 hooked up to the Pearl Mini optical sorter and using our de-stoner lift, and then the precision fill. And so by having somebody there, who's using the equipment and doing live feedback, we're able to get some really good information about it, but what it presents is an opportunity for us to do some collaboration with people who are working with grain and roasted product on a daily basis, the type of information an OEM might not normally get access to.
[00:01:33] So we really love that collaboration. It's super fun having roasters and production staff right in the space to give us live feedback on, "Hey, this sucks. This could be better." "Hey, this is working out really well. You should carry that over into future models." That's, that's incredibly valuable for us. We love to have it, but I digress.
[00:01:52] This is a day... late February I believe, maybe early March. It was a little bit of a slow day in the office. And one of the things that I had been wanting to try for a while that Kyle Ramage - I think he even mentions this in the podcast that we just released with him a couple of weeks ago - is the concept of over-sorting. And basically this is not an efficient way... this isn't something that we'd recommend anybody really do in the cafe, unless you're just playing around with coffee or you're doing like a competition level coffee. And so what over-sorting is when you set up the optical sorter, we usually just set it up to take out burnt beans and quakers and occasionally fractured beans and just, you know, weird looking stuff like patio or bone or all the weird stuff that might make it through a density sorter into the actual optical sorter.
[00:02:42] Well, it's sensitive enough that you can actually get in and start selecting the profile of bean specifically that you would like, rather than just saying, Hey, these are all light beans. Let's reject them. These are all dark beans. Let's reject them. You can get in there and fine tune it. So you've got a really specific color band that you're getting as an exclusive output.
[00:03:03] And I had wanted to try this for a while. We had the time, we had the coffee, so we grabbed our already washed process, it's a pretty well-known coffee, it's nice and clean, we're really familiar with it, and so it's a good test coffee for us because when we're discussing it, a lot of people have worked with that coffee in the past, it's got a little bit of a personal connection to a lot of people. So it works out to be pretty applicable coffee. And so what we wanted to try doing was, let's dial this in to only get the good beans in the more fully developed full color spectrum of the coffee. So basically these are ones that we would assume had a full mired reaction, had a little bit more caramelization, the more uniform, darker side of the spectrum rejecting a lot of the lighter colored coffees. And we went ahead and did that. And again, we're talking 65-70% rejects here. So, you know, you're really just selecting out the 30%, I'm not even going to say best coffee because we haven't done it for the lighter coffees, we haven't done the side-by-side on it, so that's something that we'd like to do soon, but 30% of the darkest, but still in the appropriate roast development range, coffees in the batch.
[00:04:13] And we did a side-by-side compared to just the normally sorted coffee. So we still removed all the quakers and burnt beans, et cetera, from the batch and cup them side by side. And it was a noticeable difference. It was a little bit sweeter, quite a bit more complex, just fuller body on it, felt like the same coffee with maybe a veil removed from it. Really, really pretty impressive. Definitely not worth throwing out 60% of your product, but a novel and interesting experiment, nonetheless.
[00:04:44] And so, you know, then we got thinking, well, why is that right? That's the next question? What's the difference between those darker coffees and the lighter roasted coffees... Not later roasted, but later in Agtron coffees - and my conclusion was kind of, I just chalked it up to caramelization. I'm going to bet that those were just picked right at their peak. The rest of them were either a little past or a little before, and these are just the prime, sugar content coffees. And so that's what's presenting with a little bit more, you know, you had a little bit of sweetness and it does tend to bring out other characteristics in the coffee up to a point. So that was my conclusion on it. My assumption.
[00:05:20] Fast forward a couple of weeks on an unrelated project. Another industry professional that I was connected with sent over Christopher Feran's, I believe it's "Under The Blacklight" article. Really fascinating article and photo fluorescence, and a lot of the misconceptions on photo fluorescence, how he uses it as the tool, references some great research by Tim Hill. Spectacular article, highly recommend you Google it. It's on Christopher Feran's website. Shout out to that for putting out information like that, that really is inaccessible any other way, except for coffee geeks, going out and doing it and putting it together and making that information public, you know, and looking for that information and consolidating it.
[00:06:01] So really great article. I highly recommend you read it, but there's a point in there, and I believe this is in reference to Tim Hill's work, where there's a correlation made between photo fluorescence being present in green beans and a higher Agtron score. And so let me unpack that a little bit. When I'm talking about photo fluorescence in green beans, what I mean is if you put it under a UV light, you put green coffee under the UV light, some of the beans will actually glow and others, most of them, won't. That's the presence of photo fluorescent chemicals within the green coffee.
[00:06:36] And I think this is probably a point where I put out a big disclaimer that we are standing on the shoulders of giants here. This test that I'm about to explain and go through was a, "just for fun" type test. I think someday we'd love to have the resources to have a little bit more time in a lab-like environment to go into some of these experimentation to get a higher quality output. But Tim Hill and Christopher Feran have done more research on this than I think I ever intend to do in my life. They know way more about this. Nothing that I'm uncovering in this test is in any way supposed to upturn or go against any of the research they've done.. This is just an individual single anecdote that we wanted to give a shot as a fun way to play with coffee and learn a little bit more.
[00:07:22] And so basically the presence of photo fluorescence in green beans was connected to a higher Agtron score post-roast, meaning the photo fluorescent beans dropped at the same temperature with the same amount of development appeared lighter to the eye. And of course, well, that connects back to the tests that Katie and I did, where we were taking out those lighter in appearance coffees that were developed, you know, in the same roast chamber, same roast time, et cetera, yada yada, with all the other coffee. Right. And so it's like, well maybe that's on to something.
[00:08:00] Maybe one of the reasons that rejecting those slightly lighter shades of coffee presented in a better cup is because there's some derogatory aspect to the photo fluorescence. Right. And this is something that's been discussed a ton at the, you know, there's impacts on photo fluorescence and, okay, great, that's awesome. What's a way that we could actually experiment with that to taste how those beans are impacting in the cup? And let's throw in a little bit of sorting as well. Let's figure out a way to add a little bit more, if we're going to do the tests, let's do it pretty thorough. So the tests that I devised, it was a four-sample test that winds up getting split in half for eight actual tasting cups.
[00:08:44] The Ardi washed process is one of the coffees we like to do as I described. But when I checked for photo fluorescence in the Ardi washed, there weren't any present. And that's, you know, we've had it around for a little while and I'm aware that the photo fluorescence have a tendency to fade away sometimes.
[00:09:01] So it wasn't necessarily a kill all to the premise of the test. But what I did is we went over and we grabbed the Ardi natural instead, another pretty common coffee that people are familiar with really, you know, delicious, juicy coffee. And we looked at that. Sure enough, there were some photo fluorescent beans in there.
[00:09:19] So I went and grabbed a sample tray, pulled a a hundred grams just straight out of green coffee without doing anything to it, pulled out another sample and took out all the photo fluorescent beans under a black light. So we had one just normal, straight out of the bag. One that only had the photo fluorescence removed, pulled a third sample, sent it through the Pearl Mini optical sorter as a green coffee to reject all the green defects that we can pick out and took a hundred grams of that sorted coffee, put it down. And then for the fourth sample, I just dug through the bag for like 45 minutes and grabbed out as many photo fluorescent beans, as I could find, regardless of what they looked like, regardless of defects on it just got a hundred grams of the photo fluorescence.
[00:10:03] Right? Our main goal is to taste what does the photo fluorescence do to the cup? We were expecting some pretty funky stuff. So we went in, I sample roasted them all and then post roast I split each of the samples in half and sorted half of those samples post-roast through the optical sorter, removing quakers and removing burnt beans.
[00:10:25] So that's what gave us our eight sample selection. Grabbed them threw them in the grinder, put them out on the cupping table and got some of the other staff members we have here. Robert who's the head of Torch, who's the manager at Torch. I believe Vicky participated in it, she's staging here for a little while before moving down to Guatemala to do some Torch work in Guatemala, and then to Q students that were in town that had just finished the Q course as well, just for a little bit of an outside perspective to taste some of these coffees.
[00:10:54] And we went through and tasted the coffees with just a standard silent cupping protocol. Just trying to decide which copies were preferable.
[00:11:03] Looking at the results, I think the first one that's obvious and I genuinely don't mean this as a sales pitch or anything like that, if you've ever tasted - you can do this yourself with a hundred gram sample or heck, a 30 gram sample, split it in half and pull out all the quakers - the samples that ran through the optical sorter were preferred to those that weren't right. You've got a cleaner cup, a little bit less hay, a little bit less Quaker cereal vibes going on. It's just all in all a better improvement in the cup. But as we were going through, most of the samples besides that tasted pretty similar. I think the favorites marginally, so was the green sorted that was also post-sorted, post-roast sorted, but that was a pretty marginal improvement over the rest. But there were two that stood out, you know, and you guys, this is what you're all waiting for...
[00:12:06] I remember when I was grinding and I got walloped in the face because I did this random as well. I didn't know which coffees were which, but there were two that I was grinding that just hit me in the face with crazy aromatics. And they were pleasant aeromatics they weren't bad. And they weren't weird. They were just that, you know, blueberry pie, syrup, you know, almost like the store bought cheap, canned blueberry pie coming off and hit me in the face.
[00:12:29] And I was like, wow, that's, that's kind of interesting wondering at this point that I jack up the roast what's going on there because these two were really different. We get them onto the table and sure enough, when we pour water on, the aromatics coming off two of them, same thing, just wild, vivid, blueberry, super fruity. And it, you know, it was one of those when you pour the water in, it hits the 10 feet away standing on the other side of the table. It's like, whoa. And what was really fun for me after going through and tasting them was getting some of the Q students in there and Joshua and Jesse, and just watching their face when they got to those and to see in the surprise and the expression that came off of them.
[00:13:06] You guys probably know where this is going, it's kind of, it's a hard thing to do a set up for. But the two highest scoring coffees by a long shot in the group were these two super aromatic coffees. And I lifted them up and I looked at the mark and sure enough, they're the coffees that had photo fluorescents present in the green coffee. And they were just wild, fruity, kind of exactly what you expect from a really interesting natural process. Even the non optically sorted photo fluorescent coffee was preferable to the crew over all of the other coffees with the top one being the photo fluorescence post-roast optically sorted sample that was preferred by everybody in the group, unanimously.
[00:13:55] Now this is where I got to get back into the caveats on this is a single situation with a single coffee, a single sample roaster, you know, I'm a... roaster maintenance is my background, but a sample roasting, a hundred grand batches on a 2.2 kilo roaster is not my forte. There's a lot of variables here that we didn't control, that we didn't want to control. We just wanted to throw it in the roaster and taste what it tasted like in the cup. That was the only goal that we had in this. But after finding that result, I think we're going to have to go back, do a bunch of different various coffees, dial in our procedures a little bit, dial in our protocols, figure out how to get a more accurate roast profile for our sample roast to make sure that they're sitting there well, because, you know, our bean probes not doing a whole lot for us with a hundred grams, so we're definitely not done with this.
[00:14:43] But I think what it goes back to is, I believe in the introduction of Christopher Feran's article, talking about never using photo fluorescence as a rejection tool, there's a lot going on here. A lot of chemical processes, a lot of stuff that there's not the dedicated resources within the coffee industry yet to really discover, for the most part, these super high level chemical analysis. I know some people do that and they're, they're making headway on that, but it's not the same as a PhD research program or whatever. And I do think that I'm seeing that sort of thing happen. We just did a little joint effort beta test thing with UC Davis on Saturday. I see more resources being put in that direction and it's super cool to be a part of that. But the point is I'm not considering this a true lab research type test. This was just a fun thing to try.
[00:15:35] But the point is photo fluorescence is something that traditionally people have spent a fair amount of money to try and remove from the coffee. And here we are tasting the photo fluorescence and by far preferring the purely photo fluorescent coffees over the standard sample or the samples that have the photo fluorescences being sorted out. Really a fascinating experiment for me to participate in, for me to lead and, and look at and I'm super stoked to roll what we learned into this into some more structured tests to see if there's a little bit more going on there that we can uncover. So next, we're going to do a, [?] natural process that we've got a nice and fruity one, see if we can't get similar results in a little bit more controlled environment with Katie actually doing the roasting because she's just a frickin' awesome roaster.
[00:16:26] The photo fluorescent coffees, I had actually developed them a little more than the rest of the coffees because they presented as so light. There were actually some that looked like basically barely starting first crack. And it was almost more of like a city plus-plus, or almost 45 seconds past first crack drop. Where after a couple of days, actually a couple of them had oil on it. It was pretty weird to see the super light bean with oil coming out on the surface.
[00:16:54] But the point being we also experienced the higher Agtron score for those photo fluorescent coffees and they were a little bit more of a challenge to roast as a result because it's just throwing off your entire spectrum that you're used to with a coffee. But yeah, really fascinating little study, super easy for you guys to do yourself with a $10 UV light. Throw them in you know, a Ikawa or whatever your sample roaster is. If you got a little bit of time and you want to get a little geeky, check it out, do it yourself. If you find different results or the same results, I'd love to hear about it, see what the industry is experiencing. Really fun one. Looking forward to digging into this a little bit more.
[00:17:32] One shout out that we want to get out there, we did do the hard launch on our integrated de-stoner conveyance system for Sovda. It's a really cool system. We've been beta testing it for about a year now, and we finally have something that's working for us really well in our shop here and has been working really well for the shops around that we've been working with it over at... And I think Black & White's got one, Kyle Ramage has one, got one going here at Caravan in Portland. And yeah, it's $8000 for a density based de-stoner combined with a vacuum lift system, which is... you're removing an entire piece from your workflow with the combined system.
[00:18:12] And then frankly, it's way faster than the cheaper, mechanical de-stoners and way cheaper than getting a standalone density-based system. You know, you might be spending 12, 15, 18 [thousand dollars] for a density-based system from some of the popular manufacturers. And so, yeah, really, really cool systems, super easy to control, super easy to adjust for different coffee batches, and you're killing uh, I think our website says you can "kill two birds with no stones." We're getting them all out for you. So yeah, pretty cool system that I'm pretty stoked [about]. Go to our website, SovdaCoffee.com, it's under the products page. It's a pretty cool system. We do custom mounting brackets for an extra 500 bucks. So if you got, you know, a Weigh Right or a Logical Machines or whatever, we don't care that you're not running it on the precision fill, we'll build you a custom mounting bracket to slap it right on there.
[00:19:01] So, yeah, that's my little shout out. Like and subscribe. We've got some really cool content that we're backlogging here to keep a really nice, consistent schedule. Love what we're doing. Super stoked to get to share it with you guys, looking forward to hearing your thoughts and results on the test. Cheers!