What does single dosing lose? - Page 4

December 12th, 2009, 4:34 pm

JonR10 wrote:... during the dialing-in process it was quite evident to me (and my fellow tasters) that there was no real difference in flavor or body/consistency of the shots from the two grinders. Also, the blind testing confirmed that the tasters were not able to identify which shots came from the loaded hopper and which came from the single-dose grinder.
... for now I remain unconvinced that loading a hopper has any effect of flavor or body/mouthfeel/consistency

An observation: there are gross changes in the pour when you go from a bean load in the hopper to single shot dosing. This is not a subtle effect. The single dosed shots require a significantly finer grind to avoid blond gushers. Tightening up the grinder settings for single shot dosing will produce changes in the particle size distribution. With a different distribution of grinds, you will undoubtedly get a different extraction.

Whether these changes are tastable in the cup is still an open question. So thanks for taking the time to run these tests. And I hope the results hold up.

#32: Post by **JonR10** » December 12th, 2009, 4:56 pm

RapidCoffee wrote:The single dosed shots require a significantly finer grind to avoid blond gushers. Tightening up the grinder settings for single shot dosing will produce changes in the particle size distribution. With a different distribution of grinds, you will undoubtedly get a different extraction.

I agree - so I was quite surprised that we weren't able to taste a difference.

RapidCoffee wrote:Whether these changes are tastable in the cup is still an open question.

I again agree and I am already looking forward to trying this again - with coffee we all like

#33: Post by **cafeIKE** » December 12th, 2009, 5:16 pm

JonR10 wrote:We did a little blind sampling today using side-by-side Robur grinders. One was a Robur E with the hopper loaded and the other was my doser Rober loaded per shot.

Thanks for the report, Jon. A few variables spring to mind.
[ With Macap MC4 and MXK, it may take up to 3 shots after an adjustment for things to simmer down after changing bean load, coffee or grind. ]

What's the burr age on each grinder?

How much coffee was run through each since the previous coffee?

How much coffee was in the 'loaded' hopper and what was the bean column height?

Was the throat swept clean on each?

Was the E purged for a bit before the next shot as the loaded hopper machine holds significant coffee?

How many shots were pulled / discarded after each adjustment for 'settling'?

We await your long term test

#34: Post by **JonR10** » December 12th, 2009, 5:58 pm

If I didn't know better, I would think you were suggesting that we are flamin' idiots.
We controlled all of those variables...but I'll humor you

What's the burr age on each grinder?
Both grinders are under a year old and only used in our homes.
Both have seen more than 20 pounds and less than 100.
(In short - both are prime time - similar in age - broken in and nowhere near needing a burr change)

How much coffee was run through each since the previous coffee?
We ran about 1/2 pound through each just getting dialed in

How much coffee was in the 'loaded' hopper and what was the bean column height?
A whole pound at the start. Column was never less than 3 inches.

Was the throat swept clean on each?
On the E-model? No, it's impractical. But again, we ran a BUNCH of shots in fairly rapid succession so there was no stale grinds issue in the test shots. (On the doser grinder the chute was swept on every shot as I usually do)

Was the E purged for a bit before the next shot as the loaded hopper machine holds significant coffee?
No - but test shots were less than 5 minutes apart. We saw no reason to purge at that usage rate.

How many shots were pulled / discarded after each adjustment for 'settling'?
I'm seriously starting to get offended at the implications you make. I'm one who has written volumes on burr carrier thread backlash and adjustment settling. To (again) appease your pointed inquiry I will state that we made adjustments in small increments, and only in one direction, and we ran several shots at the final adjustment for each grinder before starting any taste testing.

Instead of being critical to the point of insult - how about you invite some friends over and do your OWN blind taste test. Frankly, if the difference was even CLOSE to what you seem to suggest then it should be EASY to demonstrate. Again - Mr. Benis seems to believe it's a "night and day" difference.

I remain unconvinced and I eagerly await YOUR testing to illustrate the contrary.

#35: Post by **cafeIKE** » December 12th, 2009, 8:00 pm

JonR10 wrote:If I didn't know better, I would think you were suggesting that we are flamin' idiots.

Nothing could be farther from the truth. Reading your report put me in mind me of audio cowboys schlepping their new wiz-bang over to a pal's, plugging it in and making an evaluation in no-time-flat. The [ new ] reader needs to know the test parameters to make an informed decision. None of the parameters were mentioned, so information was requested. Sorry you took offense.

#36: Post by **JonR10** » December 12th, 2009, 8:30 pm

cafeIKE wrote:Sorry you took offense.

My apologies. Please forgive my negative inference.

I have been running the Robur on my bench for a few months now and I had a Macap M7K (same 68mm burrs) for over a year before that, so I am pretty well aware of the behavior of these beasts.

We spent about 4 hours today dialing in and testing, and even after that I did not feel as though we could make any definitive statements (other than coming to believe that any taste difference may not be as dramatic as previously inferred)

It may help to know that I have been involved with subsea equipment qualification testing for decades and I have a couple of patents as well, not to mention graduate study work in design of experiments (as well as statistics/probability and mechanics of failures). I teach industrial safety and SPC/SQC; and equipment I designed and built decades ago is still in use today.

In short, I'm pretty well qualified to test a coffee grinder.

#37: Post by **Peppersass** » December 12th, 2009, 8:40 pm

RapidCoffee wrote:The single dosed shots require a significantly finer grind to avoid blond gushers. Tightening up the grinder settings for single shot dosing will produce changes in the particle size distribution.

Yes, compared with the particle size distribution without tightening the grind on the same grinder, but here we're comparing the grind produced by a single dose with the grind produced by a loaded hopper.

Since, in theory, the weight of the bean column produces a finer grind at a given grind setting, then the question is whether there's a difference in particle size distribution between the two methods when the shot has been properly dialed in. My guess is that the particle size distribution is the same or very similar, and that's why the shots ran the same and tasted the same.

In other words, in the single dose case, the grind setting has to be made finer. In the loaded hopper case, the weight of the beans makes the grind finer without changing the grind setting.

The key is that the target for dialing in both grinders was an extraction with certain characteristics, which in turn is produced by a particular particle size distribution. The loaded hopper grinder will produce that particle size distribution at a coarser setting than the single dose grinder. But the both grinders are producing about the same particle size distribution, or it's close enough that you can't tell the difference in the cup.

#38: Post by **Peppersass** » December 12th, 2009, 8:47 pm

Here's something else that may be lost with single dosing:

Suppose I dial in a coffee with a full hopper on my Baratza Vario. After establishing the correct dose and grind, I set the timer to produce that dose. Everything stays about the same as the hopper empties, but as it gets near the bottom, the dose weight begins to increase. This is because the reduced weight of the bean column (and possibly some aging of the coffee) is producing a coarser particle distribution. When grinding for a constant time interval, a coarser grind produces a larger dose than a finer grind. Luckily, the increased dose weight usually compensates for the coarser grind, and I don't have to make any changes to the grind setting, or if I do they're minimal (this may be unique to the Baratza Vario and the particular coffees I use.)

With a single dose, I have to manually compensate as the coffee ages by either tightening the grind or increasing the dose (grinding time.) As the coffee ages, this can be a frustrating moving target. I'm still playing around with it, and I'm sure the dependability varies with the coffee in use, but it's kinda neat how the loaded hopper and timer work together to make things easier.

#39: Post by HB » December 12th, 2009, 8:59 pm

JonR10 wrote:But in this case Michael claims to have experienced a glaring difference and you have said it's a "no brainer", so the difference should be easily recognized by a blind A/B comparison. If no such glaring and obvious difference exists then maybe it starts to look like a matter of personal taste preference.

Not calling out your specific post, but as a general reference, it's helpful to read Jim's layman explanation of such testing in A note on comparison tests.

The most interesting result?
We pulled better shots of the uber-bright coffee with the MYPRESSI than we were getting from the GS3

Funny that happened to you, I posted similar results with the TWIST versus La Marzocco FB-80 yesterday. In all fairness, it was not a route by any stretch; all the espresso were very good. But the mypressi squeaked by in every round thanks to a more pleasing aftertaste and rounded texture. Go figure.

December 13th, 2009, 12:20 am

The reaction to Jon's and Nicholas's test shows why people who market expensive stuff bend over backwards trying to persuade people that blind testing doesn't work.

First off, blind testing isn't about which one is better or worse; it's about being able to tell the alternatives apart. It's a significant result if you can tell them apart, and a null result if you can't. Alas, in espresso, there is natural variability. So the equivalent of blind testing is IDing. You start doing blind paired shots. They are different. If the difference is systematic and not random, then by the third or fourth shot, you'll be able to ID which one is which; if the difference is just noise, you'll never learn to ID which is which.

(Geeky aside: The other, more standard way to handle noise is by triangle testing -- picking out the odd cup or shot out of three, i.e. the one that is most different. For technical reasons, I now prefer learning how to ID blind pairs as a better technique, since you can learn to ID a systematic difference no stronger than the noise, while not being able to pick it out in the first few triangle tests)

If the shots are not distinct or not identifiable, you have no reason to prefer one to the other.

If two cases are indistinguishable, blind testing is usually not needed. But when the two cases are randomly different, beginners always fool themselves. Since coffee is variable, you will always have a difference between two cups or shots. I cup coffees several times a week, I do blind tests several times a week. It's taken an endless series of humiliations and pratfalls to realize that most of the differences I taste are not systematic, not repeatable, and not an aspect of either the coffee or prep, but either noise or delusion.

The only way to stop having to eat crow on such things is to do lots of blind taste tests and learn about getting consistent results the hard way.