Puck prep study - Page 2

Beginner and pro baristas share tips and tricks for making espresso.
User avatar
RapidCoffee (original poster)
Team HB

#11: Post by RapidCoffee (original poster) »

Peppersass wrote:It's about time someone tried to do a study of puck prep techniques versus taste, but even though some rigor is evident in the protocol I wonder if it was rigorous enough.
Agreed. Small sample size (1?), no blinding. But a great basis for future work.
Peppersass wrote:What is circular motion of the PF? Does that mean the PF was rotated while the grinds dropped into the basket?
Scott Rao's technique: you swirl the grinds vigorously in the basket (with a funnel/collar). I didn't feel like reposting the video, but here is a screenshot:
John

PIXIllate
Supporter ♡

#12: Post by PIXIllate »

Very interesting raw data set. Would be more interesting if a basic E61 profile was tried and then maybe a run with a medium roast.

As it stands it would appear that you can't over tap/agitate the grounds and an even puck is king.

User avatar
RapidCoffee (original poster)
Team HB

#13: Post by RapidCoffee (original poster) »

Stéphane posted additional test results today, indicating less consistency with no distribution. I have updated his "main findings" with the newly posted slide.
John

User avatar
TomC
Team HB

#14: Post by TomC »

Peppersass wrote:
...After WDT, I use my mini-whisk to even the top layer of grounds, and while doing so leave a small depression or divot in the middle of the puck -- i.e., create a slightly concave surface. This has completely eliminated halo extractions. I'm using a flat tamper and have had some thoughts that going back to a convex tamper might have a similar effect.

Me too. I'm considering seeing if I can get one made up that'll fit my EasyTamp 5 Star.

Overall, I'm happy to see this information shared. It must have taken quite a bit of time and work. I do wonder how much the taste preference relative to EY might be turned on its head if they were to use a coffee that doesn't need long pre-infusion. If gentle pre-infusion and lower pressures is truly kinder to the coffee, then it would sorta imply the opposite is true as well.

If Im reading the graphs right, it appears to show about 15 seconds of pre-infusion flow before the pressure starts to ramp.

Im also curious about the swirling technique. Has anyone seen scuffing of their countertops by doing this? I imagine if it were done on any sort of pad, the friction would prevent it from moving at all.
Join us and support Artisan Roasting Software=https://artisan-scope.org/donate/

John49

#15: Post by John49 »

TomC wrote: Im also curious about the swirling technique. Has anyone seen scuffing of their countertops by doing this? I imagine if it were done on any sort of pad, the friction would prevent it from moving at all.
You can see the swirl technique video on the LWW site using the blind shaker, no countertop contact and it creates a level bed in the basket.

User avatar
Jake_G
Team HB

#16: Post by Jake_G »

Did anyone else notice the "slight" shift in protocol for the shots using the Hog?

The dose was dropped to 18g from 18.5g, but the grind was unchanged and the beverage weight was held constant.

I don't want to diminish the body of work that was done here, because it was clearly a huge amount of work to pull all of these shots and record everything. But I can't help but see the EY and earlier first drops as being rather suspect to the change in dose. Now, the higher peak pressure is rather curious...

I think the dose was likely adjusted simply because the hog required more room in the basket to work, but changing the brew ratio changes the EY% and changing the puck height changes the flow resistance. One could argue that the dose adjustment was made to keep the puck height consistent due to all the channels that the Hog creates expanding the puck, but it still seems like there should be some characteristic adjustment made to the grind setting and the brew ratio at a minimum to normalize the data set.

Such adjustments would obviously garner suspicions and challenges to the method as much or more so than a simple change in dose, but when EY% is calculated by the using the brew ratio and you changed the brew ratio in the direction that increases EY%, it seems like that should be called out when comparing EY% to the other puck prep methods.

Again, I hate to sound critical of the work, because I am most certainly not. I just think that the final conclusion needs an asterisk.

Cheers!

- Jake

**Edit**
I care because taking the 22.1% EY (the highest of the pack) and "correcting" for the change in dose (which doesn't really work because TDS changes, too but shows how sensitive the EY% is to dose) shifts the outcome by quite a lot.

22.1% * 18.0 (actual dose) / 18.5 (representative dose of the pack) = 21.5%

A good follow up, is Stéphane is willing, would be to cut the Hog shots short to maintain brew ratio. Equivalent yield would be 40g*18/18.5 = 38.9g.
Call it 39g to keep it simple.

I never would have thought that 1g output could really shift the flavor, and I don't imagine it will. But it will certainly impact the EY% in ways that make you go hmmmm...
LMWDP #704

Eiern

#17: Post by Eiern »

I'm wondering how much the secret upcoming Matt Perger distribution tool will help gentle blooming low pressure shots. Maybe that too will help more with the straight 9 bar no preinfusion shots but less gain with this?

shotwell

#18: Post by shotwell »

Jake_G wrote:I never would have thought that 1g output could really shift the flavor, and I don't imagine it will. But it will certainly impact the EY% in ways that make you go hmmmm...
At the risk of sounding like a pedant, if you believe in modern extraction theory you also believe that 1g differences in output will have an effect on taste. Whether that difference in taste is consistently identifiable in blind tasting is another question entirely. I'd find it interesting to run a blind taste test to identify the smallest changes in extraction yield identified by well trained and average tasters, but I certainly don't have the time right now to set something like that up.

I also respect the work and effort done so far, but I agree that the hog result should be at least viewed with some skepticism.

Stéphane

#19: Post by Stéphane »

yakster wrote:I bet the modified Rao "blooming" extraction profile incorporating a 20s pause after preinfusion greatly mitigates channeling, probably reducing the need for WDT in these tests.
Hi Chris, I think this might be right, even if I thought it was not the case before my tests. As blooming Espresso is said to be very difficult to dial in, I initially imagined it would be a well suited profile to highlight differences in "puck prep" (I expected more visible impact of the tested methods on the resulting pressure peak).

Furthermore, as mentioned by John, the low extraction pressure has also probably helped avoid part of the channeling of the worst shots.

Stéphane

#20: Post by Stéphane »

Peppersass wrote:It's about time someone tried to do a study of puck prep techniques versus taste, but even though some rigor is evident in the protocol I wonder if it was rigorous enough.

First on my list of questions is, was blind testing used? Seems like it would have been pretty difficult to blind taste 3-5 shots averaged for each technique, and if that was done, how were the taste results averaged?

I'm also a little skeptical that the taste results correlated perfectly with the %EY results. While my own experience has been that there's a general correlation, it's not perfect. And can the author really taste a difference of 0.5% EY? I can't.
WDT with thin needles + gentle taps has been my usual routine for the last weeks and I have been satisfied with the consistency of the extractions (visual uniformity + taste + repeatability), so my idea was to test simpler techniques and check if some of these would bring similar results.

So I considered that one single "not that good" shot would be enough to conclude that the corresponding puck prep technique would not be a good candidate to replace my current workflow.

I'm not able to do blind tasting or triangle tests. However when I record the rating of individual shots, I try to do it methodically, along different dimensions (aroma, body/mouthfeel, acidity, sweetness, bitterness and astringency (+free text comments)). I generally have no big doubts in the final ranking of shots from a single tasting session, except if it has lasted really too long (say 3h). The correlation with EY was also a surprise to me, but considering that the TDS measurements are performed after the tasting evaluation, there is no possible influence of one set of data on the other.

In other circumstances, I've had a preference for espresso shots with lower EY - lower extraction uniformity - typically when I used more developed coffee beans.

For example, a convex tamper base or a basket with a smaller perforated area (Reneka microsieve, IMS Superfine) will extract less from the outer grinds of the puck (circumference) and produce a higher proportion of underextracted coffee. This can be beneficial in the case of beans that poorly handle strong extractions (typically: overroasted).