RapidCoffee wrote:Randy is absolutely right. The right image was darkened by an overenthusiastic autocontrast setting on my photo editing software..... Thanks for pointing that out!

No problem... I've taken enough bad photos and done bad enhancements to know one when I see one!

It is clear that the extraction with the "proper" sized tamper was darker, not only coming from the spouts but also in the cup. No correlation between extraction and tamper size inferred by me, but there is a difference in those two extractions..

I would also wonder if, say after 20 or 30 pulls, whether all that loose coffee in the PF when using the smaller tamper had an effect on the shower screen and through the brewpath...

Nothing to prevent you from brushing off loose grinds with a fingertip and briefly inverting the basket prior to locking in.

But please don't get the idea that I am recommending a poorly-sized tamper. I'm just trying to defuse the myth that 0.1mm makes any significant difference in the cup.

BTW, this is the post that inspired me to question the need for a perfectly-sized piston:

HB wrote:A couple years ago I spoke to one of the engineers at MACAP manning the booth at the SCAA conference. I asked him why they didn't make a 58mm piston for their auto-tamper. He claimed it didn't matter, that an even consistent tamp applied dead center with a 53mm piston for a 58mm basket would not improve if one used a 58mm piston. I didn't buy his argument, but said I'd give it a try (never did). Anyone with very good tamping skills and a 53mm tamper/58mm basket want to test his assertion?

dsc wrote:I'm actually surprised that there wasn't even any indication of side channeling, I would bet on a donut extraction.

I think a common misconception is that all the tamping pressure is vertical. Since ground coffee is 'fluid', i.e., you can 'pour' it, some of the tamp downward pressure is vectored to the side.

When water hits the coffee, it expands in all directions. With an even initial distribution, the expansion is even. Problems arise when the initial distribution is uneven, the expansion is uneven and channeling occurs.

I think a donut extraction occurs when the PF is up-dosed, the coffee expands to the dispersion screen, is seriously compressed and the path of least resistance is the ~1mm at edge of the basket between the basket and the group.

The left column is a lower dose and the right is an up dose.

The first row is the dry puck.

The second row is at the end of pressure ramp. The coffee has expanded. In the updose, there is nowhere for the coffee to go except up around the group.

The third row is during extraction. In the lower dose , the pressure is even across the puck. In the up-dose, the disturbed coffee at the edge of the basket allows more water to flow giving a donut extraction.

That's my theory

cafeIKE wrote:That's my theory

Certainly looks and sounds plausible. It may also explain why tapered baskets (Faema style) are more forgiving than straight-sided baskets.

cafeIKE wrote:The second row is at the end of pressure ramp. The coffee has expanded.

Does this assume very gentle and somewhat long preinfusion?

Even if so, I don't know if I'm sold.
It could be right - but (though it's perhaps counter-intuitive) the infamous (and elusive) perspex portafilter video showed expansion occurring after a shot ends (ie when the 3-way releases pressure).
I don't know if, even with very gentle and long pre-infusion, you'd really see expansion like that in such a short timeframe. I suppose there could be coffees out there that just soak up the coffee and blow up quickly, I don't know.

Obviously - this is all just theory.
And your model is elegant and logical.
I just don't know if it's true.

cafeIKE wrote:That's my theory

Makes as much sense as anything else I've come across.

When it comes to tamping: just about anything goes. Some folks like a hard tamp, others advocate a light tamp (or no tamp at all). Some like a flat piston, others a convex piston, and a few oddballs insist on using both (on the same puck!). Some prefer one tamp, others two, and many advocate a four-corners (NSEW) Staub tamp. Some tap, some polish... you get the idea.

This observation suggests that none of these factors are critical to a good pour. Why not? Perhaps because, shortly after tamping with ~30# pressure, the puck gets slammed with hot water at 130psi (9 bar). Any tamp that promotes an even wetting and expansion of the coffee grinds under high pressure is a good tamp. After this initial wetting/expansion phase is complete, the tamp just doesn't matter any more.

For what it's worth (mostly my own amusement), I believe that a 58mm basket yields a puck top surface area of just over 4 square inches. That means that you're talking about ~7-12 psi applied to the puck by a 30 pound tamp.

Of course, those are just numbers. Arguably the force of water and the force of a metal piston (or any other solid) are quite different against a variably porous material, so this type of logic is certainly not conclusive about anything.

malachi wrote:The left shot actually looks better (see the less even extraction illustrated in the flow).

jpreiser wrote:To my still-in-training eye, the left shot looks better even though they both look good.
I'm still working on getting consistent tiger striping on my naked shots.

Randy G. wrote:The pull image on the right looks slightly underexposed (darker) compared to the one on the left , yet the espresso looks lighter in tone when the opposite should be the case.

Are you guys planning to compete in the 2009 NBECC?

National Barista Eye-Cupping Championships