This morning I did an interesting experiment, and I wonder if other people can test it as well.

I split the dose into 2 parts. I put a paper filter at the bottom of the basket, I grinded the first half, WDT-ed, OCD-ed, tamped, then I put another paper filter, grinded second half, WDT-ed, OCD-ed, tamped, then metal filter, extracted. The extraction was faster, and my paper filter clearly was a bit small and did not cover the middle of the puck properly (so it can be improved). I got a 1% increase in extraction yield (compared to using only 1 paper filter at the bottom and the metal filter on the top), which is pretty sizable.

Did anybody try similar experiments?

best

How did your espresso taste? How was the taste or texture different? Inquiring minds wanna know.

Taste? We don't need no stinking taste.
We got numbers!

This was a first experiment on a new coffee, new roaster. Not much data, so I wouldn't draw any conclusions. The shot with an extra paper in the middle tasted more full bodied. I much preferred it, but again, 1 data point is not enough.

I was wondering if anybody tried a similar experiment

This sounds similar in conclusion to a simpler process tested by Scott Rao which was to use two paper filters - one on top and one on bottom. Doesn't it?

This sounds similar in conclusion to a simpler process tested by Scott Rao which was to use two paper filters - one on top and one on bottom. Doesn't it?

Yes, I think so. I used 3 filters. One extra in the middle. My rationale was that if the paper filter at the bottom works by reducing channeling, an extra paper filter in the middle should do even better, because you restrict the fines moving from the top half to the bottom half. Moreover, the channels will be independent, ie, no big channel from top to bottom of the puck.

cafeIKE wrote:Taste? We don't need no stinking taste.
We got numbers!

Sorry that so many who were offended... it was not my intent. "...? We don't need no stinking..." has been used countless times for humor in films since 'The Treasure of the Sierra Madre' in 1948.

Katran wrote: the paper filter at the bottom works by reducing channeling, an extra paper filter in the middle should do even better, because you restrict the fines moving from the top half to the bottom half. Moreover, the channels will be independent, ie, no big channel from top to bottom of the puck.

FWIW, I've figured out how to get channeling, even with filter paper at the bottom of the basket: Dark roast (oily) + extra-fine (espresso) grind + egregious tamping pressure. Prolly cracked even before the water hits it.

I originally used the paper-on-the-bottom to eliminate the mess from (potential) jets coming off the naked portafilter - works great in that respect. However, I'm fuzzy on what sort of mechanism would mitigate channeling before it reaches the paper - how does that work?

I already use filter paper at the top and bottom of the puck, and now I'm inspired to try putting one in the middle. Why OCD/tamp before adding the top layer?

cafeIKE wrote:Taste? We don't need no stinking taste.
We got numbers!

That's too funny!

While I admit I haven't tried this, it's sounding awfully close to a pressurized portafilter...

I did it - I layer-caked my puck. TLDR: better.

18g of dark (Full City) roast in a VST basket, liberally RDT'd the beans, which produces a fluffy/granular texture. Extra-fine grind, which I've observed can often 'poly-channel', as in a plethora of small channels such that the basket is uniformly fully wetted but still runs faster than it should.

Loaded approximately half the grind, WDT'd to level it, and laid the filter paper on top, which extended all the way to the walls of the basket. I did not tamp, as my tamper won't make it that far down the basket without jamming. Besides, I have a pet theory that excessive tamping, especially for a thin puck, can lead to cracking and subsequent channeling effects, so not going there.

Dumped rest of grind, WDT'd again, and placed the top filter paper - then I tamped it all together, pretty firmly at around 20lbs.

My preinfusion took about 50% longer than usual, but the puck appeared to wet uniformly. 9 bar shot ran long, nearly 60 seconds - but that can happen sometimes even without the mid-layer filter paper. Taste was good, albeit strong, no acidity.

Is the filter paper acting like a pressure portafilter? The paper is designed to be highly permeable to water flow, so I think it may take many more than three layers to match the resistance of a pressure portafilter. On the other hand, the papers may add just a bit of additional resistance, which might 'rescue' a sub-optimal puck - that could be what I experienced. I get that the right grind and a well-prepared puck will produce a good shot without any 'crutches' - but are the crutches harmful?

Will try this several more times to see if it's consistently better, and worth the significantly fiddlier workflow.