Patagent wrote:IMO, correlation to either TDS or flow rate or both would give a much more compelling story. Not sure why people keep insisting on correlation to taste as that is highly subjective (some prefer sour, some prefer bitter, etc.). Also, having more of each grinders (n > 1) would be nice.

More than one sample per grinder imo is a must.
As to taste, it should be possible to describe flavor components in pretty objective manner. Taste is the final parameter I'm interested in, not yet another surrogate parameter like tds or whatever (which also lacks an established link to taste)

How do you propose they be sufficiently objective in their evaluation of "taste"? If you look at scientific research in such areas, it's usually done either using mass spectroscopy to find the concentration of chemical components known to be responsible for a given aspect of the taste or using tasting panels of a random sampling of people to find the changes in aspects of perceived taste. Both of these methodologies are expensive, time consuming, and may not have results that would be satisfying for you the high end consumer.

As far as the HG-1 distribution is concerned, I had a hypothesis that an extended particle size distribution (e.g., bimodal) might produce a more complex cup. I tested this somewhat in my Aeropress by mixing espresso grounds with drip grounds. The tatste was good initially but I didn't pursue this thoroughly.

bobdole2000 wrote:How do you propose they be sufficiently objective in their evaluation of "taste"? If you look at scientific research in such areas, it's usually done either using mass spectroscopy to find the concentration of chemical components known to be responsible for a given aspect of the taste or using tasting panels of a random sampling of people to find the changes in aspects of perceived taste. Both of these methodologies are expensive, time consuming, and may not have results that would be satisfying for you the high end consumer.

There are folks around with a pretty good palate, a very normal practice in any food industry, or look at perfume designers....I'd trust the taste description of a panel of 'palates' more than another surrogate parameter.

Patagent wrote:Not sure I understand what you're getting at. Are you saying sieves are not adequate tools for measuring particle size distributions? Why would they need to validate sieve-based methodology when it's been around awhile and routinely used to study particle size distribution?

I am not aware of any research on coffee grounds using sieves. Could you perhaps give me a few pointers where I could find similar studies?

Marcelnl wrote:There are folks around with a pretty good palate, a very normal practice in any food industry, or look at perfume designers....I'd trust the taste description of a panel of 'palates' more than another surrogate parameter.

What you propose would be biased towards the preferences and perceptions of the panel you assembled rather a truly objective measure.

Bluecold wrote:I am not aware of any research on coffee grounds using sieves. Could you perhaps give me a few pointers where I could find similar studies?

Sieves are a widely used methodology for determining the size distribution of solid particles in both science and industry. What would affect the use of sieves in the analysis of coffee particles?

Ditto to bobdole2000

Hopefully coffee grounds have no other magic properties than the ability to caffeinate us.. in a incredibly tasty package

bobdole2000 wrote:Sieves are a widely used methodology for determining the size distribution of solid particles in both science and industry. What would affect the use of sieves in the analysis of coffee particles?

Are you seriously replying with 'why not?'? Is that your whole argument? You can't think of a reason why it won't work, so it probably works? i'm sorry but I thought we were talking science here.

In any case, socratic coffee themselves have brought numerous arguments to the table why sieving can lead to inaccuracies. Kendall Mills has discussed laser sizing at length at the Titan Grinder Project, where sieving was also discussed. I suggest you get up to speed.

They are "making graphs" out of their test protocol, AND it does have meaning. Finding "meaning" in their graphs and verifying their data would be done by testing their sieving methodology to the most trusted way to test grind particle sizes which is ...... exactly nothing. This is the beauty in what they are doing. They are actually doing something and posting the results. Now we can work off of this to understand it. Testing this against laser analysis would be nice to see, but wouldn't give it a gold stamp of approval, since nobody trusts those graphs either.

There is a longstanding debate on laser particle analyzers and sieving, feel free to get in on that, but don't say it's a "meaningless" production of graphs based on their methodology then attack people who don't agree THAT'S not constructive dialogue - it's more like trolling.

Their data does have to be looked at from a standpoint of understanding the pros and cons of sieving. For instance check out what these guys have to say with respect to sieving vs. laser analysis:

http://www.coffeelabequipment.com/coffe ... esize.html

Because sieving is time consuming, messy, and difficult to accurately administer, many larger coffee grinding facilities turned to laser diffraction for their particle sizing needs. Laser diffraction has its' own set of pitfalls however. It does correlate fairly well with counting pixels using digital imaging (SEE BLACK CURVES ON RIGHT SIDE CHARTS), but laser diffraction has been found to skew towards the smaller size. Laser diffraction also makes the assumption that the particles are round, which we already know to be false.