Sieving in fact is precise and repeatable - and it's the gold standard method for particle size analysis since a long time. Green coffee is also sorted more or less same way and many other things in different industries (sand, cement and so on). Coffee grinds aren't any special. In a typical drip brew fines (say <400 average diameter) typically make less than 10% of the grounds mass, and as such have very little effect on extraction expect when affecting the flow of the water.

namelessone wrote:In a typical drip brew fines (say <400 average diameter) typically make less than 10% of the grounds mass, and as such have very little effect on extraction expect when affecting the flow of the water.

But how much of the surface area do they make up, and how much of the mass that water can easily extract flavour compounds from do fines comprise? Much more than 10% in some theories (and it is theory until it is obviously proven in some way).

For example, it's possible fines (if defined as <400um diameter for instance) give up almost all the soluble compounds, including a lot of undesirable ones, and the larger particles are what you have control over to balance the brew to how you want it to taste...

Denis wrote:Guatemala 71mm with Silver knight low fines for espresso, and on the right you see the Guatemala SB 73mm with mahlkonig unimodal burrs

Did I get it right?
The Gua SB has not anymore the SSP High Uniformity burrs as shown in the earlier posts?

Did you change from espresso to filter with Gua SB grinder?

How comes?

namelessone wrote:Sieving in fact is precise and repeatable - and it's the gold standard method for particle size analysis since a long time. Green coffee is also sorted more or less same way and many other things in different industries (sand, cement and so on). Coffee grinds aren't any special. In a typical drip brew fines (say <400 average diameter) typically make less than 10% of the grounds mass, and as such have very little effect on extraction expect when affecting the flow of the water.

The point I was making is that if you're talking a typical filter grind size, where you're something like an 800-1000ish median diameter target, using 400 as the lower "cutoff" of acceptable grind size range is somewhat useful for a general idea of a grinders quality, but isnt nearly precise enough if one is analyzing geometry differences or unimodality of a grinder. The difference between say 400 and 1200 or even 400 and 800 is absolutely massive in terms of exposed surface area for extraction and there are a ton of potential particle spreads within that window that will taste very different.

Sieving is a standard because it's available, cheap, and approachable. Thats not saying it's not useful, but it's not very precise.

I think sieving at 400um intervals, using 4 sieves, you can get a pretty decent idea of the grind distribution, not to the detail laser diffraction can do but see my current efforts below, hopefully over the coming weeks I'll be able to add some data about a stock EK43 as well.

https://cangencer.github.io/noncumulative.html

Image analysis also has its own downsides, I tried Gagne's app but found the results not very consistent and you need a minimum 10 images to get a stable distribution, and manually merge the data. It's tedious and messy.

That said, I'm not sure if maximum uniformity is really desired for filter brews, since the type of grinders that are highly rated for brew (i.e. Comandante) are not very uniform, and for example Joachim Morceau compared the stock coffee EK43 burrs with SSP high-uniformity and found that he preferred the stock burrs, which are allegedly "lower uniformity" (see his post at SSP themselves seem to recommend "low uniformity" for brewing. Unfortunately there seems no data available for comparing their low-uniformity and high-uniformity burrs, so I can't say how the grind distribution between the two diverges. So, it's still not clear to me what the ideal grind distribution is for brewing. The reported various differences between burrs just as well might be due to placebo, different adjustment/grind setting etc.

namelessone wrote:I think sieving at 400um intervals, using 4 sieves, you can get a pretty decent idea of the grind distribution, not to the detail laser diffraction can do but see my current efforts below, hopefully over the coming weeks I'll be able to add some data about a stock EK43 as well.

https://cangencer.github.io/noncumulative.html

Image analysis also has its own downsides, I tried Gagne's app but found the results not very consistent and you need a minimum 10 images to get a stable distribution, and manually merge the data. It's tedious and messy.

That said, I'm not sure if maximum uniformity is really desired for filter brews, since the type of grinders that are highly rated for brew (i.e. Comandante) are not very uniform, and for example Joachim Morceau compared the stock coffee EK43 burrs with SSP high-uniformity and found that he preferred the lower uniformity stock burrs (see his post at SSP themselves seem to recommend "low uniformity" for brewing. Unfortunately there seems no data available for comparing their low-uniformity and high-uniformity burrs, so I can't say how the grind distribution between the two diverges. So, it's still not clear to me what the ideal grind distribution is for brewing.

Where do you see what SSP Burr's he's using? Because he said it's red speed, which is recommended for espresso not brew.

It was the high uniformity ones, since that was the only one available at that time this post was made.

First post updated with info about the burr carrier from Titus which improves the alignment even more.

Really like the looks of the custom bean hopper. Is anyone making these now?

namelessone wrote:I think sieving at 400um intervals, using 4 sieves, you can get a pretty decent idea of the grind distribution, not to the detail laser diffraction can do but see my current efforts below, hopefully over the coming weeks I'll be able to add some data about a stock EK43 as well.

https://cangencer.github.io/noncumulative.html

Image analysis also has its own downsides, I tried Gagne's app but found the results not very consistent and you need a minimum 10 images to get a stable distribution, and manually merge the data. It's tedious and messy.

That said, I'm not sure if maximum uniformity is really desired for filter brews, since the type of grinders that are highly rated for brew (i.e. Comandante) are not very uniform, and for example Joachim Morceau compared the stock coffee EK43 burrs with SSP high-uniformity and found that he preferred the stock burrs, which are allegedly "lower uniformity" (see his post at SSP themselves seem to recommend "low uniformity" for brewing. Unfortunately there seems no data available for comparing their low-uniformity and high-uniformity burrs, so I can't say how the grind distribution between the two diverges. So, it's still not clear to me what the ideal grind distribution is for brewing. The reported various differences between burrs just as well might be due to placebo, different adjustment/grind setting etc.

Instagram seems to be the portal for all scientific posts nowadays ...
Looking at that post , no reference are made to extraction yields, yet anecdotally they talk of the SSP extracting more.
Of course these things are always pretty subjective, but I find these insta bait posts a little loose in premise .
Talk of bitter in brews, we don't know if the stock and the v60 EY were vastly different or not. Perhaps that coffee used just isn't suitable for a higher EY threshold. For me there isn't enough detail to make head nor tails of the preference expressed.
Was it blind taste testing, among how many people?
It's fine fine to state preference as Joachim does, because in the end, that's all you can really say. I like .prefer this over this but that post also mixes some reference to objective data ( extracts more ) without telling us what it is. They are both aligned? To what tolerance? If you are going to say aligned then why not define if they are the same ?