The variation shows differences from startup of equipment through to steady state - while you can't see it in the way I displayed it, there's clearly a progression of grinding to results in a 'startup curve'.  The fat being measured is free fat, where the cocoa cells have burst and are leaking the fat out.  The process neither destroys nor creates fat, but if it's not available (ie still in an intact cell), it doesn't help with viscosity reduction or flavor release, yet your'e still paying for it, so the goal is to release as much fat as possible w/o destroying flavor with heat or volatilization.   I'd argue that the results are quite relevant, especially for moisture.  There's a half a percent delta in moisture content, and moisture impacts *everything* including flavor.  One of the interesting things about moisture is that as it leaves, it pulls other flavors with it - sometimes that's a good thing, other times it's less good.   All of these have unimodal particle size distributions (at steady state, the startup curve shows similar differences in the distribution but evens out).

Below is an example of a flavor compound chart I made for someone recently as part of a more complex flavor management discussion.  It may be quite hard to process without the accompanying discussion around the rest of it (this is only part of 1 page and there's 50 total pages walking through it) - but essentially read it as a series of both fat and water soluble flavors in chocolate over a range of coaching temperatures, and how easily they are or aren't driven off as a function of their boiling point and vapor pressure. Things at the top leave most easily and are the first to go, and as you go down the list it takes increasingly more energy to remove them.   The scale is logarithmic. 

Regardless of processing kit - it's almost always going to boil down to managing your temperatures, your moisture, and your airflow - as all of those things directly impact the retention - or destruction - of flavor compounds.  Remember that almost all of the things you describe as flavor are actually aromas - there are very few 'tastants' (flavors that don't evaporate) - and aromas are, by definition - volatile - they evaporate - it's literally how you perceive them.  The more energy you put into your system (be it from heat or air) - the more likely those things are to be removed.  Remember that just because a system has separate healing and cooling circuits - that doesn't tell you anything about heat removal capacity. I've seen plenty of systems that have cooling capacity on paper, but in reality generated heat faster than the cooler could remove it - which means a net heat accumulation over time, or a higher heat plateau than desired once the cooling system maxes out. 

If your question is can a ball mill match the output of a melange?  hm.  I'm not sure I have a good answer for you, other than to say that matching of products made on different types of equipment is usually quite difficult.   

updated by @sebastian: 10/18/26 06:56:57

  Thanks so much for your take on this, Sebastian.  I realize there are specifics involved and it's hard to generalize broadly when there are so many variables involved in machine operation and engineering, but I imagine the presentation's chart was made targeting either general characteristics or those specific to the company's offering.  The latter I find hard to believe, especially since their ball mill has "separate heating and cooling circuits."   I presume they have decent temperature control and thus their ball mills, as you say, should be quite wonderful.
  Thanks for the chart.  That's interesting that it appears the ball mills yielded less fat.  I would have liked to see what the resulting particle size distribution curves looked like.  That's about the only way I can think the resulting fat content could differ.  Although, with n=30, would the outcome have been different with n=300?  There was a lot of variation in the results!  And even if not, I can't imagine being able to discern a difference between the averages of 54.96% and 55.43% fat.  The conclusions I'd draw from this study is that differences in resulting fat and moisture are irrelevant. 
  You mentioned there were measurable differences in other compounds, though.  A question just for fun:  If we just assume that a melanger produces better flavor and attribute it to the slower volatilization of flavor compounds (not saying this is the case!), do you think careful temp and air control during a subsequent conching program could yield roughly the same results?
  The size of the target is always important in real life:  People would notice if I went sky diving and landed in the wrong field, but no one would see if I failed to depress a blade of grass I was aiming for.  This is theoretical and purely because I'm interested.

Let me toss my two cents in here..equipment type (and even model) absolutely impacts physiochemical attributes.  I'm not sure how image posting will show here, but I've attached a snippet of a moisture and fat comparison looking at how three different mills impact cocoa nib milling - each sample ran 30 replicates (I think, I'm going from memory here) and the line is their average. It's the exact same material they're milling - same lot, same roast, same everything - and you can easily see there's differences.  While I'm not going to share the flavor or the other physiochemical attributes - those were all measurably and meaningfully different as well.

I'll share my thoughts point by point from your photo.

1. This suggests that melanges have better quality, potentially on flavor alone?  I can't tell if they're meaning to suggest that they're overall better on quality, or just on flavor.  For ball mills, these results are dependent upon things like tip speed, temperature regulation, and media composition.  Not all ball mills are created equal, nor is their media.  There's a pretty large range of metallic media composition available to you, as well as ceramic - and the specifics of those will greatly impact flavor and thermal regulation.  Your media composition selection should be tied to the tip speed (faster tip speed impacts temperature, fines, speed, etc), and the ability to remove heat via your water or glycol jacket are incredibly important to flavor. 
2. Fast vs slow - I'm assuming this is rate.  lbs(kgs)/hour.  Ball mills will almost always beat stone mills here.  Note that there are a category of vertical stone mills that I suspect you all are not familiar with that may be the exception here, as they can be quite high throughput, but are typically reserved for liquor grinding, and I highly doubt that's what they're referring to.
3. Details/notes - I've no idea what they're trying to say here.  if it's that ball mills are worse at preserving delicate flavor notes and melanges are better - I'd say that's a wild generalization, and highly dependant upon how they're operated.  I've just put together a really quite detailed compositional analysis of flavor compounds in chocolate and how they're boiling point and vapor pressure translate to equipment operation/retention - and it's pretty interesting.   Ability to regulate your temperature is everything.  Ball mills will generate a LOT more heat than stone mills. Make sure you've got the capacity to remove it and ball mills can be quite wonderful.
4. See # 2.
5. Expensive vs cheaper - details matter here.  I'd probably start with defining what product you want to make, what attributes are important to you, and how much you want to make of it.  A simply $/kg/hr metric will quickly answer this for you.
6. Does that say min cocoa butter?  My experience is different than Clays, and as a result I'd say that ball mills can absolutely run lower total fat products than melanges, but of course it depends on the details of both (motor sizing), rate expectations, and it likely requires some specialized knowledge in ingredient selection and management. 
   
   
   
   

updated by @sebastian: 10/11/26 16:05:55

Thanks, Clay.  I've sent you a link to the source video via e-mail.  There wasn't a lot of context around this part of the video, but I don't know how else to interpret it.  It was filmed as part of one of their bean-to-bar classes.  I don't recognize the instructor, but maybe he's not part of the company and this is simply his opinion.

 

I am kind of surprised that a company that sells both ball mills and melangers expresses the differences (and drawbacks) between their own products this way.(Maybe this applies to their products, not all products in the respective category?)

Of course, I am only looking at one frame of a video so it’s impossible for me to know the full context. If you want to post the URL of the video I will be happy to watch the full thing and respond – or send it to me via DM or email (I just sent you an email).

This is unlikely to be the considered opinion of the instructor – it is likely that this is based on how and what they were taught about the different technologies, especially within a competitive comparison matrix,

When I was representing FBM I spent a lot of time understanding what Selmi, Rockgate Group (Packint/Gami), Pomati, Ideo Tecnica, Tecno3, Betec, and others were doing and how they compared with FBM so I could represent everyone fairly. Even though I represented FBM I knew that FBM might not be the best option for that client for their application and situation.

It is almost impossible to do too much due diligence, unless it leads to analysis paralysis.

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@DiscoverChoc

What you say is what I suspected, but you'll be surprised to note that this was indeed in a marketing video for a manufacturer (and a very well- known one!) who sells both melangers and ball mills.  This marketing video depicted a bean-to-bar class the manufacturer offers.  I imagine no one else even noticed the details written on the pad in the video since it was present only fleetingly.

I intend to purchas a ball mill and conche from this company, but before I take the plunge, I thought it wise to see if these were just the opinions of the instructor of that course or if this was an undercurrent in the industry.  Thanks for your insight, Clay.

Skwerl: I am guessing that this marketing video was not published by a ball mill manufacturer <grin>.

Taste/Quality. If there is a metal taste from chocolate made in a ball mill the operator is likely not maintaining the mill properly, Used and maintained properly there should be no taste of metal in the chocolate. I often recommend that users of melangers lick the stones so they can identify if they are refining stone dust into their chocolate. (This is easy to do if too much pressure is applied. If there is, for example, a visible groove on the base the stone has to be somewhere – and that somewhere is in the chocolate. And stone dust can’t be removed, while small bits of metal can – with a magnetic trap.

Ball mills are not conches, so a separate conching step is required. I would never expect what comes out of a ball mill to be finished from a flavor perspective.

And quality is a different issue entirely.

Fast Food/Slow Food. Melangers are the original “fast food” machine for chocolate manufacturing. Before the melanger was the metate, So while I understand the point, I don’t agree with it 100% – it’s not one I would ever use.

Details/Notes. This is down to proper usage. I’ve tasted a lot of both under- and over-processed chocolate out of a melanger. Any technology can produce “good” as well as “bad” chocolate. This is too broad a statement to take seriously.

Faster/Slower. This is a restatement of the Fast Food/Slow Food argument. Slower is not always better – but, because time is money, slower may be more expensive.

Expensive/Cheaper. This is not a true dichotomy at a fundamental level, and it may not be true, I can purchase a 100L ball mill for ~€25,000 and a restored vintage 50L Lehmann melanger for $35,000. There needs to be a lot more specificity. You also need to consider throughput in the cost structure, not just CapEx. A 100L ball mill might produce 300L of finished chocolate/day. It could take 12-18 days for the 50L Lehmann melanger to produce the same amount.

Fat content. This is generally true, the minimum cocoa butter content in a ball mill is likely higher than the minimum in a melanger. But there is an implied value judgment here that less cocoa butter is always better – which is demonstrably not the case.

Pre-refining. It is less efficient to refine anything through > 1 order of magnitude in a single pass no matter what technology is used. Melangers are more efficient (faster) if paste is added, not nibs. But they suck as conches because that is not what they were designed for. Most makers use melangers because:

• You can get small ones for little money
• They are straightforward to understand
• Recipes tend to scale from small melangers to bigger ones
• Only one machine is required (less materials handling)

But those points do not automagically mean they are a better choice for every situation. When the Mast Bros had a line of 20 Cocoatown ECGC 65s in a rown in their Williamsburg shop they picked the least environmentally responsible way to scale production I can imagine – energy usage, heat produced, etc, etc

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@DiscoverChoc

  Below is a capture I took from a brief shot in a marketing video from a manufacturer of both melangers and ball mills for commercial chocolate production.  I was a bit surprised to see "Better taste/quality" and "Details (+) Aromas" under the melanger column and "Metal taste" under the ball mill.  This is clearly stating that there's a flavor quality discrepancy between ball mill and stone ground chocolate.

  This is not something I've heard before.  When I asked the manufacturer about this, I did not really receive a direct explanation.  Is this a commonly held (if not necessarily factual) opinion in the industry that I've simply not encountered?