Calculating Extract from Specific Gravity

[joshbuhl]

I‘ve published an article in the Journal of the American Society of Brewing Chemists on the equations that relate Extract to Specific Gravity.

Both a review of the equations historically used for this relation as well as the derivation of new, more accurate equations are given in the article. Several heretofore unknown correlations amongst the historical equations and their connection to the so-called partial specific volumes of sucrose and water in solution are shown.

Below are links to the official published version, a pre-print version that can be accessed for free, and also a TL;DR summary of the main useful result for brewers.

The official, reviewed and published version can be found here: https://doi.org/10.1080/03610470.2023.2267947

You can access a free pre-print of the AOM (Author‘s Original Manuscript) here: OSF

(TL;DR)
========

If you want to calculate Extract from specific gravity measured at 20°C, then you should use the following equation:

E(°P) = 260.4 -260.4/SG

which will give you the extract value in degrees Plato(°P) as found in the ASBC tables on Wort, Beer and Brewing Sugars corresponding to a Specific Gravity value measured at 20°C with a mean absolute error of less than 0.0198°P and a maximal error of only 0.05°P.

The Rule-of-thumb conversion which we all know as „gravity points divided by four“, which mathematically is

E(°P)= 1000*(SG-1)/4 = 250*SG -250,

is simply a linear approximation to the correct conversion equation, and although not really simpler than the equation I recommend above (and derive from physical principles in the article) is much less accurate, especially above SG = 1.040.

 

[VikeMan]

Cool. I wish I had time read this today, but a quick question...

We all know that "gravity points divided by four" is only a rough order of magnitude approximation.
Does your paper demonstrate that your new formula is more accurate than specific other formulae such as...

E(°P) = (-1 * 616.868) + (1111.14 * (SG)) - (630.272 * (SG)^2) + (135.997 * (SG)^3)

BTW, nice first post!

 

[joshbuhl]

Cool. I wish I had time read this today, but a quick question...

We all know that "gravity points divided by four" is only a rough order of magnitude approximation.
Does your paper demonstrate that your new formula is more accurate than specific other formulae such as...

E(°P) = (-1 * 616.868) + (1111.14 * (SG)) - (630.272 * (SG)^2) + (135.997 * (SG)^3)

BTW, nice first post!

Hi,

No, it‘s not. That equation (which I believe originates from „Braukaiser“) is a third-order best-fit equation. In my paper I actually also give a better (i.e. with lower root mean squared error (RMSE)) third order best-fit equation:

E(°P) = -613.9427 +1102.9079*SG -622.5576*SG^2 +133.5892*SG^3

which has a RMSE of only 5.3*10^(-4)°P.

However, that wasn‘t the point. You can get best-fit equations to be as exact as you‘d like by increasing the order. Best-fit equations, while great for automated computations, though, as above, usually have long coefficients, are impossible to remember, and are physically meaningless. The simple equation E = 260.4 - 260.4/SG is physically correct (it‘s derived from definitions and physically definitions along with some approximations), which ist why it gives astoundingly accurate results with just one simple paramenter. The number 260.4 is the so-called „modulus“ and I showed that this number is related to the partial specific volume of sucrose in water and calculated the average value for this based on F. Plato‘s original data from 1900. The main approximation in the formula above is using an average value for the non-constant partial specific volumes of sucrose and water in solution. The formula is simple to remember and should be accurate enough for any home-brewing application.

Cheers!

Josh

 

[dmtaylor]

This is super nerdy stuff. I just reviewed and digested all of it for a couple hours. I love how simple AND accurate the simple formula is! Beautiful. Also I love how the SG changed from a multiplier (in the common old "divide by 4" modulus 250 formula) to a reciprical in the new formula. Mind boggling for mathematics nerds, but it really works.

Also interesting -- when I compared the inverse in a spreadsheet with a list of data I had on hand, for which I'd previously used a third order best fit formula... very few numbers changed, and the few that did changed just 0.001.

So...... for most intents & purposes... in my opinion, it doesn't matter which of the best-fit formulae we use, as they are all quite accurate, except we should perhaps avoid the old "divide by 4" rule of thumb IF accuracy is desired. I'll stick with my existing spreadsheet for now. But if I were developing a new one, or wanted to memorize a quick & dirty... dang, that 260.4 modulus sure is easy to memorize. Pretty cool.

EDIT: By the way... For those who want to go the other way around, converting Plato (or Brix!?) to SG, looks like this:

SG = 260.4 / (260.4 - P)

Cheers all.

[/geekout]

 

[VikeMan]

[/geekout]

Quitter.

 

[dmtaylor]

Quitter.

I couldn't quit actually. I just edited.

 

[doug293cz]

Looking forward to plowing thru this.

Brew on

 

[joshbuhl]

This is super nerdy stuff. I just reviewed and digested all of it for a couple hours. I love how simple AND accurate the simple formula is! Beautiful. Also I love how the SG changed from a multiplier (in the common old "divide by 4" modulus 250 formula) to a reciprical in the new formula. Mind boggling for mathematics nerds, but it really works.

Yes! I also liked how the new equation has the same „complexity“ as the rule of thumb (both have just two terms and only one constant), yet is far more accurate! I see you really did digest it!

Cheers, DM!

 

[joshbuhl]

Hi,

No, it‘s not. That equation (which I believe originates from „Braukaiser“) is a third-order best-fit equation. In my paper I actually also give a better (i.e. with lower root mean squared error (RMSE)) third order best-fit equation:

Hi again Vikeman,

Just it case it wasn’t clear what I meant with the first sentence, I should have written „No, it doesn‘t, because it‘s not“ (i.e. „No, it doesn‘t show that it‘s more accurate, because it‘s not more accurate (than that 3rd-order best-fit...)“).

Btw, I looked it up in my notes and can confirm that your 3rd-order best-fit was (first?) posted online by „Braukaiser“. That equation has an RMSE of about 7.236x10^(-4)°P when compared to the ASBC tables, if it interests you.

cheers,

Josh

 

[ncbrewer]

To get it back to a form that us non-mathematicians can use and memorize:
Plato = gravity points x 0.26
And regardless of whether or not the more accurate 0.2604 factor is needed, many thanks for this simplified formula.

Edited 1/9/24: Ignore this post - brain fart.

 

[VikeMan]

To get it back to a form that us non-mathematicians can use and memorize:
Plato = gravity points x 0.26
And regardless of whether or not the more accurate 0.2604 factor is needed, many thanks for this simplified formula.

I might be missing something, but...

E(°P) = 260.4 -260.4/SG
-----and----
E(°P) = gravity points x 0.2604

...are not the same. In fact I think the latter gets us farther away from the former than the old simplified "divide by 4" would.

 

[joshbuhl]

To get it back to a form that us non-mathematicians can use and memorize:
Plato = gravity points x 0.26
And regardless of whether or not the more accurate 0.2604 factor is needed, many thanks for this simplified formula.

hi,

thanks for trying to simplify this, but unfortunately I must beg to differ. Your equation isn‘t correct.

The regular rule-of-thumb „gravity points divided by four“ is the same as „gravity points times one fourth“, so „gravity points x 0.25“. Your equation is basically the same but even less accurate having an average error more than twice as large as the rule of thumb.

If you have a look at my equation, it‘s not 0.2604 x gravity points, which would be

E = 260.4*SG -260.4.

(„Gravity points“ are 1000*(SG-1), so 0.2604xGrav.Pts would be 0.2604*1000*(SG-1)=260.4*(SG-1)=260.4*SG-260.4).


it‘s

E = 260.4 - 260.4/SG

these may look similar, but they are very different because in mine, you *divide* the -260.4 by SG instead of multiplying the +260.4 by it.

cheers

 

[ncbrewer]

My bad. I think dyslexia kicked in. Many apologies for interrupting an otherwise interesting and worthwhile discussion.

 

[threeoten]

This is for the most part WAY over my head, so maybe this is a dumb question, but what's the practical benefit on the homebrew scale of converting SG to plato in this way? And what does it tell you that SG doesn't?

 

[dmtaylor]

This is for the most part WAY over my head, so maybe this is a dumb question, but what's the practical benefit on the homebrew scale of converting SG to plato in this way? And what does it tell you that SG doesn't?

Well...

This can come in handy for those who like to research recipes from commercial breweries to try to brew them at home, since the industry standard many places is not SG but Plato. Talk to a commercial brewer, past or present, and they might give you data in Plato but not SG. The new formula makes it easy to convert back and forth. Previous so-called best-fit formulae exist but can be extremely complex, basically requiring software or web access to calculate. The new formula is quick, easy to memorize and punch into a calculator on your phone, AND accurate.

Also, a lot of homebrewers use refractometers instead of hydrometers, along with conversion tools like this one, which allows accurate measurements and estimation of ABV with just a couple drops of wort or beer, instead of having to fill a cylinder (traditional hydrometer) or spend more than a hundred bucks on a floating hydrometer (like a Tilt). Refractometers measure in Brix which is virtually identical to and interchangeable with Plato. So this will work with refractometers for those who want to convert between Brix (basically same as Plato) and SG.

And, like I keep saying... the new formula is super accurate. There is an old "divide by 4" estimate, which is still super handy as a quick swaggy method, but it's a little less accurate, might be off by a couple gravity points, especially at high gravities. Some people care about accuracy and don't want to be off. People who care the most should probably measure only with a traditional hydrometer. But for those of us who are somewhere in between... this is a quick and easy way to get accurate results on the fly by punching a few numbers into a calculator.

In truth... hardly anybody is going to care about this stuff except brewers who are math nerds. Tee-hee. To a few of us, this is pretty cool. Even beautiful. To the average Joe Homebrewer... please disperse, nothing to see here.

 

[doug293cz]

Refractometers measure in Brix which is virtually identical to and interchangeable with Plato. So this will work with refractometers for those who want to convert between Brix (basically same as Plato) and SG.

Brix and Plato are both defined as the weight percent sucrose in a sucrose/water solution. Brix refractometers are calibrated to read Brix of sucrose/water solutions. However wort is not a sucrose/water solution, so when measuring wort, a correction factor needs to be used get an accurate Brix or Plato equivalent to SG value. This is because the dissolved solids in wort affect index of refraction slightly differently than does sucrose. A typical correction factor is 1.04. See this calculator, for example, to see how the correction factor affects the Plato and SG derived from a Brix refractometer. Plato, which is usually measured with a hydrometer with a Plato correlated scale, does not have the same issue, since the dissolved solids in wort affect the SG in almost exactly the same way as does sucrose, so no correction factor is needed.

Brew on

 

[dmtaylor]

Brix and Plato are both defined as the weight percent sucrose in a sucrose/water solution. Brix refractometers are calibrated to read Brix of sucrose/water solutions. However wort is not a sucrose/water solution, so when measuring wort, a correction factor needs to be used get an accurate Brix or Plato equivalent to SG value. This is because the dissolved solids in wort affect index of refraction slightly differently than does sucrose. A typical correction factor is 1.04. See this calculator, for example, to see how the correction factor affects the Plato and SG derived from a Brix refractometer. Plato, which is usually measured with a hydrometer with a Plato correlated scale, does not have the same issue, since the dissolved solids in wort affect the SG in almost exactly the same way as does sucrose, so no correction factor is needed.

Brew on

My correction factor for my refractometer is about... 1.00. Verified by me after many dozens of batches. Maybe I'm just one of the lucky ones. YMMV. Cheers.

 

[doug293cz]

My correction factor for my refractometer is about... 1.00. Verified by me after many dozens of batches. Maybe I'm just one of the lucky ones. YMMV. Cheers.

Yes, I have heard of this before. My refract does have a 1.04 correction factor. I probably should have used a qualifier like "often" when talking about needing a correction factor. Luckily, most Brix calculators allow you to enter the correction factor that matches your refractometer.

I must say I don't understand the physics behind why some refracts need correction factors and some don't. Anyone know? Is it possible that some refracts are made especially for measuring beer wort, and have the correction factor built in?

Brew on

 

[dmtaylor]

I must say I don't understand the physics behind why some refracts need correction factors and some don't. Anyone know?

My guess is: imperfect prisms. I mean, these things only cost like $10 sometimes. I think mine was like $12. In that ballpark. For that price, I don't expect a perfect prism to be traceable to NIST or anything. Somebody else's correction factor is probably 1.08 to make up for mine being on the happy end.

 

[joshbuhl]

Well...

This can come in handy for those who like to research recipes from commercial breweries to try to brew them at home, since the industry standard many places is not SG but Plato. Talk to a commercial brewer, past or present, and they might give you data in Plato but not SG. The new formula makes it easy to convert back and forth. Previous so-called best-fit formulae exist but can be extremely complex, basically requiring software or web access to calculate. The new formula is quick, easy to memorize and punch into a calculator on your phone, AND accurate.

Also, a lot of homebrewers use refractometers instead of hydrometers, along with conversion tools like this one, which allows accurate measurements and estimation of ABV with just a couple drops of wort or beer, instead of having to fill a cylinder (traditional hydrometer) or spend more than a hundred bucks on a floating hydrometer (like a Tilt). Refractometers measure in Brix which is virtually identical to and interchangeable with Plato. So this will work with refractometers for those who want to convert between Brix (basically same as Plato) and SG.

And, like I keep saying... the new formula is super accurate. There is an old "divide by 4" estimate, which is still super handy as a quick swaggy method, but it's a little less accurate, might be off by a couple gravity points, especially at high gravities. Some people care about accuracy and don't want to be off. People who care the most should probably measure only with a traditional hydrometer. But for those of us who are somewhere in between... this is a quick and easy way to get accurate results on the fly by punching a few numbers into a calculator.

In truth... hardly anybody is going to care about this stuff except brewers who are math nerds. Tee-hee. To a few of us, this is pretty cool. Even beautiful. To the average Joe Homebrewer... please disperse, nothing to see here.

Excellent reply!

Cheers!

 

[doug293cz]

My guess is: imperfect prisms. I mean, these things only cost like $10 sometimes. I think mine was like $12. In that ballpark. For that price, I don't expect a perfect prism to be traceable to NIST or anything. Somebody else's correction factor is probably 1.08 to make up for mine being on the happy end.

It's pretty easy to make calibration solutions for Brix refractometers - you just need an accurate scale and some table sugar. 5 g of sucrose mixed with 95 g of distilled or RO water makes a 5 Brix solution. 10 g mixed with 90 g makes a 10 Brix solution, and so on. Below is a table for reference. The quality of the calibration solutions will depend on how accurate your weight measurements are done.

Brix​	Sucrose (g)​	Water (g)​
0​	0​	100​
5​	5​	95​
10​	10​	90​
15​	15​	85​
20​	20​	80​
25​	25​	75​

Brew on

 

[joshbuhl]

Brix and Plato are both defined as the weight percent sucrose in a sucrose/water solution. Brix refractometers are calibrated to read Brix of sucrose/water solutions. However wort is not a sucrose/water solution, so when measuring wort, a correction factor needs to be used get an accurate Brix or Plato equivalent to SG value. This is because the dissolved solids in wort affect index of refraction slightly differently than does sucrose. A typical correction factor is 1.04. See this calculator, for example, to see how the correction factor affects the Plato and SG derived from a Brix refractometer. Plato, which is usually measured with a hydrometer with a Plato correlated scale, does not have the same issue, since the dissolved solids in wort affect the SG in almost exactly the same way as does sucrose, so no correction factor is needed.

Brew on

My correction factor is also….1.00. . The correction factor that people commonly attempt to employ (and might confuse with what you’re refering to) is used to attempt to compensate for the presence of alcohol so they can use their refractometer to measure gravity in fermenting wort, as even a little alcohol drastically changes the index of refraction and makes the refractometer basically useless for measuring gravity in fermenting wort. My own attempts to determine a correction factor convinced me that this is hopeless as I found you needed a much different factor for every FG and abv.

It’s a remarkable fact that simple sucrose solutions (which as you mention are the basis for the Brix and Plato scales) so closely model the gravities of wort with all it’s complex sugars.

 

[joshbuhl]

This is for the most part WAY over my head, so maybe this is a dumb question, but what's the practical benefit on the homebrew scale of converting SG to plato in this way? And what does it tell you that SG doesn't?

Anybody who’s interested in brewing efficiency (how much of the potential sugars in the grain are you getting into your wort..) has to calculate with percent extract (the mass fraction of sugars in solution) which is what °Plato are.

The problem is that measuring °Plato directly is difficult. What is easy to measure directly is the (relative) density of the wort or beer, e.g. with a hydrometer.

Since knowing extract (°P) is mandatory for all pertinant calculations, and measuring SG is easy, it’s useful to be able to convert easily and well between the two, as there is a definite relation between the them, which is what I show in my paper.

 

[VikeMan]

My correction factor is also….1.00. . The correction factor that people commonly attempt to employ (and might confuse with what you’re refering to) is used to attempt to compensate for the presence of alcohol so they can use their refractometer to measure gravity in fermenting wort, as even a little alcohol drastically changes the index of refraction and makes the refractometer basically useless for measuring gravity in fermenting wort.

The oft cited wort correction correction factor of 1.04 (or 1.03 or 1.02, etc,) has nothing to do with compensating for the presence of alcohol. It compensates for the difference in refraction indexes between pure sucrose solutions and wort.

Refractometer calculators also compensate for the presence of alcohol (in their "FG" formulae), but that's a different correction.

It should be noted that not all worts are the same, and technically require different correction factors because of their own particular sugar/dextrin profiles. In reality, most people apply a default.

 

[joshbuhl]

The oft cited correction factor of 1.04 (or 1.03 or 1.02, etc,) has nothing to do with compensating for the presence of alcohol. It compensates for the difference in refraction indexes between pure sucrose solutions and wort.

Refractometer calculators also compensate for the presence of alcohol (in their "FG" formulae), but that's a different correction.

Yes, that was exactly my point. And the fact that trying to correct for alcohol with one factor is hopeless (online calculators for this are meaningless...).

 

[VikeMan]

Yes, that was exactly my point. And the fact that trying to correct for alcohol with one factor is hopeless (online calculators for this are meaningless...).

Was it? I thought you were saying that people try to use the WCF to compensate for alcohol, and I don't know anyone who does that.

I don't think you're trying to say that online refractometer calculators are generally useless, but just in case... Terrill's is quite good at FG for mash derived beer worts. Novotny's is quite good all the way through the fermentation. Do they give precisely the same result as a hydrometer (pretending for a moment that hydrometers and their users are perfect)? No. But they are far from useless.

And, like I keep saying... the new formula is super accurate. There is an old "divide by 4" estimate, which is still super handy as a quick swaggy method, but it's a little less accurate, might be off by a couple gravity points, especially at high gravities. Some people care about accuracy and don't want to be off. People who care the most should probably measure only with a traditional hydrometer. But for those of us who are somewhere in between... this is a quick and easy way to get accurate results on the fly by punching a few numbers into a calculator.

I can see @joshbuhl's new formula being useful for someone who wants to do the math by punching a few numbers into a standard calculator. No doubt about that. My last standalone hardware calculator has given up the ghost, so since I'm using a computer or phone anyway, it's easier for me and more accurate to punch (less) numbers into them in a preexisting spreadsheet with a full high order best fit formula. You only have to type the long formula once.

 

[VikeMan]

I must say I don't understand the physics behind why some refracts need correction factors and some don't. Anyone know?

I suspect it's like what @dmtaylor said. Imperfect prisms. (And to that, add reluctance to calibrate, or even knowing that calibration is possible.) Confusion happens when people talk about the Wort Correction Factor as a factor to correct the refractometer, when it's really a factor to correct for the wort. There really should be two up front correction factors. But since there's only one, people have to (even if not consciously) use the one factor for both purposes at once.

I believe it was Kai (Kai, forgive me if I'm wrong) who once upon a time wrote a lengthy article about using the WCF to compensate for his very bad refractometer, and I think that was the genesis of a lot of later misunderstanding about the intended purpose of the WCF.

ETA: I have thought about adding a "hardware correction factor" to BrewCipher's implementation of the Terrill and Novotny formulae. I wonder, if I also included some instructions for determining the factor, perhaps using solutions like in your post #21, if anyone would take the time to do it.

 

[dmtaylor]

Terrill's is quite good at FG for mash derived beer worts. Novotny's is quite good all the way through the fermentation. Do they give precisely the same result as a hydrometer (pretending for a moment that hydrometers and their users are perfect)? No. But they are far from useless.

After a lot of data review, I've concluded:

Terrill's formula is most accurate when SG is less than 1.014,

Novotny's formula (on Brewer's Friend) is most accurate when SG is greater than 1.014, and

When in doubt, use Novotny.

It's yet another case where using two different curves is better than relying on a single curve.

Using the proper formula, with the proper WCF, with a calibrated refractometer, and a careful and conscientious operator (you or me), I find the results for FG as calculated for a refractometer to consistently come within 0.001-0.002 of a hydrometer on average. Close enough for most intents and purposes. Again, for those who really care about extreme accuracy... it's still best to use a good old fashioned traditional hydrometer. One of those finer ones specifically intended for measuring FG is even better for that purpose.

 

[VikeMan]

Again, for those who really care about extreme accuracy... it's still best to use a good old fashioned traditional hydrometer. One of those finer ones specifically intended for measuring FG is even better for that purpose.

Yep. When I really "have to" know the exact FG, I do use a hydrometer.

One of those finer ones specifically intended for measuring FG is even better for that purpose.

Also yep. But I'd make very sure the fine hydrometer is correct first. I have only used a few of this type, but (and maybe it's just my dumb luck), they seemed to be more prone to lousy manufacture. One at a commercial brewery was awful. I had brewed a beer there, but wasn't present for the FG reading. When I saw the ABV published on the brewery's menu, I knew something was up, as I had brewed "exactly" the same beer at home many times. So I dropped by and dropped the fine hydrometer into a jar of water, and it read .05 .005 too low. What's "funny" is that it turns out the same hydrometer had been used for the FG of many beers, including a Brut IPA the brewery used for super-dry bragging rights.

 

[joshbuhl]

I thought you were saying that people try to use the WCF to compensate for alcohol, and I don't know anyone who does that.

No, I just threw out parenthetically that someone could possibly confuse the correction factor that doug mentioned (WCF) with the correction factor for alcohol content, the reason being that I misread doug comment the first time thru so just wanted to clarify it. I’ve found that there is much more mention of the alcohol correction because people attempt to use a refractometer on fermenting wort, and that’s more of a problem then the small correction for wort not being a sucrose solution.

I don't think you're trying to say that online refractometer calculators are generally useless…

Not at all, not generally, just if measuring (fermenting) wort or finished beer that contains alcohol. There is no one correction factor that will resolve that.

Terrill's is quite good at FG for mash derived beer worts. Novotny's is quite good all the way through the fermentation. Do they give precisely the same result as a hydrometer (pretending for a moment that hydrometers and their users are perfect)? No. But they are far from useless.

Well, I’d be interested in seeing the formula Novotny uses if it calculates accurately for fermenting wort….I would guess, though, that it’s using more than a simple correction factor, which is what you usually can enter in most online calculators.

spreadsheet with a full high order best fit formula. You only have to type the long formula once.

sure, if you’re using a spreadsheet, you may as well use the second or third degree best-fit polynomial, but I actually use my simple formula even in spreadsheets because it’s more than accurate enough for me and I can type it in from memory.

 

[joshbuhl]

Brix and Plato are both defined as the weight percent sucrose in a sucrose/water solution. Brix refractometers are calibrated to read Brix of sucrose/water solutions. However wort is not a sucrose/water solution, so when measuring wort, a correction factor needs to be used get an accurate Brix or Plato equivalent to SG value. This is because the dissolved solids in wort affect index of refraction slightly differently than does sucrose. A typical correction factor is 1.04. See this calculator, for example, to see how the correction factor affects the Plato and SG derived from a Brix refractometer. Plato, which is usually measured with a hydrometer with a Plato correlated scale, does not have the same issue, since the dissolved solids in wort affect the SG in almost exactly the same way as does sucrose, so no correction factor is needed.

Brew on

My refractometer reads in Brix and in SG and I’ve found that the SG reading agrees so well with my hydrometer readings (for non-fermented wort), that I have completely stopped using my hydrometer for non-fermented wort. I only use it now for the final FG reading.

 

[VikeMan]

Well, I’d be interested in seeing the formula Novotny uses if it calculates accurately for fermenting wort….I would guess, though, that it’s using more than a simple correction factor, which is what you usually can enter in most online calculators.

If you're an AHA member, you can read all about it here:
https://www.homebrewersassociation.org/zymurgy-article/revisiting-the-refractometer/
It's copyrighted, so I don't feel comfortable posting the formula. I use it (with permission) in BrewCipher, but it's hidden there.

 

[joshbuhl]

If you're an AHA member, you can read all about it here:
https://www.homebrewersassociation.org/zymurgy-article/revisiting-the-refractometer/
It's copyrighted, so I don't feel comfortable posting the formula. I use it (with permission) in BrewCipher, but it's hidden there.

Not a member...

The formula is copyrighted?!

Something about that goes against my understanding of science and knowledge...just imagine if Einstein had coprighted his formulas...then no scientists could make any relativistic calculations...

 

[joshbuhl]

If you're an AHA member, you can read all about it here:
https://www.homebrewersassociation.org/zymurgy-article/revisiting-the-refractometer/
It's copyrighted, so I don't feel comfortable posting the formula. I use it (with permission) in BrewCipher, but it's hidden there.

As an aside, I hadn't heard of BrewCipher before, but at a first glance, it looks pretty cool! Good job!

 

[MaxStout]

Not a member...

The formula is copyrighted?!

Something about that goes against my understanding of science and knowledge...just imagine if Einstein had coprighted his formulas...then no scientists could make any relativistic calculations...

No, but their publication of it and the accompanying text is. Better to be on the safe side, rather than reprinting it w/o permission. OTOH, the argument can be made that using it in a forum like this would constitute a Fair Use exception (educational use), but again, better to be careful if you're not comfortable posting something like that.

 

[dmtaylor]

I didn't have to pay anything to find the formula, honestly wasn't aware any of this was copyrighted. There's links within the link below to sign up for a free account on the Czech site to access more than one version of Petr Novotny's formulae. Once you've got it, pump it into a spreadsheet and you're good to go.

And it DOES account for fermenting wort. Nice...

https://www.homebrewersassociation.org/forum/index.php?topic=28544.0

 

[joshbuhl]

If you're an AHA member, you can read all about it here:
https://www.homebrewersassociation.org/zymurgy-article/revisiting-the-refractometer/
It's copyrighted, so I don't feel comfortable posting the formula. I use it (with permission) in BrewCipher, but it's hidden there.

Well, if these are the formulae:

Linear formula:

Quadratic formula:


(where Bx_i is the OG in °Bx and Bx_f is the current (fermenting or final) SG in °Bx)

then Novotny’s published it himself online (http://www.diversity.beer/2017/01/pocitame-nova-korekce-refraktometru.html) , so no longer any need to make a secret of it.

The results look good (see graphs on Novotny’s page) and I’m gonna give ‘em a whirl, but stuffing Novotny’s page into google translate (s. link below) seems to indicate that Novotny himself might have something of a Napoleon complex or similar. Here’s the first paragraph of Novotny’s page translated:

refractometer.cz

It's the second time in a short time that I boldly and brazenly engage in cross-hairs with the "computer gods". First it was Glenn Tinseth and his formula, which I took the liberty of generalizing and modifying into a new methodology for calculating hops (IBU), which can then be used in general for any hop (even for Whirlpool, see HERE ). Now, on the other hand, it is Sean Terrill, whose refractometer correction equation is very popular today. It is used by calculators such as Brewersfriend.com, or recently also Homebrewmap.com. Unfortunately, it sounds a lot worse for Sean today as we show why his equation is just a dead end. In addition, I will offer you a much better and more accurate equation, which is also based on professionally collected data, so its relevance will be much better. So let's get to it!​

​

And continues in that vein….

Anyways, Thanks Vikeman and DM for pointing me to this interesting resource.

https://www-diversity-beer.translat..._sl=cs&_x_tr_tl=en&_x_tr_hl=en&_x_tr_pto=wapp

 

[dmtaylor]

...Novotny himself might have something of a Napoleon complex or similar. Here’s the first paragraph of Novotny’s page translated:

I took the liberty of generalizing and modifying into a new methodology for calculating hops (IBU), which can then be used in general for any hop (even for Whirlpool, see HERE ).​

Heh... call me Napoleon then as well, because I too have independently developed my own IBU calculator which is a simplification of Tinseth and includes whirlpool.

 

[VikeMan]

Heh... call me Napoleon then as well, because I too have independently developed my own IBU calculator which is a simplification of Tinseth and includes whirlpool.

I did similar in BrewCipher back in 2013, but it was/is a complication (different curve spliced in above 65 IBUs) rather than a simplification. For post boil hops I beat BeerSmith to the punch. It helps to be able to develop and release on the same day, with a triple your money back guarantee.

 

[joshbuhl]

Heh... call me Napoleon then as well, because I too have independently developed my own IBU calculator which is a simplification of Tinseth and includes whirlpool.

I've no problem with improving Tinseth...and I notice you didn't say that you 'boldly and brazenly' developed your own IBU calculator, engaging the 'Computer Gods' whose methods are 'dead ends...'