Brewing software

[IPA]

I have just run a couple of recipes through GW's BEER ENGINE and BrewMate and in every case the EBC is lower in BrewMate by about 25%. Does anyone know which one is right?

[Pinto]

depends on the recorded EBC units for your particular ingredients that are entered into software - the defaults are often not the same as the ingredients that you have, nor the same between software (the values of the various crystal malts and the AA% of hops being particular examples)

[Monkeybrew]

I've noticed this too, even if you edit the EBC of your ingredients to the same in both programs the final brew colours are very different!

My current bitter is 26 EBC in Beerengine and 19 EBC in Brewmate IIRC

[Graham]

The reason that the colour calculation in BeerEngine does not match other software is mainly because most software, particularly American software, is reliant upon a thing called the Morey equation, which is flawed. I have no knowledge of Brewmate, but I suspect that it also uses Morey, even though it is written by an Aussie. The Morey equation perpetuates a misconception that beer colour is not linear; that is, that it assumes that if you double the ingredients you do not get twice the colour. In fact, for all practical purposes, you do get twice the colour.

This misconception goes back to 1991/2 when the late Dr George Fix performed an "experiment" whereby he took a dark American beer and measured its absorbance (colour) as-is and at several dilutions. Fix ended up with a strange-shaped "curve" and from this he concluded that the Beer-Lambert Law, commonly known as Beer's Law, did not apply to beer and that beer colour was non-linear. Beer's Law is a law pertaining to spectrophotometric measurement and, confusingly, Beer is a person in this context. The idea behind George Fix's "experiment" was that home brewers could measure the approximate colour of their beer by diluting a dark commercial beer of known colour until it matched the home brewed beer, and then calculate its colour from the dilution required.

Other people tried to make colour prediction formulae using Fix's data, or at least incorporating Fix's non-linearity assumption, but these were somewhat unsatisfactory. They had obvious limitations and different formula covered different colour ranges. Then another worker, Dan Morey, came along and combined the various formulae into one universal formula. This became known as the Morey equation.

Unfortunately, George Fix did not know how to use a spectrophotometer properly; he was trying to use it outside of its reliable range. His laboratory technique was somewhat school-boyish and his interpretation was flawed. The flaws were noticed at the time and highlighted, but it became quite controversial because George Fix, and some of his followers, doggedly defended his results and methodology to the hilt; despite the fact that people far better qualified pointed out where he went wrong, and despite the fact that several people performed similar experiments using the same reference beer and found no deviation from Beer's Law.

So the Morey equation is wildly wrong because it is based on bad data that has had its errors compounded by other workers who tried to make the data fit the real world. It is unfortunate that these formulae still persist some twenty years later, but I think it persists because has been incorporated into so much software. If it was not for software perpetuating these ideas, they would have been dead, buried and forgotten years ago.

[IPA]

Many thanks Graham for your comprehesive reply. I have been using your Beer Engine for a couple of years now and have been very satisfied with it. It was pure curiosity that made me compare it with the Aussie programme. It started when I was trying to formulate a clone for Eldridge Pope's Royal Oak. Beer Engine gave me the colour I remember whereas BrewMate came up with a very pale beer. I see Amazon are saying your new book will be published at Christmas are they correct?
Thanks again
Ian

[darkonnis]

Brewmate is a right let down for colours. I cannot honestly remember a single time I've brewed and it has been correct, For ABV it's usually about right, but thats about it. Mind you, colour isn't something I usually worry about unless it comes out stout looking in which I have a serious look at whats going on (usually missed a decimal point)

[micmacmoc]

Grahams right about spectrophotometry and liquids. I was a colour matcher in the print trade for 20 years, we used spectrophotometers as a back up for colour testing. The spectro will shine a light at a sample, the reflected spectrum is recorded. If this sample is translucent the results will vary massively on the same sample. As the samples darken the re will be more light reflected relatively thus giving a false impression how how much darker the change of samples are to each other...from what I can remember.

The eyes will always be the best judges!

[rootsbrew]

Re: Morey's error:
Are the values produced using Morey's equation "out" but a known amount - i.e. is there a means of correcting them or is the difference better perceived like "human error": thoroughly unpredictable!?

What is interesting, is how many people clearly use Morey's methods and are completely unperturbed, judging by the size of the American HB scene and the fact Morey's assumptions are still being used!

edit: used "clearly" twice in one line - sounded a bit odd on reading it back when post was quoted!

[Pinto]

What is interesting, is how many people clearly use Morey's methods and are clearly unperturbed, judging by the size of the American HB scene and the fact Morey's assumptions are still being used!

Blind, unquestioning adherance in the USA to a theory disproven by evidence and research ??? Its the new Bible Belt !

[boingy]

What is interesting, is how many people clearly use Morey's methods and are clearly unperturbed, judging by the size of the American HB scene and the fact Morey's assumptions are still being used!

It think much of this might be explained by many folk not really worrying about the exact colour of a recipe. When I put a recipe together I know from experience whether it will be pale, dark or "middling". I don't really mind about the exact shade and if I did then I'd adjust the recipe the next time. I guess it matters more if you are trying to brew to a competition style guide but I think for many folk the colour is a fairly minor consideration compared with taste and aroma.

[Graham]

Are the values produced using Morey's equation "out" but a known amount - i.e. is there a means of correcting them or is the difference better perceived like "human error": thoroughly unpredictable!?

I see no reason why the Morey equation could not be worked backwards and its effect cancelled out. It would be better if it wasn't there in the first place though. Colour calculation is a simple ratio - nothing complicated.

It isn't helped by the fact that the Americans (ASBC) still use the old Series 52 (brown-scale) colour-glasses in their tintometer. These were devised in 1885 and are not good at dark beers/ malts or any beer/ malt with red in it. This side of the pond we devised a new set of glasses in the 1950s, known as the EBC glasses to cope with this and to follow photometric laws more closely.

Also, American home brewers tend to treat Lovibond, which is a visual method and used for malt colour, and SRM which is photometric and used for beer colour, as being equivalent. They are not really equivalent.

The differences here, however, are not as significant as those introduced by the Morey equation.

What is interesting, is how many people clearly use Morey's methods and are completely unperturbed, judging by the size of the American HB scene and the fact Morey's assumptions are still being used!

As boingy said.

Perhaps they assume that if the software says that a beer is 20 SRM, that that is what a 20 SRM beer looks like and do not question it. Also, I suspect that the majority of American home brewed beers are towards the pale end of the colour spectrum, where these errors are not so noticeable.