I was going to dovetail on the most recent SRM post below but didn’t want to hijack the post. Question: I use Brewer’s Friend or even BrunWater to calculate SRM, which I am in general taking to mean not just lightness of the beer but also hue. I ask because sometimes I brew something and don’t achieve the exact hue I was aiming for. For instance, a beer with an SRM of 16 is supposed to be reddish, but this is not always the case?
John Palmer (I believe it was him) did an experiment on this several years ago and reported results in an issue of BYO magazine. SRM is a numerical value determined in a laboratory. It was discovered that depending which malt is used, how it was roasted and kilned, mash pH, finished beer pH, boil length, concentrated vs. dilute boil gravity, etc., all made a significant difference between the actual lab SRM versus the actual color or hue of the beer, and the two expected values do NOT always line up at all. There were examples of like 5 SRM beers that were dark brown, 20 SRM beers that were pale in color, etc.
So yeah. It’s totally possible to get oddball color beers that don’t line up with SRM calculations in any way. It happens to me every once in a great while. For the most part we can predict color with pretty good certainty. But sometimes under some conditions, the color will come way out of whack and can leave us wondering why. And there are probably dozens of different variables so it can be extremely difficult to figure out why it happened. When it happens to me, I just shrug and say oh well. It usually doesn’t happen twice, or I can jot it down in my notes to help correct the recipe for the next time it is brewed.
SRM is only a measure of color density – how light or dark the sample is, based on how much light it lets through. It does not indicate hue in the sense of red, yellow, gray etc. tint. This is its big failing. Two samples of exactly the same SRM but different color spectra can look very different to the human eye, and there’s no system of beer color rating that accounts for this.
SRM stands for Standard Referance Method. This method was developed many years ago as a means of determining the color of a beer. It is meant as a guide more than anything. The actual results of your brew are determined by the Lovibond of the malts, and adjuncts in addition to the Maillard reaction contributed by these ingredients along with the effects of boiling the wort. Slight variations from prediction is not unusual.
There are many good articles on line on this subject.
Lovibond was a system developed 1893 but also just intended to measure color density, not hue. For some reason the scale is still used by maltsters to describe malt color (based on the color density of the wort from a Congress mash,) but not beer. SRM was developed in 1958 by the American Society of Brewing Chemists as a replacement for the Lovibond scale in actual measurement of beer color. In place of Lovibond’s optical filter slides to measure color density against a sample, it electronically measures absorbance of 430nm wavelength light in a 0.5" sample jar. It just happened that when they multiplied the original SRM value by 10 it closely matched °L, so that’s how it’s now expressed. The EBC scale uses a similar methodology to the SRM. So all of these systems are designed to express how much light is absorbed by a sample, not what color it looks like to the human eye and brain. And they are intended for actual measurement, not prediction.
The problem we are up against is that the scale used to rate malt color is not strictly compatible with the one for beer color, and we are misusing them. No system has ever been devised to a) describe actual beer color as opposed to density or b) predict beer color based on ingredients. This is what we as homebrewers want, but it doesn’t exist. It is similar to the situation with bitterness. There is no actual system for prediction of IBUs because IBU is purely a laboratory value measured after the fact. The professionals will never develop the kind of prediction formulas we want, because they are not interested in anything but their QC labs being able to simply measure each finished batch, to confirm that it is on spec or how to blend batches to meet spec. For both color and bitterness, there are plenty of hombrewer-devised “calculators” and “formulas.” But we are still trying to make a tool do something it isn’t intended to do. So we may as well use these calculators to the best of our ability, and accept that they simply won’t always exactly reflect reality.
I remember an article years back in BYO where Briess had made beers or wort with the same SRM but different hues.
Did a search, found this paper from Breiss. One picture that illustrates the difference. Haven’t read it yet.
I see the picture, great illustration. You could put a sample of blue water through the test and it would give you an SRM value identical to a given beer, but would not hint that one is blue and one is yellow. Look forward to reading the paper.
Excellent find. I’m gonna read that in detail later. So dang busy this month…
There should be a beer color rating system.
There is, from BJCP. They ignore the scientific laboratory method and developed a color card that defines what color each number on the SRM should look like according to them. It’s a bit better than what we had before, certainly a more objective standard.
Or just describe the hues that are perceived with word descriptions, rather than instrument testing? I find thatI rarely use the numbers anymore other than at the very ends of the spectrum, which are straw and black, anyway.
In any event Cheers to those who think this deeply about beer color, though. And there is a whole nother issue about oxidation and long boil effects on the same wort produced from batch to batch…
Exactly. These analytic methods were developed for industrial brewers who basically make just one beer, and need it to be identical in every glass. They don’t need to describe hue, as it’s all the same, just slight variations in intensity. They recognize what we should just accept, that due to variation in materials and process beyond our control, you can brew the same recipe over and over, and it’s a little different every time. So they need to quantify those variations, in order to calculate how to blend batches before packaging to even them out. The same goes for bitterness. We shouldn’t be hung up on numbers (especially those that are meaningless the way we misuse them,) and should just pay attention to our senses. They, along with experience, are the only tools we really need for recipe development. But any mathematical shorthand you might adopt is fine, as long as it means something to you.
Well I learned something for sure. Thanks for the links. What I’ve learned today, in a nutshell:
There are two distinct ways in which a specialty malt are colored. Lightly kilned malts and crystal malts have the typical range of visual color from straw, gold, amber, red, brown, to black. Briess calls these the “red” malts, and rightly so. However, in malts roasted or kilned to >120 L, the red is destroyed and the range goes more from straw, burnt orange, brown, to black more quickly and will appear darker for the same amount of malt as for a lighter kilned malt. Briess refers to these as the “black” malts, even though some are clearly light to dark brown in color. The spectra for the red vs. “black” malts are two entirely different things, and any particular malt product produces mainly just one or the other color spectrum, not both. The SRM scale and calculators are thus flawed and consistent only with use of the red malts and not the roasted “black” malts.
Fascinating.
Also, FYI, the BJCP color guide appears to be somewhat of a combination of the red and “black” malts, with leanings more towards the red side. Thus, recipe calculators which all appear to be based on use of red malts will be pretty consistent with the BJCP color guide, while beers using a higher proportion of roasted or “black” malts will tend to be darker than predicted by software.
That’s what I get out of all of it. I’ll start putting it to practical use immediately. 
Briess may not be the go-to malt for my beer but they sure have taught me a lot about malt flavor and now color in articles like these and presentations like those given at NHC.
Hey, it seems most of the important hop research we benefit from was done by Miller back in the 90s. So these guys are good for something! ;D
That is similar to what I remember in BYO. The link I referenced was a MBAA paper, so more technical.
Thanks.
Ditto! I don’t love their malt, regardless of the fact that they’re located only 1 hour away from here… but they sure are sharp folks anyway. When my nuclear job goes kaput a few years from now, I’m seriously considering applying for a job there. Win-win.
I vaguely remember going to a seminar at NHC on this subject by a speaker from Briess. As I recall it was at least ten minutes of excellent information crammed into an hour session.
You recall correctly.