E-pistle #17/01 – The Colour of Whisky

By Michel van Meersbergen, Holland (first published February 2006)

Or: ‘Our Adventures with E150-a’…

KONICA MINOLTA DIGITAL CAMERAA lot has been said about E150-a, better known as spirit caramel. About its usage to adjust colour to bring various batches within the same visual standard year-in, year-out. About the fact it will not affect the taste at all. About the purists who say caramel destroys the taste completely so all coloured whiskies should be avoided at any cost. Average Joe can’t be bothered and because he represents about 90% of the whisky market the conglomerates will continue to use caramel with legitimate commercial reasons: ‘Costumers are very attached to their brands and expect them to be the same for years and years and years.’ How right they are. The purist camp will not go with that, saying: ‘Caramel, or better E150-a is not generated by distilling or wood maturation and is therefore a non-natural thus a destructive element in whisky.’ How right they are.

The maniacs however are not that decisive and more cautious before coming out and spreading their opinions.
That’s why we took up the challenge and created ‘The Big Caramel Experiment’. Maniacs Klaus, Thomas, Serge, Charlie, Alexander and me, as well as malt sponsor for this experiment Dirk van Staden of ‘Liquid Gold’ have challenged their tastebuds in an attempt to clear up those hazy skies around the subjects on caramel. In this experiment we want to find out what caramel exactly does to whisky. Is it, as the chemists say, ‘organoleptically inert’ – not detectable by nose and taste? Does it ruin your precious dram of let’s say: an Arbeg 31yo OB Manager’s Choice for France? Is there a threshold for the use of caramel?

Liquid Caramel

Let’s start with E150-a itself. As the code suggests it’s a generic colourant.
E150 is available in four different versions ranging from E150-a to E150-d. Each is created for specific purposes. E150-c is the best known, it’s what gives your Coke or Pepsi its colour. E150-a, important for us, is quite special because it has to do its work in an extreme environment: a liquid which contains a minimum of 40% alcohol and on top of that it has to be light-proof for a longer period of time. You can’t have your whisky separating or losing colour after a few weeks on the shelves in your shop. It’s made by controlled heat treatment of sugar, with or without the presence of alkalis or acids.

Another ‘special feature of E150-a: it is not obtainable for mortals like us.
Fortunately Diageo PLC was very happy to provide us with a 200ml bottle of E150-a. E150-a is a highly viscous liquid, it took several days to settle down in the bottle. The colour is almost opaque black. Against the light a thin film of E150-a will show as a deep, brownish red colour. The smell reminded me of roasted raisins, ground coffee and burnt bread. A minute quantity of E150-a is enough to give a one litre bottle of water the looks of a ‘Dark Sherry’ maturation, so our 200ml could be used to colour a small loch!

The Test

Now for the test – in itself quite a simple one.
Three different liquids, of which two are coloured (one medium and one heavy), should be put in a certain sequence.
From not coloured to heavy coloured or the opposite way, from heavy coloured to not coloured. This has to be done three times for the nose and three times for the taste, so any lucky guesses will be smoothed out in a greater total. All nosing and tasting is done blindfolded and an assistant is to give the different coloured samples (of the same liquid) in random order. For the colouring we had to dillute one part of E150-a with four parts of water to create a more workable fluid. To make life a bit harder the taster gets one chance and one chance only to put an at random presented sample on the right spot in the sequence.

The liquids for colouring are plain water, a Lowland malt, a Highland malt, a Campbeltown malt, an Islay malt and a blend of the former four malts. I think it’s obvious that the selected malts have to come from refill casks so we can be as safe as possible to have malts that have a preserved distillery character. The honours for the Lowlands are taken by a Rosebank 1990/2003 (46%, Helen Arthur, cask #486). The highlands are represented by a Clynelish 13yo 1990/2004 (43%, Van Wees ‘Ultimate’, cask #12733). Campbeltown by Springbank 10yo 1993/2004 (50%, DL OMC, cask #628) and Islay by Bowmore 11yo 1992/2003 (46%, SigV UC, cask #4229). With this line-up we think we have a solid, general coverage of Scottish malts. In the tradition of the Malt Maniacs points are given in this test. 3 points if the arrangement was put in correct order. 1 point if only one sample was given the right place in the sequence. For 5 sessions that makes a maximum of 30 points for each taster.

Klaus has calculated an expected average score, it is shown at the bottom of the outcomes from the sampled clusters.

‘There are 6 different combinations of neat (1), medium (2) and heavy (3) caramelized samples (123, 132, 213, 231, 312 and 321). Therefore the chance to guess a bullseye is 1/6 (=1/2 point). The chance to get the position of one sample right can be calculated as follows. 1/3 (3 different samples for the first position) + 1/2 (2 different samples for the second position) +0 (no choice for the last position) – 1/6 (for the bullseye combination which has to be subtracted) = 2/3 (2/3 points). For a series of 5 runs we get an average score of  5 x (1/2 + 2/3)= 25/6 = 4.167.’

Due to excitement Alexander forgot to read these lines and renosed and retasted the samples over and over before putting them into a sequence. Even with this foul play he got one sequence false! Good for him I’d say… It means however that the scores from Alexander are not included in the results of the experiment. Luckily for all of us Alexander was the only maniac to present extensive tasting notes of the different coloured samples. You will read them at the beginning of each malt cluster.


1st Cluster: Neat and coloured springwater

First let’s try to find out what E150-a does with the smell and taste of water.
Every taster has coloured his own samples to the same specifications used for the malts.
This will also give some information about intensity of colouring and the amount needed to reach a certain level of colouring.
The following are the results from nosing and tasting springwater:

            For nosing:   For tasting:   SubTotal:
Charlie       13 pts        15 pts        28 pts
Michel        13 pts        13 pts        26 pts
Klaus          7 pts         8 pts        15 pts
Serge         13 pts        11 pts        24 pts
Thomas        11 pts        15 pts        26 pts
Expected    4.17 pts      4.17 pts      8.33 pts

Total score: 119/150 pts (Expected: 41.67/150 pts)

Comments from several tasters:

Serge: ‘The influence of caramel on water is totally obvious on the nose, and very easily detectable. For the taste it’s almost the same. Very easy to detect caramel – just a little harder to make a difference between light and heavy concentration.’

Klaus: ‘My results are slightly above average. The samples performed differently but I was not able to nail it down. The main reason is almost certainly that my sensory equipment is not so sensitive. A cigarette every now and then might do no harm, but I am a heavy smoker. Another reason to stop smoking!’

Thomas: ‘This was almost too easy. Except for the first sequence when I mixed up sample 2 and 3 I had a perfect score. There was no way to miss the smell of burned sugar in the nose and while tasting the water the extremely bitter finish (sounds funny to use that word in connection with water) gave away the coloured samples.’

Charlie: ‘The dark coloured sample becomes more obvious/bitter after ca. 10 minutes while the medium coloured sample is quite difficult to detect by nose and mouth. All this is spite of my scores…’

So, obviously big influences on water by the caramel. The key word is bitterness.
Especially in the finish I found sharp notes as well, it reminded me of overly oaked spirit.

2nd Cluster: Neat and coloured Lowland (Rosebank)

Now that there is a concept about the smell and taste of E150-a the test continues with the first malt.
A light-hearted Lowlander, delicate and easily approachable. Surely caramel should be easily detectable…
Alexander made the following notes: ‘Nose: (1-neat) Herbal, Citrus, Lemongrass, A bit sour, Fat make-up powder.
(2-medium colour) Dead, Like the first sample,but seems to be covered by a blanket, Citrus, The herbal part isn’t there anymore, Fat powder is still there, but the Lemongrass is gone. No sour note too. (3-dark colour) Sorry, but I can smell the caramel, Full, Burnt, Bit raisiny, Oat. Taste: (1-neat) Strong Alcohol, Slight bitterness, Most complex version, Fresh. (2-medium colour) Bitter, More vivid, Warm. (3-dark colour) Bitter, Burnt toast, lacks complexity.’
The following are the results from nosing and tasting Rosebank:

            For nosing:   For tasting:   SubTotal:
Charlie        6 pts         4 pts        10 pts
Michel         6 pts         9 pts        15 pts
Klaus          7 pts         2 pts         9 pts
Serge          0 pts         0 pts         0 pts
Thomas         6 pts         4 pts        10 pts
Expected    4.17 pts      4.17 pts      8.33 pts

Total score: 43/150 pts (Expected: 41.67/150 pts)

Comments from several tasters:

Alexander: ‘The more caramel, the more dead the nose was. It flattens the nose. Sample 1 was the fullest and the sharpest. The raisins I occasionally taste turns out to be E150 and not something coming from sherrycasks…SHIT !’

Klaus: ‘Unfortunately I did not write down my impressions right after the test. I only remember that it was a fine and delicate lowland malt (maybe c/s) where the influence of caramel should be obvious. Could it be that the samples with caramel tasted fuller, more dark fruits, less stingy, – I don’t remember exactly. My nosing results slightly above average. The tasting results below par, but I finished the sample with 3 sips.’

Thomas: ‘Although sure to detect the same differences with the same additional aromas I still failed to score better. But after reading the results I detected that obviously I was on the heavy caramelized samples’ trail for a change. Why could I find those 4 out of 5 times but couldn’t tell the clean sample from the medium one? Even at the risk of tainting the scientific value of our test I decided to to some additional tests. I asked Anke to hand me the glasses with samples 1 and 3. I had no problem at all telling those two. Four times I sniffed and four times I was right. Hmmm… let’s try the same again only this time with samples 2 and 3. No problem again. Not a single wrong guess in a few runs. Now for the „tough” one. Would I be able to find the clean one between sample 1 and 2? The first try….wrong! I did it again – and again I picked the wrong one!! Now I opened my eyes and sniffed at them alternately when it finally hit me. I had mistaken the fruitiness in the whisky for sweetness caused by caramel while actually sample 2 was more damped and muted. On the other hand sample 3′s added bitterness seemingly had prevented me from making the same mistake with this one. Now that I knew what to look for I went without any more wrong answers through five more runs with samples 1 and 2. I had learned something!’

Serge: ‘Now, there’s clearly a difference with caramel, but I thought the heavily caramelized version was the neat one (4 times out of 5). That’s weird! In fact the caramel sort of took some sweetness off! The results for tasting are at complete random. The caramel’s influence was almost undetectable for me.’

Okay, a minor case of mayhem caused by the caramel. No-one has difficulties detecting caramel.
To decide if it’s the medium coloured or the heavy coloured is a different case altogether. As with the springwater a very distinct sharpness in the finish made it fairly easy to detect it as ‘heavy coloured’.

3rd Cluster: Neat and coloured Highland (Clynelish)

A more ‘sturdy’ malt for this session. Would it be powerful enough to withstand the overwhelming caramel?
Alexander made the following notes: ‘Nose: (1-neat) Flowery, Perfume, Maximum depth. Strongest Alcohols, Fresh, Sour. (2-medium colour) Chocolate, More complex, Some herbs, Livelier, Perfumed. (3-dark colour) Warm, Flat, A lot of dark chocolate. Dead. The caramel from the previous samples killes al lot of those tiny tastes. Taste: (1-neat) Sweet, Light, Tarry, Most complex version, Chocolate mousse. (2-medium colour) Stronger, Sweet but less burnt. Like Cognac, More complex, less bitter. (3-dark colour) Bitter, Flat, Thick, Sweet.’  The following are the results from nosing and tasting Clynelish

            For nosing:   For tasting:   SubTotal:
Charlie       13 pts         8 pts        21 pts
Michel        10 pts        13 pts        23 pts
Klaus          9 pts         5 pts        14 pts
Serge          7 pts         0 pts         7 pts
Thomas         7 pts         6 pts        13 pts
Expected    4.17 pts      4.17 pts      8.33 pts

Total score: 78/150 pts (Expected: 41.67/150 pts

Comments from several tasters:

Serge: ‘The caramel is more detectable in this Clynelish, which is a bit more neutral malt than the rather citrusy Rosebank. Nothing too spectacular, though. On taste: the caramel is heavily present now, it’s just that again, I took the most caramelized one for the ‘neat’ version 5 times out of 5. Maybe I should have learnt how spirit caramel tastes before the session ;-).’

Thomas: ‘Hey, am I getting better at this or what? While the score doesn’t look much more impressive than my previous attempts for the first time I managed to detect all the clean samples with my nose! And with this whisky it wasn’t difficult at all. Instead of fresh seaspray, salt and grassy notes the caramel had „managed” to absolutely kill the nose. It was totally bland, unimpressive and non-descriptive when colour was added. It didn’t make that much difference how much caramel was in it: once it was spoiled it was gone. Now, why that same findings didn’t hold true while tasting the samples again is beyond me. However, this was the most impressive tasting in this series so far!’

Klaus: ‘For my sensory equipment it seemed as if the aroma intensity increased, if more caramel was added.  However the results from the blind tasting only showed the average results. If I remember, I liked the caramelized samples better than the neat one.’

Charlie: ‘The neat sample is surprisingly bitter. Therefore I had difficulties identifying the coloured samples. No sooner did I think I was on the case with the nose, than the correct judgement of flavour eluded me! Overall, I was rarely able to judge the samples with complete confidence.’

Alexander: ‘Whereas in Rosebank it was like a blanket, in Clynelish it changed the character a lot. E150 dulled it a bit and destroys the complexity of the whisky, mellows it out.’

Clynelish and caramel definitly don’t go well together. Huge inbalances and more nastiness happening in the glass.
A case of putting things in sequence from ‘not so good’ to ‘plain awful’. Or the other way around of course…

4th Cluster: Neat and coloured Campbeltown (Springbank)

The 4th cluster will see a much higher ABV, perhaps this will have some influence on the behaviour of E150a…
Note from Alexander: ‘Nose: (1-neat) Flowery, beautiful depth, Salt, Tar, Very good, Unmistakeably a Springer (2-medium colour) Complex, very perfumy, Fresh, Mint, green notes. (3-dark colour) Clay, A bit flat but nice nevertheless. It’s hard to cover up this great whisky. Taste: (1-neat) More organics, Natural sweetness, Farmy, This is a Springer! (2-medium colour) Sweet organics, More complex than previous sample, The alcohol is sharper, Sweet, Less bitter, Beautiful. (3-dark colour) Even with lots of lots of E150, this still is a great whisky. Tarry, Harsh, Warm, Bitter, Sweet & Strange finish, Organics, Cleans mouth.’
The following are the results from nosing and tasting Springbank:

            For nosing:   For tasting:   SubTotal:
Charlie        8 pts         8 pts        16 pts
Michel        13 pts        13 pts        26 pts
Klaus          9 pts         3 pts        12 pts
Serge          6 pts         3 pts         9 pts
Thomas         3 pts         6 pts         9 pts
Expected    4.17 pts      4.17 pts      8.33 pts

Total score: 72/150 pts (Expected: 41.67/150 pts)

Comments from several tasters:

Alexander: ‘This is such a strong whisky, that the E150 has problems covering it all up. Still E150 changes the bitterness, adds harshness and covers up some tastes, but it can’t push out all the tastes.’

Klaus: ‘I am no big Springbank fan, but this one I liked very much. I enjoyed the neat sample most. The heavy caramelized sample showed some bitterness in the finish. The nosing results were slightly above average, the tasting results below par.’

Serge: ‘Again, some results almost at random… I got the neat version  the three last times; but couldn’t really make a difference between the lightly and heavily caramelized versions.’

Thomas: ‘Right from the start I knew this one was going to be much more difficult than the previous Clynelish cluster. Somehow the caramel aromas were way closer to the Springbank character than with other whiskies. I really had a hard time to tell any differences. It was as if my nose was already tired and though I thought I’d be able to finally tell the different samples apart I obviously wasn’t able to. A new theory crept up my mind. What if caramel works differently with every whisky?’

Mixed results and comments now. To me it was almost too obvious. Differences were minute but oh so clear!

5th Cluster: Neat and coloured Islay (Bowmore)

Bring in the peat. Will E150-a finally see its defeat? A very interesting thing happened when I prepared the Bowmore samples. Altough I used the same amount of caramel Bowmore coloured much darker. I have no explaination for this strange behaviour. Alexander’s notes say this: ‘Nose: (1-neat) Added organic element that must be covered up in the other samples. Still mild peat, more balanced, Fishy Islay (2-medium colour) More perfume lifted by sharper alcohol to carry it, sea, less peat ? Bitter, Mild Islay. (3-dark colour) Dead, Peat, Sea, Perfume, Sweet, Not bad. Taste: (1-neat) Sweet, Liquorice, Cleanest taste, Most balanced, Syrupy finish, Nicest finish, Still not the most complex around. Could have been a Caol Ila at some points. (2-medium colour) Deeper and more complex. Sweet, Peat, Sweet again, Warming, Improved finish. (3-dark colour) Flat, Peat, No complexity whatsoever, Licquorice, Ruined finish.
The following are the results from nosing and tasting Bowmore.’

            For nosing:   For tasting:   SubTotal:
Charlie        8 pts         3 pts        11 pts
Michel         5 pts         6 pts        11 pts
Klaus          6 pts         5 pts        11 pts
Serge          3 pts         9 pts        12 pts
Thomas         5 pts         6 pts        11 pts
Expected    4.17 pts      4.17 pts      8.33 pts

Total score: 56/150 pts (Expected: 41.67/150 pts)

Comments from several tasters:

Klaus: ‘This was the worst malt. All 3 samples showed a disgusting bitterness in the finish. The results were almost what you could expect when you just guessed.’

Thomas: ‘Well, if you look at the first results you’d have to say I had no idea if there was a difference or not. But all the time I had the feeling I could tell the caramelized whiskies from the clean sample. My „assistant” Anke took a blind sniff as well and she, too, had the feeling that it would be impossible to NOT notice a difference. So what went wrong? The first thing I did was to tell Anke to pick one of the clusters without telling me which one it was and to hand  me one of the glasses so I could take a guess.  First impressions were smoke, a little bit sweetish . With it came a hint of something slightly burned – almost with a Lagavulin touch. I liked it! So my guess went with the only Islay the test had to offer, the Bowmore, which was right although the sweetness made me consider Clynelish as well for a little while. That wasn’t too hard but but it proved that my nose was okay that day which was what I wanted to find out. As I learned later on Anke had picked the heavily caramelized sample which explains the „burned” aromas. I have to admit that at that point I kind of liked the added dimension. It made it a bit sweeter and rounder. At least that’s what I thought at that point. But that still doesn’t explain what had happened. My nose was fine, I was sure to smell and taste differences and yet the results were so off. My misgivings even before I started were that I didn’t know what to „look” for: I might detect something and erroneously take it for caramel. If that were the case here there should at least be a pattern, but nothing like that. Very frustrating in a way. At that point I was scratching my head….’

Serge: ‘Most funnily, I could find ‘a difference’ again four times out of five, but I took the heavily caramelized version for the neat one again. Decidedly! On taste: things were a bit more obvious in this series, as I got it correct three times.’

Charlie: ‘I found all these samples unusually bitter, which confused me, and made me feel my results were pretty random. When I saw the results, it confirmed my suspicion that phenols serve to cover caramel well.’

Mayhem shifts to havoc! This was really strange.
All the samples were very different from each other but almost no-one was able to come up with the correct sequence.
This was the sequence Dirk and I enjoyed the most. Almost impossible to ‘put your finger on’.

6th Cluster: Neat and coloured Blend

For this part of the experiment a vatting had to be made from the four malts provided. The reason for this cluster is a line from Charlie. He heard from blenders that caramel is a very important ingredient to bond different spirits together. I found this too interesting to ignore. In their own words, this is what the panel found out:

Thomas: ‘After creating the  vatted .. err… blended malt I first tried the clean sample. What a weird mix that was! Very imbalanced with all kinds of aromas that didn’t work together very well: heather, tar, fishing net, pine, paint and dishwater. The palate was a little bit better when grassy notes were duelling with some herbal tea aromas. The finish, however, spoiled any pleasure again: flashes of plastic and used sneakers. Pretty disgusting! That came quite unexpected to me since I more or less liked all the whiskies when I tried them seperately. So they are probably right when they say, blending is an art! Okay, now let’s add some caramel to our creation. And I have to admit it actually improved my product by adding something liqueur-ish to the malt. Nose: Tia Maria, a bit musty, sandalwood, furniture polish and cough medicine. Not too shabby! On the palate: memories of years long gone by: Jägermeister! Really weird!! The finish, however, couldn’t hold the interesting level. Despite some polished leather it was too bitter overall to still be pleasurable. But what about the some more colour? In my opinion this took it over the hill: sherry and smoke in the nose, very syrupy and overall dominated by the spirit caramel. Even worse: it seemed one-dimensional, sweetish at first, bitter in the end. Not much development at all.’

Serge: ‘Malt Vatting (Rosebank, Bowmore, Springbank, Clynelish) + 1 drop of water. Colour: white wine. Nose: fresh, clean and a bit spirity. Gets quite yeasty, on mashed potatoes and yoghurt. Some green apple and pear. I really smell the mash. Mouth: sweet and green at the same time. Rather grassy, with some peat from Bowmore. Notes of apple juice. Lacks complexity but it’s easily drinkable. 78 points. Malt Vatting (Rosebank, Bowmore, Springbank, Clynelish) + 1 drop of caramel solution. Colour: straw. Nose: again, yeasty and spirity, but with a little less ‘green apple notes’ and more fruits such as apricot and plum. Markedly ’rounder’. Some whiffs of liquorice I didn’t get in the ‘neat’ version. Mouth: no doubt it’s sweeter and rounder, with some fruit jam, apricot pie, caramel (ha-ha!). I think it actually is better, even if perhaps a bit bitter and drying. 79 points. Malt Vatting (Rosebank, Bowmore, Springbank, Clynelish) + 4 drops of caramel solution. Colour: orange/amber. Nose: extremely marked by the caramel, almost like a sherried whisky. Bold notes of crystallised fruits, wine (???). Also more fragrant and more on cooked fruits. Interestingly, I find it quite better! Mouth: round, creamy, sweet… The texture seems different. More notes of dried fruits and caramel, milk chocolate, light toffee. Also more liquorice. The profile has changed, no doubt. The whisky didn’t get sweeter at all, but most certainly rounder, with much more liquorice and a nice creaminess. And I liked it better, imagine! 83 points.’

Klaus: ‘The neat sample was very stingy, strong alcohol influence, everything very sharp with a lot of edges. The sample with just 1 drop of caramel was a lot better. Every nuance seemed to have been integrated. I don’t know why, but this dram was considerably better. The full coloured malt had gone over the top. The caramel influence was detectable. Sour and burnt nuances were detectable, dark bread. In the taste category the differences were not that large. The neat sample tasted sweet with citrus notes, then a sour impression and finally creamy bitter chocolate. The sample with 1 drop caramel was a pixel better. The unpleasant sour note in the middle was missing. Additionally everything seemed to be rounder. With the heavy coloured sample I found that the sweet notes at the start had turned a little bit dull. I think here the caramel had done too much. I think that caramel cuts some of the edges and integrates them. The neat blend sample had the most nuances, but the lightly coloured blend was friendlier and definitely more delicious.’

Charlie took his chance and made air-tight notes according to industrial standards:
‘I nosed and tasted ‘straight’, then added one third as much water, to bring down the ABV to approx. 30%.

Sample A (no caramel)
Nose (straight): Mellow to start, then developing some nose prickle. Sweet biscuits.
Then a charred aroma takes over (from the Bowmore?); trace of vaporous acetone; mossy in the development.
Flavour (straight): Hot (especially swallowing). Light sweetness, then light acidity. Centre palate delivery; dries in finish.
Nose (dilute): More estery (vinyl); now some fruity notes (boiled sweets  – pear drops, acid drops).
A musty, mossy note behind, also a faint artificial scent (Bowmore FWP?)
Flavour: (dilute): Considerably sweeter, with a peppery engagement on the tongue; still warming to swallow.
Light acidity; dries in the medium length finish. Pleasant.
Development: Nose more fragrant, delicate. ‘Natural’ compared with the other samples. Flavour same as above.
Noses/tastes much better at reduced strength. The best of the three.

Sample B (one drop of caramel in 15ml)
Nose (straight): Sweet vanilla sponge cake; softer and not nearly so aggressive; possible trace of coffee grounds.
Not as complex as A; other aromas difficult to isolate.
Flavour (straight): Sweet and smooth; less acidity; acceptable level of bitterness in the finish;
Warming; lingering bitterness in aftertaste.
Nose (dilute): Estery, acetone notes increase; traces of very dark (i.e. bitter) chocolate. Less complex than A.
Flavour (dilute): Pleasant soft mouthfeel; sweetness greatly enhanced, now also a trace of saltiness;
Dries in the medium-length finfish, but not bitter. Warming.
Development: Back to sponge cake, now possibly Madeira cake i.e. joined by Maraschino cherries.
Flavour same. An acceptable dram.

Sample C (four drops of caramel in 15ml)
Nose (Straight): Vanilla toffee, with moss/damp earth behind.
Dried fruits or constituted raisins/sultanas; trace of sweet tobacco.
Flavour (straight): Mouth-filling; dry overall with some salt; lingering bitter aftertaste after a long finish.
Nose (dilute): Nose very tightly integrated; hint of treacle toffee; soon distinct coffee grounds dominate the entire aroma.
Flavour (dilute): Pleasantly sweet to start, with a good mouth-feel. Bitter finish and lingering bitter aftertaste.
Development: Develops towards coated cardboard; little aroma. Same flavour. This amount of caramel at this strength (ca. 30% Vol), clearly has the effect of binding in any other aromas and subsuming them to the aroma/flavour of bitter coffee.’

Thank you Charlie! I think we can draw some sort of a conclusion: caramel is important in blends.
All the tasters agree that the coloured samples are better than the neat ones!

Overall conclusions

Time to come up with some overall conclusions.
I was pleasantly surprised with the outcome of the test. Especially the blending part openend my eyes and made me think the other way around: not as a malt-consumer, but as a malt-producer. It’s so obvious you need caramel to bond different casks together it’s beyond belief. I can only imagine the same goes for different casks from the same distillery. Caramel breaks up the ‘tension’ from a single cask and makes it approachable for another… If I had 10.000 casks in my warehouse of which only 1.000 would make it as ‘fitting the distillery profile’ and a drop of caramel would bring 8.000 casks into that profile I know what I would do… with the greatest care that is… That does not mean I have great objections against the colouring of a single cask and bottle it as such. Caramel does have influences on a malt and can destroy it completely. My final thought is about the producers. Only days ago I had a little chat with a person from the the same certain company that provided us the spirit caramel. He kept on raving that caramel has no influence on whisky. He did nosing and tastings and was not able to detect any differences. I suggested it was alright to say that he actually did notice differences, even then he denied… Isn’t it about time for them to come out of the closet and tell the world how and why they use caramel? Credibility is also a very powerful PR tool…’

And now some conclusions from the other tasters:

Alexander: ‘E150 ruins the finish ! Conclusion: E150 has the ability to cover up a lot of specifical characteristics, which is nice if you are a blender looking to put out the same product you did last year. It also allows you to use a wide palet of whiskies, because most characteristics will be masked by the caramel. And gives you a nice dark colour. Luckily this Springbank showed us that caramel can’t kill it all. Which shows us again what a good whisky Springbank can be.  There is a difference how caramel reacts with (regional) characteristics, since caramel behaved differently in a lowlander and an Islay whisky. Especially in the Bowmore is ruined the finish. In the Rosebank (and in a lesser manner the Clynelish) it covered up most traits… When I stand in the corner of the whiskyproducers I surely can understand the use of caramel. When owning a lot of mediocre casks at best, caramal can be the cement between the bricks (when used with care). It allows to build a tasteprofile very easily, allowing you to use multiple sources. And gives a nice colour too ! At the same time it can cover up flaws and tones down the character of most whiskies that would be too outspoken for most consumers anyway. Since they make up 95% of all whiskyconsumers you should use E150 in blends. Another story for the use of E150 in Singles. What kind of consumers are they? I would say they are more aware of what they drink and when they would know how whisky can taste like without the caramel, I would hope they would look past the colour of a whisky… But I don’t know. I understand the use of caramel in the big sellers like Highland park 12, Laphroaig 10 etc. since they do perform a crossover function to pull consumers from the blend to the HIFI world of Single Malt Whisky. Still the true whisky for the specialist are the single casks, non chillfiltered, not coloured…and not reduced ;-)’

Klaus: ‘For me caramel has only marginal influence on the malt.
They are within the range of batch variations. I think they can only be detected in a head to head tasting. I am curious what my fellow researchers, with maybe sharper sensory equipment, will find out. The most astonishing thing for me was, that a drop of caramel can sometimes improve a malt. The effect is marginal but detectable. For caramel in springwater there is a very strong indication that small amounts of caramel can be detected by nosing and tasting. The results for the malts were not that clear. I would say that the influence of caramel in the Rosebank and Bowmore  is questionable because the results were very near to ones which could be expected. On the other hand Springbank and Clynelish achieved overall scores which were almost two times higher than expected.  Unfortunately I cannot express the probability in numbers that this result comes from just by chance. But I think that it indicates that the influence of caramel can be detected here. To conclude with: I think our little experiment clearly shows that something happens, when caramel comes into play. But we were not able to nail it down with certainty.  Therefore I want to encourage everybody, to make his own experiment. Buy a bottle of E150-c (that is not much different from spirit caramel E150-a)  for 1 euro in a supermarket and voilà you can colour a whole cask of whisky.’

Thomas: ‘As much as I would like to say it, caramel can NOT be detected in every case! I do believe, however, that it DOES influence whisky although I do not have scientific proof for it (and certainly my scores don’t support this statement..). In my opinion there’s three basic findings in this:

a) Caramel does not work the same with every whisky but it depends on how it interacts with the character of that specific malt. Some malts might show characteristics that have similarities with caramel in the first place.  For example, I could imagine it to be very difficult to find it in Lagavulin. Some whiskies will seriously suffer, others might actually profit from the addition (making them warmer, more rounded)

b) There is obviously a certain level necessary for caramel to be detected.

Once this level is reached it doesn’t make that much of a difference if you add some more IMO.
c) I think your nose has to be „educated” to tell the difference. You have to know what to look for and I believe the more often you have the opportunity to test different samples like we did  the better you’ll become at this.

As I said before spirit caramel might actually improve what’s in the glass (or bottle) like the blend I mixed.
So would there be any harm in adding colour to your product? Despite my findings I say yes. Not only might it kill the characteristics altogether but caramel in most cases will lead to a standardized taste and lessens the individual aspects of the malts. If I were to sell whisky would I use caramel? Is my whisky of good quality and sells well enough, then definitely not. Am I stuck with an inferior product that’s gathering dusts on the shelves I might be desperate enough to do it. Why should I? But think about it: you invest all your knowledge and money to produce an individual  new spirit, then you select casks which seem promising and wait for 10 years ore more to get something special. Then you go on and add caramel and what might you achieve: the equivalent of what MacDonald’s is to food for drinks. Doesn’t make sense in my book.’

Charlie: ‘For me this was a fascinating, and somewhat humiliating, exercise.
But then, I am constantly humiliated in blind tastings! Here are my general thoughts:

a) Spirit caramel is not ‘organoleptically inert’. This is perfectly clear when you add it to water.
However, when added to whisky it behaves in a different (and even a beneficial) way.

b) As one would guess, stronger flavoured whiskies disguise or overcome the aroma/flavour of caramel – but not in nearly as predictable a way as I had formerly thought. I found it remarkably difficult to rate the samples with confidence. My notes are litered with ‘maybe’ rather than ‘probably’ – and few ‘certainly’.

c) The claim that it ‘binds’ flavours in a blend is proved, for me. It becomes increasingly difficult to separate out and identify aromas and flavours – which is, after all, the role of the blender, to increase the ‘integrated flavour complex’ – and the overall effect is generally pleasant. UNTIL too much caramel is added when that flavour becomes bitter [It ocurrs to me that maybe this is lost when the beldn is married for a while?]. I will never again be able to taste industrial Scotch with the same gusto!  But since we are assured that only small amounts are used in decent blends…no problem.

d) For me there is stil the sadness that, when single malt whisky is tinted, you 1) cannot hazard a guess as to what kind of casks (U.S. or Euro; first fill or refill) have been used in the mix, and 2) have some of the aromatic and flavour complexities ‘smoothed off’. But this is a minor complaint, compared with the overall enjoyment of the whisky.

e) Single cask bottlings should definitely not be tinted – but mostly, they aren’t!

KONICA MINOLTA DIGITAL CAMERAAll this experimenting was not enough for Serge
‘My baffled conclusions: no doubt caramel does change a malt’s profile. I’d say it might make a mundane one creamier and rounder, but I still have to check what would happen with a great one. Oh, yes, why not try that now? Let’s see what I could do…  Okay, now that I have my bullet-proof jacket, let’s try the great Ardbeg 1972/2004 Manager’s Choice for France with and without caramel (with apologies to all the Ardbeg fans!) Nose: ah yes, the caramelised one definitely has much more coffee notes, with much more dried fruits, sultanas and chocolate. But get this: it isn’t any worse than the ‘naked’ original version! Just a little less ‘ultra-clean’ and more ‘sherried’ (whatever that means from now on in my book).  There’s also more ‘jammy’ notes, such as marmalade and quince jelly.  On the palate, the caramel’s effect is very obvious too, with mainly some  heavy notes of liquorice that sort of overwhelm the great lemony notes  the ‘original’ had. It also got more dryness, with more burnt rubber… But on the whole, the malt resisted the caramel like a champ and perhaps it lost two points, max. Overall conclusions: caramel does change a whisky’s profile, obviously. It does so with even a small amount. On the other hand,  it might help some MOTR whisky getting better, while the great ones might loose a bit of freshness and subtlety. The main effect isn’t to make the whisky sweeter, but rather to add some coffeeish, sherry-like notes to the spirit.  It also brings more liquorice, and makes the palate a bit creamier. Okay, now you can shoot at me, I’m ready!’

As a matter of fact, we’re all ready to be shot at, dear reader! We’ll all await the inevitable fallout!
What? You’re now sipping that wonderful sherry matured malt… and feel no need to shoot us? Are you sure it’s sherried. Are you??
This is dedicated to all the distillers who realise E150a, used in small amounts, isn’t that bad!

Dramming happily ever after!

Michel & the Coloured Gang

Previous post:

Next post: