How do you actually make perfume? For AbdesSalaam, the fundamental starting point is with an idea or thematic concept for your creation, one that is as simple as possible. Again and again, AbdesSalaam returned to the Guerlain quote discussed in Part I on the importance of simple ideas, and emphasized that you should not get lost in your own fantasy or over-complicate things. Once you have the idea, then you try to render it concrete by blending materials in accordance with a formula that is centered first on the main accord, then on secondary elements.

AbdesSalaam slowly walked us through this process, starting by asking each of us our personal idea for a future fragrance, the possible notes that may be involved, and a potential name for the scent. My classmate, “M,” thought of a fragrance called “Lost Innocence,” encapsulating scents of her childhood with incense, rose, and blackcurrant as the mainstay notes. My table mate and lab partner, Manuel, also had incense and rose as part of his tribute to a chapel on his family’s old estate in Cuba, but he wanted the greenness of oakmoss and neroli somewhere in there as well. For Zenobie, a lover of gourmands, her concept centered around ice-cream, while Matthew sought to bottle the smell of his mother’s horses, right down to their poop and the peppermint used to keep flies or mosquitos away.

My idea was partially derived from the lingering effects of AbdesSalaam’s outstanding ylang-ylang soliflore, Tasneem, that I had tried earlier that day and that I’ve talked about in Part III. I wanted to make a floral oriental with an underlying gourmand streak, but it would be my personal idea of “gourmand” notes, rather than the conventional ones. AbdesSalaam’s superb Peru Balsam on the organ before me had truly blown me away with its rich, spicy sweetness, while his vanilla had surprised me with its woody aromas and an unexpected dryness floating from the top of the bottle. So, both of those would accompany the beautiful Ylang #1 in “Golden Luxe,” the possible, working title for my future scent.

AbdesSalaam listened and took notes on each proposal, but came to firm stop after Salman’s idea on making a fragrance for his mother, encapsulating the scents that he associated with her and with his childhood. Salman’s mother would always wear jasmine in her hair, but also smelt of roses and the spices that she cooked with. He recounted memories of coming home from school to be enveloped in her warm hug, fragrant from the treats she had made him, as well as the flowers in her hair.

AbdesSalaam kept circling back to Salman’s concept, honing it further and further down to its simplest form. What we hadn’t initially realised was that he was asking us for an idea that could be simultaneously narrowed and broadened to have wide appeal on a theoretical basis. With Salman’s proposal, he’d found something which could be broadly accessible to a larger audience: the scent of a mother with spicy and floral oriental notes, a story that would appeal to people beyond simply the specific ones in the story. It was almost a universal idea when slashed to its core basics, and one that would now become the basis for our first foray into perfume creation, because each of us would be making this scent. First, AbdesSalaam worked with Salman to come up with a name, settling on “Sindi Mother” (or “Sindhi Mother”) with “Sindi” being the name of the region of Pakistan where Salman was originally from.

After that, there was the matter of settling on the perfume’s main olfactory bouquet. A perfume should have a driving force or central character stemming from a main accord. For AbdesSalaam, that accord should consist of three ingredients, maximum. In the case of Sindi Mother, jasmine and rose were clearly two of them, based on Salman’s memories. What should be the third? They discussed various notes, their applicability to Salman’s memories but, more importantly, to the profile of the region in general. They settled on cinnamon as the third core note. One reason why is that AbdesSalaam believes cinnamon is the main archetype for spices in perfume language, in addition to being a big part of SE Asian cuisine.

Once you have the main accord, you must think of what would work for the secondary elements. For AbdesSalaam, when you’re pursuing a theme-based perfume, the symbolic character of the overall scent should drive you in choosing these additional ingredients. For example, what else is strongly evocative of either South East Asia (India and Pakistan) or their cooking? Obviously, sandalwood fits the former, while cardamom was a strong candidate for the latter. That gave us 5 notes. We needed one more.

In general, the blending formula that AbdesSalaam suggested to us was a very basic one that was designed to keep things as simple as possible (and not to overwhelm us). There should be three notes for the “main accord,” and another three as the “secondary elements.” AbdesSalaam settled on almond as the sixth and final ingredient, because it was very common to Pakistani desserts or cooking, and therefore something which would symbolically represent Salman’s core idea of his mother’s scent.

We had a perfume concept, a name, and the notes, so what now? AbdesSalaam told us to write our notes in order on the left side of a page in our notebooks, each note on a separate line starting with the three that would compromise our main accord. So, to demonstrate, in the case of Sindi Mother, it would be:

Jasmine: ____
Rose: ____
Cinnamon: ____
Sandalwood: ____
Cardamom: ____
Almond: ____.

The amount for all six elements should be roughly 80 drops in total, at the absolute maximum. The trio in the main accord would have the lion’s share, while each of the secondary notes would consist of much smaller quantities. AbdesSalaam suggested that the main trio would have 40 to 44 drops out of the maximum total of 80. The reason for the latter number was because, theoretically, one would eventually double the final formula before adding alcohol, and the small, roughly 5 ml glass atomisers that we used for blending held approximately 155 drops, give or take a few. Doubling was the theory but, all too often, most of us blew our 80 drop ceiling limit. For example, I once ended up with something like 112 in trying to edit and perfect my blend. In those cases, we simply added alcohol to what was there. (More on that later.)

It was up to us to decide how many drops of essential oil would be used for each ingredient but one thing that was critical in coming up with our individual formula was using a comma. Yes, a comma. As we played with the proportions for each element, we needed to keep our numbers as clear as possible as we went along. To that end, it was important to use a comma after each edit to demonstrate and keep track of the new, extra amount added. For example:

Rose: 5, 2, 1.

That shows we started with 5 drops of rose, then added two more, followed by yet another 1 drop for a total of 8. In actuality, we generally used far greater quantities for each of the main accords, unless it was something very powerful like seaweed, oakmoss or cinnamon, but that line should give you a basic idea of what I’m talking about. One thing that AbdesSalaam warned us was to be extremely careful with the cinnamon, and not to use more than a single drop at first because it could quickly overwhelm the entire balance of the composition. We could always add more later as our formula developed. He was right in stressing caution because his cinnamon had a major punch.

That was even more true of his almond essential oil which, as a few of us later learnt, was so incredibly powerful that the tiniest amount almost overwhelmed and blitzed the other notes, resulting in some people scrambling like mad to increase the quantities of their core trio to prevent their blends from turning into almond soliflores. A small wrinkle in all of this is that a few of the bottles had droppers which emitted oils with painful slowness (you could be waiting up to a minute or more for a single drop!), while others rapidly streamed out their contents almost like a flood. AbdesSalaam showed us how to hold the bottles over our atomiser/decant in a way designed to give us maximum control over their emissions, but there were still a few accidents here and there.

It was now up to each of us to decide how we would personally complete the vision of Sindi Mother. No two formulae would be alike, as each person decided by scent and individual preference how much or how little they wanted to add. The very first difference involved the choice of jasmine. AbdesSalaam had said we should use Rose Absolute, the same Bulgarian rose damascena oil mentioned in Part II that had blown us away on the night of our arrival, but we had two options for the jasmine. There was the Sambac which smelt indolic, opening with a blackened camphorous note before quickly blooming into something more spicy, sensuous, and full-bodied. Then, there was Egyptian jasmine that was much sweeter, fruitier, lighter, softer, and without an indolic blast. I normally love my jasmine to be as lush, blowsy, and heady as possible, so the Sambac is usually my thing in blended, finalized perfumes, but I was surprised here to consistently gravitate to the Egyptian jasmine instead. So, while everyone else opted for the Sambac, Zenobie and I chose the Egyptian variety instead.

We got to work on the perfume, focusing on our main accord first. My early formula looked like this:

Egyptian Jasmine: 5, 15, 2 (22 total)
Rose Absolute: 5, 15 (20)
Cinnamon: 1, 3 (4)

I’m not someone who is very keen on rose fragrances, so I was surprised by how much I liked the trio in those quantities. A number of others were drawn to it as well, with Salman far preferring the impact of the Egyptian jasmine in my formula to the Sambac in his. The more I smelt my trio, the more I rather liked it just as it was.

That brings me to a few other, related points. From the very first, AbdesSalaam had stated that “making perfumes is easy,” at least when dealing with beautiful, natural, raw materials. I was skeptical, not as to the truth of his comment, per se, but as to whether it would apply to me with my particular personality. Right from the start, I knew my issue would be editing. How do you know when to stop? How do you know when something can’t be improved with one more drop of, say, a secondary element in order to add just that final, necessary degree of body, nuance, or complexity? As all my regular readers have gathered by now, I’m not only obsessive-compulsive and love details, but I’m also a perfectionist. I always believe my own work can be edited or tweaked to be better, better, better. I’m not an artist by any means, but plenty of those have struggled with the question of when to put down the brush versus when that final dab makes their creation really sing. For a lot of people, I think it’s hard to let go and stop. Even some of my classmates wondered later on if they “should just add one more drop,” sometimes hearing that they should, sometimes being told, “Stop! It’s great as it is.”

For AbdesSalaam, only God can create perfection, and we must be happy when something is simply “good.” He believes that when you smell a blend and think that it’s good, you stop. “It is enough.” I struggled to wrap my mind around this concept. In all honesty, I found “simply good” to be incomprehensible; I just don’t think that way! It didn’t help that, as noted above, I love details — and the more the better, as these posts amply demonstrate. For me, simplicity of composition and a calm, serene acceptance that something is merely “good” is wholly antithetical to my character, which brings me back to my formula for the main accord in our “Sindi Mother” project. To my surprise, I actually liked it, everyone else thought it was wonderful, and I was oddly loathe to ruin it by adding more, particularly the explosive almond which had bulldozed over several classmates’ compositions. Was my formula “Sindi Mother”? No, not even remotely. But it was a very pretty, appealing take on a simple, spicy floral. For AbdesSalaam, the key was that I liked it, despite continuously being my own worst critic. So he told me to leave it as it was. Finished, end of story. He said I should give the blend time to settle by leaving it overnight, since time makes a difference as to the final balance through maceration.

In the end, however, days later, I actually didn’t like my fragrance all that much at all. It was okay, nice I suppose, but overly simplistic, terribly boring, and, yes, a little flat. I know I’m always overly critical of my own work, but I truly thought it had no pizzaz at all, probably because there were only three notes in it and I’m someone who loves layers. My classmates had all created full “Sindi Mother” compositions using the secondary notes to varying degrees, and my table partner’s creation was simply beautiful! To this day, Manuel’s “Sindi Mother” remains one of my favorites from the class, filled with complexity and spice atop a strong sandalwood base. I won’t share his formula lest he decides to do something with it later for his business, but it’s a far better composition than a good number of orientals from established houses that I’ve reviewed.

Yet, his blend actually hadn’t seemed so robust, bold, or exciting the day before. One small reason why was the quirky character of the sandalwood, Santalum Spicatum from Australia. No matter how much everyone applied, the wood seemed oddly subdued in my classmates’ blends. It was only the next day that the wood bloomed, coming out of the shadows to suddenly show its face and to add character to each composition.

The larger reason why Manuel’s perfume improved so much overnight is because of a general principle that I briefly referenced earlier, though it does relate to the situation with the sandalwood: time changes all things. You have to give a blend plenty of time to adjust and for the notes to recompose themselves before you finalise the formula and mix it with alcohol. AbdesSalaam said he commonly left his compositions a few days before coming back to see how they had changed or developed.

Obviously, we didn’t have that sort of time, but I think it’s a lesson to keep in mind if you start to make perfumes for yourself: how something smells today is not how it will be tomorrow or even the week after. That’s not always a bad thing, though, because some blends can actually improve over time. For example, sharper top notes can become rounded out or counterbalanced by the other elements; or the underlying nuances in the base might develop further. Be that as it may, the way my quasi “Sindi Mother” turned out was so boring (to me) that I became determined to add as much nuance as I could to upcoming experiments.

Our next project was to work was to work with our table mate to make each other’s original perfume idea. For Manuel, it was the scent memory of his family’s old chapel in Cuba before the revolution forced them to leave. He knew he wanted incense, oakmoss, and neroli, but we worked together to find the suitable levels. He was, in effect, my “client,” so I made suggestions for what might fit his vision. Given the intensity of AbdesSalaam’s Somali frankincense oil, we both agreed that it should be part of the secondary elements along with the opoponax (or sweet myrrh). Since he loves citruses passionately, and the neroli wasn’t sufficiently crisp or fresh in the mix, I pointed him to petitgrain bigarade and mandarin as ways of amplifying, brightening, and lifting up his main accord. After he sniffed them individually and gave his approval, we had our finalized list of elements.

I added each ingredient drop by drop, taking a conservative approach to the numbers, especially for powerful materials like the frankincense and oakmoss which could easily overwhelm the composition if we weren’t careful. Throughout it all, we worked together seamlessly, often finishing each other’s sentences and repeatedly having the same idea for how much more was needed for a particular element. Step by step, I would add a few drops, close and shake the atomiser, open it to draw out the inner spray for a sniff, then pass it to Manuel for his thoughts. From there, we decided which elements needed shoring up or had gotten lost in the mix, editing slowly until the overall bouquet fit Manuel’s mental image. Then, we left it to steep overnight, and it was my turn to be “the client.”

My idea was the one that I had mentioned earlier, a ylang-ylang floral oriental with a quasi-gourmand streak, and with a possible working title of “Golden Luxe.” (I know, I know, it’s terribly cheesy, but I was hopelessly sleep-deprived and jet-lagged!) I had a definite idea as to the notes I wanted starting first and foremost with the Peru Balsam, the oily sap from the bark of a Central American tree called Myroxolon. I’d always loved Peru Balsam but I’d now become utterly obsessed with it, in part due to its unexpected contrast to AbdesSalaam’s Vanilla. Both from the top of the vial and on the scent strip, the vanilla was surprisingly woody, unsweetened, and almost dry. It really didn’t smell at all like the vanilla extract or the raw paste from inside the beans that I had expected. While that subsequently changed and the scent strip eventually displayed all the aromas with which we are so familiar, it took more than a day to reach that point. In contrast, the Peru Balsam was sweet from the start, with a quiet spiciness that sometimes resembled cinnamon, in addition to an almost syrupy, dark richness.

I knew what I wanted for the other notes as well. Previously, in a different context, AbdesSalaam had mentioned that tonka adds a golden smoothness that works really well in ambered compositions. So that would be one note, possibly even as a fourth main accord, despite his formula preference for having only three notes there. I also wanted the patchouli and mandarin that I love so much, thinking that the latter’s tangy brightness would add a necessary “lift” to the heavier notes in the same way that it did for my beloved vintage Opium.

The final thing was the Buchu that I’d discovered earlier in class, and whose scent I’d loved both from the top of the vial and on the scent strip. It is an African herb that is supposed to be intensely animalic, far more so than even hyraceum. On the paper strip, however, it smelt aromatic, green, minty-camphoraceous, woody, slightly piney, and a bit like anise and licorice. Unfortunately, that was not how it smelt when blended! Yikes! I recoiled at its intensely aggressive, pure cat pee sharpness in the blend, even though Manuel had merely added two tiny drops. This made civet look like a water! Strong litter box odors didn’t fit my perfume’s intended profile or the other notes, so Manuel and I quickly agreed we should obliterate the Buchu in any way possible, as though it had never existed. (Lesson #2: how something smells in the bottle or on paper is often not how it smells in a blended, finalized scent!)

Manuel and I kept tweaking the formula for the fragrance until we finally arrived at the following:

Ylang #1: 8, 8, 10, 6 (32).
Peru Balsam: 8, 8, 4 (20).
Vanilla: 5, 5 (10).
Mandarin: 5, 5 (10).
Patchouli: 4, 3, 2 (8).
Tonka: 3, 4 (7).
Buchu: 2.

To be precise, though, some of those later additions were edits done the next day, once I had smelt how the blend had settled. For example, the ylang was no longer powerful or central enough, so I added 6 more drops. The patchouli needed shoring up, too, so that was increased by 2.

The total number of all drops was now 89. Sometimes, AbdesSalaam would have us double a formula if the overall quantity of drops was low, but 89 was sufficient to finalise the scent by adding alcohol. This was a stage where I had a little difficulty, because it was hard for me to believe that the alcohol would not dilute the overall blend. AbdesSalaam said that it wouldn’t but, if we felt a fragrance was now lighter, we could just spray ourselves a second time. It felt counter-intuitive to me that dispersing the oil in liquid would not weaken it, but he was the expert and I was the dummy, so I followed his directions. I carefully poured all the blended oil into a larger bottle, and then added alcohol up to the very top.

The very final step is to seal the perfume bottle. AbdesSalaam gave me an atomiser spray top, then pointed me to a silver contraption that looked like a fancy wine opener. In reality, it was a crimping machine that folds down over the metal base of the atomiser top, tightening it and binding it so that no liquid seeps out. While AbdesSalaam was expert enough to use it solo, we always teamed up; one person held the bottle so that it wouldn’t accidentally tip over, while the other (usually Manuel with all his muscles) pressed down on the handles to seal the bottle.

I now had my first full, proper perfume, at least in terms of having a complex set of notes. Manuel came up with the name “d’Oré” (Golden) which he wrote on the label and which seemed better than the cheesy “Golden Luxe,” but I’ll be honest, deep down, I wasn’t rapturous about any of it. The scent didn’t smell like the olfactory bouquet that I’d originally had in mind, the name wasn’t a perfect fit, and something just didn’t seem… perfect.

There it was again, the perfectionist’s gnawing doubts, the endlessly self-critical, nagging feeling that things can always be better and be improved upon. Part of it was me, and part of it was the issue of detail. When I fall head over heels for a scent, it generally tends to be for complex, bold, Wagnerian and/or divaesque fragrances, with the singular exception of my “comfort scents” where simple snuggly coziness fits the intended purpose and is ideal (so long as there is also opulent luxuriousness to accompany it).

Here, I kept thinking that an additional ingredient or two (or three!) might have given “d’Oré” a spark or head-turning ooomph. AbdesSalaam tried to reassure me that when something is good, it’s good. I clearly looked dubious, because he added that there will always be someone who likes what we ourselves do not, a lesson that he tried to teach us when we made perfumes in accordance with client briefs. And he did have a point, because there were things I made and didn’t like that others thought were good. Manuel even said he wanted my bottle of one of them, kind soul and good friend that he is. As a whole, though, I was hyper-critical of every perfume I made, particularly in the early days, and they would have been ripped apart had they been the subject of a review on this blog.

AbdesSalaam soon realised that it was the nature of the formula that I was struggling with. Something about the process that it entailed and its limitations on notes didn’t feel organic or natural to my personality. And I struggled even more when using that formula to work with materials far outside my scent preferences or comfort zone for client briefs. (Seaweed, hay, and chocolate in a perfume, were they mad??!!) I quickly concluded that I could never be a custom perfumer, forced to follow client’s strict set of notes, and unable to make changes (or to stare at them as if they’d grown two heads). Others in the class could, but I’m well aware of my own limitations.

Later, AbdesSalaam suggested an approach that better suited my love of details, multiple notes, and layers. I was still overly critical of my efforts, and I still thought that others (like Manuel) had an innate talent for perfume-making that I lacked — but at least I now had an approach that I felt really comfortable with and, equally important, that resulted in fragrances I liked much more. One of them was even pure love, though the scent subsequently changed quite dramatically when I got home after it had settled for two weeks. I’ll talk about that in Part V, after explaining the language of perfume, how olfactory archetypes let you know what stories to tell through your creations, the sorts of things we made in accordance with client briefs, some of the new materials we used like Rosewood and the “curiously nasty” Violet Leaf, and our “signature scent.” Part VI will cover the animalics and what they smelt like. Part VII will conclude the series with a look at perfume psycho-aromatherapy, distillation, final thoughts, and information on the next perfume seminar.