The IKEA Hemsjo candle

It is now officially the season of Advent. For most of my earliest years, candles at Advent played a big part in the lead up to Christmas. (They still evoke many fond memories at this time of year.) Mainly, they were red taper candles for the Advent wreath -- one new candle would be lit each Sunday, until we had all four burning on the last Sunday before Christmas. (Of course, my sister and I had our Advent calendars too!) But I digress.

The four candles I want to mention come from IKEA in a pack labelled Hemsjo. (The O should have an umlaut above it.) All four cost me a modest £1.75. Each candle is 5.7 centimetres (57 millimetres) in diameter and stands 8 centimetres (80 millimetres) high, including the conical top. They are all unscented and white. An individual candle is rated at 15 hours of (continuous?) burning, so the whole pack should be good for at least 60 hours when lit consecutively. On a price per candle basis, they appear to be the cheapest pillar candle (IKEA calls them block candles) in their range, working out at £0.44 each.

As budget pillar candles, how do they perform? Well, as you might expect from the price, they need a little bit of effort to get the best out of them. I have found that they are definitely not light and forget, as this results in an excessive candle flame and a tendency to burn down too quickly ("funnel"). I would say that they are slightly under-wicked -- but it is safer to have them under- rather than over-wicked. With a little care you should certainly be able to exceed the stated burn time.

Going on the 30 minutes of burning per centimetre of diameter rule-of-thumb, and rounding up, this gives a single lighting duration of near enough three hours. I always trim the wick before starting, and Hemsjo seems to work best when I start with the wick at 3 millimetres (1/8 inch). (From experience, it is a bit of an art to set the starting wick length based on both the thickness of the wick and the diameter of the candle: it is usually somewhere between 3 and 6 millimetres on most household candles!)

Although it looks aesthetically pleasing, the conical top makes initial lighting a little more challenging. Due to the slope, the candle burns down quickly, resulting in a long wick that consumes wax at a faster rate -- and then it starts to funnel! When starting a new candle, I have tried limiting the first burn to one hour, but even this proved to be too long. I shall try 30 minutes next time! (A more drastic solution might be to slice the conical top off?)

The wick is not anchored to the base of the candle in any way. (Some more expensive candles sometimes have a metal tag attached to the end of the wick.) What I have discovered is that further down the wick is markedly off-centre, and eventually it will collapse and drown in the melt pool. This is a good thing (safety again), but it does mean that the last centimetre or so of candle won't be consumed.

Unscented candles normally have no noticeable smell to me, but I have sometimes detected a faint paraffin odour after burning a Hemsjo candle for three hours. I don't find it offensive but I thought I should note it. I suspect it is down to the price: you get what you pay for, and all that...

Otherwise, I think that the IKEA Hemsjo candle is good value, and with a little attention burns reliably and with few surprises. This must be true: I have bought four packs so far and have already started on my second. Now if only they sold them in red for Christmas...

Ken Bromley Practice Paper

One slightly daunting aspect of watercolour painting for a beginner is the cost of the paper. Even at competitive prices, a quarter imperial (11" x 15") piece of quality watercolour paper will cost around a pound sterling (UK), so a serious mistake hurts a little in the financial department. Risk can be lowered by dividing the paper into eighth and sixteenth sheets, but this is not conducive to a looser painting style using larger brushes.

I have been a customer of Ken Bromley Art Supplies for a few years and it didn't take me too long to notice that they sell a 50-sheet quarter imperial pack of practice paper, at around 30 pence a sheet. The paper is Fabriano 280 gsm watercolour paper in a Not surface, which works out to 134 lb thickness. This is about a third of the price of popular 140 lb watercolour papers and is worth considering as a practice surface if you are not too worried about the outcome. Even the reasonably-priced Bockingford paper still works out to be more expensive in this quantity.

One mystery -- and what attracted me to the paper in the first place -- is the cotton content? Going by the Ken Bromley catalogue at the time, it appeared to be a full cotton paper. But I think this referred to a paper that is no longer sold. A couple of internet sources suggest that there might be a 25 percent cotton content, but I'm still not sure!

I duly ordered a pack and it soon turned up in the post thanks to Ken Bromley's usual efficient and prompt service. I chose to divide up some of the sheets into halves and quarters. I used the sixteenth (imperial) sheets for experimentation and testing, as well as a few finished paintings. The larger eighth sheets were used for paintings only, including a test to see how it behaved when stretched.

The first thing that struck me was the very regular pattern left on each side by the paper making process. It looks as though a wire mesh (screen) was used on both front and back, unlike the woollen felt normally used on the front of other papers. It is therefore very difficult to determine which side of the paper is the front, and which side is the back. (It probably doesn't matter very much!) The texture is also why Jacksons Art doesn't recommend this paper for finished works.

Notwithstanding the regular texture, how did it respond to paint? Well, as someone who is still very much at the learning stage, I found that it was a challenging surface. It did not appear to be heavily sized and was more absorbent than other papers I have used. Consequently I had difficulty lifting some paints, especially staining ones. Too much water and over working caused the surface of the paper to break up. And yes, it did cockle substantially after heavy washes if it wasn't stretched.

The absorbency and reaction to water meant that achieving a plausible cloudy sky was an effort in itself! (Mine have been hit and miss -- mostly miss.) I think that pre-wetting, then a heavy wash followed by quick work with absorbent material may yield the best results.

Perhaps I was lucky with my method, but the paper did seem to behave better after it had been stretched. (I used standard brown tape on a plywood board.) I felt more confident using larger amounts of water and wetter washes, and even though the paper would still lightly cockle, it would always return to its stretched state once dry. This seems to validate the recommendation to stretch paper of 140 lb weight and lighter.

This practice paper definitely fills a niche, but may be best for painters already confident with their technique and wanting to get some brush mileage behind them. I would still recommend something like Bockingford for beginner and inexperienced painters, as it is a harder wearing surface that stands up better to repeated working.

Watercolour paper: rag or wood pulp?

I recently discovered that the watercolour paper that I had been using (Ken Bromley's Practice Paper) -- actually Fabriano 130 lb Watercolour paper -- did not contain the proportion of cotton fibres that I originally believed to be the case. (Depending on the source, it could be either none at all, or 25 percent.) I know that most references and commentators stress the use of 100 percent rag (or cotton fibre) papers for watercolour, but is this just another oft-repeated mantra when there are some less expensive but almost as effective alternatives?

Wikipedia says that cotton is nearly 100 percent cellulose. Paper made from wood pulp should also be nearly 100 percent cellulose, as the claim is that it has been processed to remove the lignin and acid that causes it to turn yellow. The main difference between the two types of paper is that the cellulose fibres in the rag paper are longer, therefore making the paper stronger, more resistant to water, and a better surface to paint on. (It is not obvious that the cost of making a pure cotton paper explains the significant difference in price to consumers, however.)

The "sizing" of paper also has a bearing on how it works as a painting surface. Sizing is a substance used to control the absorption of a paper: gelatine is frequently used in manufacture. The most expensive papers appear to be both internally and externally sized. There is hard sizing and soft sizing. Hard-sized papers are most resistant to water and are best for some wet-in-wet techniques, whereas a soft-sized paper may not stand up well to repeated working.

Labelling on watercolour papers will give dimensions, weight and surface texture as standard. Most seem to claim 100 percent acid-free. Some will say archival, but what does this actually mean? Details such as composition and type of sizing can be harder to discover, although the degree of cotton fibre content is usually a selling point. And the price compared to other papers is usually a good clue.

The reality is that there is no (international) standard for watercolour paper, so the painter has to trust that the paper is fit for purpose and is what the manufacturer claims it to be. Consumers have to go on trust that the artwork will look as good on it a century (or more) later. How many people have the necessary equipment and knowledge to test that a paper is made of 100 percent cotton fibres and will not lose its original brightness after a hundred years? Do watercolour paper manufacturers actually do this? From a price point of view, I would expect Arches Aquarelle paper to be a better long-term investment than St Cuthbert's Mill Bockingford paper, but do we know if this is true?

I am still learning and experimenting. It would certainly be foolish to use a paper that is not labelled for use with watercolour. When I find a paper that seems to suit the way I (try to) paint, I will probably go with that -- assuming I can afford it in quantity. The chances are that it will be 100 percent cotton, but I am not ruling out something like Bockingford either. When I'm gone, do you think that the people disposing of my effects will care whether the paper is derived from cotton or wood pulp? ;-)

A watercolour secondary palette

When I first got interested in watercolours in the late 1980s, one of the many things that I found bewildering at the time was the question of which colours I needed? My sole reference book back then listed a number of manufacturers supplying a multitude of paints, but had no clear recommendation about what a beginner should start with!  I remembered from distant school days that red, yellow and blue were (supposed to be) primary colours, so why did I need a palette containing a dozen (or more) different paints?

I have since learnt that in watercolour painting, the use of the split primary palette is fairly well established. This takes my three primary colours listed above, and splits them into cool and warm -- so a cool red and a warm red, a cool yellow and a warm yellow, and a cool blue and a warm blue. But then it seems to get complicated when you start mixing them together. For example, I guess that green should be mixed from a cool yellow and a cool blue, but what if a warm colour is used instead?

Presumably with practice it becomes second nature to mix colours using a split primary palette? However, this does seem to raise the bar -- unnecessarily? -- for beginners, who may struggle with colour theory as it is often presented. Perhaps this helps to explain why some people use palettes with a large number of colours, as it is too much of a chore to mix them from primaries? (I have ignored the discussion of staining and granulating pigments, as that is another dimension!)

In the course of browsing the Handprint pages, I discovered the idea of using a secondary palette instead. This made a lot of sense to me. No longer did I need to worry about whether I needed to use the cool or warm primary colour, because the primaries had been changed to three that I recognised from computer colour printing (cyan, magenta, yellow [CMY]), and three secondary colours had been introduced (red-orange, green, blue-violet). Furthermore, the gamut of the palette had been increased -- especially for greens.

I was then pleased to find out that the selection of paints in my Cotman Sketchers' Pocket Box readily lent themselves to a secondary palette: Cadmium Yellow Hue would be primary yellow, Cadmium Red Pale Hue would be secondary red-orange, Alizarin Crimson Hue would be primary magenta, Ultramarine would be secondary blue-violet, Intense Blue would be primary cyan, and Viridian Hue would be secondary green.

Further reading of the Handprint pages led to more specific recommendations of primaries and secondaries. Restricting myself to the Winsor and Newton Cotman range, I replaced Cadmium Yellow Hue with Cadmium Yellow Pale Hue, and Alizarin Crimson Hue with Permanent Rose to closer match the suggestions. The Cotman Intense Blue is PB15 rather than the PB15:3 (Phthalo Blue Green Shade) that is suggested, but I was happy with that. If I was fussy about single pigment paints, then Winsor Yellow and Winsor Orange (Red shade) from the W&N Professional range would need to replace the Cotman Cadmium Yellow Pale Hue and Cadmium Red Pale Hue respectively. (I have since found that the Daniel Smith Transparent Pyrrol Orange gives a closer match to Cadmium Red Pale Hue than Winsor Orange (Red Shade).)

I have been regularly using my Cotman secondary palette since late 2014. Even though I was a little unsure at first as to whether the [CMY subtractive] primaries would work as expected, I have been delighted with the outcome. I have had no trouble at all in mixing the colours I want. (See the crude colour wheel at the top.) My mind has not been cluttered by "rules" of what I should (or should not) mix with what. By alternating the arrangement of primaries and secondaries in the palette, I can easily find the complementary when I need to tone down a colour or mix a chromatic grey. (The row of pans at the bottom corresponds to the row of colour swatches at the top.)

So my question has now become: "Why is the split primary palette still often taught to beginners in exclusion to anything else?" (Or to put it another way, why don't computer and commercial printers use split primary colours? Hmm...)

Mind the gulls!

A few weeks ago I spent a pleasant extended weekend in Swanage, Dorset. The weather was fantastic, the place was buzzing with visitors, and the jazz festival was on. I could have been somewhere on the Mediterranean coast. It was my first proper visit to Swanage, as I discount the day trip I made on the Swanage Railway in the 1990s as I never ventured outside the station...

So, strolling along the seafront amongst the crowds in the sunshine, it occurred to me that one common feature of the British seaside that can otherwise spoil the experience, was thankfully only playing its normal background role. I am thinking here of the gulls, whose calls and wheeling flight do so much to evoke the coastal location.

For the last decade, I have visited St Ives in Cornwall on at least one day a year and I am well-acquainted with the behaviour of the resident gulls there. (The photo of the dubious subjects was taken there a few years ago.) Those in the know are aware that the unsuspecting visitor enjoying their ice cream or pasty en plein air are fair game for the predatory sea birds, and I have witnessed a number of victims being robbed at bill-point! (Indeed, I have had a couple of close calls myself.) These birds are so notorious that there was even a recent BBC TV documentary about their (bad) habits.

It was refreshing then not to have to be watching my back while enjoying a few chips on the beach at Swanage. I don't think that this was a coincidence, and at least some credit needs to go to Swanage Town Council -- or is it Dorset County Council? Signs have been posted asking people not to feed the gulls. And I was most impressed to see that the waste bins along the seafront were even being emptied on a Sunday: bravo! It seems as though this could be helping.

However, I did then see something that made me mutter with disappointment. What I assumed to be a mother and son, were feeding the remains of their fish and chips to an appreciative collection of juvenile gulls on the beach. I walked by incredulously. To their credit, a couple of young girls walking in front of me were moved to action, and approached the gull-feeding woman with some friendly advice. (They knew that gulls were a real problem in Brighton.) I was afterwards told that the advice was (predictably) not gladly received and the reaction was along the lines of "I'm not retarded." Just a bit thick then, madam...

Programming references?

I must have been fortunate in that my higher education took place about the same time as the personal computer was starting to make an impact on everyday life and the world was starting to turn digital. It was a good time to be in computing -- despite what my parents thought, especially my father (bless him). Over the years I have been able to make a modest but comfortable living from IT, and so it continues.

When I started out I wanted to be a programmer, and my wish was fulfilled immediately after I had finished university. Even though I then moved into (computer) Systems Administration, I have always enjoyed programming and try to keep my hand in, even if it is only simple shell scripts which are then subsequently discarded.

Pre-university, probably like thousands of other people, I taught myself BASIC, no doubt with some common bad habits. In university, I was taught -- or was expected to learn at various times -- FORTRAN, Pascal, PL/1, SNOBOL, PDP Assembly language, and C. I subsequently picked up Intel 8086/8088 assembly language , Bourne shell scripting, and Perl. (Some of the programming reference books I used can be seen in the photo.) I briefly dabbled with C++ and PHP, but Java and JavaScript completely passed me by. Either shell scripting or Perl are my weapons of choice for repeatable systems administration tasks and projects.

In recent years, I have heard much about the scripting language Python. It is the favoured language for a number of my users, and the Debian Linux packaging system is written in it. I have received a number of requests to install new Python packages, or update existing ones. (This does not require much Python knowledge, fortunately.) Perhaps it was time that I learnt more about Python?

With this in mind, I sought a suitable reference book -- as I have found a portable, printed, read-nearly-anywhere guide had been essential in the past. I have long been a fan of the O'Reilly series of books, so this seemed like the obvious place to look. My copy of Programming Perl had served me well over the years; I wanted an equivalent written for Python. So, rather than check the wisdom of the internet and look more closely at what I was asking for, I put O'Reilly's Programming Python on my Christmas gift list...

My sister was kind enough to get it for me. What arrived was this monster volume of 1600+ pages that can be used to prop doors open. It is not such a pleasant experience to carry it any distance, and its size makes it difficult to easily keep the book open on certain pages. Had I done just a little more research, I would have noticed that it is big by the page count: 1632. And also that it should have been named more appropriately as Using Python. (Perhaps I should have been looking at Learning Python instead, all 1648 pages of it?...)

To use a bit of (Monty) Python: Stop, this is silly! The general consensus is that it is easier to learn Python than it is C, so why does it take 1600 pages to cover what the seminal The C Programming Language seems to do in 200? (I know this is being provocative, as the C book assumes a single version of the language, a Unix OS, a command line and a limited number of libraries. Things have moved on since the 1990s.)

In praise of the tea light

It would have been the mid-1970s when I first came across the small candle known as a tea light. My parents had bought a filter coffee making set from the Ideal Home Exhibition, and shortly afterwards a metal warming stand was purchased to help keep the coffee warm. Heating was provided by a single tea light. It did the job -- but I suspect that tea lights weren't as readily available or as reasonably priced then as they are today -- so use was for special occasions only! I remember being intrigued by the way that the wax completely liquified and was contained by the foil cup, as well as how the wick was anchored to the bottom of the cup.

Many years later I wanted a source of room candle light, and all the local supermarket had available was a pack of tea lights. Not the box of household candles that I might have found three decades earlier. All right, I needed to use two tea lights where I would have used a single candle, but it was fine. (Household candles are still available if you know where to get them -- yes, I remember the UK power cuts of the 1970s and the shortage of candles...) In general tea lights are easier to use, safer and cleaner. And of course they are now more widely available, including scented varieties if that's your thing.

The classic tea light has a cup made out of thin aluminium containing unscented white wax -- which becomes transparent when molten. The cup is normally around one and a half inches in diameter (37 millimetres). The wick is fixed to the bottom of the cup using a metal (steel?) "tag ": a collar gripping the wick attached to a circular base. This should burn continuously for at least four hours, and then self-extinguish when the pool of melted wax runs out. The empty cup can then be (safely) disposed of, or even recycled?

My internet searches have not yet come up with who first developed the tea light as we know it today, and when. I suspect it might have been as far back as the 1960s and in the U.S., but I would be delighted to be corrected. (The diameter suggests that it would be somewhere using Imperial measurement units?)

As well as the original design, tea lights are also available in longer-burning versions (sometimes referred to as night lights), and in larger diameters. Another variation is the replacement of the standard foil cup with a transparent hard plastic version, which is especially useful when the tea light is used for illumination. (These appear to be harder to find in shops, and more expensive.)

Tea lights have come into their own as dining table candles when placed in decorative tea light holders -- mostly made from glass or transparent material. They have especially become a common feature of restaurants and at wedding receptions in the UK. A holder fulfills at least two functions. It helps protect the tea light from being accidentally knocked over or touched, and it (usually) enhances the light from the flame. There are now dozens of different designs of tea light holder available, so there should be something to suit everyone.

A tea light in a good holder can be lit and enjoyed for as little as half an hour without showing signs of candle funnelling -- so it is ideal if you don't have the time to do a larger diameter candle justice. If you find the light of a single tea light to be a bit feeble, two or more can be clustered together to compensate for the smaller flames. Candles in general are now considered to be luxury items (with some prices to match), but the wide availability and reasonable price of tea lights -- especially when bought in bulk -- means that you don't need to pay luxury prices to enjoy them!