One of the main things on my desk at the moment is work on the Madder Baselines experiment - at least a bit of a preliminary write-up has to be done, and, well, I am more and more convinced the stuff is called madder because it makes me... the same.

Also, there's such a huge difference between looking at colours on the fabrics in real life and in daylight, and in looking at them on a photograph on the screen. Print will probably be different again, too, but for now, I'm frustrated enough by not getting nearly as much of a difference on the pictures than they show when yo look at the objects.

On that image, four should be very similar to each other, and the other two quite different from the group, with some difference between them as well. That's one of the better images actually, where it's relatively easy to distinguish at least on my screen. The others... ah well.

There's also quite a stack of results (dye-wise) that I'm not sure how to make sense of. Samples that should be exactly the same colour are sometimes (almost) the same, sometimes not. The samples from the pots running in parallel are quite well-matched to each other, the samples from a second run with the same ingredients don't match those of the first run.

As I said - quite surely, madder comes from crazy!

At least it's a pretty red. Or some of them. Mostly.

So after finding out about how difficult it is to judge, compare, and describe (or colour-list-match) colours by eye, I'm now trying to see if a colour checking tool is helpful in this regard. Which means... hearing a lot of beeps.

Because every time this thing here takes a measurement: 

...there's a little beep.

I'm going for averages, since the samples are not all completely evenly dyed, some more than others. So I've been taking 20+ measurements of each sample so far, and now I have enough to go and see if the data looks useful, or not.

For today I've definitely heard enough beeps!  

Here's a better picture of the outcome of our dyeing tests for the variations between different madders and different waters: 

So. Can you see differences? What do you think? 

When you're working with colours, definitions or descriptions of them are always an issue. Green as grass, blue as the sky, dark as the night - even these descriptions, time-honoured and always nice, include a large variety of shades.

It would be nice to have a good, solid way of describing colours, wouldn't it? Well. There's electronic colour checkers, there's colour space charts, and then... there's human beings.

One of the little side shenanigans we did at the Forum was test how well a colour checker fan works. These are printed colours with a defined number according to a system - you might be familiar with the RGB system, or have heard of CMYK. There's a number of other systems as well, but in this case, that's only relevant insofar as the number you'll write down as your colour name will change.

We asked people to sit down and try to describe the colours of some of our madder sample dyeings with help of the fan. (Thank you to everyone who did!) Everyone who I asked about how it went agreed on it being difficult, as the printed colour patches were not really matching the actual textile colours.

Well. Having had quite the experience with how dissimilar measurements done on the spinning angle were if taken by different persons, I was a bit suspicious from the start on how well the colour fan might work (or not). It would be really, really nice if it did, making the description of colours easier, and more reliable, and hope springs eternal. 

This is what we had as one set of samples:

It's a rather bad picture, but hopefully you can see that there's a little difference between some of the samples. Make up your own mind on how many groups you'd see regarding darkness/colour depth - myself, I see three in the originals and four in this (bad) picture, with the one to four pieces in the group having no discernable difference.

Here's the clincher: The cloth patches were not sorted by darkness when participants were doing the colour checking and matching. And... as a result, we had the same colour number assigned to the darkest shade and the lightest shade by different people.

Which, to me, means that this simple method is, unfortunately, out - and that you should not trust the colour numbers given by "comparison-by-eye" using a colour fan tool. 

Sigh. It would have been so nice, wouldn't it?

Heaps of things to take to the Forum, some larger, some smaller, are turning up at different spots hereabouts - sometimes three in a room, sometimes one. There's a stack of books, there are tools, there's a few canisters now to take the liquids along without them exploring the car illegally... there's materials and extra stuff and the whole documentation logistics like the camera and tripod and so on and so on.

I've started a second batch of bran water, using the starter from the last go, which will be fermenting for a few days now. I am also, since this morning, the proud owner of a lot of 250 ml lab glasses - of which a lot can fit into a single canning pot, and thus we can run a good number of sample dyeings at once.

 A waterbath is the tried-and-trusted method to have the same temperature curve for all the samples, and to do them at the same time. We've used that method before when doing dyeing experiments, but with the 1 l beakers, and only 5 or, at the most, 6 will fit into the pot.

If my math and rudimentary drawing skills are correct, we'll be able to fit 20 of the 250 ml beakers into the same pot easily. And though using the small pots means that we'll only be able to do very small samples, and run the risk of them being dyed not completely evenly, having the large number for a run seems, to me, to be an acceptable trade-off.

This, by the way, is one of the bits that I find both exhausting and fascinating when planning and doing archaeological experiments: The myriad decisions that you have to make. Smaller jars or larger ones? Sometimes things and processes don't scale well, and there's a point at which it becomes difficult to measure the very small amounts, plus the inevitable tiny differences will have a much larger impact. Natural substances (like raw, fresh tartar) or processed? The natural stuff or raw stuff may have components that change the outcome, but the processed pure form means it will be possible to repeat the experiment better. What to document, and how often? Sometimes taking measurements will disturb the process, but not taking them means you'll stand there lacking data. How many variables to test, and which ones are the important ones?

So, so many questions. There's a decision to make at every single step, and an experiment has a lot of steps, even if it's a simple one. So you make the decisions, and then you stick with them, and you hope they were the right ones!

(No blog post tomorrow, by the way - it's a holiday here.)

Well, the plan for the Madder Baselines is about done - the rest of the planning will have to be done as soon as we know how many samples there will be to handle, and then go for a last check through the long list of steps - before following them through to, hopefully, a nicely colourful end.

Meanwhile I have a second template to turn into a plan and protocol: the template for a mordanting experiment. Common Horsetail is said to contain quite a bit of alum, so it is in theory a replacement for mineral alum which may not have been available everywhere. However, it's not really clear if the horsetail is really suitable for this, mostly because there's no recipe that tells us about amounts necessary, or if there was any other preparation done before using it. Or at least I have not been able to find any...

So the idea was to try out if the plant will work as a mordant, and if yes, how much of it is needed. Because even if it's available, if you need ten times the amount of wool weight to have enough alum, well... that would mean 10 kg of the dried plant for a single kilo of wool, and if you've ever woven fabric, you know that a kilo is a puny amount.

I can think of three different methods of using the plant straight away: as it is (just dried, then soaked, and maybe boiled a bit previously to better get out the contents), fermented, or (which would also reduce the bulk of it) burnt to ashes (which should still contain the metal, though maybe in a different form). 

Obviously, the burning and the using as is do not need extra prep time, but the fermenting does. I'm happy to report that the 100 g of plant that I put into 7 litres of rain water are doing what they are supposed to be doing: Making bubbles and working on changing their smell.

It's not an unpleasant smell (at least not yet), but it is definitely much different from the smell it had at the start (like dried horsetail, but thank you, Captain Obvious). I'm very, very curious already to find out how (or if!) all this will work!

Planning an experiment is like planning a reconstruction - it's really, really easy to forget something. Where something can be a minor detail or, if one is a bit unlucky, something that seems like a minor detail at first but turns out to be rather crucial. 

 I've also found, in the number of experiments that I've either run, or co-run, or been somehow a part of, that it is quite easy to forget that one had planned to make photos at a specific point, or that something has to be changed and this all results in a number of hectic notes that are scribbled somewhere.

Part of that is just business as usual. Not every experiment runs as planned, and sometimes adjustments have to be made on the fly. But I found that good prep is really helpful... and that having a good template for the planning makes the good prep easier.

I've recently updated the experiment planning template on the Textile Forum website, and if you're interested, you can find it here. In case you are using it, please let me know if you have any comments or feedback - I'm always happy to hear about this!