Jono Shields

Casting Bronze

Over the last few years I have moved away from buying gifts and into making things. This is us fighting the man on consumerism, but also a cool way to make gifts really personal.

Let’s start with the finished product…

I started after seeing one of Alex Steele’s casting videos. And decided I would start in much the same way, by carving wax.

I should also mention that I was very lucky to have the help of a professional to guide me through the journey. This is the reason everything turned out so well on my first go. It also explains the next step. So rather than try to cast straight away, we made silicone molds of my wax originals.

This way we could make many more if anything went wrong in the casting process. And it all turned out so well that now I want to try to sell them, so that’s a bonus.

Here you can see we attached all the wax carvings and casts to a wax cup that we will use to cast from. This first time I used a combination of both the green carving wax and some special brown wax, this way we can see the preformance of them.

Now we coat the wax in a ceramic shell, the first two layers uses a finer slurry and sand to keep the detail, then thicker ones are used to build up the strength.

When we did the burn out (heating it up to not just melt, but evaporate any remaining wax), we noticed that the green carving wax expanded more than the softer brown wax. This sadly broke the shell around these pieces and we needed to seal over the cracks. We ended up losing a piece this way and two others had noticeable seamlines where we sealed them back up.

What you see above is proof of the professional help I was getting. We used the same ceramic shell mortar mix to join the ceramic crucible to the ceramic shell and with enough bronze inside (10 times the weight of the wax). Notice that there is also a little peep hole that we use to check if the bronze is molten.

When it is ready we take it out of the furnace and flip the contraption into some sand. The bronze flows into the molds and the crucible on top helps to insulate the remaining top loaded bronze. This insulation helps to prevent shrinkage because the bronze cools in the molds first can can pull from the above molten bronze through the sprue.

Taking off the ceramic shell, these came out really well. Much better than my first wax casts in silicone interestingly enough. As you can see the detail comes out really well. Next time I will have to work more on my wax carving. So much easy to get it right first time in wax, makes for much less time sanding and filing bronze.

After some clean up and some polishing they look really good. Very excited to make some more.

Island Generation

So I saw another cool post by someone that was generating islands and I wanted to give it a go.

Since I haven’t really used three.js for anything before I thought it would be a good way to learn that too.

Finished product first…

Starting out by using Perlin noise to generate some terrain, there were tons of examples of this and three.js had one on their site.

Then once I had that working I had to shape it into an island. After looking around for some ideas I finished up borrowing a technique from here. I essentially create a dome and multiply it with my terrain I generated in the first step.

This is the result…

Next I use the height of each part of the island to determine what colour the vertices should be.

Add some finishing touches and boom!

Fake News

Test it for yourself here.

What are Markov Chains?

Markov Chains are a way of modelling states of something and using probability to determine a likely next state. For example if you are sitting, your next state might be as follows… 50% continue sitting, 30% standing, 10% lying down, 10% running.

As per usual someone else already has an amazing explanation with animations. So if you want to learn more check that out here.

How do we use this to generate text?

So first we need to have a data set. Any text will do, and the bigger it is the better. In this example I use Donald Trump’s last ~30,000 tweets.

You could try this with any other data set, maybe the bible, news headlines, some shakespeare or a favourite poet.

Next we go through the text and for each word we have we create a list of words that came after it. This is an example of the word Fake.

{
    "word" : "Fake",
    "followedBy" : 
        ["News","News","News","News","Whistleblower?","Impeachment!","News","Whistleblower?","News","News","News","News","Hearing","Whistleblowers","Washington","News","News","News","News","News","News!","News","Washington","as","News!","News!","News","News","News","News.","News","News","News","News","News","News","News","News!","News!","News.","News","News","Whistleblower","News","News","News","News!","News","News","Witch","News","News","News","News","News","News","News","News","News!","News","News!","(Corrupt)","and","News!","News","News","News","News","News","and","Poll","News","News","News","News!","News","News","News","News","News","News!","News","News","News","News","News","News","they","News","or","News!","Interview","story","News!","News","News","News","News","and","News","News","News","News","News","News","News","and","and","News","News","News","News","News","News.","News","News","News","News.","Media!","News","News","News","reporting!","News","News","News.","News","News","News","News","News","News!","News","News","News.","News","News","News.","News","News","News","Polls.","News","News","News","News","News","News","News","News","News","unsourced","News","News","News","News","News","and","News","News)","News","News","News","News","Polls","News","News","News","News","News","(Corrupt)","numbers","News","Polling","(Corrupt)","numbers","News","Polling","News","News","News!","and","News","News","News","News!","News","News","News","News","News","News.","News","Stories","work","News","News","News!","Media","News","News","News","and","News","News","News","News","News","News","Dossier)","News","News","News","News","and","News)","News","News.","News","News","News","News","News!","Dossier’s","Story","Story","News","News)","News!","News","News","News","News","News","News","News","News","News","News","Dossier","News","Dossier","News","News.","Dossier","News","News","Science.","News","News!","News!","News","Dossier","News","News","Dossier","Media","News","News","News!","Fact","News","News.”","News","News","News","Media","and","News","News","News","News","News!","News","News","News","News","just","News","reporting","reporter","News","News","News","News","News","News","News","sources","News","News","News","News","News","News","News","News","News","News","News","News","News.","News","News","News","News.","News","News","News","60","Media","News","News","News","News","News.","News.","Suppression","News","News","News","News.","News","News","News","News","Story","News!","News.","News","News.","Dossier","News","News","News","News","NBC","Dossier","nothing","News.","Reporting","News!","Reporting","News!","News","Dossier)","News","News","CNN","News","CNN","New","CNN","News","News","News","News","News","Story","News","reporters","News","piece","News","as","Dossier.","News!","News","Dossier","News","News","Dossier","News","News","News","News","News","News","News","News","News","News","News","News","Dirty","News","News","News","News","News","News","News","News","News","News","News","News","News......","News","News","News","News","News","Media","News","ABC","News","News","News","News","News","News","News","News!","News","News","News","News!","News","News","News","News","News","she","News","News","News","News","News","Mainstream","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","Russia","News","News","News","News","NBC","Washington","News.","News","News","Memos?","Dossier","News","News","News","News.","News","News","News","News!","News","News","News","News","reporting","News","News","News","Book","News","News","News","Book","News","News","News","News","News","News","News","News","News","Polls","News.","News","Mainstream","News","News","News","News","News.","News","News","News","News","News","News!","News","News","News.","News","News","News","News","News","News","Dossier","News).","News","News","News","News","News","Dossier","Media","News","News","News","News","@NBCNews","News","News","News","News","News","News","most","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News","News.","News!","News","News","News","News","Media","News","News","News","News","News","News","News","News","News!","News","News","News","News","News","News","News.","News","News","News","Media","News.","News","Media","News","News","News","Media","News","News","Trump/Russia","News","News","media","news!","News","Tears","News?","Twitter"]
}

So when we come to generate some new text we can start with a word, then randomly choose a word to come after it from the list, then repeat until we have a sentence (I like to finish on a full stop).

So in the above example the most likely word to follow “Fake” is “News”. Who would have guessed.

How accurate is it?

Not great. This solution doesn’t consider Natural Language and this means the results can just be gibberish.

However in saying that a lot of the results are quite amusing.

Source code available here.

New Camera Day

I recently acquired my grail film camera. The Leica M2. Leica is a heritage brand and has a long history with press and documentary photography. The build quality is beyond excellent and their cameras have always been built solid. This bad boy is roughly 60 years old.

Why not the M3? It doesn’t support 35mm framelines, so shooting with a 35mm lens means that I wouldn’t be able to see the whole frame. Why not the M4? No good reason. I am just a little less fond of the plastic trimming and the fastload take up spool is a bit harder to know if you have the film secured properly. Why not the M5 or newer? These started getting out of my price range very quickly. And from the M6 you get a bunch of features I don’t really need. Especially because I am quite keen to learn to shoot without a lightmeter.

But Jono, you don’t even really like using a range finder! This was true. Before this the oly rangefinder I had used was the Yashica Electro 35, and I really didn’t like having to use the rangefinder patch because it was too hard to see.

My digital camera though is a Fuji X-Pro2, kind of like a hybrid rangefinder. Okay, yes I know it isn’t a true rangefinder, but I love the feeling of shooting it.

Besides the thousands of people shooting Leica M series cameras can’t all be wrong. I particularly enjoyed EduardoPavezGoye’s youtube channel, I spent hours watching his videos with his M3 before I made this decision.

I received it on Thursday and the next day I was off to Wellington. The following photos were taken over that weekend. These were shot on a combination of Superia 400 and Kodak Color Vision2 500T.

So how am I finding the M2?

I love it! It feels so good in the hand. The film advance lever is so slick and smooth. The shutter is so quiet, perfect for street photography.

I got a Voigtlander 40mm f/1.4 with it because I couldn’t afford to buy this and a summicron at the same time. But honestly after seeing these photos coming back I might never upgrade.

And I am slowly learning the Sunny 16 rule so I can quickly pick my settings without needing a lightmeter. I thought these shots would come out a lot worse than they did. I’m stoked.


Learning About Me

Recently a friend reached out to me to help him with a project he was working on.

He is studying Theology at St Johns and wanted to highlight the role of women in the Gospel of Luke.

I am not religious in the slightest, but I liked the idea as most texts of this age downplay the impact of female characters.

We met at St Matthews to talk about it and plan each of the shots on location. And wow, what a beautiful and interesting place to take these photos.

The idea was that I would take two series of photographs, the first depicting women as they were seen in the Gospel of Luke. And the second series to recognise women that continue to or are about to undertake a role of service today.

“Women were among those who observed the crucifixion” (23:27, 49)



“Mary listened while Martha worked” (10:38-42)



“Angels told the women that Jesus had risen” (verses 4-8)



“Women were the first to tell the other disciples” (verses 9-11)



Michelle Hollings - Fire Protection



Latisha Fia - Medicine



This was my first time doing any proper portraiture work. To be honest this was the closest thing I have ever had to a proper photography gig. It was daunting. But it was also lovely working with everyone and I got to meet a lot of cool people.

This has been different to my other photography experiences. Normally I take photos for me, because I want to and because I find the subject interesting. I am not saying the subjects or the project wasn’t interesting, because it was. But it was a differently feeling, its hard to describe.

I understand that this is how people make a living from photography. But perhaps I’m more content to just do my own thing. At least for now.