 Hello everyone, my name is Rob Barrett, I'm from Queen's University in Belfast. Today I'm going to be talking about game engines, more about 3D reconstructions and a lot about the theoretical background to 3D reconstructions and simulations in archaeology. I'll be giving you some examples at the end of when I implemented some custom scripts in the Morty's landscape to get more information and further on interpretation of the archaeological context. So when we talk about 3D reconstruction and archaeology, this is often what we have in mind, 3D reconstruction is mainly used for presentation, especially to the general public. So you get a lot of this, a lot of nice renders, a lot of pretty pictures of very interesting sites, and this is a great tool for transmitting our knowledge, our understanding of archaeological evidence, especially to the public. But I think we can do a lot more with 3D reconstructions. I think through the use of simulations and gaming engines, we can actually advance our interpretation of archaeological sites. And that's because 3D reconstructions are a great system for experimentation. They possess physics, they have some kind of physicality in a way, and that means that they can be accessed as a system with variables that you can change. And adapt in order to get some conclusions, get some new raw data that can be used to better understand the archaeological context. And this new data, this new conclusions have actual applications in the real archaeological evidence. So when I talk about 3D construction, gaming engines, what I'm actually thinking about is more this kind of image, which is not as pretty, it's not as nice. But this is actually creating new data that can be used to interpret archaeological sites. And the use of custom scripts, a piece of software written specifically to answer specific questions, can help us answer these research questions with more data. And we'll be looking at these two examples a bit later. By firing fails, I want to go a bit more to a theoretical background to 3D reconstructions and simulations, because it's drawn a lot of criticism, both by 3D modellers and archaeologists in general, due to accuracy and inaccuracies. When you see a 3D model like this, it's very difficult to determine which of these elements is hypothetical and which one is based on actual archaeological evidence. And this obviously can have a lot of problems because, first of all, it's subjective. Some people think it's unreliable. And that's because you can't really trace back and know exactly what's going on. But I think that this is probably not just the 3D reconstruction simulations, it's also a problem that is a wider archaeological problem. And that's because of hyper realities. And hyper realities is this very big term, but actually it's quite simple. It was introduced by Bordruller, and it's basically the idea that there are some realities which are a mixture of hypotheses of fiction and reality. And these two are mixed so well between one another that it's difficult to determine which is which. So, if you think about reality TV, it's very difficult to see. It's still an example, but it proves the point. You know, reality TV, you look at it and you go, well, that could be genuine acting, or so it's a genuine emotion, that could be acting, that just could be editing. So, the fiction and realism, it's so intermingled that it's very difficult to determine what is what. And this is a big problem of 3D reconstruction, as you can imagine, because the hypothetical and the archaeological evidence are mixed together and they are indistinguishable. Now, when we talk about archaeology, this is kind of a system of archaeological board that brings from the archaeological evidence to the archaeological interpretation. So, I'll go through this very quickly. What we're trying to do as archaeologists is we're trying to access what's in original reality. This is a culture, a past culture, which to us is lost, you know, the Roman culture, the medieval culture. And that original reality is in itself a hyperreaptor because you've got the archaeological evidence, which is intermingled with our own culture, our own understanding, our own ideas. And they're so mixed together that it's difficult to determine. We also have archaeology, which is the archaeological evidence, so archaeological finds. And through this, we can access that original reality and have some limiting constraints that give us a better understanding of the past culture. The archaeological records are a copy of those archaeological evidence. And then we use those archaeological records to provide constraints to archaeological interpretation. But the archaeological interpretation is itself a hyperreality because our hypotheses and the archaeological evidence are concerned to me, that just by reading any archaeological book, it's very difficult to determine what is hypothesis and what is real. And this system is very similar to what we use in freedom reconstruction and assimilation. You've got the original reality, the partial reality again. And then you've got the base model, which is just the freedom reconstruction you do based on the archaeological evidence, which is itself a copy of the archaeological evidence. And then the base model finds constraints to the hypothetical model, which is when you add all the elements to embellish it and to complete it. So what I'm saying is the similarities between the two methods. While there's hyper realities in freedom reconstructions and these could be problematic, we've got the same problems in archaeological interpretation as a whole. But obviously there are ways to help justify these disanguilabilities, and that's through the use of metadata and paradata. If you keep a good record of how you go from that stage to that stage, from that stage to that stage to that stage, then you can trace it back and you can see how you can go from that from all the way back. And that gives you reliable information on the actual accuracy of the data you are presenting. And that's the basics of the scientific method. So in conclusion... Well, not conclusion, Tyffain. A theoretical conclusion is that although subjectivity and inaccuracies are inevitable in freedom construction and archaeology in general, this doesn't mean that we can't trust it as a whole. If you use that metadata and paradata carefully, then you can trace it back and then you can replicate it and you can analyse how unbun it is. And that's very important. Now, going back to custom scripts. This is what I mean with custom scripts. It's just a little piece of code written to answer a specific question. And you work with game engines that work with Unity 3D, and that's because you can import a 3D model into it. You write a piece of code and you can export data through that piece of code. And this is very important because you're generating more data, so if you've got a very specific question that you're trying to ask and you're trying to answer, stuff like this can help you answer those specific questions. So going on to a few examples. I work a lot in mortar, specifically in a lithic mortar. So mortar is these Twilands, I'm not sure you're aware, but mortar was very, very efficient in the lithic because they built enormous megalithic structures between 3,600 and 700 BC, some of them printed out. And these enormous structures are, imagine Stonehenge, but much bigger and nicer, essentially. They're great. But they're very important because we have a lot to help material evidence, but a lot isn't about the culture, what is people. Although we have a good understanding of some of their beliefs. For example, we know that they were very strong on their binary between life and death and light and dark, these were very strong concepts. And they used screens and they used certain tricks to funnel light in certain directions and visibility in certain directions. So there's a lot of interesting concepts that we can try and analyse on different sites. So this site here is the site of Chigentier. It's a temple site. I mean, it's a free reconstruction of it, obviously, but we've got two temples, one next to the other. And this is kind of what we've got today, but that's pretty much what's remaining. So one thing I wanted to look at Chigentier is solar alignment because solar alignment is very important in Malta. Most of the temples seem to be aligned with winter or summer solstice. So I wanted to check if it was true that Chigentier was aligned with winter solstice. But rather than just check the one date to see if there was an alignment, I wanted to check for up to the entire year to see where this alignment happened and how that affected the use of the temple. So I wrote a piece of code that calculated solar alignment for every minute of the year. And it does this in about 20 seconds. And what I got up to is this little graph here. So I'm exporting actual new data, actual raw data. I can then be modified and you can get to ask us that part of it. So one thing you may have to see is that the 21st of December, the winter solstice, that's when it has peak alignment with the temple, with both temples in fact. But the thing that really led to me is that it's not only on that day that you've got the alignment. It's actually, the entire week is a peak alignment and a full month and a half before, a full month and a half later, there's still alignment at that temple. So rather than being one singular event, one singular important day, these are what was much more torn out. You can have an entire week or entire months of solar alignment associated events. Oh, yeah, sure. OK, so this is an other site. This is the Sagrobotl circle. It's seen from the top. It's a bit of a weird angle, a Swedish model. So this is Hypergea. It's an underground cave system which was used for the burial of the dead. They would take the bodies of them here and then separate them into different chambers. But one of the important things about, I think, about the Sagrobotl circle is that it uses screens. So this stuff here, those are actual megalithic screens. They're big pieces of rock which are placed. And the idea is that they're placed there to follow your view into certain areas, to limit your view into certain other areas and maybe to also impact lighting. So I was interested to see in this how screens affected local visibility of this site. So I've written on a piece of code which produced scenes that we saw earlier. So this is just a visibility map of the site, but it shows it from human hide within downwards. And in this one here, this is without a megalith here, which is missing in the site, but is suggested that it's there. And in this one, you can see there's quite high visibility. Interestingly enough, there's a big visibility corridor in the middle and that's similar to what we have in temples. It's showing parallels between the two. And also there's an area which was called a display area in the original excavation, and that's that one there, which is highly visible. So it's sort of confirming that that is a high visibility area. And then by changing very small variables, by adding one screen here, you can see that, and you're not sure it comes across because of the lighting, this area is much darker, so that area is much less visible by the addition of that one small screen there. So by manipulating space, by adding screens and megalithing screens, you can drastically change the visibility of a site. And again, this is a custom strip that's just creating new data which helps us understand better this site which is called SiteWolts. Obviously there are limitations to this. Even with power data, metadata, if you don't have enough information to reconstruct the models or you don't have enough archaeological evidence, then obviously it's difficult to... The data you're going to obtain, the information you're going to obtain is going to be... The reliability is based on that archaeological evidence, so the more you have the better, the less obviously the less accurate it is. But also I think it's very important to have a specific question to ask the software. A lot of people do free-due construction just for the purpose of doing a free-due model to make the nice thing. But I think if you actually write a custom strip, you should be doing it to answer a specific question you have in mind. And that's not always going to come across. But in conclusion, I think that free-due construction and simulation are basically congruers with archaeological practice. They have very similar priorities. I only mentioned a bit about it, but there's a lot more going on as well. Custom tools can be used to answer a specific archaeological question and they can create data that has implications on the why of archaeological evidence and interpretation. Yeah, that's it. This is one of the people on the... I've also got a blog that you can follow. Yeah, cool. Cheers.