 So, the stronger reliance on image and range-based modeling techniques to document Rokat, the Rokat heritage in Scandinavia, has enhanced our ability to study the images on the bedrock panels and gain new knowledge about those who applied them. So, research. The Swedish Rokat Research Archive was able to show, for example, just making an example, that an early Bronze Age boat was carved on one of the slabs in the famous Bronze Age can in Chile by being able to identify the curvature of the crown. It's very technical. If you all would be here, you could shoot you with that stuff. Several case studies have shown that in many cases the cup marks and the individual spears are the earliest engravings to which then later warrior figures have been applied, sort of the biography we heard from Martin already. Recent new documentations of large-scale panels that were documented previously increased the number of known petroglyphs significantly. For example, through a collaborative laser-scanning project involving the Swedish Rokat Research Archive and the Regional Heritage Board of Estonia. So, there's a lot of work done, but as promising as encouraging as these results are, it would be remiss not to discuss some of the shortcomings of 3D imaging techniques. So, in the following, we aim to discuss some of these issues and propose a possible solution. The proposed method is relatively fast, effective and easy to carry out. Since the output looks very similar to the visualizations produced by the old technique Fotage, we call it digital Fotage, but that's just a label. Also, if anybody knows of the people who do similar work, please tell us in the comment section about it, because we would like to hook up with them. So, we're not pretending to do anything new, but just show our approach to it. First, I will discuss a bit the traditional rubbing or Fotage method, which has in Scandinavia, as I learned from... We've got some most famously introduced by a journal Rokat researcher, Dietrich Ivers, who was also in Italy and Spain, as I learned, in the 1960s and 70s. I suppose we're all familiar with Fotage, so I'm sort of skipping how it's done part. There are some problematic aspects with Fotage, and Marta has touched on it. I would just summarize them. So, Fotage is a very protected process, and for the large families in Scandinavia, it can take several days up to a week to actually get the documentation done. The second problem is that usually multiple sheets of paper are used, which need to be refitted later, which invites problems. You see this, actually, if you go to the archive, you see some very ill-fitted Fotage, which sort of split figures and put them on different levels. Fotage has difficulties in picking up depth differences, and of course in the 2D plane, depth differences are not recorded. The storage requires special conditions for the paper, which can be costly, as the archive has experienced. And the tactile first examination of the rock surface and the intensity of the graphite application depend on the expectations, experiences and perceptions of the person who documents. So, there's a high amount of bias inscribed into the documentation. And lastly, dissemination can be difficult unless you scan it and put it online. But Fotage has a very exciting advantage. It creates a very striking visualization. It's a visualization where you immediately are able to grasp what's on the workbook panel. So, with that, I go to the 3D documentation after some pioneer projects of image-based and range-based modeling. In the early 2000s, these techniques have gained real traction in the past five to six years in Scandinavia. This has a lot to do with increasingly cost-efficient equipment and software, but also with much, much more user-friendly applications. Internally, the new techniques are slowly replacing tracing and Fotage as standard documentations. And we will discuss this with James, I know, but I think it's for a good reason. So, since they are 3D models, they record depth differences. Investigations can take place from an endless variety of viewing and lighting angles, including those not possible in the field. We have a more or less correct representation of spatial relations in the 3D model. And since the methods record everything that is on the panel, if they're employed correctly, there's no bias in the recording itself. There's obviously some bias in the interpretation, but not in the recording itself. But there are problems in 3D modeling. So, one of the problems is we saw a couple of movies showing 3D models, and this is again for a reason, because the human eye can't perceive 3D unless there is motion. And this is a problem for the dissemination of 3D models or our 3D documentations. Because usually we put in images. And if you use meshlib and use different lighting angles, as soon as they're static, the lighting angle will show you some aspects of the 3D of the rock carving. But it will obscure other ones in which the light goes parallel, for example. And this is a problem with publication, because usually you only have a limited amount of figures you can put in your paper. And even if the journals allow to upload models, they get quite anxious if you tell them that your model has one and a half gigabytes. And then you want to put in five of those. This brings us to another problem. Sites like Sketchnap had until now a limit of five megabytes. They will change this. And they have an example up. We tried it with our processing computer in Gothenburg, which is quite potent. But models that reach the gigabyte size, it's really difficult to process them. Because the file sizes are too big and the computational power is just not there yet. At least not, I don't know, probably in some very well-funded large institutions there are. But for smaller institutes or private persons or private researchers or whatever, it's a real problem. Yes. So we applied a landscape approach to the 3D rock art data. To enhance the visualization of the 3D rock art data, we turned to the landscape approach because they struggled with similar problems. New high resolution data sets such as LIDAR have challenged landscape ecology to provide new means for visualization of ecological remains. On a landscape scale, ecological features typically represent a micro topography, which is much more subtle than the geographic background. And this is pretty much what we have with the rock art. It is, of course, reduced in scale, but the need to highlight faint topographic features and filter macro topographic structures remains the same. Local relief and slope modeling, short LRM, have significant potential for the visualization and analysis of petrographic imagery. What LRM does is it removes the macro topography by de-focusing a topographic ruster image using in RGIS the focus statistics tool. This is sort of the workflow I'm showing here, but I can also send this around if anyone's interested. These ruster images you use, digital elevation models you can create in photo scan or other 3D modeling softwares. And then this is subtracted from the original ruster image, the focus statistics output. And it takes away the large-scale topography and leaves you with the smaller-scale variances. You don't see them there, but you will see them in a second. So this contains the small variations, the output of the subtraction. This process is effective on both shallow, subtle pictographs as well as on deep incisions, which tend to be natural in Scandinavia at least. The visual image can be tweaked by adjusting the color range and its stretch, and this is what you see here. We use a color stretch usually from 1 to 0.0, and you can use color ramps. Below you see an example in red-yellow-green, but it might help in some cases, but we didn't find it that more helpful. So we just use the black and white usually, because that gives you a really clear image. And as you can see on the upper image, it's close to the old footage, so this is why we came up with the digital footage label. So it's not just visualizations, it's also research. But before I address some of the advantages we found in this, I have to say there's one big advantage. This is not a 3D model any longer. You can't twist and turn it, and you can't do all the fancy stuff with it. So the visualization we create using an ARC map for the full-size model is a 600 dPa, and the size of the output ranges between 12 and 32 megabytes. This is easily distributed. You can put this in apps and all that. Additionally, in the Resta created in IBGIS using digital footage, it's possible to zoom in from the large-scale to individual votives. So we have done this here. This is usually the same DEM. Are we going into this way and let's choose this example specifically for you, Marta? And we don't use much resolution, so we can go from this, I think, 14 by 6 meter, this panel, and then we go into this single motif and see all the details. It also applies color equally, because you don't depend on somebody applying the graphite doing it differently in different regions. You get an equal color distribution, even on complicated surfaces, like the upper one, which is a bowl which is decorated with a small bowl with a curvature. And it's still clear. And because the color is equally distributed, it really represents depth differences. So we can judge depth differences on the digital footage, even seeing multiple superimpositions on this boat below. So you have these two warriors sitting on the boat, which was cut across by this line, and then later on this was applied to make a round-shoot boat. So there's this whole story about biographies again. We used our approach to great effect discovering new rock art and features on a large scale and a small scale. So this recent work done at the Runohellen in Jero in Talon. This is a panel that's 9 meter by 6 meter, and it has been documented and re-documented for 150 years. And we were quite surprised by looking at this, that we discovered, at least when looking at the latest publications, that we discovered 13 new boats, 7 new anthropomorphic figures, 5 new animals, and a whole total of 68 new cup marks. So you can go back to very old sites and find your stuff. The other example is just published in a nice volume edited by this nice fellow there, which is this is the page where you see in Talon if you go to the famous Wittgenpegel. And you see these two warriors, this one Bernini has two bones, and this one are both atomic. But if you go and use the 3D documentation and put it into digital footage, I guess it's sort of more of a hairdo than holes, there's several things go out there. And this one, I know it's kind of hard to see, referring to the publication. This is not a phallus, but it is an animal that has been cut across by a human figure, and then sort of the animal is reused to implicate not only the phallus, but also the soul sheep that got here. So this is the head of the animal going here, this is the tail, one leg, the second leg, the third leg is kind of missing the fourth leg bone. So, to conclude, we are not proposing that our method can in any way replace any of the other methods. We don't want to replace old footage, we don't want to replace 3D modeling because this is obviously based on 3D modeling. And digital footage is just an easy to use and effective visualization tool for research and publication, combining most of the advantages of the 3D models with the striking visualization of traditional footage images. The small file size of the visualization also makes images useful for dissemination because they can easily and cost-effectively be integrated into, for example, mobile apps and similar outreach tools. And its main purpose is for research in combination with the 3D models and perhaps older documentations to provide figures, images for publications that can be easily read and the content of the panel can be understood with one glance. Thank you.