Luxury and Art: A common framework for metadata in practice
Kelly Vero with Gabriel G Fidalgo
Overview
I’ve been writing and talking about metadata in luxury and art for a year and I wanted to be able to provide a format for how metadata should be defined and populated for the purposes of purchase, investment, arbitrage, and onward value. It is here that I will break down the use case for the framework, it is an open framework for everybody working in the luxury or art (NFT) industry and can be applied simply to objects created or being created within this sector.
It’s important to mention at this point that it is not a framework for product life-cycle management: that is already in place and works very well. I’m more interested in developing a common framework between worlds between the physical and the digital. So, article number, or SKU etc is given for this discussion but it isn’t as important as some of the other points I will expand here in my digital use case.
Abstract
The key motivators for the metadata backend or stack will support the sale or distribution of your product as well as PLM information. Every digital asset in the luxury industry should have it, but not every object does.
- Geometric Data
- Export Format
- Digital Rights Management
- Value
Practice
Fig 1. Agree upon a model for what the object or asset contains before it is produced.
Digital is the future of luxury and art and it is no longer a concept or theory, but has acquired a status of required implementation for all creators and houses wishing to deliver luxury directly to the consumer, So what’s the problem? The main problem is that luxury houses and art creators are effectively giving away individual intellectual properties, innocently, and without a real understanding of how easily they can be repurposed or cloned, rendering them worthless to the creator of the original design. Physical objects carry authenticity proof as a matter of course at the time of the transaction, something which many luxury/art consortia and families have carried into blockchain as a way of digitising the physical. However, with assets and objects now digital, it can make everything from authenticity to traceability too fluid and intangible to truly control the IP. Meaning that products being developed quickly by changing seasons and trends, are largely missing IP data, and transaction protection because no one really understands how it works to apply this methodology in the first place. Digital assets by virtue of the term “digital” should always live as a data product first before they are rendered into something more physical, tactile, or tangible.
The use-case for the purpose of this paper is going to be a handbag. Handbags are ubiquitous in luxury and range, and in price. My use-case handbag is like most in the luxury sector; it is hand-finished and can be bought from any boutique or online store. But that’s physical, it’s the norm and it’s here that the first problem arises in terms of how we could handle and maintain the transaction process as a digital or NFT object. You can refer to any of the previous articles I’ve written about this, but the experience online should be exactly the same as experience in-store, that’s not what this paper is about, however. So purchasing the handbag online assumes that all of the attributes of the handbag are displayed for the consumer to select and purchase. But what about if the object is purely digital or is sold as physical but has an optional add-on so the object is available for digital use? Let’s get started.
Dimensional, or geometric data, is the first part of the metadata framework. Size, weight, shape, volume is the footprint of the entire luxury experience when we choose physical items. How something fits or feels in our hands is as important in digital as it is in physical terms. So in creating an initial metadata gateway, dimension in this particular handbag is the basis of not just branding and finishing but also scalability for the purpose of digital objects being used in everything from video games to the metaverse.
We must also consider the shape of the product as part of our data point on dimension. When we talk about shape we’re not talking about whether this object/asset is a shoe, or a dress, or a handbag, (which in this case it is), that doesn’t matter; what I’m actually talking about here is the polygon mesh, which is a collection of vertices, edges, and faces that define the shape of a polyhedral object. This geometric data is its most basic and universal form. Most digital objects consist of quads, triangles, or other simple convex polygons. This allows us to render objects simply for use in everything from e-commerce to video games. The advent of digital object creation or digital asset creation means that we can look at objects as specific base shapes which are measured by count and complexity (of polygons) allowing us to wrap texture(s). Additionally, this gives an opportunity to explore the possibility of describing and embedding or presenting the state of the object at the time it is finished or rendered. And so being able to identify whether the asset is rigged so that it can be easily animated, or that the asset has a PBR (physics, or physically-based rendering) layer adds versatility (and utility) to the asset. Many people who currently work in fashion and luxury term this photorealism.
Physically Based Rendering allows 3D productions to obtain these “photorealistic” renders in real-time engines. Real-time engines are the backbone of all the “Metaverse” projects. Whether they are web-based or offline engines, they allow users to build simple and complex environments which can be put online as standalone experiences or complex metaverses.
What many pipelines strive to do is create an approximation of photorealism. However, feasible and quick conjectures point towards photogrammetry as the basis with which PBR is achieved (and often with little attention to detail or quality). PBR principles must therefore also include shading, texture properties, and the depth of photogrammetry used to attain real-photorealism (opposed to bought-in texturing and PBR).
In the case of an NFT file, if the production is aiming for a “photorealistic’’ render, all the materials composing an object must be listed. In this use case, a handbag, the data would contain detailed values on the type of leather used, its specular value, color value, pattern, logos, anything that is visual. However, it’s important to note that the data contained in the texture information of a PBR model is different if the render is more stylized than photorealistic.
So to simplify, any type of visual information must be accounted for. But in the case of a photorealistic object, information regarding every single material composing the object must be accounted for.
Our brains are wired to think spatially. Any form of creation aims at representing an object or a place in space. So in the context of NFTs, we must think of them as spatial objects. They may not share physical space but they are somewhere beyond numbers and file extensions. 2D objects should be converted into 3D objects with data mimicking their real-life attributes. This might mean that though the amount of geometric data is less (polygons etc are not as relevant here), that in order to maintain the standards for this common framework approach, the dimensions and supportive art data must still be in place. So a jpg can still be an NFT provided it is supported by the other three factors in the metadata stack.
Another overlooked facet of digital asset creation is in the object’s ability to be exported simply into formats that are compatible in everything from e-commerce to physical product development. I have always tried when creating tools for the fashion/luxury industries, to ensure there is an ease of adoption of the object for the client and the consumer. The eye should see as much of that photorealism I talked about in the paragraph above. This means graphical volume, and that affects the asset export massively. Too little and it’s a forgettable JPG, too much and it’s a behemoth TIFF. Think about the appropriateness of the file format in comparison to the platform. Are you creating for VR or AR? Perhaps a try-on? Low-poly is the format that will work best here (in most cases), building for e-commerce or high-resolution lookbooks, then nothing else will do but super high poly — it’s about the detail of deliverables. And finally, if we are using proprietary formats to create digital objects we have no opportunity to scale out of a proprietary marketplace, it’s really that simple.
As we are increasingly aware, there currently is no governance outside of regular IP laws that point towards the infringement of digital luxury goods and I hope this section will change that; because, when a consumer buys a luxury handbag as a digital object they should adhere to the end-user licensing agreement or some form of digital rights management as one might find in a paid download of a movie or an MP3. The growth of interest in this area particularly leads me to ensure that, in order for there to be a smart contract between the luxury house and the consumer, this EULA or DRM information is apparent at the time of purchase. What if the information is omitted? That’s a question for IP lawyers who work specifically in this area of art and luxury, however, I would err on the side of unless there is a EULA or DRM that is provided by one party and either checked and signed by the other party (as a checkbox) and added to the metadata stack for the preparation of a smart contract, then the transaction as a point of authentication is voided.
The value of digital assets, of course with some exceptions, is not as high as the value of the physical object, that goes without saying, however, value is almost priority #1 when we are talking about luxury. Following value, quality definitely provides interesting motivations for people to buy luxury goods in the first place. That could be as an investment piece or as an heirloom, it could also be a way to bring about change in a specific economic space or emerging marketplace. Whatever the reason, value is an incredibly important part of this metadata stack. But where does it sit against dimensions or DRM or formats? There are two clear areas of value that need to be addressed by the metadata we’re creating in this common framework.
1. Arbitrage. In economics and finance, arbitrage is the practice of taking advantage of the price difference between two or more markets: striking a combination of matching deals that capitalise upon the imbalance, the profit being the difference between the market prices at which the unit is traded. Luxury is a commodity that is observed at price points throughout the world, and the price of a luxury handbag from the same luxury house or brand should cost the same at a local level regardless of the location. Many luxury houses abide by this principle. However, as I previously described, these new economic spaces or marketplaces or perhaps a third party involvement in the distribution of luxury goods presents a possibility of arbitrage that is out of the control of the luxury house or artist/gallery. What metadata can do for a digital object is secure the price points fluctuation of the digital asset for the purpose of sale, or even resale. When thinking about arbitrage it’s easy to simply stop at the price point, however markets change regularly, hourly, and the luxury price point must reflect that. Therefore, it would be advisable in your metadata creation to ensure that you have written an algorithm into your price point which reflects potential arbitrage.
2. Futureproofing. Moving from that point very quickly and into onward value, I’m always fascinated by how much a luxury item could be worth, not today, but five years from now. Metadata allows us to be able to insert other algorithms that monitor market growth, fluctuation, or movement in currency from the aspect of the luxury house in question rather than from the aspect of exchange rates. This means that luxury houses and artists have the capability to set prices for the assets not just for today but for the next five or ten years, regardless of the buyer and reflective of the market only. This must happen as part of building a metadata stack to invoke the value of the asset later on (rather than retrofitted).
Finally, descriptors, tags, #, SEO, and anything concerning the story of the digital asset or object, where it can be either easily identified or easily searched for, adds an extra level of utility to it. If you’ve ever read any of my other papers about digital fashion you will understand the importance of utility when creating both digital fashion or NFTs (Non-fungible Tokens) for consumers to use. It is not enough to assume that if we build it they will come. We have to always give our consumers a reason to buy from us, but also allow our brands, designs, and quality standards to assist consumers in their buying choices, behaviours, and loyalty. So please use this final part of the metadata to ensure that the description of the handbag as a handbag which was originally developed as a physical object “in coated canvas with patinated handles” is there for the consumer to enjoy at the point of sale. Product life-cycle management descriptors have long been able to help merchandisers to merchandise, this is the last opportunity that a marketplace has to be able to draw the consumer into the world or their aspirations, on both sides this is incredibly important.
When building a smart contract for luxury or art, it’s important to not be overwhelmed by where the digital object goes. Provided that the information set out here is embedded in the digital object itself, by virtue of creation, the smart contract, in this case, should not be altered (so there are opportunities to extend or replace parts), the smart contract should not be distorted (so there is no room for interpretation), that information can be protected, not necessarily by using blockchain but that there is a clear checks and balances approach by the luxury house to the consumer. After all, the information contained within the digital object potentially exists within the digital object until the expiration of the digital object; the idea that this would be a unilateral decision by the luxury house is not an adequate enough approach to fully understanding digital assets as commodity items.
Conclusion
A common framework for the creation of digital assets for the luxury and art industries is clearly undefined. By providing reason and motivation to agree upon a solution for what the digital asset framework could be, we can begin to standardise the quality of digital assets created for these industries in a variety of use cases. In doing so we can also protect our creations and futureproof the value cases of luxury and art in a changing digital world. From transactions to exhibitions, the need to protect our IP is extensive and evolving; failure to act now may result in lost, stolen, or counterfeited properties. Agreeing to create a benchmark in how we adapt luxury and art for digital is the first step to a mainstream approach to digital luxury for creators and consumer's conformity of ownership.
Acknowledgments: Thanks to my amazing exes at SO REAL Digital Twins for the opportunity to explore and debate this, especially Gabriel G Fidalgo; thanks to Mark Harrop for supporting these ideas that extend product lifecycle management and functionality. Julie Zerbo at The Fashion Law for identifying a need for standardisation of NFTs and digital luxury. Julie A Evans for associating a link between analytics in luxury and fashion and the data which drives it (because it’s flawed). Finally, Felicitas Morhart and Cristina D’Agostino from the Swiss Center for Luxury Research and Luxury Tribune respectively for knowledge and insight on the luxury and NFT industry from Switzerland to the world.
References
10 popular 3D file formats used in various industries (2021). (n.d.). Retrieved June 29, 2021, from https://professional3dservices.com/blog/3D-file-formats.html
Boucher, J. L., DeCapo, T. A., Drylewski, A. C., Fisch, E. J., Levi, S. D., Liu, J. K., Toms, S., Ali, A., Cannuli, S., Hobson, C., Giroux, G., Huynh, M., Moore, M., Shadarevian, V., Urbina, J. A., & Yoo, S.-A. (2021, April 20). The distributed ledger: Blockchain, digital assets and smart contracts. https://www.lexology.com/library/detail.aspx?g=34d6a621-7d64-4660-b8bc-3d02d439c00a
Celia. (2015, June 24). 3 questions to Mark Harrop, founder and managing director of WhichPLM. 3D PERSPECTIVES. https://blogs.3ds.com/perspectives/3-questions-to-mark-harrop-founder-and-managing-director-of-whichplm/
CoreLedger. (2019, October 9). What are smart contracts? A breakdown for beginners. CoreLedger. https://medium.com/coreledger/what-are-smart-contracts-a-breakdown-for-beginners-92ac68ebdbeb
Cunningham, S. (2020, May 4). What is a low poly style in video games design? Digitalste. https://www.digitalste.com/post/what-is-a-low-poly-style-in-video-games-design
Morhart, F., Wilcox, K., & Czellar, S. (2020). Research handbook on luxury branding. Edward Elgar Publishing.
ms-off1. (n.d.).
Opinion: Will fashion and luxury pass the digital stress test? (2020, October 26). https://luxurysociety.com/en/articles/2020/10/opinion-will-fashion-and-luxury-pass-digital-stress-test
Physically based rendering: From theory to implementation. (n.d.). Retrieved June 29, 2021, from https://www.pbr-book.org/
Shenk, T. (2005). Introduction to Photogrammetry [PDF, The Ohio State University]. https://www.mat.uc.pt/~gil/downloads/IntroPhoto.pdf
Silva Araujo, G., & Piazza Gaglianone, W. (2020). Machine learning methods for ináation forecasting in Brazil: new contenders versus classical models [PDF, CEMLA]. https://www.cemla.org/actividades/2020-final/2020-10-xxv-meeting-cbrn/Session%202/3.%20Machine_Learning...%20Wagner%20Piazza.pdf
The role of IP rights in the fashion business: A US perspective. (n.d.). Retrieved June 29, 2021, from https://www.wipo.int/wipo_magazine/en/2018/04/article_0006.html
Webster, N. L. (2017, January 2). High poly to low poly workflows for real-time rendering. Informa Healthcare. https://www.researchgate.net/publication/316435393_High_poly_to_low_poly_workflows_for_real-time_rendering