Dave McD is back this week with the final part to his mini-series which has explored an environment-artists perceptions of “accuracy” in heritage. Of particular interest will be the call for standards and clarity – something which Archaeology has been striving for (look especially at the great work being done by the ADS), however the lack of transparency, interoperability or accessibility still seems to be problematic for those wanting to approach data from outside of the field. Artists, game designers and developers are looking to use this information, but without the metadata being accesible in ways which matter to them it is likely that the models will continue to be used and abused out of context. I will pass over to Dave now to give his concluding remarks – hope to hear your responses and opinions!
Right….where was I? Apologies for the gap in between posts, but I’m back with the final part of my mini-series looking at 100% art/visual accuracy in heritage.
After exploring some of the potential pitfalls/issues and reaching a conclusion, I’d thought I’d end this with exploring what could be done to help keep accuracy in the high 90% mark.
Please bear in mind this will be more UK-centric so if you’re reading this in the rest of the world, you may be doing some of these already. However feel free to share amongst your peers.
Also, they are thoughts, suggestions. Easy to put down but in some cases difficult to put in practice. Nevertheless, worth pursuing, if important.
Accept you can never be 100% accurate.
Simple really, however there needs to be a range. You can be 0% accurate theoretically. So straight away there’s a scale….0% to 99% as an “industry” you need to (as best as you can) quantify that scale. What are you looking for, what do you expect to see/accept in a 90% accurate model compared to say a 99% accurate model
Work with the best artists and scan technicians (super-important) in determining the above issue. What quantifies a 99% accurate model? If a global ruleset is created then this can have positive effect when handing out projects to artists down the line. They know their boundaries and you’ll be speaking their language. Over time a consistent quality mark will be achieved. Also look into “standardising” any equipment that is used. So for example, all photogrammetry projects must use a certain DSLR camera or a certain group of lasers need to be used for laser scanning. This will be updated as and when tech ology evolves.
Once a ruleset/code has been established, a brand/historically accurate classification can be established. This could be used on videogame packaging or in films/TV. Bit like the BBFC PG, 15, 18 ratings which can be found HERE!
The same “heritage classification” could be used in museums displays. Work and educate with the public in helping them understand what they’re getting. Over time, with greater understanding by the public and within heritage/digital, the standards will go up; they know what they’re aiming for to get that “gold” historically accurate kite mark.
Work with the Best
If you want high standards then you need to work with the best. Scanning/photogrammetry technicians live and breathe their work 24/7. Games artists are obsessed with materials/textures, polygon counts. Build up a good list of companies and organizations. That way you can put them in touch with each other. So for example, a games company is approached to do a VR experience of a Roman fort, however they have no experience with scanning. Simple, you know one of the best scanning companies in the UK, connect them up.
See it as heritage becoming a matchmaker
There are already hundreds if not thousands of scanned models gathering dust on servers around the world. Now say I was asked to look at one of them, maybe put it into a game. I then become an archaeologist, albeit a digital one. Naturally I will have questions about the scanned object e.g.:
- What equipment was used to capture it?
- Is it any good (quality wise)?
- Is there anyone I can talk to who originally scanned it?
… And many more
So the question I pose to you dear reader is that information there? If not then why not? Surely having as much metadata attached to a scan as possible will help future generations of digital heritage groups as well as present day ones. Again, this needs to be standardised either worldwide or by country.
Now the metadata could be digital or paper or both, but essentially you are tagging the digital object with information such as
- Laser Scanning
- Laser scan used
- Company hired to laser scan
- People who actually did the scan
- What dates did the scan take place
- Who signed off on the scan?
- What software was used to process the scan?
- Accuracy rating
- Cameras used
- Any additional equipment used
- Name of company
- Name of individuals responsible for photogrammetry
- Dates when photogrammetry was used
- Time of Day (helps with lighting/texture info)
- Who signed off on scan?
- Accuracy rating
Now I expect there’ll be many more questions that can be answered, but do you see my point now? Imagine in 2035 and you encountering a digital scan from 2010, do you see how this information, this metadata can help you understand, what, where and how this digital scan to be be?
Hopefully some of these suggestions will help towards quantifying, improving and maintaining accuracy not just within digital heritage but in film and videogames. Some will require more work than others but If you are interested in employing any of these then cannot stress enough how it needs to be adapted by everyone, either in your country or worldwide.
This means someone, a body, taking ownership of making these rules alongside videogame makers and laser scan technicians which everyone then adheres to. Not an easy feat, but the alternative is wasted money, badly scanned models of objects no longer available and a “Wild West” approach to scanning and digital implementation that does nobody any favours especially the future generations.
Please provide feedback, questions and ideas below!
Have your say
The following polls are to guage the audiences view on the topic – have your say by voting or commenting below!
Standards for meta-data, accuracy and process exist (for example the London Charter and the ADS Guides to good practice), but they are largely confined to the Heritage Industry, are not enforced and rarely offer specifics for how to deal with the borader questions of accuracy and authenticity as related to external applications or implementations in specific scenarios:
If we are to produce standardised metadata, what do you think the most valuable entities would be for recording systematically?