4/29/19 - The Waiting is the Hardest Part

I left Dianne a message this week seeing how things were going.  As I mentioned in my post from a couple of weeks ago, I don't want to put too much pressure on her.  You don't randomly call a stranger about stuff they may or may not have in their house and just demand "drop everything in your life and look for this stuff".

I've been thinking about the future possibilities for T7G and T11H.  I'll note that these are just my thoughts and do not reflect the business plans of Trilobyte LLC nor of Attic Door Productions.  I'll also note that if the original footage does still exist, I have no legal rights to it and the footage will likely go directly to Trilobyte, LLC as it is their property.

If The Footage Exists

Once again, I have no real say in what happens if the footage does exist.  However, I would talk to Rob and see what he thinks about the following:

1. Digitally Capture the VHS Tapes

This is an obvious and a must.  I'm pretty sure that if the tapes exist, this is a high priority for Trilobyte and fans alike.  There would be no debate on this one, although, you'd need to find a good company that will ensure they use a VHS deck with similar tracking encoding as the original VHS camera.  They also would need to capture the audio and video in LOSSLESS quality and not in the compressed versions often used with personal digital video capture services.

2. Publicly Release Some or All of the Footage in Original Quality

This is a dicey area.  The footage is Trilobyte's property.  It isn't something that fans have a "claim" to.  It represents possibilities for future revenue for the company.  Just "releasing" it to the public on Archive.org might not be the best thing for business plans.  That being said, I see two potential options.

- Release ALL the Footage in Original Quality on Archive.org

This is UNLIKELY but, is something I wish would happen.  Despite the footage being Trilobyte's property, it was "lost" for 20+ years.  Putting the footage on Archive.org would ensure that this piece of gaming history is preserved.  It also would give fans the ability to see all of the outtakes and deleted scenes in their original VHS quality.

- Release Select Outtakes and Behind the Scenes Footage on Archive.org

This is a more realistic ask of Trilobyte.  The fact is, there may be deleted scenes the company could use in new/revised versions of the games.  We have technology today that wasn't even a thought back in 1993!  However, there would be no risk to the company bottom line if they released outtakes of known scenes and random behind the scenes conversations, director instructions, and other fun things like that.  This would at least provide fans with unique detail into the filming process for the game(s) without actually giving away the stuff that could turn a profit.

3. Upscale the Footage and Textures (if they exist) to HD

This is another OBVIOUS move.  If the video footage and possibly the original textures for the games exist, they need to be upscaled to HD because this would allow for the creation of a remastered version of the games.  The question would be, what is the best way to do this?  Here are some options:

- A Commercial Solution

Trilobyte hires a company to upscale the video footage and old textures using some level of proprietary technology.  The advantage here would be the expertise of the company in doing this kind of work and would likely have good results with minimal effort on the part of Trilobyte itself.  The problem is, it costs a lot of money to have a company properly interpolate and upscale old video footage and textures.

- An Open Source, AI-Driven Customizable Solution

The ESRGAN AI Engine is open source and is used to upscale game textures.  It is easy to use and relatively easy to train for custom applications.  It requires time and a medium to high-end computer.  The only down-side is that it would require hardware and someone to train the models, both for upscaling the video and for upscaling game texures.  Granted, ESRGAN has some amazing models already for texture upscaling but, it is always better to customize a model to fit the specific kind of textures or video that is being worked with.  This might be a cheaper option, even with the cost of custom hardware and the time necessary to train custom models.  The custom models would yield better results as they would be trained to do SPECIFICALLY what Trilobyte needs them to do.

4. Re-Create Updated Versions of the Games

This would be a potential end goal if the footage could be found and upscaled.  It would be possible to literally re-create the original games, possibly with the inclusion of new or alternate scenes.  The games could be kept point and click or could be moved into a "free-move" environment.  Either way, the games could be re-created as they were originally intended using modern technology.  Think of the idea as T7G and T11H director's cut editions.  Of course, there could also be play modes that remove the extra or changed scenes and just allow the player to experience a modern version of the classic games.

If the Footage Doesn't Exist

Right now this is a high possibility.  What can be done if the footage doesn't exist and is in a trash or recycling center?

1. Call the Medford Area Transfer Stations

It's worth a shot.  However, I doubt that anyone would know where a box of old tapes and CDs were buried 2-3 years ago.  If they were sent to a recycling center, they would have been wiped and shredded.  But, hey, maybe someone saw the box, saw Trilobyte's name, and took the stuff.  It would be worth trying anyway.

2. Try to Upscale What Exists

It is possible to export all of the video and audio assets out of The 7th Guest and The 11th Hour!  The problem is getting the stuff back INTO the games.  Even then, the games would need to be re-coded/re-written, which is a daunting task to say the least.

It MIGHT be possible to take the original game assets and move them into a more modern development engine/environment.  That would still mean re-coding all the puzzles and some of the animations from scratch.

- But What About the BlueScreen Quality on T7G?

Getting better images out of the T7G data we have is a unique challenge.  While modern video compositing applications could work miracles with the old VHS blue-butcher-paper bluescreen footage, without it we are stuck with 640 x 320 video with very little detail.

The only way I can think of to improve that footage is to train a very specialized ESRGAN model.  To do this, you'd take the better footage from The 11th Hour and footage from other better bluescreen FMV games from the same era.  You'd then intentionally create a mask around ONLY the actors in the FMV videos and pixelate/distort them as much as possible.  This would train the model to attempt extrapolating/inventing the missing data of the actors.

It wouldn't be foolproof or guaranteed.  It likely would take a LOT of iterations to create the model and it would take manually doctoring/messing up thousands of FMV images.  In short, it would be a TON of work, and might not even get great results.  It still might be worth trying.

The rest of the game footage without FMV scenes is easy enough to attempt AI upscaling.  The models/textures are fairly normal by today's standards.  The AI engine might even be able to add more detail by training it on better wood, marble, and other such textures used in the games.

3. Just Be Happy With What We Have

At some point it is just best to let a sleeping dog lie.  While I'm personally tempted to try getting more detail out of the existing T7G FMV scenes using AI, it would be a ton of work.  I'd be willing to do it just to try it out, however, if the process didn't yield good results I wouldn't want to go back and try again.  Even if it DID yield OK results, it would be a large effort to upscale every frame of every video.

At best, I could stitch a "movie" together of the footage but, it likely wouldn't be worth trying to re-code or re-create the original games.  Yes, we may be able to upscale the video footage from the original games but, a "movie" is probably all we'd get out of the experience.

----

So, those are my thoughts as to the potential future regarding The 7th Guest and The 11th Hour.  With the video footage a lot of interesting doors get opened.  Without the original footage there is still some possibility of AI-assisted upscaling but, with very limited applications.

I'll keep everyone updated with any new developments regarding the footage.  As always thank you for your attention and support.

Update 4/12/19 - The Beginning of the End?

Hi Everyone,

As you know, this quest is a long on and off journey into gaming history obscurity.  The quest, to find the original video footage and any other assets from The 7th Guest, The 11th Hour, and other Trilobyte Inc. properties.

20+ Years is a LONG Time

It is difficult to find or even expect people who retain work from over 20 years ago.  I was 15, going on 16 in 1999.  I can't think of one item that I've retained from those days.

That being said, I've made some progress moving forward.

Earlier this year I tracked down some of those who worked most closely with the original footage.  None of them had any copies or materials related to the work.

I didn't do much searching after that.  Life gets in the way, and this is purely a one-person volunteer effort.

A Renewal of Interest

Last week the 25th Anniversary Edition of The 7th Guest launched.  It inspired me to continue digging.  Specifically, I wanted to track down Diane Moses/Watson.  Dianne seemed to be the potential key to everything.  She was, among other things, the office manager for Trilobyte, Inc. during its entire existence.  If anyone would know where stuff was, it would be her.

So, I dusted off my Google skills and decided to have another run at things.  After several hours and some weird free people search sites later, I tracked down what I thought were some good leads.  95% of them were wrong and lead to disconnected phone numbers, disconnected email accounts, and wrong numbers.  Out of curiosity, I decided to try one last phone number before I went back to the drawing board.  I left a message, expecting nobody to return the call.

The Beginning of the End

The call I made on a whim paid off!  Dianne was nice enough to call me back and leave me a voicemail.  It WAS the correct Dianne, although, she has a different last name now.  I called her back and we had a good chat about Trilobyte, and that I wasn't the first person to ever call her out of the blue about the good old days.

In the course of our conversation, she gave me both some amazing and devastating news.  She was the one who had the videos and a bunch of other Trilobyte materials!  The bad news, she thinks she threw all of them away a couple of years ago.

Now, Dianne wasn't sure if she did or didn't throw away the box of Trilobyte treasures.  She is going to look and see if she still has them and get back to me.  

On one side of things, I can 100% understand someone getting rid of something over 25 years old, especially when nobody had asked about it.  (Yes, NOBODY had asked her about the footage in 25 years)  On the other side, I REALLY hope that she didn't throw it away and just thought that she did.  The material she had is a piece of gaming history and wasn't the only treasure in the Trilobyte box.

If she does have these assets, they will end up going to Rob Landeros at Trilobyte.  I hope to continue talking with him about preserving this footage both for the fans and for Trilobyte use.  Obviously, with today's technology, it's easier to apply masks and keys to the worst bluescreen footage.  It's also possible to upscale the video to HD, especially with the advent of new AI-based tools like Waifu2x-caffe.

If the footage still exists I hope Rob might include me in the preservation efforts, and allow me to make the full contents, outtakes, and deleted scenes, available to the fans.  It might not be that simple due to IP rights, but, that was my original goal.

I'm going to give Dianne space in the meantime.  I'll check back with her in a couple of weeks or a month.  It's been 25 years, surely we can wait a while longer.

But What If...

What if everything was thrown away?  Well, the only thing I can do is contact the dump in the hopes that they might know where something like this would be buried 2-3 years ago.  That is presuming the footage didn't go to an electronics recycling facility and that a garbage dump would know where they were dumping things.

This isn't like the Atari ET cartridge find, where thousands of cartridges were buried and someone remembered where they were.  This is a random box in a garbage dump, and over 2-3 years who knows if the contents would even be worth finding.

Yes, I can dream that someone at the dump saw the stuff, knew what it was, and set it aside.  But, there is a high chance that all of the footage is lost forever.

Let this be a lesson to game developers and companies.  Archive everything.  If you get sick of holding on to it, put it on Archive.org or somewhere that will preserve things for you.  If you are someone from the 90s era of gaming and have raw footage or audio, get it preserved for the sake of the fans of your work.  You never know how it might be able to inspire others and possibly be a road to a better tomorrow for you. 

Update 1/1/19

HAPPY NEW YEAR!

I got some responses to my emails from Christmas eve!  The first was from Keith Ranney, who indicated that he didn't have any access to the original game footage.  However, he did state that some of the masking work was farmed out to a team in Ashland, OR.  I did look at the credits in the game again and tracked down some members of that team. (I think it is CyberROM Technologies)

Stephen Clarke-Wilson also got back to me via Facebook and said that Virgin never got access to the original video footage.  However, he did find some still photos from the shoot/production itself!  He's trying to find a way to scan/upload them and when he does I'll be happy to share them here.

Today I reached out to Andrew Anderson who did some of the video work on the game.  Andy still does graphic design and video work.  He has taught some of the best people in the industry today as you can see from his website: http://www.realandy.com/index.html .  Andy got back to me and said that he doesn't think that he has much, if anything from the original game but, he is going to check his attic and see what pops up.

I did reach out to Verin Lewis who worked with Andy at CyberROM Technologies Inc.  That was back on the 24th of December.  I haven't heard back from either of his email addresses but, will continue to try in the new year.

One person I can't track down is Dianne Watson.  She would likely be the key to unraveling a lot of the mystery behind Trilobyte and other "lost" video assets that might exist.  However, there is little or no trace of her online that I can find.

Debra Ritz-Mason played Martine Burden and was the Talent Director on The 7th Guest.  Sadly, she is deceased.  She belonged to a company called the Rogue River Motion Picture Company.  I haven't been able to find anything about that group so far.  They may no longer exist and I don't know if the company would know anything about the tape itself.

So, a lot to think about for 2019!

Thanks again for following this blog and my progress!

-Chris