Advice on creating a Digital Gettysburg Experience

[mhands]

Hey all,

Grew up loving Civil War history and, in the last 2 years, become obsessed with Gettysburg history. Love reading primary source accounts and earliest photos with the goal of trying to picture 'what did it look and feel like during the battle?'

I also love and work in technology. Around a year ago, I had the dream of making a digital experience of Gettysburg.

• 1:1 scale
• real-time (i.e. you can move around at will like a video game)
• photo-realistic
• representation of 1863 battlefields
  • ...and, eventually, likely troop movements (also 1:1 scale down to the regiment level ~150k animated people moving around)

I finally got the chance to sit down and power through a proof-of-concept recently and am happy enough with the progress where I'd like to dedicate my next 8 months to completing this project: watch a quick recording here.

Planning to do it in multiple phases:

1. Still working on the basics: (done) 1:1 scale landscape with accurate landscape proportions, basic terrain type blocking (e.g. forests, fields, etc.), (in progress) outline major rivers and roads
2. As attention span dictates, alternate between
   1. Outlining
      1. minor rivers, roads and lanes
      2. fences (post & rail, wormwood, cow high, stone walls)
      3. generic buildings (houses, farms, etc.) > then refine buildings to match photos
      4. unique buildings (saw/mineral mills, quarries, lutheran sem, college, etc.)
      5. terrain types like orchards, grassy knolls, forests, etc.
   2. Detail out specific areas by interest/importance (e.g. LRT, Wheatfield, Peace Orchard, McPherson's Ridge, Devil's Den etc.

First question, anyone know of good resources to help in this endeavor (e.g. websites collection of old photos by area of interest, personal projects/blogs where people have compared maps for accuracy, sketched areas of the battlefield based on primary accounts, etc.)? I anticipate posting questions here about things like 'Warren has this as a stone fence, Cope as post & rail' for help.

Second question, I'd like to start working on the first detailed-out section of the map this weekend or next. Ideally, starting with an area that has the highest concentration of photos as close to July 1863 as possible--or those areas of today's battlefield that are generally considered to closest match 1863. Any recommendations? Any advice would be appreciated. I'd love to share results in 2-4 weeks for feedback on any inaccuracies to fix.

 

[RayDoolittle]

This looks really cool, with tons of potential. I would love to eventually see it with troops from both sides.

As far as photos, I would guess that the Rose Farm probably has the most photos from within a few days of the battle. There are many available on the Library of Congress websites. Those photos would allow you to incorporate casualities into your views. William Frassinito's works would give you amazing photographic background.

There are some maps available (Visitor's Center, etc.) that detail exactly what was planted in each field at the time of the battle.

 

[Tom Elmore]

I see a lot of potential with your concept. Before I ask Hal Jesperson prepare a map for me, say for example of Cemetery Hill, I review available participant sources to see how well they match up against what is shown on old maps like the Cope survey or what Bachelder used, or compared with the more modern versions like the Desjardin map. Sometimes I have to make minor modifications based on those original sources. Also pertinent to your plans are to accurately identify the type of crops (wheat, oats, corn) planted in a given field, plus orchards (apple, peach) and so on. For the crops, how high was the wheat or corn at the time of the battle, and for that one can't beat accounts from actual participants even if they don't always precisely agree. If you plan on including night actions or views, the phase of moon, moonrise and moonset times, azimuth, etc. have already been accurately determined.

In answer to your question, the Devil's Den area I would suppose attracted the most attention from the earliest photographers and there are so many readily identifiable reference points, plus that area is entirely preserved, although a bit marred by past intrusions like the electric railroad.

 

[Kurtlg]

Hey all,

Grew up loving Civil War history and, in the last 2 years, become obsessed with Gettysburg history. Love reading primary source accounts and earliest photos with the goal of trying to picture 'what did it look and feel like during the battle?'

I also love and work in technology. Around a year ago, I had the dream of making a digital experience of Gettysburg.

• 1:1 scale
• real-time (i.e. you can move around at will like a video game)
• photo-realistic
• representation of 1863 battlefields
  • ...and, eventually, likely troop movements (also 1:1 scale down to the regiment level ~150k animated people moving around)

I finally got the chance to sit down and power through a proof-of-concept recently and am happy enough with the progress where I'd like to dedicate my next 8 months to completing this project: watch a quick recording here.

Planning to do it in multiple phases:

1. Still working on the basics: (done) 1:1 scale landscape with accurate landscape proportions, basic terrain type blocking (e.g. forests, fields, etc.), (in progress) outline major rivers and roads
2. As attention span dictates, alternate between
   1. Outlining
      1. minor rivers, roads and lanes
      2. fences (post & rail, wormwood, cow high, stone walls)
      3. generic buildings (houses, farms, etc.) > then refine buildings to match photos
      4. unique buildings (saw/mineral mills, quarries, lutheran sem, college, etc.)
      5. terrain types like orchards, grassy knolls, forests, etc.
   2. Detail out specific areas by interest/importance (e.g. LRT, Wheatfield, Peace Orchard, McPherson's Ridge, Devil's Den etc.

First question, anyone know of good resources to help in this endeavor (e.g. websites collection of old photos by area of interest, personal projects/blogs where people have compared maps for accuracy, sketched areas of the battlefield based on primary accounts, etc.)? I anticipate posting questions here about things like 'Warren has this as a stone fence, Cope as post & rail' for help.

Second question, I'd like to start working on the first detailed-out section of the map this weekend or next. Ideally, starting with an area that has the highest concentration of photos as close to July 1863 as possible--or those areas of today's battlefield that are generally considered to closest match 1863. Any recommendations? Any advice would be appreciated. I'd love to share results in 2-4 weeks for feedback on any inaccuracies to fix.

This is very interesting and I'm sure in the future a visitor to the battlefields will be able to use a device to see what the area looked like during the war. Good luck with your project.

 

[Josh The Lighthouse Guy]

You might want to reach out to Gettysburg NPS about this, if you haven't already.

 

[peacefulanimalguy7]

I did do something similar last year with Roblox Studio, but I have abandoned it due to me getting bored. Here's some photos from what I did.

The Brian Property

The Angle and Codori Farm
I would love to see your project be done honestly.

 

[Gettysburg Greg]

Hey all,

Grew up loving Civil War history and, in the last 2 years, become obsessed with Gettysburg history. Love reading primary source accounts and earliest photos with the goal of trying to picture 'what did it look and feel like during the battle?'

I also love and work in technology. Around a year ago, I had the dream of making a digital experience of Gettysburg.

• 1:1 scale
• real-time (i.e. you can move around at will like a video game)
• photo-realistic
• representation of 1863 battlefields
  • ...and, eventually, likely troop movements (also 1:1 scale down to the regiment level ~150k animated people moving around)

I finally got the chance to sit down and power through a proof-of-concept recently and am happy enough with the progress where I'd like to dedicate my next 8 months to completing this project: watch a quick recording here.

Planning to do it in multiple phases:

1. Still working on the basics: (done) 1:1 scale landscape with accurate landscape proportions, basic terrain type blocking (e.g. forests, fields, etc.), (in progress) outline major rivers and roads
2. As attention span dictates, alternate between
   1. Outlining
      1. minor rivers, roads and lanes
      2. fences (post & rail, wormwood, cow high, stone walls)
      3. generic buildings (houses, farms, etc.) > then refine buildings to match photos
      4. unique buildings (saw/mineral mills, quarries, lutheran sem, college, etc.)
      5. terrain types like orchards, grassy knolls, forests, etc.
   2. Detail out specific areas by interest/importance (e.g. LRT, Wheatfield, Peace Orchard, McPherson's Ridge, Devil's Den etc.

First question, anyone know of good resources to help in this endeavor (e.g. websites collection of old photos by area of interest, personal projects/blogs where people have compared maps for accuracy, sketched areas of the battlefield based on primary accounts, etc.)? I anticipate posting questions here about things like 'Warren has this as a stone fence, Cope as post & rail' for help.

Second question, I'd like to start working on the first detailed-out section of the map this weekend or next. Ideally, starting with an area that has the highest concentration of photos as close to July 1863 as possible--or those areas of today's battlefield that are generally considered to closest match 1863. Any recommendations? Any advice would be appreciated. I'd love to share results in 2-4 weeks for feedback on any inaccuracies to fix.

Good luck! Sounds like a very interesting and worthwhile project. I certainly am no authority on Gettysburg photos, but have followed in Frassanito's footsteps by searching out all the 1863 photo locations and taking now photos. I accumulate much of my work on my Facebook page which you are welcome to take a look at here: https://www.facebook.com/profile.php?id=100066644171244
If you see any material that interests you, feel free to use it in your project any way you wish. I like to share my work, so I make no restriction on anyone using it. Would love to answer any questions you may have so don't hesitate to ask. Again, good luck!

 

[johncla]

So is the basic idea to build a digital model of the battlefield, pick a specific time -- for example noon on the 2nd day, place all the regiments in their locations at that time, and allow users to explore the battlefield?

That will be very useful! One of the big benefits might be providing a 3d view of some spots that are really complicated, like the Houck's Ridge/Plum Run/Wheatfield area.

Best wishes!! My guess is you'll find the coding and terrain assembly part fun, but figuring out locations for all those units (approx 90 brigades = maybe 300 regiments ??) pretty tedious. I think there's digitized versions of the Bachelder maps online now, which will be a shortcut.

 

[johncla]

Oh, and one further remark. An annoying part for me of looking at photos of the battlefield is that I often can't figure out which way the camera is pointed. For example, a view looking along Chambersburg Pike -- are we looking E or W? A picture taken on the top of Cemetery Ridge -- are we looking West, North, South?

Maybe there's some way to build in a compass-like thing which tells the viewer which way they're looking?

 

[mhands]

@johncla and all, here's a rough snippet of Southwest corner of wheatfield. This week I'll be working on some terrain textures and meshes for rocks. I think with that, will really make Wheatfield/Houck's/Plum Run area feel more realistic, so hope to have that to share next weekend!

 

[johncla]

I love it! I'm assuming you have in mind a particular scenario - for example regiments A and B are on this ridge and the other guys' regiments 1 and 2 are behind a stone wall over there. And then maybe something happens...
So how do you represent the units ... regiments A,B,1, and 2? No hurry, just curious. Are they going to move, maybe following a clock? Or is it a static scenario, at for example 2pm on July 2?

Believe me, I'm on your side. It's so easy to stand on the sidelines with questions.
Shame on me, but I'm bad...

 

[mhands]

All, apologies, had rather lengthy responses to everyone but ran afoul of 'spaghetti' posting rule. Lesson learned. Will reach out send direct replies in DMs.

@johncla, welcome questions from the sidelines, especially great ones that hadn't occurred/thought through yet--and especially from someone who's 'in the area,' done this before and know what pitfalls lie ahead!

You've got the core intent.. 1:1 terrain and troops movements (focusing on days 1 and 2 first, and to whatever extent possible with historical ambiguities), photorealism (to whatever extent eventually lets me share the project with others so they can use it), and would love to initially include things like animations, fire-smoke-sounds, indicate causalities of time, and basic method of orienting users (like you said.. what regiment is this, what's the time of day, etc.).

I think there are a lot of exciting methods to achieve the above.. but generally the more exciting, the more complex & time consuming, resource intense, expensive, etc. So the goal is to have an end-to-end, first iteration that accomplishes the above by end of year, which imposes some (healthy!) constraints. Then, it's just a matter of prioritizing which features to add/enhance. All that to say, I'd love to share details and get your thoughts on the various trade-offs.

But to your exact question/example:

• 'regiments A and B are on this ridge and the other guys' regiments 1 and 2 are behind a stone wall over there'
  • More below but showing movement will be tablestakes, but ^100% obtains at a given point in time
• 'And then maybe something happens...'
  • The low res thought was at least have them move and fire at each other if they were in range (if that's what they did). But it's occurring to me while thinking about your question that it might be useful to abstract/indicate other types of 'something' (e.g. once the units disengage, is it enough to just show Regiments A and B moving--evening running--backward, or is it worth indicating that the unit is routed/retreating. Or say flanking was a particularly meaningful tactic for a brigade/regiment engagement, is it enough to show Wofford just shooting at Zook's flank or is it worth signifying. Perhaps counter-attacks, confused orders, etc.?
• So how do you represent the units ... regiments A,B,1, and 2?
  • The Cadillac option in the back of my head would be something like: you got a blue or red plane that runs under peoples' feet--and the line expands/contracts/snaps based on zoom level. The further you zoom out (conceivably seeing the whole battlefield), the troops would look like ants, but you would see things at a 'corp' level--every person in same corps would have a red/blue plane under their feet and you'd see giant lines, maybe using corp-level labels/names/insignias to differentiate; then you zoom in closer and your view snaps to a brigade-level, all the troops of the same brigade would have a contiguous line/plane under their feet--maybe insignias/icons for the their corp, maybe brigade commander labels, maybe infantry/cavalry/artillery, maybe that's all configurable. If you zoom out, you're back to the corp-level view and zooming in or jumping in as a character takes you to the regiment view, etc. I'd like to at least spend a couple days trying that to get a sense for whether it's feasible/how much time it'd take but unfortunately, guessing won't have time for the first iteration, so initial thought is just do the brigade or (if time permits) regiment view (rough mockup below--just proof of concepted static soldier models an hour ago, hot off the presses!). But would welcome any ideas and feedback on this before or during development--about 2-3mo out from starting on this (though have started on something similar for terrain, will DM you screenshots when I sit down to work on it this weekend).
• 'Are they going to move, maybe following a clock?'
  • Yes, that was the exact thought. But the model for navigating the animation across time is pretty architecturally fundamental not think of something and to change it later--major hassle and time sink. Really would appreciate any advice now or later 3-4 mo out. Initial thought is 'animations' are broken out by day, so you'd load the program or go to website, choose day and that day's animation would start @ 6 or 7am--troops either in camp or marching. Since size/movement speeds are 1:1 scale, thinking the base time would be too to keep things easy--if boring (i.e. troops would march ~3mph, charge ~6mph, it'd take Kershaw 15m to reach rose farm, playing at base speed you'd watch the wheatfield fight unfold over 2 hours but most of that would be watching repetitive animations of troops firing and retreating--though lots of exciting ideas that don't make the initial cut for end of year that could potentially using base speed more enjoyable--would welcome any others). Somewhere, maybe top right of screen, you'd have the 1) time of day ticking by minute by minute, the 2) options to fast forward or rewind at x2, x5, x10(?) speeds 3) pause 4) maybe a timeline you could click a point and jump to it.
    • The other basic pattern might be--instead of plotting the animations and navigation around a timeline--going with an 'event-driven model'. Got to run but would love to discuss more--other models, the trade-offs, peoples opinions, etc.

**Here's that screenshot referenced above:

Also, believe I share these on another post but not this one, for anyone curious on progress:

• v1 demo: Wheatfield
• v2 demo: Bushman Farm to Wheatfield (last week)

Will take to block off time this weekend and share any new progress!

 

[Molchman]

This looks absolutely amazing! One reference that I love besides Laino is this page. https://gettysburganimated.com
I am not affiliated with the site, but I often use this as a reference when I do my readings or listen to lectures. What you are doing is amazing. I can't wait to see what you come up with and I will be among those willing to give feedback and assist in any way. God Bless!

 

[mhands]

Hey all, have a new demo to share for those interested in progress updates: Gettysburg Demo v4

Quick notes:

• Unfortunately, lost lots of good progress (especially on farms, walls and fences, and crops--including the beloved 90 ft. tulip poplar)--still not sure how I bungled it (not good with version control and couldn't figure out how to recover it), but had to take step back out of frustration. Lesson learned: Will need to figure out how to prevent or recover something like that again.
• For all the progress, the goal remains 1:1 accuracy with 1863 references and photorealism--so there is a lot of tedium remaining to remove any trace of current day roads, resize and retexture the 1863 roads, lanes, streams; clean up trees, crop and rock density/color/types; etc. Pacing for the long haul!
• The photorealism vs. performance tension is becoming unavoidable. Goal is eventually to make this available to anyone (so it must work for the average, budget computer). This will likely be a constraint going forward--if time allows, might do a high graphics vs. low graphics setting.
• Plenty of tedium to do, but may take a pass at animations earlier than expected. Will be excited to share any progress there.

@Molchman, thanks! Great reference- was planning on reaching out to them to see if they could share any data/advice when the time came. Unfortunately, still a ways out from the large scale animations. Which is a shame, because the battle is most of our favorite part--the rest of this has been a long means to that end (though I will say, never had an interest in the terrain or town, but have learned a lot and newfound appreciation for it as a result of being forced to do it as a 'stage' for the battle animations.

@johncla, you may be interest in 1m through 1:30m mark. As we discussed above, rough first draft of the vision of having the regiment marker bars morph into the troop models at a certain level. (the demo had ~100 models representing Ward and Smith, will need to drastically scale).

 

[mhands]

Hey all,

Going forward, I could really use some help collecting 1) earliest (as close to the battle as possible) known and 2) highest possible resolution photos of key locations and buildings of Gettysburg as possible.

• Happy to share those photos I've collected, upscaled and colorized for anyone's convenience: Colorized, Upscaled Photos of Gettysburg
• In return, if you have high resolution copies of key Gettysburg locations, features and buildings, it'd be greatly appreciated if you shared those either as a link or uploading to this Google Doc Folder (if I upscale and colorize, you're of course more than welcome to download from the above link): Original Gettysburg Photo Inbox
  • If easy, please share the 1) approx. date it was taken 2) subject(s) of the photo 3) photographer's name -- either in the filename or in a DM

In particular, if you know of any reasonably early photos of 1) Steven's, Willoughby, or Weinbrenner Runs 2) Almshouse, Jail House, Rail Station, Kuhn's Brickyard, South brickyard off the old Brickyard Lane, 3) photos of any specific houses/farms (including barns) 4) any rare/obscure photos in general-- would especially appreciate as these are things I'll be working on soon and would take me a lot of time to track down