How long would it take you to make a level of your favourite retro game in Unity?

Interesting topic. I don't think I could give a legitimate answer, as different retro games will take vastly different amounts of time to make.
I reckon i could fully make Scramble, for instance (if someone else did the art) in less than a day. Elite? er, a little more time.
Having made some games on old computers, we'd probably spend 90% or more of the development time actually getting things to work quickly enough, and in the memory space available.
These aren't issues in Unity, so at the very least it should take ~10x less time. Then factor in having to work in asm vs c#, assembly times of 10 mins or more, sometimes even no Undo functionality in your editor.. it should take 100x less time to make something in Unity!

 

Your right and a lot of retro games were made by single developers or very small teams and made in relatively short periods (weeks-months). If you check out some of the old timer interviews, they were often able to port a game in a matter of months to new hardware with only a few team members.

Note on Art: Check out 8 bit (ZX Spectrum, Commodore 64) games they have very simple art, and sometimes an entire 'level' was just one screen.


Manic Miner (a retro classic and programme art).



Elite would be trickier to do a level of, as it was one of the earliest open world/universe games. What would you class as an Elite level a flight from Lave to Finji?

So you would include Launch / Flying Out / Fighting Pirates / Jumping / Fighting Thargoids (50%) / Jumping / Fighting Pirates / Flying In / Docking.



I've written a Elite Dangerous Ship Database (with flying), so it's doable.

The radar/UI and feel might be tricky to get right.

 

Talking of ports.. I did quite a few ports, Amiga -> ST -> PC / Genesis, etc. Line by line code conversion stuff. I could get a conversion done in a week or two using that technique.

 

You'd think so and that was my expectation as well but in my experience at least it doesn't work that way. Some of those old games were made literally by one person in 3 to 4 weeks. Assuming they worked 5 days per week for a month averaging 10 hour days that's about 220 hours. 1/100th of that time would be 2.2 hours.

Realistic I think is reducing the time down to about 1/5 to 1/4 so what took 1 month back then may be completed in 1 week now. IF the person knows what they are doing. A lot of the older games look the way they do simply because these people were primarily code warriors not artists. Superb programmers and that is why they were able to produce the games so quickly. Certainly just anybody couldn't have done it then or now.

This is why I firmly believe the major bottleneck is in the presentation layer. The creation of the graphics, music and sound fx. Especially today.

It's basically like your ports. How fast would it be to port only the codebase compared to having to also create the graphics, music and sounds from scratch even using the original work as references? I bet it would add a significant amount of time.

The same is true now. If you had a huge library of assets (say the Asset store actually had everything you need to complete your game except for all of the programming) making a game would be incredibly easy and fast at least compared to having to create all of the content as well which will almost certainly take far more time than any other aspect... particularly if you get caught up in the continual iteration, tweaking of the graphics as so many do.

There are other bottlenecks such as the greater complexity of the tools used. If you are using something like Unity Animator defining state machines that is way more time consuming than simply defining an animation in code even for the more complex cases where you define an animation table. In code that is dead easy and fast to do. Visually in a tool with state machine it is cumbersome and time consuming in comparison.

EDIT: That's an awesome background you have. Can you share the names of some of the games you ported?

 

Oh I agree, the presentation is the slow part - unless you are someone who can do it really quickly of course - there are those people around too

Edit:
McDonald Land PC
That's one of mine.. 68k to 086 conversion. That didnt take long, primarily because my mate Derrick;'s original 68k code was so damn good. Art was just passed through a converter to my format, so that took no time. I already had my library of sprite / scroll / audio routines.

 

Definitely. I have worked with many pixel artists and one guy in particular can knock out very high quality pixel art animations (for the size and color limitations given) incredibly fast. Literally like a set of 6 images for a character animation in about 20 minutes or so.

There is a lady I hire out work to as well who does excellent work and again is very fast. Now the crazy thing is neither of these people do this professionally. They do it simply because it is their passion. They make graphics constantly in their free time just for fun. So that is the kind of people I always look for to hire. Such people are usually highly experienced and very fast.

 

Awesome! How long do you think it would take to recreate in Unity assuming you could simply do the same kind of conversion process for the graphics, sounds and music? Less time, more time or about the same?

 

Well to make it identical would be very hard indeed! As the original was just code conversion. Translating asm to c# would be trickier, i would assume. So in this case it'd take longer.
However to get something complete which looked and played similarly would take a week perhaps? (for the full game).

It's not something I'd like to attempt!

 

OK Doom Level 1: Low Poly 3D + Low Rez Textures

Spoiler: Time Estimate

3h 3D Map Model
4h Map Textures
6h 2h Monsters x 3 ? I think
2h UI Bottom Bar
1h Gun
1h Shotgun
1h Armour / Health pickups
1h Doors
1h Door Buttons
2h Basic SFX
2h Basic FX e.g. Gore
3h Very Basic Music Track
5h Putting it together and Playtesting (Should be FUN)
32 hrs Total Time

Spoiler: Actual Time Taken

About 40 hrs over 5 days

Spoiler: Finished Result

Try it here -> https://arowx.itch.io/doom-retro-remake-level-1

Start Time Now - Wish Me Luck!

 

Doom 1 is one of my favorite games. Be ready for your example to be torn apart and criticized into oblivion.

 

Making slow progress.

 

Very interested in seeing this. One thing I notice though is you don't seem to be using quads / sprites for this and instead are doing it all in true 3D. Just noting it. Doesn't matter one way or the other. Other than I am just a bit curious if there is a reason you made this choice.

I think this could be your most popular game if done right. Fast action. Excellent feel. Tight controls.

 

Mr Arowx I believe you that you can make a game that fast. I have seen a YouTube video of someone making Minecraft in a day. So its possible. And in the old Spectrum era people used to churn out games in a week or less.

 

Either they used an existing asset that implemented the bulk of the game for them or they only achieved alpha era Minecraft.

Building games for older 8-bit machines was easy until you started to push against the boundaries, but I don't know of any examples off the top of my head where they took a week or less outside of a port of an already developed title.

If there are examples though it's because developing for those platforms only required you to be a good programmer. Being good at artwork, music, sound effects, etc only became important once we started moving beyond that era.

By the way one of the shorter development times I'm aware of was E.T. for the Atari 2600 which took about six weeks.

http://www.digitpress.com/library/interviews/interview_howard_scott_warshaw.html

 

About 13 hours in basics working UI, gun hand, potions/armour, switches, doors, acid pools working.

Note using physics so no steps all ramps. Not a 100% match of first level but getting closer.

Still not got:

• Enemies going to try and model them in 3D (the originals were repainted photographed models) with fixed poses they can switch between.
• Shooting.
• Textures.
• Ammo.
• Shotgun Hands.

Having fun though. Calling it a day for now.

 

What's stopping you from using the same collider for the stairs that you would use for the ramps? I doubt anyone would care or notice on a fast paced first person shooter though the ramps might give a better ultra low poly feel than stairs would.

 

If I remember correctly the steps in Doom pretty much feel like ramps anyway.

 

Interesting what you find out when your looking for details on an old game - doom sprites started out as real world models...


https://www.rockpapershotgun.com/2014/12/11/john-romero-doom-art-archive/

 

Some did. A lot were just made from scratch because working with clay in that context was ridiculously cumbersome and time consuming when you needed to change poses. Like, I hesitate to call the Pinky demon a scan when it was more completely drawn over.

 

Zombie Guard and Trooper (Modelled and Posed, Not Active Yet) along with programmer art Doom Guy selfie (looks like hell!).

 

And the Imp (Just posing not active).

Still lots to do!

 

ArrowX, what you're doing is not a recreation of doom level. It is a tribute at best. But not recreation of the original.

If you're recreating original level, enemies must be sprite based. The gun must be sprite based. The face of the marine must be sprite based as well.

What you're doing right now is not doom.

If you want to use 3d models, try quake 1 level instead.

 

Sprites are made of pixels right and 3d models render to pixels, therefore I'm almost using sprites?!

I've got the monsters chasing the player! It's getting there!

PS: Can you get a shotgun from the shotgun zombies or do you need to use the secret room?

 

The makers of doom photographed real 3d models. I could take my 3d models and photo them then put in sprites but that's a who load of extra work just to show the models in the game!

 

So? Deal with it. You're recreating something, not reimagining it entirely.

 

See this is why you need to choose a simple game (eg Pitfall II) to recreate. Picking one of the most popular retro games and only half heartedly recreating it was doomed to get people complaining.

 

Look at goleden eye source.

 

I've even made a spinning intro screen!

I'll just say it's a demake of the new one!

 

What if I have invisible to the player monsters that take lores selfies to a sprite doomguy can see?

 

does parsing doom.wad in unity count as a remake?

been working on this with a friend, but pretty sure would still take several months to add missing features..

 

What if you stuck to the brief?

 

No.

If you're recreating the first level, then visuals of original Doom are your specification.

Arrowx, what you're doing is ignoring original specification completely. You quickly throw together a result that sorta vaguely resembles the original, say "Yay, I'm done!" and call it a day.

In contract work it is a no, and it is a good way to ruin relationship with a client and have your contract terminated without getting paid.

Basically, you should adhere to the specification down to every pixel. Visuals, lighting, lack of texture filter, sprites instead of 3d models, etc. Then it is recreation.

This, by the way, applies to ALL screenshots you posted in this thread. You ignore original specification in all of them, don't pay any attention to detail, stop at 5% of work and then call it a day. Which is actually quite disappointing.

 

I'd say no, although it is an interesting project on its own.

A recreation would be something similar to freedoom project:
https://freedoom.github.io/screenshots.html

 

Stars emit light and my monitor emits light so clearly my monitor is a star, right?

 

Speaking of "Almost using sprites", there's this example:

This is a level from saints row 4 where they emulate look and feel of 2d beat em up game. They ARE using 3d models, but the game looks exactly like spirtes would look. Down to every detail - including animation made in just a few frames, sound, music, etc.

This is a proper "recreation of the original".

 

The monster looks really weird try giving some exponential red fog to the scene so the monster blends in

 

Perhaps we should rename this thread to "How long would it take you to make a vague approximation of your favourite retro game in Unity?"

 

Do max payne or goldeneye next.

 

Why else did you think I had such a strong interest in stop animation a while ago?

 

At first I was perplexed by the OP's insanely low estimates for each task. If I can even open up Visual Studio and find my code in 1 hour, then I am lucky, haha.

Then I read through and realized these are atari clones. The type of games newbies make in tutorials. Those really DONT take very much time at all. That is why they are in tutorials.

A single level for a space shooter like asteroids shouldnt even take a day. A total newbie who has never even touched gamedev before could go through these tutorials in 3, 6, 9 hours.

This is verifiable too, as there are entire video series that are 3, 6, 9 hours long where the creator explains and teaches while doing exactly that: make a retro atari game with one level.

For myself, it just depends.

I've created entire game engines that, for certain games, rival that of Unity3D in just 9 hours. In fact my estimates in an experiment showed that engine to save time in development for one game I was making- making it better than using Unity.

I've also worked my ass off for 16 hours straight getting nowhere because nothing worked. Just enormous time sinks wasted for absolutely nothing.

My usual for any one feature is 1-3 work days (4-16 hours). Any task, no matter how small, seems to always consume at minimum 1-2 hours. Rarely will any task or even feature take longer than 6-8. Thats a rarity.

And these atari games are really elementary in featurelist. To see even Unity employees acting like arrowx is underestimating is strange. They must be thinking of more complicated games.

I keep rigid track of how long features, tasks, and category types take me.

Anything that deals with Unity always takes an incredible amount of time. Far more than anything else, except multiplayer. Of all categories, Multiplayer and GUI are king, topping every form of measurement (difficulty, minimum time, average time, wasted time/failure time, brain tumors, etc.)

And while AI and intelligent programming (ex. complex proc gen or learning fancy maths) also reach the top, they are very enjoyable, less frustrating, and for AI/ProcGen much easier- just requiring time still.

Retro games dont have any of that stuff. GUIs are super simple buttons or scores, AI isnt really AI, and no multiplayer unless on-screen.

For my devtime, all other features seem very consistent. Brand new systems from scratch, about 4-12 hours each, average. New systems derived from previous ones, about 4-9 hours each. But nothing I do is ever simple.

One level is easy. What consumes my gamedev time is all these tiny tasks that seem to have a high minimum, added together hundreds of times. Something as simple as reorganizing files on my HDD or not writing a script to automate changing Unity settings can take a surprising amount of time because I have to move my mouse and click on files- it all adds up. Hell, the time I have to wait after hitting "Play" or compile times alone have probably resulted in hours of my life wasted when it's the 10,000th time you do it.

There is a speed to ultra simple games that larger projects dont have. In fact, larger projects snowball into inefficiency. Longer load times, eternal recompile times, endless systems testing, numerous bugs popping up, putdated systems or code, and so many systems interconnected it becomes a confusing jumble of engineering and code. Even when you organize well and use the strategies people have created to deal with complexity.

And it makes sense why so many developers choose small games and the overused genre of Platformers. That is the genre of most retro games all the way up to (and sometimes past) the invention of the CD.

Thats for me, anyway. My experiences, observations, and time tracking data.

 

Looks good, I see nothing wrong with taking a few shortcuts and not following exactly the original.

The essence is there and you seem to have a good ratio of outputs / time invested.

We all know that the devil is in the detail and the polishing, fine tuning, optimizations required for a commercial game is the stuff that can make the development quite frustrating at times. Nothing wrong in cherry picking the fun parts, since you don't plan a release.

I like your development posts (better than some of the crazy ones, though they are also entertaining

 

Few days pretty much regardless of the game.

Prototyping is easy.

 

Good thing no one ever says their favorite retro game was Pong.

 

Few hours.

 

But to answer the question, idk because all my favorite games dont really have levels.

I'd have to get a general idea to define an "area" or chunk of content.

But since one level means youre making a MVP... I'd imagine it would take me months working full time. Just like any game.

My favorite retro games include

Starflight (Sega)
Shadowrun (Sega)
Earthbound (SNES)
Willow / Fester's Quest (NES)
Dragon Warrior (NES)
Maniac Mansion (NES)
The Immortal (NES)
Young Merlin (SNES)
Simon the Sorcerer (PC)
Interplay's LOTR Rpg v1 (PC)
Parody LOTR Text games (Retro PC)
And I wont even get into retro multiplayer games (Warcraft 1, UO/EQ, Age of Empire) since those, although newer than atari, are still retro

All of these would probably take me many many months at least to create a MVP. They are the harder types of games to create. Year(s) to complete.

I really like the idea of text based games. They were among my first ideas for what I could do myself. But when I was a newbie, I went for the big dreams instead. If I knew better I'd have went with a text game. I'd be in the same spot I am now, but with a release or two behind me.

Honestly I wish I got into gamedev as a teenager, rather than as a full grown adult. Or I wish I got to do that segment of my life a second time (infinite freetime, no responsibilities, fully supported financially, extremely motivated & enthusiastic, thirst for knowledge). I'd have made a Text based adventure game, then a MUD, then finally a Starflight or Shadowrun.

I had the opportunity as a teen to make an RPG, but I passed on that after playing around with a SNES RpgMaker rom for a few hours. It just wasnt my thing. Dreaming of making FF7 was though, for some reason.

At heart, I'm a game designer, not a programmer or artist. My youth was spent staying up all night with my best friend creating PnP RPG's and my own card games. And as much as I love programming and feel so naturally at ease with code, I'd never become a programmer working on someone else's game.

Text Based games feel alot more open with design when you want to be a designer, arent much of an artist, but would enjoy programming the engine.

Now though? I'd say text game would take me a week or two to get a MVP. Maybe a month max. And I wouldnt use Unity, I'd write my own engine. But I'd actually end up taking 3 months for MVP as I expand it to become some kind of open world sandbox "where you can do anything", rather than a "Choose Your Own Adventuee" story, which I LOVED as a kid, lol.

 

This ^.

I did mess about with game development in my early teens but then found beer and the rest is history (that I don't recall).

And this...

If I had time to play I'd say Star Quake, Starion (just so I could play with line gfx) or The Hobbit (all Speccy games).
Although the challenge for me would be to dust my speccy off and have a go in Z80.

As for the Doom level, its looking good IMO but as others have mentioned the authenticity isnt there. Perhaps you could use billboards.

 

The camera FPS rig needed for doom!

 

Umm...

What?

Why aren't you using multiple cameras and layers?

And why do you have Doom guy modeled out at all?

 

-_-

Those screens make me want to spend next two weeks redoing this properly. Then I remember that I have another project that needs attention.