Metal Gear Solid Delta: what a faithful remake actually means in production
A faithful remake is a deceptively hard product. The marketing line sounds simple: keep the original design, rebuild it on a modern engine, ship it on current hardware. Inside a studio, that sentence becomes a chain of decisions about camera, controls, AI behaviour, audio mix, art fidelity, certification, and player expectation. Metal Gear Solid Delta sits inside that chain. It is a full rebuild of the 1998 stealth-action game Metal Gear Solid, intended to preserve the original level design and narrative beat structure while using contemporary rendering, animation, and input conventions. For game developers, producers, and technical artists, the interesting questions are not whether the project exists, but how a team of that size scopes the work, what the engine migration actually costs, and where the line sits between faithfulness and modernisation.
This article walks through the production realities of a project like Metal Gear Solid Delta. It treats the title as a case study in faithful remake craft: the preproduction questions, the art and animation pipeline, the audio rebuild, the AI and gameplay simulation work, the QA strategy, the porting considerations across PC and current consoles, and the risks a producer should price in before green-lighting the schedule. The goal is to give an experienced GameDev reader a clear view of how a studio plans, staffs, and protects a remake of a beloved single-player game.
What “faithful remake” means as a design brief
A faithful remake is not a remaster and not a reimagining. The distinction matters because each term sets a different scope for the team. A remaster keeps the original code and assets, typically raising resolution, frame rate, and asset filtering. A reimagining keeps the brand and themes while rebuilding mechanics and level design from scratch. A faithful remake keeps the original level layouts, camera beats, narrative structure, and gameplay feel, but rebuilds art, animation, audio, and underlying simulation in a modern engine. Metal Gear Solid Delta targets the third category.
That brief creates two opposite failure modes. The first is over-fidelity: the team reproduces the original camera angles and input quirks so precisely that modern players find the game awkward. The second is over-modernisation: the team rebuilds the controls, adds open-world padding, or restructures boss fights until the product no longer feels like the game players remember. The production challenge is to keep the recognisable skeleton of the original while letting lighting, animation, and input meet modern expectations.
Scope decisions a producer has to lock early
Before a single asset is rebuilt, a producer working on a faithful remake typically has to answer a small set of scope questions, because each answer reshapes staffing and milestone planning.
- Is voice acting re-recorded or kept from the original masters? Re-recording changes licensing, casting, recording budget, and lip-sync work.
- Are cutscenes rebuilt in-engine or kept as compressed video? In-engine cutscenes improve aspect ratio handling and accessibility but cost animation and camera time.
- Are boss fights redesigned for new control schemes, or kept as scripted sequences that tolerate the original’s stiffness?
- Is the original score preserved, remixed, or recomposed? Each option has different music supervisor, licensing, and integration work.
- Are difficulty and AI balance numbers tuned for modern players, or kept as shipped in 1998?
These are not creative flourishes. Each answer drives headcount, milestone length, and the QA matrix. A team that decides too late ends up re-planning mid-production, which is the most expensive moment in a remake schedule.
Engine migration and the cost of a new renderer
The visible side of Metal Gear Solid Delta is its new real-time renderer. The less visible side is everything that comes with it: asset conversion, lighting rebuild, shading model rewrite, and platform certification against a new graphics API. Moving a 1998 game’s data to a modern engine is rarely a one-step export. Levels designed for fixed pre-baked lighting have to be re-baked, often with physically based lighting and a higher dynamic range, and the team has to decide what to keep from the original art and what to rebuild from reference.
For a faithful remake, the engine choice usually follows three constraints: the team has to be able to import legacy geometry, the renderer has to support modern post-processing without breaking the original mood, and the engine has to ship on every platform the publisher wants to reach. A team that picks an engine without an asset import path for the original file formats ends up rebuilding geometry by hand, which quietly inflates art scope.
Asset conversion pipeline in practice
Asset conversion is one of the largest silent costs in a remake. The original Metal Gear Solid shipped with rigid hierarchical animations, fixed-camera triggers, and precomputed lightmaps. A modern pipeline expects skeletal meshes, PBR materials, and runtime lighting. Bridging the two usually means a converter tool written early in preproduction, with the following responsibilities:
- Read the original geometry and convert it to a mesh format the modern engine accepts, with tangent space and UV channels preserved.
- Map the original palette-indexed textures to PBR-friendly albedo, normal, and roughness sets, either by reauthoring or by procedural generation.
- Translate fixed-camera triggers into the new camera system’s volumes, including detection radii, blends, and look-at constraints.
- Re-bake lightmaps for the new dynamic range, then validate that the original mood survives the brighter base lighting.
Without a converter in place, the art team recreates every asset manually, which turns a two-year art schedule into a three- or four-year schedule. With a converter, the art team still has to validate, fix, and reauthor large portions of the data, but the marginal cost per asset drops.
Art and environment rebuild under a faithful brief
The art brief for a faithful remake is unusually narrow. The team is asked to produce a 2020s-quality image of a 1998 layout. That is harder than it sounds, because the original layout was tuned to a specific camera, a specific texture budget, and a specific lighting model. Replace any one of those three and the level reads differently. Replace all three and the level reads as a different game.
The pragmatic answer is to set visual reference targets per level, not per project. For an outdoor jungle or a facility interior, the team picks two or three screenshot references from the original and rebuilds toward them. The reference is not “the original game” in general, it is the specific image of a specific room at a specific camera. That keeps the rebuild from drifting into either a remaster-style polish or a free redesign.
Materials, lighting, and the readability problem
Modern rendering introduces a tension the original did not have to solve. Physically based shading and high dynamic range lighting make textures more realistic, but they also reduce the kind of contrast the original art depended on. A dark corridor in the original was painted dark because the engine had no choice. The same corridor rebuilt with PBR shading and bloom can fall into a washed-out mid-grey unless the team deliberately keeps contrast at the material level.
A few practical techniques tend to recur in faithful remakes:
- Authoring materials with a slightly lower albedo in dark areas, then adding emissive or rim lighting to keep the silhouette readable.
- Keeping the original level geometry but tightening the texture work so the close-up view holds up at 4K resolution.
- Using fog, depth haze, and grain to preserve the original’s mood without making the scene literally dim.
- Reintroducing the original “hot spots” – the small bright objects in dark rooms – because they often carried the gameplay readability that the original lighting provided for free.
The goal is not to mimic the original renderer’s output. The goal is to preserve the gameplay readability the original renderer accidentally provided. That is a separate, harder problem and it is one of the reasons faithful remakes take longer than remasters.
Animation, mocap, and the question of character fidelity
Character animation in a faithful remake splits into three categories. The first is locomotion: walking, crouching, crawling, climbing. The second is interaction: aiming, grabbing, melee, item handling. The third is performance: cutscenes, codec calls, scripted beats. Each category has a different production answer.
For locomotion, mocap is the obvious choice, but the team has to decide whether to re-capture against the original control scheme or to capture against modern controls. Modern players expect smoother turns and a wider stance when aiming. If the team captures only the new controls, the original’s stiffer turn rate has to be reanimated by hand. If the team captures only the original controls, the animation feels dated on a modern pad.
For interaction, animation usually ends up split: hand-animated for tactile moments like first-person weapon handling, mocap for combat and movement. For performance, the answer depends on whether the original voice is preserved. If voice is preserved, animation is locked to the audio. If voice is re-recorded, the team has to reanimate mouth and body to the new takes, which is one of the most expensive parts of a faithful remake.
First-person aiming and the legacy of fixed cameras
The original Metal Gear Solid used a mix of top-down cameras and first-person aim. That camera split has to be re-decided for a faithful remake. The team can keep the hybrid, which preserves the original’s signature feel but reads as unusual today, or it can move to a modern third-person camera and accept that some original sightlines no longer work. A studio that chooses the hybrid has to spend extra animation and camera-design time making the transitions feel intentional rather than dated. A studio that chooses the modern camera has to redesign several stealth sections because the original corridors were tuned for fixed angles, not for player-controlled cameras.
Audio: voice, score, and spatial rebuild
Audio is the area where a faithful remake has the hardest trade-off between authenticity and quality. The original’s voice acting is part of its cultural footprint. Replacing it risks alienating long-time players. Keeping it risks a noticeable quality gap between voice and the new 3D environment. Most faithful remakes compromise by preserving the original dialogue masters and rebuilding the spatial mix around them.
That compromise has a concrete production shape. The audio team re-records foley and ambience from scratch, often at a higher sample rate, to match the new engine’s spatial audio. Voice stays in stereo or a lightly processed surround bed, with mild convolution reverb to seat it in the new room. The result is recognisably the same performance, in a room that sounds like the rebuilt room, without the actor returning to a booth.
For the score, the choice is usually either a faithful re-recording with the original composer or a remaster of the original sessions. Re-recording gives the team flexibility to adapt the mix to the new soundstage, but it changes the relationship between the music and the player’s memory. A remaster keeps the performance but ties the mix to the old production.
Accessibility and the new audio contract
A faithful remake is also a chance to add accessibility options the original could not support. Caption styling, separate volume sliders for voice, music, and effects, mono audio output, and visual indicators for off-screen audio cues are all standard additions. The accessibility work is not separate from the audio rebuild. It is part of the audio brief, and it has to be specified before the mix is locked, because adding a separate voice track at the end of a project is significantly more expensive than leaving headroom for it from the start.
Gameplay simulation: AI, damage, and balance
The original Metal Gear Solid used scripted guards with simple state machines. Rebuilding the simulation in a modern engine means deciding how much of that scripting to preserve. A pure faithful rebuild keeps the original behaviour: guards on fixed routes, triggered alerts, scripted boss patterns. A modernised rebuild uses a behaviour tree or utility AI system that can react to player actions in ways the original could not.
The risk of modernisation is that the original’s design depended on its own AI limitations. A guard who never checks a corner the player has just visited is not a bug in the original design. It is the design. A modern AI that does check that corner can break sections of the level that the original team carefully balanced. The safest production answer is a hybrid: keep the original route scripts as authored data, but add small reactive layers on top for behaviours the original never needed, like responding to thrown magazines or environmental damage.
Balance and the question of damage values
Players’ expectations of damage and healing have changed since 1998. A faithful remake that keeps the original damage values often feels punishing to a modern audience, while one that rebalances them loses the “hard but fair” reputation of the original. A common approach is to keep the original values as the default and offer an optional modern balance preset, gated behind a menu the player can revisit at any time. That keeps the faithful path intact and gives a producer a defensible answer when QA or community feedback flags the original balance as a barrier.
QA strategy for a single-player remake
A single-player faithful remake has a different QA profile from a live-service game. There is no server to monitor, no patch cadence to defend, and a finite number of levels. The QA effort concentrates on regression coverage, because the original game has thousands of recorded playthroughs online, and any deviation from the remembered path is immediately visible to the community.
QA in this context usually runs in three passes. The first pass validates the rebuild: does the level match the original layout, do the triggers fire in the right order, does the cutscene match the original beat structure. The second pass validates the new systems: does the new camera behave, does the new control scheme respond, does the new audio mix hold up across output devices. The third pass is compatibility: PC hardware variants, controller remapping, ultrawide aspect ratios, and platform certification regressions.
| Pass | Primary goal | Typical coverage | Exit signal |
|---|---|---|---|
| Layout regression | Match the original level design, trigger order, and scripted beats | Every level, every alert state, every boss phase | Signed-off layout diffs against reference |
| Systems validation | Confirm new camera, input, and audio mix work as designed | Cross-section sample of encounters, plus edge cases | Bug burn-down against the systems matrix |
| Platform compatibility | Pass certification on each target platform and PC configuration | Per-platform certification suite, plus PC hardware matrix | First-party submission approval |
Reference testing as a discipline
One QA discipline specific to faithful remakes is reference testing: a tester plays the original and the remake side by side, on the same section, and logs every behavioural difference. The output is a diff, not a bug list, and it has to be reviewed by a lead designer who can decide which differences are acceptable modernisation and which are regressions. Without that discipline, the rebuild quietly drifts away from the original level by level, and the studio discovers the drift too late to fix without re-scoping.
Porting and platform considerations
A remake of a single-player stealth game usually ships on PC and on the two current console families. Porting a modern PC build to consoles is a known engineering task, but a faithful remake carries two specific complications. The first is input: the original was designed for a D-pad and a face button layout that no longer maps cleanly to a modern pad. The second is aspect ratio: the original ran at 4:3, and the rebuild has to decide whether 4:3, 16:9, or both are first-class.
For input, the common answer is a default modern mapping with an optional classic mapping for players who want the original feel. The classic mapping is not a one-line remap. It often means separate animation state for crouch transitions, different camera behaviour in tight corridors, and a separate aim sensitivity curve. The cost is meaningful but smaller than re-tuning the whole game.
For aspect ratio, the studio has to decide whether 4:3 is a presentation choice or a quality compromise. If 4:3 is presented as a period-accurate mode with appropriate letterboxing and film grain, it can read as a deliberate artistic option. If it is shipped as a fallback that simply crops the new rendering, it reads as a regression. The first option protects the faithful brief, the second damages it.
PC-specific concerns
PC adds two specific concerns a console-only remake does not face. The first is the graphics settings matrix: the rebuild has to expose sensible defaults for ultrawide, HDR, ray tracing or its equivalent, and upscaling, without overwhelming a player who just wants the game to look like the remake trailer. The second is input device coverage: the game has to behave with a pad as the primary input and degrade gracefully when a player uses keyboard and mouse or a generic controller. Both concerns are solvable with standard PC engineering practice, but they are not free, and they have to be planned into the schedule from the first vertical slice.
Production planning and milestone structure
A faithful remake has a long preproduction and a comparatively short polish phase. The reason is that the design space is constrained from the start. The team is not green-lighting features; it is green-lighting fidelity targets. Once those targets are set, the bulk of the schedule is asset production, integration, and validation, not design exploration.
| Phase | Typical share of schedule | Primary output | Major risk to manage |
|---|---|---|---|
| Preproduction | 15-20% | Engine choice, asset converter, fidelity targets, vertical slice | Choosing an engine that cannot import legacy data |
| Production | 55-65% | Full level rebuild, art pass, animation, voice and audio, AI pass | Scope drift in cutscenes or boss fights |
| Polish and QA | 15-20% | Reference testing, balance pass, accessibility, platform certification | Discovering the rebuild has drifted from the original |
| Release and patch window | 5-10% | Day-one patch, known-issue triage, post-launch hotfixes | Day-one patch scope creep |
The exact percentages vary by studio and by the original game’s complexity, but the shape is consistent. A faithful remake underestimates preproduction at its peril, because the cost of choosing the wrong engine or the wrong asset pipeline compounds through every subsequent milestone.
Risk register for a project like Metal Gear Solid Delta
Faithful remakes carry a specific set of risks that greenfield projects do not. A risk register for a project of this shape usually includes at least the following items, each with an owner and a mitigation.
- Reference drift: the rebuild quietly diverges from the original level by level. Mitigation is the reference-testing pass described earlier, plus a design lead who signs off each level diff.
- Original actor unavailability: the original cast is not always available, and licensing can fail. Mitigation is to record preservation masters early and decide the voice policy in preproduction.
- Legacy data loss: original source files are sometimes incomplete or unreadable. Mitigation is a data archaeology pass in preproduction, with a written record of what was recovered and what was rebuilt from reference.
- Brand expectation mismatch: the marketing material sets a fidelity bar the production cannot hit. Mitigation is a marketing-production review at the end of each milestone, with shared reference targets.
- Scope creep on accessibility: the team keeps adding accessibility features past the planned scope. Mitigation is a documented accessibility matrix in preproduction, with a freeze date before certification.
None of these risks is exotic. All of them are predictable. The differentiator is whether the team names them in preproduction or discovers them in polish.
How a studio staffs a faithful remake
Staffing a remake of a single-player stealth game is not the same as staffing an open-world action game. The work is concentrated in art, animation, audio, and engineering, with a comparatively small gameplay programming team. A representative shape for a project of this scale is below.
- A small design team focused on reference accuracy, balance, and the small set of modernisations the project allows.
- A large art team, split between environment rebuild, character and weapon art, and materials.
- An animation team of meaningful size, because the original had a distinctive camera and the rebuild has to re-author many of the original’s pose-driven moments.
- An audio team that includes a dialogue supervisor, a music supervisor, and a sound design lead for the new spatial mix.
- A central engineering team owning the engine, the asset converter, the camera system, and the platform ports.
- A QA team large enough to run the three-pass structure above, with at least one dedicated reference tester per level group.
- Production and production engineering to track milestone burn, dependency chains between art and engineering, and certification deliverables.
The team’s centre of gravity sits in art, animation, and audio. That is unusual for a modern action title, where engineering often dominates, and it is one of the structural reasons faithful remakes are scheduled differently from greenfield projects.
Modernisation trade-offs: where the line moves
Every faithful remake draws a line between what stays and what changes. The line is not a single decision. It is a series of small decisions, each made against a reference target. A few recurring examples illustrate how the line moves in practice.
| System | Original behaviour | Common modernisation | Trade-off to manage |
|---|---|---|---|
| Camera | Fixed angles, hard cuts | Hybrid: fixed camera logic with smoothed blends | Smoothed blends can hide original sightline puzzles |
| Controls | D-pad, two action buttons, no strafe | Modern pad mapping with optional classic scheme | Classic scheme has to be a first-class system, not a hack |
| AI | Scripted routes, simple alert states | Route scripts with a thin reactive layer | Reactive layer can break original stealth balance |
| Damage and healing | Punishing by modern standards | Optional modern balance preset | Default must still feel faithful on first playthrough |
| Codec and dialogue | Triggered audio calls, no captions | Captioned calls, optional subtitles for ambient dialogue | Caption timing has to match the original pacing |
| Save system | Fixed save points | Save points plus an optional modern save-anywhere | Save-anywhere can trivialise sections tuned around resource scarcity |
None of these trade-offs is settled by a single design memo. Each is revisited during polish, and each is a candidate for post-launch community feedback. A producer who has named the trade-off in preproduction can defend the choice. A producer who has not, has to defend the choice while the studio is still deciding it.
The role of community and historical reference
Faithful remakes ship into a community that has memorised the original. Every speedrun route, every skip, every sequence break is documented. The studio’s QA team has to internalise that documentation, because the community will find any unintentional change within days of release. That is a productive constraint. It raises the QA bar, and it forces the studio to keep reference material close at hand throughout production.
For a project like Metal Gear Solid Delta, the reference material is unusually deep. The original game is one of the most analysed stealth-action titles of its era, with a long history of re-releases, ports, and community documentation. That history is useful to the production team, but it is also a warning: the gap between the original’s reputation and the rebuild’s execution is a public, well-lit surface. Every visible regression will be compared to a remembered original.
The reference for the original game’s critical reception and review history, including the kinds of design notes the press and players recorded at launch, is preserved in places like the long-running record of contemporary Metal Gear Solid critic reviews and in the broader public history of the series, which is documented in detail on the Metal Gear Solid video game Wikipedia entry, which provides useful background for this point. A studio that builds its reference testing pass on top of those public sources starts ahead of a studio that does not.
Post-launch: patches, ports, and the long tail
A faithful remake does not end at certification. The first thirty days after release typically bring a day-one patch, a small hotfix, and a longer patch that addresses balance feedback. The post-launch plan should be part of the production plan, not a reaction to launch.
A workable post-launch shape is a small live team that can address regressions, a balance pass planned for the first major patch, and a port plan if the studio holds back a platform release for stability reasons. The studio should expect a community-led discovery process in the first two weeks, where previously unknown original-game quirks resurface as remake bugs. A documented triage path keeps that discovery process from derailing the patch plan.
Frequently asked questions
What is Metal Gear Solid Delta in production terms?
It is a faithful remake of the 1998 stealth-action game Metal Gear Solid, rebuilding art, animation, audio, and gameplay simulation in a modern real-time engine while preserving the original level design, camera beats, and narrative structure. The studio commits to a specific fidelity target per level and validates the rebuild against reference material from the original release.
How is a faithful remake different from a remaster or a reimagining?
A remaster keeps the original code and assets and improves presentation. A reimagining keeps the brand and themes but rebuilds mechanics from scratch. A faithful remake keeps the original design and rebuilds the production underneath it. Each option has a different scope, a different schedule, and a different risk profile.
What is the largest production cost in a faithful remake?
Asset conversion and art rebuild, taken together, usually dominate the budget. The original’s data has to be converted to a modern pipeline, the art has to be reauthored at modern resolution, and the lighting has to be re-baked for a different rendering model. Without a working asset converter in preproduction, the art team ends up recreating geometry by hand, which inflates the schedule.
Does a faithful remake keep the original voice acting?
Most do, because the original performance is part of the title’s cultural footprint. When voice is preserved, the audio team rebuilds the spatial mix around the original masters, with re-recorded foley and ambience to seat the dialogue in the new environment. When voice is re-recorded, lip-sync and body animation have to be reauthored against the new takes.
Why do faithful remakes take longer than remasters?
Because the rebuild touches the simulation as well as the presentation. A remaster touches textures, resolution, and frame rate. A faithful remake touches the camera system, the AI scripts, the animation state machine, the audio mix, and the level triggers. Each of those systems is a separate engineering and validation surface, and each one needs its own QA pass.
How does a studio decide which original systems to modernise?
By setting a fidelity target per level rather than per project, and by keeping a reference test that compares the rebuild to the original section by section. Modernisations are allowed when they preserve the original’s gameplay readability, and they are rejected when they would change a section the original team carefully balanced.
What is the biggest risk in a project like Metal Gear Solid Delta?
Reference drift. The rebuild can quietly diverge from the original level by level if there is no formal diff between the two versions. A reference-testing pass, with a designer signing off each level’s diff against the original, is the most reliable way to keep the rebuild faithful.
Do faithful remakes need a separate accessibility pass?
Yes. Accessibility features like caption styling, separate audio sliders, and visual indicators for off-screen cues are easier to design in from the start than to retrofit at the end. The accessibility matrix should be specified in preproduction and frozen before certification, to keep the feature from creeping the schedule.
Leave a Reply