Was Metro Exodus downgraded? E3 2017 trailer vs Final/Release version Comparison Videos [UPDATE: PC vs E3 2017 trailer video added]

Metro Exodus has just been released and a lot of gamers have been wondering whether the game has been downgraded from its E3 2017 reveal trailer. And while we haven’t reached the outdoor level that was showcased in the E3 2017 reveal trailer, we’ve decided to share below two videos that compare the final version of the game with its E3 2017 build.

The first video compares the E3 2017 trailer with the Xbox One X version and we can clearly see that the Xbox One X version does not look as good as the E3 2017 build. However, I should really note that the Xbox One X looks way worse than the PC version. For example, the PC version supports Hairworks (the fur on the monsters is present on the PC version), Advanced PhysX and real-time ray tracing. Still, and since Metro Exodus’ trailer was showcased during Microsoft’s Xbox conference, we can all agree that the Xbox One X version is inferior to that trailer.

The second video is unfortunately in Russian so I don’t know what the YouTuber is actually saying. Again, this comparison is between the console version of Metro Exodus and the E3 2017 trailer. However, there is a significant difference between the time of day that was used in this comparison, thus making this comparison a bit unreliable. Still, this video shows some level design changes and we can all agree that some scripted events (like the annoying head bobbing and the scripted sliding mechanic) are absent from the game though I’m not sure whether someone would call the absence of these scripted events a downgrade-ation.

As said, there are some level differences between these two versions. The E3 2017 trailer basically uses two completely different areas from the final game. The underground area is from the beginning of the game and the outdoor is from a later area. The good news here, and contrary to other games, is that these scenes are indeed featured in the game. Moreover, the bear encounter is completely different in the final version.

It would have been better if someone could compare the final PC version – with all its bells and whistles enabled – against the E3 2017 trailer, with the exact or similar Time of Day. Unfortunately, there isn’t such a comparison, however, we’ve decided to share these comparisons as you will easily notice some differences between these builds.

So what do you think? Was Metro Exodus downgraded? Are these level design/gameplay changes as significant as those for Watch_Dogs or Crackdown 3? And most importantly, was 4A Games able to deliver something that was close to what it showcased or did the team deceive its fans with that trailer?

Metro Exodus | E3 2017 Reveal Gameplay VS FINAL VERSION | Comparison

НА E3 НАМ СНОВА ЛГАЛИ? METRO: EXODUS

UPDATE:

YouTube’s ‘Nick930’ has shared a direct comparison between the PC and the E3 2017 trailer!

Metro Exodus E3 vs Retail | Direct Comparison

104 thoughts on “Was Metro Exodus downgraded? E3 2017 trailer vs Final/Release version Comparison Videos [UPDATE: PC vs E3 2017 trailer video added]”

  1. If you believe that this game was running on an Xbox One at E3 2017, then you are a fully blown idiot. Not even Microsoft runs their Debut gameplay footage on their own console. Sony also does the same sh*t, especially considering the time period in the dev cycle. sorry folks, they all do this sh*t. They typically run these debut footage from a PC.

  2. To add to that list Digital Foundry showed tessellation to be missing from all console versions of the game. Also, and somewhat amusingly, the crappy old base Xbone has a more stable 30fps frame-rate than Xbone X or either Sony console.

  3. we still have to be getting these downgrades for the sake of parity even though our systems run these games at +100fps.

  4. crackdowngrade 3 is much worse cuz of the scaled back destruction/physics, especially with all that development time and cloud power shiz

  5. Watching it in 4k, I thought there were a number of things that looked better on the X than the E3 trailer. However I don’t remember trailer claiming to be in game X footage or in engine footage since the draw distance is noticeably pared back and the large amount of changes

  6. Hey John you should make the same article but about Just Cause 4 this time. Even the steam page is full of videos that don’t represent the actual graphics.

  7. To add to that list Digital Foundry showed tessellation to be missing from all console versions of the game. Also, and somewhat amusingly, the crappy old base Xbone has a more stable 30fps frame-rate than Xbone X or either Sony console.

    1. I wouldn’t say amusingly.

      Metro games have tremendously great optimization and even amazing optimization CPU side, leveraging as many CPU cores/threads it can find. The fact X1 has slightly faster CPU than PS4 and the fact the devs said DX12 did wonders for the X1 helped as well.

      Plus, all platforms use dynamic resolution. Used properly, if the PS4 is locked to 1080p, the X1s when properly leveraged has overhead when rendered in 900p, that’s why the X1S performs so well.

      1. That’s not true because it doesn’t have the overhead, that’s the reason why it’s 900p, we’ve seen XBox do 900p and still drop below 30fps or perform just as bad, it’s dynamic 900p which means there is no overhead, otherwise there wouldn’t be a dynamic resolution in place if there was overhead, it would be fixed 900p resolution always.

  8. If you believe that this game was running on an Xbox One at E3 2017, then you are a fully blown idiot. Not even Microsoft runs their Debut gameplay footage on their own console. Sony also does the same sh*t, especially considering the time period in the dev cycle. sorry folks, they all do this sh*t. They typically run these debut footage from a PC.

    1. Sony runs their exclusives on their own consoles at least. There hasn’t been any downgrade so far. Yes, I am a self confessed sony fanboi and Destiny lover.

      1. “Hasn’t been any downgrade so far”, Sony is the first publisher to downgrade their games years before Ubisoft (Killzone, Motorstorm…) and a some of their games on this gen were downgraded too (Uncharted 4, Spiderman) So you must correct yourself : Confessed BLIND Sony Fanboy and Destiny lover… Destiny lover, how can you go that deep…

          1. Go watch Digital Foundry’s analysis videos on both Spider-Man and Uncharted 4 and come back later.

          2. I still think the lighting looks much better in the E3 version because it always is, they light everything so assets look good in all conditions but in the actual final game they can’t and it’s passed off as an “artistic” change or “different time of day”.

          3. Oh that good old “artistic change”. Same thing was used in The Witcher 3, Watch Dogs and countless others.

          4. Yeah, a few hundred lights less in the scene is “artist change”. they change the time of day or season simply because they can’t nail that E3 look as it’s too complex.

          5. Same for Spiderman, both games are beautiful i know, i just shared those pics to that guys defending Sony and saying there are no downgrades on their side

      2. I hate to disappoint you self confessed fanboy, no they DO NOT always use their own consoles. I work in marketing and have worked many Trade Shows, Expos & Conferences for Microsoft. I’ve personally seen it myself that they run from a PC, ESPECIALLY SONY. Sony is most famous for this actually. Back in 2012ish during the PS4 itself debut they had Assassin’s Creed Black Flag running on a PC on A GTX 670. That’s right a GTX 670, I saw it myself. This is a well commonly known practice in the industry. Because the last time I check consoles can’t develop games. That’s why I’m equally pissed when games are unoptimized on PC because I know and seen what they have consoles running on. A freakin PC…

        And furthermore, Buddy please don’t confuse the conversation/narrative. I never said anything about Sony downgrading anything. Let’s not start with having people respond back and forth with strayed narrative. Stay the topic at hand.

          1. The smell of bullshit coming off you is strong. Unlike you I actually work in the industry. I have done for years.

            Now you claims of “well known in the industry” is very interesting as demonstration certified builds tend to be the norm for trade shows. These by definition are certified by the platform holder as publicly viable display builds for that specific platform. Which is actually the norm. These are builds that run on native hardware which are at least repräsentative of the SKU they are officially labelled as, if not the SKU itself.

            Developer consoles are a thing, in fact they are a vital part of the process. While true you do not develop directly on these devices they are used for vigorously for testing and in fact must be used prior to submitting for any certification.

            In order for a game to be badged as supporting a specific SKU it must be certified for it. Which means that the game or at least display build had been signed off by the platform holder as working on their platform.

            There have been cases where a build was running on a different platform (including PC), but for the most part they tend to run on the devices being advertised.

            So please wow us some more with your fictions, lies, and bizarre need to represent yourself as some sort of informed insider. You’re obviously not. Weird faux flex, but maybe dont try so hard.

          2. You mean the lies you smear all of the place in here? Not what I would call “in the industry”. That’s just talking sh*t. The fact you’ve done that for 15 years really isn’t all that relevant.

          3. You can’t verify your own claims, I mean your profile is private for one start. “Game Dev and Gaming commentator” means nothing.

          4. We’ve actually seen photos of what he is claiming, like the pictures of PC’s covered up at the bottom with the Xbox consoles on the front. Also, The Metro developer reluctantly said the E3 demo of Metro Exodus was running on a PC.

          5. I am not saying it does not happen. Only that it is not the norm like the liar above claims based on limited information and internet hyperbole.

            I dislike the term downgrade because it over simplifies a complicated process. There are literally uncountable reasons a released title uses different rendering and/or post-processing solutions, meshes, textures, or any number of in-game assets. Generally it’s down to optimisations across an entire product rather then a small slice on display at a trade show, or demo reel. Devs always want to produce the best product possible, but sometimes when you put the whole game together the solution you used in pre-alpha/beta builds does not work, or work consistently once all the other parts are added to the complex machine a game is.

            It’s unfortunate, and I am not claiming that some shady marketing does not happen at time, but for the most part what is shown pre-release is the best example of what the game is planned to look like at release.

            It is interesting that “downgrades” are jumped on, but games that perform or look better than pre-release footage never get any attention. Something that is actually more common. Hell, we even had outrage from uninformed gamers about Spider-Man last year only for them to be told that not only was there no downgrade, but what they’re seeing was a visually superior game.

            I would prefer people to assess the game on it’s merits at release. Hell Metro Exodus is considered one of the best looking games on the market and we’re still seeing garbage like this on the internet. The Witcher 3 was also raked over the coals. A game that was stunning for the time and a lot of people did not appreciate it for what it was, but instead ranted about relatively minor changes made in order to ensure performance was consistent across the whole open-world. CDPR didn’t downgrade, they made optimisations to present the best looking and performing game they could.

          6. Well ,the E3 demos are not real, they’re intentionally created for marketing the game and to say the game will look like that. Few of us know it’s just a render target and they try to get to that target with the actual game. It’s really an ethical one, E3 demos sell preorders, they’re literally selling something that isn’t real on the screen.

            It seems to me these E3 demos or vertical slicers somewhat blur the lines, I understand they want to show us the game but can’t in it’s real state because it’s in no way in any good state to show but Star Citizen’s Chris Roberts does it, they even hit bugs showcasing the game. At least it’s not faking the game but these other developers and their publishers want maximum perfection and marketing, which can seems shady and have proven that they can’t hit that E3 render target time and time again.

          7. 670 was much faster GPU compared to PS4 1.8 tflops.

            GTX 660 (1.9 TFLOPS)
            GTX 670 (2.6 TFLOPS)
            GTX 680 (3 TFLOPS)

            I had 680 GTX and I could run PS4 games at 2x higher framerate and better details at the same time (gta 5 for example). The only issue was vram, my 680 GTX was vram limited in many multiplatform games later on. I would go even as far to say also 660 GTX was faster than GPU in PS4 because Nv TFLOPS was way more efficient back then.

          8. PS4 GPU is somewhere between 7850 and 7870 and both are no match for 680 GTX and even 670 GTX. Of course I can agree on consoles they can always use tweaks (dynamic resolution, low shadows details etc.) and optimize games better, but from the pure technological point of view showing PS4 games running on PC allowed them to show performance in much better light then it was on PS4 GPU. When I had 680 literally every game run much better compared to PS4, and the only problem was vram. Can PS4 run gta5 in 60fps? 680 GTX does that with even higher settings so clearly PS4 GPU was far behind. If someone bought GTX 680 model with 4GB then card like that should still run all multiplatform games with better results than consoles even today.

            When it comes to metro exodus, what’s interesting Nv has gimped kepler performance in that game. You can see hardwareunboxed benchmarks on their YT channel and even 7970 crush 780ti, and that’s something totally impossible.

          9. GTX 670 has 2.7 TFLOPS, PS4 GPU 1.8 TFLOPS, and let’s not forget AMD TFLOPS are far less efficient, so there’s nothing to suggest PS4 GPU would ever match that card.

            780ti performance was always comparable to 970 GTX (Maxwell GPU’s), but for some strange reason in new metro 780ti is totally crushed (and like I have already said even 7970 is faster than 780ti in this game). It looks to me like Nv simply no longer optimize drivers for their old cards, while AMD still support them, but it’s not like 7970 had more future proof features compared to kepler.

          10. Maxwell has a massive improvement in compute performance compared to Kepler but people put it down to driver gimping because they couldn’t understand the technical reasons, just, mah GPU was gimped by NVIDIA.

          11. And what you said about AMD Teraflops being less efficient? Where did you hear this?

            Nvidia GTX 480 (1.3 TFLOPS card) was comparable to AMD HD 5870 (2.7 TFLOPS). Nvidia TFLOPS was clearly much supperior years ago, and even today the gap is still there (Vega 64 should crush 1080ti if you look at TFLOPS numbers, but that’s not the case)

            Teraflops mean nothing. I mean. They have a definition, but it’s mostly marketing bullshit. They have no impact on real world performance

            If you compare totally different architectures, then maybe, but on the same architecture TFLOPS is a good estimation. PS4P with it’s 4.8 TFLOPS GPU is more than twice as fast as PS4 1.8TFLOPS in games. On PS4 developers can not only increase resolution by 2x, but also improve framerate a little at the same time exactly as TFLOPS difference suggest.

          12. You are wasting your precious time with this Paul86 guy. He is completely clueless, delusional, and just pretends to have some tech knowledge.

            Also, you can’t debate with this guy, because he is stubborn as hell, and won’t admit his own mistakes, when others try to correct him..

            A complete dumbo.

          13. If developers want to increase FPS from lets say 30fps to 60fps on PS4P games compared to PS4, then because of CPU problems there will be dips below 60fps (it will be not perfectly locked 60fps, more like 45-60fps) and TFLOPS will not scale perfectly in this case. But in normal situation when developer want to only increase GPU workload by increasing resolution 2x times over standard PS4 model, then CPU is no longer limiting anything and TFLOPS increase scales perfectly. That’s a fact because you can see 2x resolution increase on PS4P compared to PS4 in the vast majority of games. According to you TFLOPS have no impact on real world performance, but PS4P vs PS4 example literally destroyed your argument because it shows perfect TFLOPS scaling. It’s why you started acting rude towards me because you failed to prove your argument?

            As I have said, it’s hard to compare AMD card to Nv TFLOPS because of totally different architectures, but what is well known fact (since ancient times 😉 AMD TFLOPS are less efficient (in real games) compared to Nvidia TFLOPS. I have provided for you a very good example with radeon 5870 vs Nv GTX 480, because Nv GPU could match and even surpass AMD card with twice as much TFLOPS (so obviously Nvidia TFLOPS were much more efficient, exactly as I have said). So although you cant compare AMD and Nv architectures directly it’s still a sure thing Nv TFLOPS are faster, and that’s important point to consider in our PS4 (AMD GPU) vs Kepler discussion, because then we are looking not only at massive TFLOPS diffrence (1.8 vs 2.7), but also TFLOPS efficiency difference.

            My old 680GTX (3 TFLOPS) could EASILY run all multiplatform games (made with PS4 in mind) with 2x much higher resoulution and higher graphics settings on top of that and it was not just because of better CPU as you suggest because I also remember how bad 7850 results looked like next to my old 680GTX so even paired with good CPU 7850 never was a match for kepler top dog cards. PS4 GPU was simply much weaker compared to 670 and that’s exactly why sony have used GTX 670 PC to showcase PS4 games :), they sure knew framerate would be worse on real PS4 so they have used much more powerful PC because they didnt wanted to dissapoint people. IMO you were totally wrong telling people here 670 GTX was comparable performance wise to PS4 GPU and I have provided my detailed perspective on that. You dont need to agree with me if you dont like facts, but at least please act like a normal guy and dont start acting rude towards me, because then I will start talking with you accordingly, and that fool “Jackyies” can attest I’m not a nice guy if someone want to fight with me.

          14. The PS4 Pro also has games at Native 4K. That’s an increase of 4x in resolution over 1080p which is the highest resolution games on the standard PS4 run at.

            Simple games yes, but that’s not really something that should count. The rule of the thumb is, the vast majority of PS4P games renders only 2x many pixels natively (1440p most of the time) and developers go for 2x resolution increase for a reason, there’s simply no more performance left to render games in higher resolution. They can just upscale internal resolution to 4K, but that’s no longer 4K native and should be not considered as such.

            Oh but also. The GTX 980 is 5 Teraflops. And the GTX 1080 is 8.28 Teraflops.

            That’s already totally different architecture unlike PS4 vs PS4P GPU. Pascal was massive architecture improvement over maxwell, so, TFLOPS numbers no longer reflects real performance difference in your 980 and 1080 example. You should only compare pascal architecture cards, to pascal cards if you wanted to post a good response to my argument with PS4P vs PS4.

            So let’s sum it up, first you wrote a delusional comment and compared GTX 670 performance to PS4 based on your imagination alone, and when I have provided my detailed perspective on that based on TFLOPS facts (and my experience with PS4/P and 680GTX as well) and then sh*t happened :D. Besides Metro Exodus (where it’s evident Nv no longer optimze their drivers for kepler, because even GTX 970 is almost twice as fast compared to 780ti) you cant find more example of games where 670 and 680 provides worse experience than PS4 game with the same settings. I have played many games on GTX 680 before I bought 1080ti 1.5 year ago and I could run ALL PS4 games until then with much more headroom than PS4. If you would be correct as you think you are, then my old GTX 680 would struggle in literally all PS4 multiplatform games, but it was just the opposite.

            Because you continued to post rude comments towards me even when I have asked you politely to stop being an a*sh^le, so now I will be no longer nice to you. Man fool like you can upvote Jackyies comments and even lick his a*s because you both are full of sh^t, but because you stink with sh^t now I’m not even going to talk with you more :D.

  9. we still have to be getting these downgrades for the sake of parity even though our systems run these games at +100fps.

  10. Because this isn’t using like-for-like devices, it is a pointless comparison. The article even acknowledges the considerable reduction in quality for the console versions, so why post this at all?

  11. Comparing the two is useless, since the game goes thru a lot of changes ant optimizations. Thats what development is. PC is always used to showcase games since you can bait people.

    1. It’s probably the most beautiful game we have right now. Gives everything a solid run for their money. RDR2? It doesn’t look as good as this game does in my opinion.

      4A did an amazing job technically and artistically.

      1. You work in games marketing, and you’re using the term “downgrade”? The smell of bullshit is very strong. In the future claiming to work in a field and then using a term specifically NEVER used by anyone in that field because they know better is a dead giveaway.

        1. Hey Idiot! Am I at work right now? If this my place of business? Dude STFU, you are trying so hard to make a case and trying to tell me that my experience is invalid as if you are the have all/end all and you oversee all things. Please Go and Take the Toaster Shower Test. You’re garbage, enough of you already.

          1. Angry child has a tantrum at being called out for telling lies. No one is surprised. Telling someone to commit suicide really does demonstrate just what you are now doesn’t it. This is why no one should take you seriously.

      1. You clearly don’t know the difference between a vertical slice from E3 and something real which users can play.

        1. alpha build is not final game

          there is always demos and alphas and betas that get changed by the end of development

          1. For the third time, Which is not what we’re talking about. Alphas and beta’s released to the pubic are actual games which don’t get downgrading during public testing.

          2. it happened before

            as the scope of the game gets bigger it might be easier for devs to lower it for their convenience or streaming issues that might come up further down the road so it is still not the final product

            let’s end this discussion

    1. The new video added that show the difference between E3 and PC retail is showing that this is the best version so far!

    2. Beyond the fact that this is, without a doubt, the worst type of comparison you could have chosen, explain what the ‘downgrades’ are…

      1. The church in the background is clearly simplified and a much lower mesh quality, the lighting in the final version is inferior and flat looking, the sky in the E3 is superior in it actually has a proper storm cloud formation. The materials like on the car are rusted and aged better with the lighting.

        I find it hard to believe they came to the conclusion that the autumn look looks worse than the more bushy, one tone lighting look of the final version.

        1. Your name it compatible to your writting, i don’t know how can anyone even say that retail was Upgraded over E3, to me is so stark the difference, that makes me think how people are easily influenced.

    3. Watch the latest comparison video in this article – the time of day shifts so abruptly and becomes this one tone mess as if someone hit the off button. And the lighting is definitely less advanced in the release version.

      No matter really as it’s still a boring console game with dumb AI, lifeless static levels, lame gunplay and movement and a whole lot of scrounging two resources from the same looking hidden caches.

  12. It is but this bait and switch method of false advertising is how they do it these days. They show how it could look in all its glory on PC and then succumb to the realization that they have to make it work on consoles. The Witcher 3 suffered the same fate, im sure there are many other examples too.

    1. Basically E3 vids do not specifically represent the final products Those showcase engines that look similar but at the same look drastically different are mostly made worth to of the line,, state of the art gpus.

      At the end the final product had to be chipped down a bit to make it viable to a big number of pc s, not just 100k computers
      I still remember when W3 released, how harsh it was on my gtx980. Both where contemporary products. There was no way I could run West the same fidelity it was displayed in E3.!

    2. “These days”? For as long as I can remember this isnt new. Dazzle the imbeciles with glitz because that hype sticks all the way up to the sale.

      1. You’ve got a point there actually, I remember the back of Amstrad CPC game boxes showing Amiga screenshots.

        It seems so blatant these days though when you see someone stood on a stage pretending to play a game with a controller and acting out a script with their coworkers only to take out entire features and lower the graphics quality and then say they just changed the TOD.

        1. Yeah that the excuse, artistic change or different time of day, I mean it can’t be a “artistic change” when the meshes have clearly been simplified and the lighting has looks flat and washed out because either they couldn’t put that amount of lights in the scene like the perfect E3 showcase or they never could make it look like that in the first place to sell preorders.

  13. Why the hell you show us an XBOX footage? The demo was played on a PC(they said it and everyone was laughing at them),
    search before you post something…. Of course its a downgrade since
    your argument is invalid.Microsoft would be stupid enough to show us console gameplay from a game like that…

  14. In e3 2017, we can clearly see hairwork on monster in tunnel. This is more than enough to say that e3 2017 was on PC. Secondly. Ray trace was active. Because, the in the house which Artyom entered was superior AO, same as PC ray trace AO. And another thing is, in E3 2017, flashlight used ray trace GI too. It is more visible, when Artyom opens red door in tunnel, we can see red light bounce from door in everywhere. So, what I think is, Game downgraded like this: Flash light doesnt use GI, assets are less dense. Iron sight lights, the lights on Artyom watch and so on, were more bloomy. The artistic changes is, the position of the gun and Artyoms hand are less cinematic in final game but now it is easy to see what is in front of you.

    But when I compare other places of that map and time of day, I cant see any major downgrade. Almost exact same thing. But what I want to mention that, the Aurora trailer which released after few mounth, almost one to one same with final game. Even all the destroyed cars at same exact place in Moscow map. I compared all of them one by one 😀 So they didnt changed any asset during 2 years and this is already the huge thing for me. Game is amazing. Thank you 4A for that.

    So, e3 2017 captured from xbox one x was a lie. Or, they just made that section of game separately, and put all that crazy GI things in it using all console power. But in reality, it is almost same tech as final PC versionwith minor downgrade. But after e3, there is almost not any downgrade at all.

  15. Considering how absolutely amazing this game looks with RTX (ray-traced global illumination) nothing is downgraded, it was upgraded. This is one of the best single player FPS since Bioshock.

  16. The PC retail version (updated article) is beautiful and sharp!!! Can’t wait to try this on Steam (if they come to their senses..)

  17. it’s still a $hitty console port with consolitis game design

    seriously, there is barely any “Metro” in this game

Leave a Reply

Your email address will not be published. Required fields are marked *