Welcome Stranger to OCC!Login | Register

Middle-earth: Shadow of War Full Experience Review



The one-word descriptor I would use for the graphics in Shadow of War would be good, but I do not really want to say anything better than that. The game definitely does look good, but there are imperfections you can see, and while some are definitely easy to disregard, others can be annoying.

Among the imperfections is the occasional softness of textures on orc captains. For the most part you can only notice this during the brief moments when the camera flies in to let them introduce themselves, and then later when it flies in to show off their death, which is one reason it is easy to disregard. Another reason, for me at least, is that Shadow of War is an open world game and texture sharpness is something I consider a fair sacrifice for the genre. To have as sharp of textures as are found in more linear games would make an open world game a massive install, or require the same textures be repeatedly used. Of course Shadow of War is taking up 105 GB on my 120 GB SSD (111 GB usable), so it has kind of already lost when it comes to saving space. (That 105 GB value is with all of the DLC.) It could be worse though as the Ultra resolution texture pack would be another 15 GB if I checked off that DLC in Steam, and there is also a 4K cinematic DLC. (As an aside, I like seeing things like this as separate DLC, specifically so you can decide to not download them if you wish.) I am not using the Ultra resolution textures, in part to save on space but also because my monitor is 2048x1152, so approximately 1920x1080 and those ultra textures are meant more for 2560x1440 or higher.

Most of the other imperfections coming to mind are somewhat standard, like fire and fluids not looking very realistic, though they hardly look bad, and the seams for gore also do not look very good. Addressing the fire and fluids first, while there definitely can be a lot of fire, I honestly never paid too much attention to it. Maybe I am just losing my eye for this stuff, or it is because typically when there is fire around it is because I detonated something during a fight, so my attention is on my enemies and not the graphics. There is some water in the game and it is unimpressive and almost completely lacking interaction when I ran through some of it. The primary fluid you encounter is definitely orc blood, which appears when a limb or head is removed from a body. It looks like sprites just being thrown about, which can be fine, but since it is so often shown during a slow-motion moment of a captain's death, you get time to examine its appearance. This is similar to the seams for the gore, which to be fair is not something easily made realistic in video games and is something open world games should be given some latitude on. It is just that after seeing an orc or troll beheaded for the nth time, you wish there were more variety in the appearance.



One imperfection that I do not find so easy to dismiss is the relatively short draw distance of some effects. One of the abilities you can unlock is to remotely poison barrels of grog, so when an orc or troll drinks from it, they will suffer damage, at least, but potentially more than that will happen. Anyway, the effect to indicate a barrel is poisoned is a green mist over it and around it, but this mist is not rendered if you are too far away. The issue is you can poison the barrels at a greater distance than this mist is drawn, so you actually cannot tell if you did successfully poison it or not. (It is also an issue that the prompt to poison a barrel remains present even on those you already have poisoned.) The distance for shadows being drawn and some objects like foliage is also rather short, which is quite unfortunate given the settings I am running at.

Another technical issue is the orcs popping in. After completing a mission that took you into a special instance of the world, when you return the world around you needs to be repopulated with enemies. This can happen shortly after you arrive, so you will see the enemies just appear, I think the algorithm for placing them bugged a little because I count at least 19 enemies all in a bunch, and they spawned in like that a moment before I took the screenshot. It was a little convenient, because they were next to a barrel of grog, which will explode when shot, either creating a fire explosion or poison explosion, depending on if you have tampered with it. I still feel it should not happen, but at least the enemies were easy to deal with.



I wish I could comment on the lighting and other aspects of the graphics, but honestly I failed to notice a lot of it. As I have stated many times before, that can be a good thing because when done well, some graphics should not be noticed because they are as you expect. As I look through the screenshots, everything else, like the lighting, does look good and I lack any cause for complaint or criticism.

One thing I do want to mention before moving on is that the pre-rendered cutscenes, which are how a lot of the story is told, have a fixed frame rate of 60 FPS and show Talion in his base gear. As you play you find and will equip new gear that have different appearances, but this will never be reflected in these videos. I understand the necessity for pre-rendered scenes like this, but it is still something broadly unfortunate to encounter.



Time to talk performance, so here are my specs:

  • Processor: AMD Ryzen Threadripper 1950X @ 4.0 GHz or 3.7 GHz
  • Cooling: Enermax Liqtech TR4 240 AIO
  • Motherboard: ASUS Zenith Extreme
  • GPU: AMD RX Vega 64 (Stock @ 0.965 V +50% Power Limit)
  • Memory: G.Skill TridentZ 4x8 GB (32 GB) @ 3200 MHz 14-14-14-28
  • PSU: OCZ Fata1ty 750 W
  • OS: Windows 10 Pro 64-bit

Before getting to the numbers, I want to explain something you may or may not have noticed in my listed specs list: I have my CPU at either 4.0 GHz or 3.7 GHz. When reviewing a game I try to run at 4.0 GHz, unless I am recording video as there seems to be some odd software quirk with my system, causing it to hard-lock when running x264 at 4.0 GHz. Twice in one day my system hard-locked while I was playing, and after the second time I decided to use the 3.7 GHz profile in Ryzen Master, which then did not suffer any issues. Unfortunately, I cannot offer any explanation for what caused this crash, only that I had spent many hours playing at 4.0 GHz without issue, but potentially by coincidence my system crashed twice at that speed, and then did not crash again when at 3.7 GHz.



For what it is worth, I have not noticed in this or any other game a significant difference in gameplay experience between 3.7 GHz and 4.0 GHz, so every comment I am about to make applies to both evenly. (If you are wondering why 3.7 GHz, because it works and is what the second custom profile in Ryzen Master is. I have been meaning to try pushing it up to see where the limit of x264 stability is, but have not done so yet.) Unfortunately because I was so focused on getting through the gameplay, I failed to note the time this happened to then differentiate between the OCAT data at 4.0 GHz and 3.7 GHz. However, I am fairly certain all of the recordings were from when I was still running at 4.0 GHz, and, as I said, I have never noticed a significant difference in gameplay experience resulting from a clock speed of 3.7 GHz and 4.0 GHz.

(For anyone curious, I am quite confident the x264 issue is a software issue more than hardware because my previous build with an AMD A10-5800K clocked at 4.4 GHz also suffered hard-locks when encoding video and I copied things over to this Threadripper system. Whatever software issue caused the crashes before I assume was carried over.)



With that covered, I have been running Shadow of War at its highest settings, save two. One is I am using the High textures instead of the Ultra textures and the other is I have not changed the resolution scaling from 1.00x. This setting would allow you to have the game render at a higher resolution and then be scaled down to your monitor's resolution, potentially improving image quality, but I did not mess with it. There definitely was enough performance overhead I could have bumped this up, but I kind of forgot to try… How about the data!

I made five OCAT recordings, and as usual the recordings lasted for 300 seconds of just normal gameplay. The third recording did capture when a pre-rendered cutscene was playing, so a little more than two minutes of it is at a flat 16.7 ms, which will naturally skew the results toward that 60 FPS frame time. In any case, here are the averages:

Average Recording 1 Recording 2 Recording 3 Recording 4 Recording 5
Frame Time (ms) 10.47 8.90 13.70 9.10 10.93
Frame Rate (FPS) 95.49 112.35 73.01 109.80 91.48

Here are the percentiles now, so you can see the distribution of performance.

Percentiles (ms) Recording 1 Recording 2 Recording 3 Recording 4 Recording 5
0.1 6.07 3.83 7.84 5.93 6.59
1 6.30 5.46 8.25 6.29 7.11
99 13.84 16.36 20.45 12.22 17.54
99.9 16.50 17.84 21.77 14.98 19.59

And again but in the easier-to-read FPS unit.

Percentiles (FPS) Recording 1 Recording 2 Recording 3 Recording 4 Recording 5
0.1 164.85 261.24 127.52 168.73 151.74
1 158.81 183.02 121.21 158.88 140.62
99 72.27 61.11 48.89 81.86 57.01
99.9 60.62 56.05 45.94 66.77 51.05

As you can see the two minutes of fixed 60 FPS significantly impacted Recording 3 but Recording 5 is also a special recording. This recording covers a fortress assault, which involves a lot of enemies being in the area, which is a lot to process. I specifically wanted a recording from this event, and as you can see there was an impact with performance dropping below 60 FPS. Next up are the reverse-percentiles, so we can see just how much of the data did fall below 60 FPS/16.667 ms.

Percentile of FPS (%) Recording 1 Recording 2 Recording 3 Recording 4 Recording 5
60 0.093 0.721 25.971 0.030 1.818
50 0.031 0.033 1.433 0.012 0.087
30 0.007 0.006 0.005 0 0.011

Remember, almost half of Recording 3 is a fixed 60 FPS video. Looking at Recording 5, we see that not even two percent of the time was the frame rate at or below 60 FPS, while the other three dropped that low not even one percent of the time. What this tells us is the game performed very well and consistently well for my hardware at the highest settings. If your monitor has a higher resolution or frame rate, you might need to tweak your settings for an optimal experience, but there does appear to be a healthy amount of performance overhead here to work with.

I do tend to mention any bugs I encounter here and I did run into two. One was sometimes the game would keep the visuals in Wraith Vision mode, which will highlight enemies and other objects you can interact with, but makes the world otherwise disappear. Just re-entering and leaving it fixed this. The other issue required restarting the game, and therefore the mission, and it was the game getting stuck in one place. Normally the third-person camera will follow Talion around, but apparently there is a bug that will cause it to become fixed in position and orientation, making it very difficult if not impossible to control Talion. (This happened during a fortress assault, so a particularly chaotic time.) Restarting the game is all I could do, but thankfully it was enough.



Overall, the graphics are good but not exactly impressive. The performance is also quite good, letting me run it at max settings with a healthy amount of performance to spare. It is always nice when this is the case, even more so for a modern title.

Data graphs:






  1. Middle-earth: Shadow of War Review - Introduction
  2. Middle-earth: Shadow of War Review - Graphics
  3. Middle-earth: Shadow of War Review - Story
  4. Middle-earth: Shadow of War Review - Gameplay
  5. Middle-earth: Shadow of War Review - Conclusion (Base Game Only)
  6. Middle-earth: Shadow of War Review - The Blade of Galadriel Expansion
  7. Middle-earth: Shadow of War Review - The Desolation of Mordor Expansion
  8. Middle-earth: Shadow of War Review - Additional Gameplay Screenshots
  9. Middle-earth: Shadow of War Review - Conclusion (Full Experience)
Related Products
Random Pic
© 2001-2018 Overclockers Club ® Privacy Policy
Elapsed: 0.1580288410   (xlweb1)