Jump to content

Performance VS looks: LOD settings examined


Recommended Posts

Being one of those gamers with a brand new high-end rig that runs, say, 'Shogun: Total War' at max settings without my machine even breaking a sweat I've been a little disappointed with the performance of the demo on my machine. ;) However, I'm not really the complainy type, so I figured I'd just lower some settings a little and take it from there.

To that end I've been trying to figure out what the 'sweet spot' is between performance and looks (unit/terrain detail) using the different LOD settings one can access in-game using the { and } keys. These settings are:

Fastest

Faster

Fast

Balanced

Improved

Better

Best

Without going into specific FPS gain/loss (which would be pointless, as it all depends on size of map, density of terrain features, wind speed and number of units anyway) so far I've noticed the following things:

1: LOD settings have no impact on soldier models, foxholes or trenches.

2: The 'sweet spot' for small arms, mortars and the like is at the Fast setting. Going from Faster to Fast adds a noticable extra layer of detail to these. However, higher LOD setting do not seem to have any additional effect there. LOD settings have no effect on Panzerfausts, but, get this, the level of detail is already so insanely high, you can actually read the user instructions printed on the thing on whatever LOD setting you use! Provided you can read German, that is.

3: This 'sweet spot' on the Fast setting also works for tanks: lots of pretty extra detail, but not in every demo scenario! That's what confuses me a little. The Sherman models in the tutorial also look their best from the Fast LOD setting and up, but for some reason LOD settings have NO impact on the Shermans in Busting the bocage or the German AC's in Closing the gap at all! I don't know about the AC but the Shermans are at their highest level of detail regardless of LOD settings. Bug or demo thing, I wonder?

As with small arms, higher LOD settings do not have any additional effect.

4: LOD settings do not seem to have any effect on buildings or flavor objects (wood piles, carts, barrels and such).

5: Haven't really looked very deep into this but the higher LOD setting have the most impact on the terrain. Especially the quality of distant trees, and the visibility of grass at medium viewing distances. Also finer detail like edges of roads and bocage seem to be a little more defined when viewed from longer distances.

So far I'm finding that -for me- using the Balanced LOD setting is the best compromise between looks and performance. Especially when playing with the 'show trees, nearby trunks only' setting.

But, really, the elegant thing about the system is that you can adjust LOD on the fly with a few simple keystrokes, without the need to quit the game and go into the game's option to adjust LOD.

My system:

Asus P7P55D-E,s1156 mobo

Intel Dual Core i5-660, s1156 @ 3.33Gh

4 GB Mushkin Radioactive DDR3 @ 1600Mhz

OCZ Solid State Drive Vertex 2 - 120GB

Asus Geforce GTX 580 w/1.5GB DDR5

Thoughts? Other discoveries?

Link to comment
Share on other sites

What's your system? So that we might compare.

I have a Phenom IIx4 clocked at 3.8 GHz and a slightly overclocked 5870. I set everything to max at a resolution of 2560x1600 in CM demo and noticed no performance problems.

I also have Shogun:Total War II and can max the settings (although there isn't any AA available)

Link to comment
Share on other sites

What's your system? So that we might compare.

I have a Phenom IIx4 clocked at 3.8 GHz and a slightly overclocked 5870. I set everything to max at a resolution of 2560x1600 in CM demo and noticed no performance problems.

I also have Shogun:Total War II and can max the settings (although there isn't any AA available)

Here goes:

Asus P7P55D-E,s1156 mobo

Intel Dual Core i5-660, s1156 @ 3.33Gh

4 GB Mushkin Radioactive DDR3 @ 1600Mhz

OCZ Solid State Drive Vertex 2 - 120GB

Asus Geforce GTX 580 w/1.5Gb DDR5

All GPU settings on default in the control panel, except anisotropic filtering @ 16x

All gamesettings maxed, except LOD to balanced.

This gives me 15 to 60 FPS @ 1920x1200 resolution depending on, well, a lot of things, as you know.

Link to comment
Share on other sites

I also have Shogun:Total War II and can max the settings (although there isn't any AA available)

Actually there is now, since the latest May 9th patch all graphical features are available for that game. Just so you know ;)

I myself find the balanced/balanced to be the best compromise. Which thankfully makes sense considering it is called balanced!

Link to comment
Share on other sites

Oddly enough, I was just running the same tests on the latest version of the game. In CMSF I always used best/best. Since most scenarios featured no trees, it worked well.

However CMBN features lovely detailed trees and forests which can devastate your FPS. Running a variety of tests, it also appears that balanced/balanced offers the best compromise or "most bang for your buck". On a large forested QB map, those settings gave me FPS of around 35 when looking straight into the forest vs 15 at best/best. Furthermore balanced/balanced gives up little in terms of noticeable graphical quality.

My system: Q9550, ATI 4890, win 7 64, 1920x1200, no FSAA/AA, 27" monitor.

Link to comment
Share on other sites

I have a pretty cruddy setup in my opinion, except for the video card, and I'm running it at 1920x1080 using HDMI on a 32inch LCD TV with max graphic settings on every option and haven't noticed anything that I consider to be performance loss.

My System:

CPU: AMD Athlon 64 X2 5000+ - Dual Core - 2.6 GHz

Memory: 2GB DDR2

Video Card: ATI Radeon HD 5870

Now I haven't tried it at lower settings or even checked the FPS, but I usually don't do that unless I notice below average or worse performance. I find that when tweaking graphics its easier for me to start out at max and then lower settings until I stop seeing "chug." I have yet to see chug in this game... But I've only played single player, so performance could be different during real time multiplayer when I can't pause, and also when playing larger scenarios in the full release version.

Link to comment
Share on other sites

BTW, I'm running EVGA Precision, and just noticed the GPU usage is only at 40 to 50% at best. Memory usage maxes out at 375MB. Clearly CM:BN isn't using the full potential of my card.

Any ideas why?

Link to comment
Share on other sites

Actually, 2560x1600 is a very common resolution for 30" monitors. There aren't that many on the market, but I've seen some from Nec, LG, Dell, HP and Eizo at least.

That's the problem. In my area I've never even seen/found one. 28" is the best local shops can do, but those 2" just isn't enough to warrant an upgrade for my 26" Samsung.

Link to comment
Share on other sites

BTW, I'm running EVGA Precision, and just noticed the GPU usage is only at 40 to 50% at best. Memory usage maxes out at 375MB. Clearly CM:BN isn't using the full potential of my card.

Any ideas why?

An application can't use what it doesn't need. If you're still experiencing a "slowness", then that's due to other bottlenecks.

Link to comment
Share on other sites

That's the problem. In my area I've never even seen/found one. 28" is the best local shops can do, but those 2" just isn't enough to warrant an upgrade for my 26" Samsung.

Very few shops carry stock of 30", they are usually IPS panels, and very expensive compared to the 'mainstream' 24" TN panels (1920x1080 or 1920x1200) - think $1,500+ USD. They are available on dell / HP / samsung websites.

Also, about 5-7 years ago 27 - 30" monitors were more popular then they disappeared from supply for a while, I assume there just wasn't demand. The focus on the market seemed the 19-24" inch range (1080p), however they are starting to come back to popularity - I assume because now the machines can start pushing decent framerates at that resolution.

Link to comment
Share on other sites

Running a Core i5 2500K @ 3,7GHz

MSI 67AGD65

8GB DDR3 @ 1333mhz

SDD OCZ Vertex 2

X-Fi XtremeGamer

HD5770 (to be replaced)

have 8x FSAA and 16x AF, 1440x900 res, running best/best, absolutely no slowdowns in the CMBN demo. Oddly enough, CMA did feel a tiny bit sluggish at these settings. Haven't tested CMSF as I can't be arsed to reinstall it after having experienced CMBN ;)

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Unfortunately, your content contains terms that we do not allow. Please edit your content to remove the highlighted words below.
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...