Graphics configuration wizard

[zmz125000]

20 minutes ago, upyr1 said:

The Su-25T in the Marinas would work

You would need the tools to monitor system performance, and as stated before the fact performance can vary widely is exactly what I am asking for a benchmark tool to help with the graphics As the entire point of the tool would be to test under different conditions

you can use any saved track as a benchmark. 

[upyr1]

2 hours ago, BIGNEWY said:

A benchmark is something we discuss often internally, but I have no news to share sorry. 

I would be in favour of having a benchmark in DCS 

thanks

It's always good to find out the topic in question is being discussed by ED.  I hope that one of ya'll can chime in about some news on the topic.  I'm sick of shooting blind trying to pick the right settings

[SharpeXB]

45 minutes ago, zmz125000 said:

you can use any saved track as a benchmark. 

Right. If you’re just looking to set your own graphics you don’t need an in-game benchmark. Just make your own. 
 

The advantage of an in-game standard would be for ED to troubleshoot performance problems. The trouble there is that these could occur in situations not included in the standard track. In that case just submit your track. 
 

Either way having an in-game standard doesn’t seem to be much help. 

[upyr1]

2 minutes ago, zmz125000 said:

you can use any saved track as a benchmark. 

it doesn't matter what they use as long as the utility gives us a good baseline. I really don't care I just want to have a reasonable balance between good performance and graphics. Though I think an import/ export settings might be nice too

6 minutes ago, SharpeXB said:

Right. If you’re just looking to set your own graphics you don’t need an in-game benchmark. Just make your own. 
 

The advantage of an in-game standard would be for ED to troubleshoot performance problems. The trouble there is that these could occur in situations not included in the standard track. In that case just submit your track. 
 

Either way having an in-game standard doesn’t seem to be much help. 

The tool would be helpful in setting your own graphics since it will provide a good baseline and optimum settings. 

[Tank50us]

26 minutes ago, upyr1 said:

The tool would be helpful in setting your own graphics since it will provide a good baseline and optimum settings. 

A good example would be how well does your settings handle a full battery of M270s eliminating a grid square? All those rockets and bomblets raining down on an area and cause frame rates to absolutely tank if your computer can't handle them. Or how well does the game handle a particularly busy airfield with dozens of planes lined up to taxi and take-off? Or how well does it handle a CVBG launching a wave of Tomahawks at a target area. Things like that. Things that will put the graphics settings under stress.

[Tippis]

53 minutes ago, SharpeXB said:

Right. If you’re just looking to set your own graphics you don’t need an in-game benchmark. Just make your own. 

…except that it would not actually test the different aspects that go into determining what settings make a difference. You would be vastly better served with a standardised script and corresponding analysis tool that is directly linked to the existing in-game settings in a way that only an in-depth understanding (and profiling) of the engine would allow for.

You can always roll your own, of course, but it doesn't offer what is being requested here.

53 minutes ago, SharpeXB said:

The advantage of an in-game standard would be for ED to troubleshoot performance problems.

Another advantage would be that it would let players troubleshoot performance problems, and with a bit of help even target specific settings that would alleviate the specific problem the player has given their setup. So that's two benefits, and zero downsides.

53 minutes ago, SharpeXB said:

Either way having an in-game standard doesn’t seem to be much help. 

Arguments from incredulity are fallacies for a reason. Your lack of imagination does not provide any intelligent or logical reason why a benchmarking tool should not exist or why it would not help identify optimised settings.

[SharpeXB]

36 minutes ago, Tank50us said:

A good example would be how well does your settings handle a full battery of M270s eliminating a grid square? All those rockets and bomblets raining down on an area and cause frame rates to absolutely tank if your computer can't handle them. Or how well does the game handle a particularly busy airfield with dozens of planes lined up to taxi and take-off? Or how well does it handle a CVBG launching a wave of Tomahawks at a target area. Things like that. Things that will put the graphics settings under stress.

The point of a benchmark track isn’t to completely crush your system. Given the sandbox nature of DCS that will always be possible. Just overload the mission with too many units and action. You could just make a situation like this for yourself in the ME. There’s no point in ED making a track to just crush everyone’s PC. It would discourage anyone from buying the game.

1 hour ago, upyr1 said:

The tool would be helpful in setting your own graphics since it will provide a good baseline and optimum settings.

Yeah but you can do this yourself with your own track. You don’t need ED to do it for you. 

[_Hoss]

20 hours ago, Tank50us said:

While we're at it, how about a control setup wizard as well for all the 'common' commands? After all, when someone changes to a new system or new control setup, having *something* to make it easier to set up the dozen or so planes, while also making sure everything else is good to go would be freaking awesome. Heck, if a more arcady sim from that company with a snail as its logo can pull it off, I'm sure ED could add it to DCS

JoyPro........

[Tippis]

1 hour ago, SharpeXB said:

The point of a benchmark track isn’t to completely crush your system.

Good thing, then, that no-one suggested that it is.

But it is the point of a benchmark tool to stress the system and to find where it is lacking — you won't achieve that if you allow for too great an overhead.

1 hour ago, SharpeXB said:

Given the sandbox nature of DCS that will always be possible. Just overload the mission with too many units and action. You could just make a situation like this for yourself in the ME. There’s no point in ED making a track to just crush everyone’s PC. It would discourage anyone from buying the game.

No. No-one would be discourage from downloading the game for free just because it comes with a built in tool to help optimise settings. Quite the opposite. And the reason you can't just do it in the ME is because as you have demonstrated, you have no idea what additions, actions, and in-game events (in the plethora of meanings that word holds in DCS) stress what components, or why, or how. You also don't know what settings are at your disposal to alleviate those stress points.

ED, thankfully, clearly disagrees with you, and it's no surprise seeing as how it would be hugely beneficial to everyone and a significant improvement to the game to have that kind of universally available and applicable analytical tool. No wonder, then, that you're dead set against it.

[upyr1]

4 hours ago, SharpeXB said:

Yeah but you can do this yourself with your own track. You don’t need ED to do it for you

As I keep saying a built in benchmark utility would be able to configure the graphics settings automatically while external ones won't. Someone from Eagle has stated they want to add the tool

Edited December 13, 2021 by upyr1

[SharpeXB]

27 minutes ago, upyr1 said:

As I keep saying a built in benchmark utility would be able to configure the graphics settings automatically while external ones won't. Someone from Eagle has stated they want to add the tool

 

Those are two different things. A benchmark track is just a track that runs so you can test your system. The automatic graphics optimization does just that, it doesn’t use a benchmark. ED mentioned wanting the former, not the latter. 
You don’t need ED to make you a benchmark track, any replay from the game will work. Unless it involves AI, that needs fixing. 

[upyr1]

2 hours ago, SharpeXB said:

Those are two different things. A benchmark track is just a track that runs so you can test your system. The automatic graphics optimization does just that, it doesn’t use a benchmark. ED mentioned wanting the former, not the latter. 
You don’t need ED to make you a benchmark track, any replay from the game will work. Unless it involves AI, that needs fixing. 

No they are not, as I am asking for a bench mark utility which would set graphics recommendations based on the results.. That was the freaking title of the thread.  

[SharpeXB]

34 minutes ago, upyr1 said:

No they are not, as I am asking for a bench mark utility which would set graphics recommendations based on the results.. That was the freaking title of the thread.  

I think you are asking for something that no other game has. I have never seen a game with what you’re describing.

[upyr1]

28 minutes ago, SharpeXB said:

I think you are asking for something that no other game has. I have never seen a game with what you’re describing.

So it would be a unique feature. A graphics optimization wizard Norma need a list of hardware to set an optimized configuration. Using the benchmark utility which Eagle is discussing would mean don't need the list as it can test settings

14 hours ago, BIGNEWY said:

A benchmark is something we discuss often internally, but I have no news to share sorry. 

I would be in favour of having a benchmark in DCS 

thanks

 

If you add the benchmark tool I hope you have a stress test mode to set optimal graphics 

[SharpeXB]

45 minutes ago, upyr1 said:

So it would be a unique feature

How about we don’t waste ED’s time asking for features that PC games don’t have. 

[SharpeXB]

47 minutes ago, upyr1 said:

If you add the benchmark tool I hope you have a stress test mode to set optimal graphics

A benchmark tool is just a track or gameplay sample which runs and then shows you the min, max and average frame rate. You are the one that sets the graphics. It doesn’t do this for you. Since it’s just a matter of taste or opinion how would it know what to set? Now Nvidia has some giant cloud computing data center that comes up with it’s optimized settings. But your PC isn’t that and ED isn’t Nvidia. 
 

And the track replay in DCS has been broken for years and is apparently a big deal to fix.

[upyr1]

1 hour ago, SharpeXB said:

A benchmark tool is just a track or gameplay sample which runs and then shows you the min, max and average frame rate. You are the one that sets the graphics. It doesn’t do this for you. Since it’s just a matter of taste or opinion how would it know what to set? Now Nvidia has some giant cloud computing data center that comes up with it’s optimized settings. But your PC isn’t that and ED isn’t Nvidia. 
 

And the track replay in DCS has been broken for years and is apparently a big deal to fix.

This  won't require some massive super computer. All that it will require is for the utility to have a routine where it slowly cranks up graphics on the track being benchmarked until the frame rate drops bellow an acceptable threshold. For example let's say the min FPS setting is 30, then the tests stop at that point and you have the maximum settings for your computer.  

[SharpeXB]

17 minutes ago, upyr1 said:

This  won't require some massive super computer. All that it will require is for the utility to have a routine where it slowly cranks up graphics on the track being benchmarked until the frame rate drops bellow an acceptable threshold. For example let's say the min FPS setting is 30, then the tests stop at that point and you have the maximum settings for your computer.  

You should look at other PC games to see how this works. It’s a lot more simple than you’re imagining. And even something this simple in DCS is probably not possible without fixing the replay function. 
You could have set up your graphics in the time you’ve spent posting here…

[upyr1]

11 minutes ago, SharpeXB said:

You should look at other PC games to see how this works. It’s a lot more simple than you’re imagining. And even something this simple in DCS is probably not possible without fixing the replay function. 
You could have set up your graphics in the time you’ve spent posting here…

I know how it works with other games, they use a look up table.

[SharpeXB]

40 minutes ago, upyr1 said:

I know how it works with other games, they use a look up table.

The benchmark track requires you to set the graphics. It doesn’t determine them. It just tells you an FPS result. Do you own any games with this feature? It’s actually not very common. 

[Tippis]

5 hours ago, SharpeXB said:

I think you are asking for something that no other game has. I have never seen a game with what you’re describing.

Again, argument from ignorance is a fallacy for a reason.

Finding games that determine some sort of “best” setting depending on detected hardware is utterly trivial. Finding ones that come with a built-in 3DMark-style benchmarking mode, while a bit less common, is still pretty darn easy. There are even games that were, or over time became, a kind of benchmark standard because of this (think the likes of Crysis or Serious Sam or even good old Unreal, just to span a decade or two of different testing suites).

49 minutes ago, SharpeXB said:

The benchmark track requires you to set the graphics. It doesn’t determine them.

So maybe you should stop being so focused on just this supposed track and the supposed ability to cook your own, then, and join the actual conversation about a tool that does actually set your graphics (and other settings) depending on the benchmark, hmm…? You know, the topic of the thread?

Just because you've chosen to not discuss the topic and instead want to only talk about some very tiny subset of the feature doesn't mean that the feature cannot exist or that games haven't done this since forever. It just means you're not really qualified to participate in the discussion because of your lack of experience with the genre.

[Boosterdog]

Not arguing for or against but how would “optimal” be defined?

My optimal is likely different from yours. Some think flat shadows are better, some don’t  some are happy with vsync, some are not.  An optimal setting in Caucaus will likely crush a machine on the Marianas, channel or Syria maps. Different maps utilise the settings such as visible range and detail factor sliders very differently. A cap mission is hugely different in its  demands from a helo cas. Modules have different demands. Even different missions using the same module and map can vary significantly dependent on what the AI is doing and what is depicted. 

I’m finding it difficult to understand how a game like dcs with so many variables necessary and otherwise can have a meaningful benchmarking tool. 
 

 

[sirrah]

1 hour ago, Boosterdog said:

Not arguing for or against but how would “optimal” be defined?

My optimal is likely different from yours. Some think flat shadows are better, some don’t  some are happy with vsync, some are not.  An optimal setting in Caucaus will likely crush a machine on the Marianas, channel or Syria maps. Different maps utilise the settings such as visible range and detail factor sliders very differently. A cap mission is hugely different in its  demands from a helo cas. Modules have different demands. Even different missions using the same module and map can vary significantly dependent on what the AI is doing and what is depicted. 

I’m finding it difficult to understand how a game like dcs with so many variables necessary and otherwise can have a meaningful benchmarking tool. 
 

 

I agree with @Boosterdog

Until all maps cause similar strain on performance (so probably never ), there is no such thing as one optimal setting. While running butter smooth when playing PG map, my pc is sweating severely when I'm flying over Syria at the same settings.

Not that I'm against the idea, I just don't think it's feasible.

Edited December 14, 2021 by sirrah

[Lurker]

19 hours ago, SharpeXB said:

Either way having an in-game standard doesn’t seem to be much help. 

This is not even remotely true in most games that feature it, and repeating it ad-infinitum will not make it true. Not least of the problems with a general statement like this is that it's stated as factual while in reality this feature doesn't even exist in the game yet, and has no basis in reality. The only downside to the OPs suggested feature is the time and manhours it would take to implement in DCS World. 

Edited December 14, 2021 by Lurker

[cfrag]

4 hours ago, Tippis said:

There are even games that were, or over time became, a kind of benchmark standard because of this (think the likes of Crysis or Serious Sam or even good old Unreal, just to span a decade or two of different testing suites).

Hey I seem to dimly recall Unreal's "Castle Flyby" built-in benchmark. That was 1999? Nowadays 'Guardians of the Galaxy' and 'Cyberpunk' also sport similar benchmarks, so yeah, I guess they are still common for AAA titles. Not that I'm into benchmarks (they to me feel uncomfortably close to some juvenile comparing of various body parts), but benchmarks can be useful to manage your expectation and determine if you need to upgrade your rig (with DCS, it seems, that answer is always a resounding "YES" ).