Jump to content
Mib2347

I have devised a way to demonstrate heavy aim

Recommended Posts

Holy fuck this is like "the discovery" movie. This is ground breaking. I was skeptic until this very moment. Good job sir you have proven heavy aim.

  • Upvote (+1) 1

Share this post


Link to post

Heavy aim is input lag.

 

Why would an input lag cause inconsistent distances moved by the reticule?

 

 

Also, this could just as well be proving that Halo 5 doesn't always use the same rumble pattern.

 

I'm not denying the existence of heavy aim, however this isn't proof of much until you can show that the amount of rumble applied is always the same. The fact that it seems to always go to one of 3 spots is weird. If it were due to a random latency issue then you'd think there would be more variance.

 

A better control would be to test Halo 4 as it is more likely to use the same rumble patterns as H5. Even then that's not an exhaustive proof.

  • Upvote (+1) 2
  • Downvote (-1) 9

Share this post


Link to post

If 343 didn't know about this, then what did the latest aiming patch we had last year actually 'fix'?

This amazing thread that the OP did are some of the reasons why this should be the case. I wouldn't be mad if 343 said they don't know how to fix something as long as they acknowledged the problem and properly explained the technical hurdle they are having to overcome to fix it. Making/fixing a Triple-A game isn't exactly easy by any means, but I'd imagine rebuilding the trust between dev and community is just the same.

Share this post


Link to post

Heavy aim is input lag.

 

 

Why would an input lag cause inconsistent distances moved by the reticule?

Input lag is input lag

 

Think about distance covered by reticule. The more distance covered in that demonstration could be seen as the default - snappy and fast. The slower or shorter distance in that example can be seen as the heavy aim. Basically the variations that cause it to slow down over a similar input is what heavy aim is

  • Upvote (+1) 1

Share this post


Link to post

The guys who did this are fucking heroes in my eyes.

 

So did 343 know about it and just ignore it, couldnt fix it, or thought it was fine as is ? We will never know

  • Upvote (+1) 1

Share this post


Link to post

Input lag is input lag

 

Think about distance covered by reticule. The more distance covered in that demonstration could be seen as the default - snappy and fast. The slower or shorter distance in that example can be seen as the heavy aim. Basically the variations that cause it to slow down over a similar input is what heavy aim is

Input lag is the time difference between when you press a button on the controller and when the game receives that input.

 

If you hold a button for 50ms the game is always going to receive an input which lasts 50ms, regardless of whether you have 0ms or 100ms of input lag. The difference is how long it takes for that input to be displayed on the screen. Input lag won't cause an inconsistency if the amount of time a button is pressed is always constant.

 

To use the aiming example, let's say that the reticule moves 45 degrees if you hold the stick for 50ms. Assuming the controller always rumbles for 50ms (which isn't something we should assume and the reason this doesn't prove all that much), the reticule will always move 45 degrees, regardless of how much input lag there is. 

 

The reason heavy aim makes the aiming feel inconsistent is because our brains are using the movement on screen to inform our thumbs of when to move. Varying input lag will make it it difficult to judge when to stop/start deflecting the stick and so you'll move the reticule different amounts depending on the amount of input lag. However, under controlled circumstances (ie: a button is held consistently for the same amount of time) input lag won't cause there to be a difference.

 

 

Again, I'm not saying heavy aim doesn't exist, I'm saying the op isn't proof of it.

  • Upvote (+1) 1
  • Downvote (-1) 3

Share this post


Link to post

To OP, u could also test this with multiple people in the game, to see if there is more of a drastic difference in the crosshair placement. More people in the game could also put more stress on the CPU or whatever, just a thought

Share this post


Link to post

Thanks for creating such an excellent post! I can't make any promises, but I can tell you that our Sandbox team has seen this post, and is looking over it as we speak. Thanks again, Mib!

Hey, here's a few more observations on the issue (and I'd be surprised if someone on 343 wasn't considering this already):

 

1. Outside of the various elements that Mib described could potentially be taxing the system's memory and the CPU, could Halo 5's dynamic resolution scaling be causing an issue here? Is it possible that X and Y coordinates are not being translated properly across resolution up/down scaling?

 

2. If so, running Halo 5 at 1080p on a platform easily capable of handling it (PC, scorpio) should theoretically eliminate this issue since dynamic resolution scaling shouldn't be needed. Does anyone know if heavy aim occurs on powerful PCs (using H5 forge on arena maps) as well?

  • Upvote (+1) 7

Share this post


Link to post

Hey, here's a few more observations on the issue (and I'd be surprised if someone on 343 wasn't considering this already):

 

1. Outside of the various elements that Mib described could potentially be taxing the system's memory and the CPU, could Halo 5's dynamic resolution scaling be causing an issue here? Is it possible that X and Y coordinates are not being translated properly across resolution up/down scaling?

 

2. If so, running Halo 5 at 1080p on a platform easily capable of handling it (PC, scorpio) should theoretically eliminate this issue since dynamic resolution scaling shouldn't be needed. Does anyone know if heavy aim occurs on powerful PCs (using H5 forge on arena maps) as well?

1. That's an intriguing thought.

  • Upvote (+1) 8

Share this post


Link to post

If someone has time (I can do it after work), you can confirm the duration of the stick command input (and therefore the vibration length) by measuring the time it takes for the aim to move. For example, start at the frame the horizontal angle stats changing and stop at the frame the angle settles into the final value. If the number of frames in that timeframe is consistent, then we can be confident the input signal is the same.

Share this post


Link to post

Hey, here's a few more observations on the issue (and I'd be surprised if someone on 343 wasn't considering this already):

 

1. Outside of the various elements that Mib described could potentially be taxing the system's memory and the CPU, could Halo 5's dynamic resolution scaling be causing an issue here? Is it possible that X and Y coordinates are not being translated properly across resolution up/down scaling?

 

2. If so, running Halo 5 at 1080p on a platform easily capable of handling it (PC, scorpio) should theoretically eliminate this issue since dynamic resolution scaling shouldn't be needed. Does anyone know if heavy aim occurs on powerful PCs (using H5 forge on arena maps) as well?

 

I honestly can't imagine a 3D game would be translating aim input into X/Y screen coordinates, unless 343's aiming implementation is even more ass-backwards than I think it is.

Share this post


Link to post

So, the real question is, if you were to pause the game between deaths will the movement be more consistent as many claim it to be? Or is the ritual of mapping START to your elite controller fruitless. 

Share this post


Link to post

Heavy aim is input lag.

 

Why would an input lag cause inconsistent distances moved by the reticule?

 

 

Also, this could just as well be proving that Halo 5 doesn't always use the same rumble pattern.

 

I'm not denying the existence of heavy aim, however this isn't proof of much until you can show that the amount of rumble applied is always the same. The fact that it seems to always go to one of 3 spots is weird. If it were due to a random latency issue then you'd think there would be more variance.

 

A better control would be to test Halo 4 as it is more likely to use the same rumble patterns as H5. Even then that's not an exhaustive proof.

The varying rumble patterns have been brought up and is currently being discussed. I agree that variance would make more sense. However, it is also possible that there are strict parameters as to how the information is handled when it is being processed. The game may not allow variance. Instead the controller input may be put into a que behind other processes. If it falls in between point two and three, the game waits until time^3 to being the input.

 

I am by no means stating that this is how it works. That example by itself doesn't circumvent the argument. Just pointing out that it's possible dynamic variance has been excluded from operation somehow as a contingency or stability measure.

 

 

 

  • Upvote (+1) 1

Share this post


Link to post

The varying rumble patterns have been brought up and is currently being discussed. I agree that variance would make more sense. However, it is also possible that there are strict parameters as to how the information is handled when it is being processed. The game may not allow variance. Instead the controller input may be put into a que behind other processes. If it falls in between point two and three, the game waits until time^3 to being the input.

 

I am by no means stating that this is how it works. That example by itself doesn't circumvent the argument. Just pointing out that it's possible dynamic variance has been excluded from operation somehow as a contingency or stability measure.

 

 

 

How would that explain the lack of variance in the distance covered by the reticule?

 

If we assume this isn't due to varying rumble patterns then it implies that the input is being rounded to one of the three values which:

 

a. Makes no sense.

b. Wouldn't cause the aiming to feel heavy.

  • Downvote (-1) 1

Share this post


Link to post

Hey, here's a few more observations on the issue (and I'd be surprised if someone on 343 wasn't considering this already):

 

1. Outside of the various elements that Mib described could potentially be taxing the system's memory and the CPU, could Halo 5's dynamic resolution scaling be causing an issue here? Is it possible that X and Y coordinates are not being translated properly across resolution up/down scaling?

 

2. If so, running Halo 5 at 1080p on a platform easily capable of handling it (PC, scorpio) should theoretically eliminate this issue since dynamic resolution scaling shouldn't be needed. Does anyone know if heavy aim occurs on powerful PCs (using H5 forge on arena maps) as well?

 

Halo 5 on my PC feels amazing on any map, on any host.  This has been well documented by other players as well (lethul for one) I have an I7-3770k.  Even a laggy host my aim feels perfect.  This lines up perfectly with CPU stress being the root cause...that or the fact my GPU doesn't struggle to run Halo 5 at max resolution so it never has to use dynamic res resizing.

 

Either way, I'd almost gurantee the scorpio will not have this issue for the exact same reason but I don't think we all want to wait 6+ months on that.  Sounds like they could just lower the resolution of H5 on consoles currently to relieve some CPU load, or figure out a better way to optimize CPU bandwidth for the game.

Share this post


Link to post

How would that explain the lack of variance in the distance covered by the reticule?

 

 

If we assume this isn't due to varying rumble patterns then it implies that the input is being rounded to one of the three values which:

 

a. Makes no sense.

b. Wouldn't cause the aiming to feel heavy.

How would that explain the lack of variance in the distance covered by the reticule?

It doesn't. Which is why you shouldn't take it as an explanation. If the command is delayed it is just delayed. Therefore the distance covered should be the same.

 

I'm simply making the point that it is worth exploring further until varying rumble problems are proven. There is probably more to it than just a simple delay. It could be that the initial input gets trimmed and lost or the acceleration doesn't proceed correctly. There are many possibilities worth thinking out while we have this data. Which is why I specifically said "I am by no means stating that this is how it works. That example by itself doesn't circumvent the argument."

 

 

  • Upvote (+1) 5

Share this post


Link to post

With regards to the rumble signal I have a few things I'd like to say. The aim seems move to either a or b and seems to do so for a period of time. So if it was because the signal is different then it would surely be more random instead of a for 5 tests in a row and b for another 5.

 

I have an old 360 and a copy of halo 2. What I would like to do tomorrow is take the vibration from the 360 controller instead and use that to trigger the movement. If I did that on both Halo 1 MCC it's consistent then see what happens with Halo 5.

 

And thanks for responding Unyshek. A lot have people have complained about aim for a long time and it's great to know we are heard.

 

I think the reason this has taken off is because we've all experienced heavy aim but no one has been able to really prove it, even though those of us that have experienced it would bet our lives on it's existence. I did this because I just want to enjoy halo.

 

This coincides with my idea that the array that you used isn't suited to deal with a signal that is sending information to a rumble motor processing unit that deals with a signal that has info for stuff like angular acceleration, frequency, etc. 

 

From looking at the specs of the array you linked, it looks like a simple pin digital switching circuit, meant specifically to be used with simple hi-low signals that allows it to function well for a hobbyist, mechanical actuator circuit, etc. 

 

If the digital to analog signal that is being sent from the game to the controller to mimic a certain type of vibration by rotating the rumble motors at a certain speed is more complicated than just a hi-low/low-hi set for a certain # of cycles, then that could explain the repeatable errors of the aiming in H5 when it comes to your tests; the relay is sending minute errors in terms of a convolved signal that cause it to last longer or shorter than it probably would otherwise. 

 

As for CE, the vibration signal has to be much more simple and more like a pulse. This would explain why the relay works perfectly with CE because the vibration signal from a game as old as CE will be fairly archaic and simple compared to that from a next gen game designed with the concept of vibration feedback as a way of making the player "feel" the game more.

 

Ex. CE vibration will be akin to a pulse lasting a certain number of cycles = accurate representation through your design.

 

Halo 5 vibration will have a much more nuanced digital-to-analog signal to utilize the new age vibration tech = pattern-like error patterns.

 

This also coincides with the weird behavior of the rockets when you were killing yourself with them. The deflection was much more varied, either much less to one side or actually to the right on the death cam. I know that may be the tendency for the camera to immediately follow the body for a split second on death, but you might want to check that out. 

 

Also, last and final question, why did you decide to use the carbine for the H5 test? I ran through the weapons and their respective vibrations and the magnum and BR felt like the best candidates to possibly simulate a pulse. The carbine felt like it had a truncated waveform of vibration to make it less annoying when spamming. 

 

Anyways, great job either way. Look forward to your responses.

  • Upvote (+1) 4

Share this post


Link to post

If they just lock H5 at 720p instead of the current 900/1080p, that might solve this problem ... not sure if that can be done through an XBL update though

Share this post


Link to post

The last time I played Halo 5 on PC was when that "fix" for the aiming as patched, the mouse aim still felt really bad. I wish I had tried the controller to test. 

 

I even tied it with a FoV and FPS mod. I had my frames locked at 120 no problem but I have a beast PC. 

 

Yeah the mouse aiming is absolute trash. Probably the worst aiming on a PC game to date. The controller however, felt amazing. If you're going to use the mouse+KB for something on there, only use it for the actual forge stuff, but that's about it.

  • Upvote (+1) 1

Share this post


Link to post

Also, last and final question, why did you decide to use the carbine for the H5 test? I ran through the weapons and their respective vibrations and the magnum and BR felt like the best candidates to possibly simulate a pulse. The carbine felt like it had a truncated waveform of vibration to make it less annoying when spamming. 

 

This makes me wonder as well. I'm kind of worried that he tried with the pistol, didn't get the results he wanted, then tried the carbine and found it "worked."

 

Even though I very much believe that the aiming in H5 is hot garbage, I'm not sure that this has proven anything yet. I would be very, very interested to see the results in H4. I suspect that they'll match H5's. Hope I'm wrong though and that maybe this will make 343 do something.

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×

Important Information

By using this site, you agree to our Terms of Use.