FOV and Input Lag/Latency

[Bobby Pennington]

I just made a video of what using the correct FOV per all the calculators and the math and how it can destroy your experience. I've been chasing my tail with my G27 settings and it turns out the biggest problem was the stupid FOV.
Visit my boobtube page MulliganF1 for a full description of my setup and explanation.
This is the only way I can give back to the Sim community since I don't make skins, tracks etc., but I'm working on it.
Please let me know if the video helps. Have a great night and as always Opposite Lock to 'ya.

P.S.: I hope I posted this in the correct section. Still learning.

 

[Bobby Pennington]

Just changing your FOV makes your car handle that much differently?

Yes it does. It doesn't change the physics, but does affect the latency so by default it effects how smoothly or jerky you drive. You'll notice how close the signs on the left and corners seem to appear so close to you when I use the FOV value of 31 per the calculators and by doing the math myself. Do the trial yourself. You will see how it is. I made no change to my car or wheel when I did this yesterday. I drove just far enough for you guys to see the time difference. FOV that is to low will haunt you forever. Get it right and lap times start to drop considerably.

 

[Bobby Pennington]

I've had it happen too, it was very noticeable in R3E. In AC I can't change my FOV because I use triple screens, but I have noticed when I go into the triple screen app and adjust the distance between the monitor and my eyes (probably doing the same thing as FOV) it can fairly drastically change the way my car behaves. If I decrease it a lot the car feels very pointy and twitchy, if I increase it a lot the car feels very sluggish and floaty. I thought it was a fluke or something to do with my system, but apparently not.

Pointy and twitchy!!! You just hit it on the head my man.

 

[Bobby Pennington]

Well dude, nobody said anything about the physics or track being different, we just said it can make the car feel/behave different, which you just admitted is a thing, so I'm not sure what point you're trying to make here.

Do you have triple screens? If so open the triple screen app and put in a value of 1,000 for the distance spanner, then drive a few laps. Then put in a value of 300 and drive a few more, tell me how it feels.

Brandon you are correct. It does effect the input and how much we feel the pointy and twitch you mentioned earlier.
Burrito is correct in saying it doesn't change the physics or how the car handles.
My point in all of this is that an incorrect FOV has a HUGE impact on OUR inputs. Ever had a mouse with a low resolution? Notice how it moves in large CHUNKS and you can't be as accurate? Same thing here.

 

[Bobby Pennington]

You are right. Feel is different. Physics are not. Nor is the track.

You can think and feel whatever you want but the facts are there. Happy driving.

Burrito I agree with you completely, but if you watch my video you will see how just changing the FOV took me from driving smoothly to not hardly being able to make the first two corners at the 6:53 mark of my video.
Spatial awareness has a lot to do with it, but with the FOV set to low I barely have to turn the wheel and I get a huge input on the track and as Brandon said "Pointy and Twitchy" best sums it up.
I use the FOV for the correct aspect ratio that makes the car react like it does in the Bonnet view. I then move my seat back so I can see my wing mirrors if possible.
Have a great day and I appreciate your input. We're all just trying to make our experience better.

 

[burrito]

Burrito I agree with you completely, but if you watch my video you will see how just changing the FOV took me from driving smoothly to not hardly being able to make the first two corners at the 6:53 mark of my video.
Spatial awareness has a lot to do with it, but with the FOV set to low I barely have to turn the wheel and I get a huge input on the track and as Brandon said "Pointy and Twitchy" best sums it up.
I use the FOV for the correct aspect ratio that makes the car react like it does in the Bonnet view. I then move my seat back so I can see my wing mirrors if possible.
Have a great day and I appreciate your input. We're all just trying to make our experience better.

I saw it and from personal experience I found a lower FOV (the correct mathmatical FOV) to be best. It feels immensely slow to drive but offers the most time to setup and execute a corner whereas a high FOV means everything looksfaster and is much harder to judge braking points and apexes.

You can get used to both but a low FOV, for me, was an obvious advantage in getting faster and more consistent.


Question? Did you record all the tests at once? Back to back? And what was your normal FOV before the video recording?

Because all I see is you hitting a kerb at 6:53 and then not saving it. Nothing to do with FOV apart from the fact you couldn't see it once you were on top of it (which I wouldn't be able to see in the rift either).

 

[Bobby Pennington]

In your first post you said the car behaviour changes.



And what I am saying is that your inputs change because your judgements change (for better or worse) due to the visual FOV change. The way the car reacts to your inputs is the same however.

And no I don't own triples. I have a DK2 because I am a believer in using as close to a mathmatically accurate FOV as possible

I just don't like the original post and it's reasoning and wanted to chime in on how best to approach it

Reasoning? It's plain and simple and there for you to see. I do not claim to be an expert at any of this stuff, however, I do claim to be a practical person and when my experience is there for you to see I've done all I can and want to do. I'm not trying to convince anyone the my approach is the best, but it's just my approach and the results that came out of it.
It all started with me Watching Empty Box aka Matt Orr. I used to drive with a FOV that was entirely to high, but I never had any input lag. I just couldn't see the corners coming up if you know what I mean. Once I did the math and used the FOV calculators I get set on my FOV being around 35. Then I noticed the latency and dealt with it ad nauseam. Then I thought it must be a controller issue. When I could not reach the results I wanted and I finally took my head out of my butt (after all I've got to be hardcore don't I lol) I used the arcadeish Hood view and it was like night and day to me. That's when things merged and I came to my conclusion. Mathematics and practical collided to share this with everyone.
I respect your opinion and I hope you do mine as well. Have a great day.

 

[unknwn]

Definitely no difference in physics or latency.
What you experience is overcompensation with wheel inputs based on the screen output. With low FOV it appears that things move faster therefore you overcompensate with your inputs as you are used to higher FOV and less movements. You use your reflexes to adjust your car based on previous FOV, you need to slowly adapt.
Probably you can't just jump to that low FOV instantly. Do that in steps. I used to run 45, then 40, now 35 FOV.

Edit: I guess there is a possibility to experience more latency with lower FOV - that would be high LCD/TV latency. Normally lower FOV allows you to notice movements quicker and you will be able to react and adjust the car more precisely. However if your LCD/TV has high inbuilt latency that latency will also feel more noticeable/exaggerated and might feel out of sync with the wheel.

 

[Bobby Pennington]

Yes, as in the behavior feels different, as in with one setting I need a very small steering input to get around a corner but with another setting I need sometimes 180º of input to make the same turn.



Ok, I'll make a video so you can see what happens. I understand the point you're making but it's different to the point I'm trying to make.

I like it. Tag me in the post if you can. I would love to see someone other than myself do this and their results.

 

[Bobby Pennington]

I think if I have my 55" screen right in front of my face I can use wider FOV. Being about 45" away I can't see most of the apex if I set it to 32 deg. I am not sure how having the camera turn towards the apex is any more rear since my head doesn't move. I'd rather move my head than the camera tells me where to look. I've tried in the past in some older games that can do that and that messes with my feel even more.

I too used to use look to apex and it led to me over correcting the car because the view is tied to your inputs. JMHO.

 

[Bobby Pennington]

I saw it and from personal experience I found a lower FOV (the correct mathmatical FOV) to be best. It feels immensely slow to drive but offers the most time to setup and execute a corner whereas a high FOV means everything looksfaster and is much harder to judge braking points and apexes.

You can get used to both but a low FOV, for me, was an obvious advantage in getting faster and more consistent.


Question? Did you record all the tests at once? Back to back? And what was your normal FOV before the video recording?

Because all I see is you hitting a kerb at 6:53 and then not saving it. Nothing to do with FOV apart from the fact you couldn't see it once you were on top of it (which I would be able to see in the rift either).

Yes I recorded all of this in real time in one 10 minute period. Started with hood view, then at 2:10 mark I just cycled the camera to my 46 FOV, at 4:21 I used Auto Pos and moved the seat back, at 6:53 I used the Mathematically correct FOV of 31 (by the way I continued on to the same point on the track as I did with the other views). Then at the 9:11 mark I used the Auto Pos again to show the same guy that couldn't drive at the 6:53 mark could drive a little and it was just by changing the FOV.
Before I started this experiment my FOV was 33 and I would move the seat back (no effect on FOV).
Thanks for getting back to me. Just trying to help.

 

[Bobby Pennington]

Even a simulation has to make compromises. I mean, we're all instantly compromised by the fact we lack the feeling of movement and g-forces on our bodies, so we have to make do with FFB through the wheel telling us what the car is doing. FoV falls into the same boat for me. Sure, you can make it exactly 1:1 so you get real perspective, but then you're looking though a tunnel (on a single monitor) and compromising your ability to see to each side... something which is rather important when actually racing other people.

One thing I really like about AC is the ability to move the cockpit camera anywhere, and that it saves per car. I set my FoV to 45, then move the camera back so that it's a foot or so in front of the seat headrest. That and a tiny bit of look-to-apex and I have a good compromise between perspective and situational awareness.



It definitely takes time to get used to, but once you do it can really help with nailing apexes when you're a single monitor user. I don't like racing in sims which don't have this feature now... it feels far too static.

Hey Ross, remember the other day when I was complaining about the Ruf Yellowbird and some other cars I couldn't catch if they stepped out a little? FOV was the issue not my wheel. Once I started getting FOV in the cockpit correct (in relation to my input lag/latency like in the Hood view) I could all of a sudden drive much smoother and didn't lock brakes up nearly as often. Never had a throttle issue because I've driven F1 sims for years with no TC so you learn to feather it. I still can't drift for crap, but I've grown old diving F1 sims and GT cars. I like braking points, apexes, milking the throttle and lap times. BUT if I could drift I would. These kids make it look like childs play and it frustrates the crap out of me soooooooo I just watch them.

 

[Bobby Pennington]

I'm totally disagree with FOV experiments. Why ? Because its a sim, and i want to drive with correct calculated FOV closest to real live view, instead change it to more easy driving

Sorry for my english

Your English is great. My second language is jibberish so hats off to ' ya.
I've always prided myself on using the cockpit instead of hood view because it is a more realistic experience. Same way with FOV. But after years of fighting this (wheel felt to sensitive in AC only, no matter what I changed in the sensitivity settings in Logitcech Profiler unless I took my steering axis sensitivity below 20%).
Then the other day I gave up and tried the hood view out of frustration. Suddenly I could drive smoothly and didn't lock the brakes up as much. It doesn't change the physics, the track or how the car handles. But if we all keep in mind that the only thing that connects us is our hands and eyes and how we use the wheel to react, well that's the point.
Sorry for my English as it's the only language I know lol. Have a great day Denis. Thanks for your input.

 

[Denis E.]

Here is comparsion between FOV 47 (calculated) and 75 (calculated x 1.5) with the same seat position. Looks like i use wrong side of binoculars
I'm done ~ 10-12 laps with each FOVs with Praga R1 at Highlands. And don't see big difference in the lap time. All objects looks smaller and far away from me. Sometimes its maybe helpful for turns, but probably i can't see any crests and dips on the track with FOV 75. My display is 27" and distance ~ 35cm

 

[RacingManiacGT1]

http://www.projectimmersion.com/fov/index.php

I use this to calculate, whats the what is your FOV supposed to be? I guess this doesn't go down to 35cm, at the minimum it'll do(20") its says 37 deg.

 

[Denis E.]

I use this calculator. You need to double your distance and display size. For ex. If you have 27" and distance 35cm, you need to set in calculator 54" and 70cm distance

 

[RacingManiacGT1]

Missing a link?

 

[BhZ]

Obviously changing the FOV, especially if you are not used to the new value, will make you drive very badly for a while. Still, what changes is just your perception. But latency? That's another problem that is not related to the FOV. If you experience latency is either because you have v-sync activated or your monitor/tv sucks.

 

[Denis E.]

Missing a link?

I mean the same from projectimmersion

FOV will be the same for double values
FOV for Xcm distance at X" display = FOV for *2Xcm distance at *2X" display

 

[Kenny Paton]

And in conclusion, like most things in life, there is no one size fits all result

 

[RaceWasGood]

And in conclusion, like most things in life, there is no one size fits all result

Agree. With a single 40" screen on a distance of about 75-80cm i've been using the mathematically calculated FOV (~35 degrees) + ~7 degrees. Yesterday i tried with the mathematically calculated FOV x 1,5 as someone here suggested. This results in ~52,5 degrees and my first impressions based on a short test is that it just might be the sweet spot in my case. Lap-times improved quite a bit and will be sticking to this for now.