Podchaser Logo
Home
220: Hit That Optimize Button

220: Hit That Optimize Button

Released Sunday, 4th February 2024
Good episode? Give it some love!
220: Hit That Optimize Button

220: Hit That Optimize Button

220: Hit That Optimize Button

220: Hit That Optimize Button

Sunday, 4th February 2024
Good episode? Give it some love!
Rate Episode

Episode Transcript

Transcripts are displayed as originally observed. Some content, including advertisements may have changed.

Use Ctrl + F to search

0:00

My internet cord done come loose. Wait, what?

0:02

We were sitting here mic checking and doing

0:04

our, we do a little beep to

0:07

sync this podcast before we start for people who don't

0:09

know. It's how Thrift Heart made

0:11

a thing that beeps with the

0:13

internet times. Our beeps are in

0:15

sync, which makes it easy. Turns out

0:17

for us to sync up the conversation. Yes. Thanks

0:20

to an enterprising member of our community. Anyway,

0:22

I was tilting my head to

0:24

hold the phone up the microphone to beep as we did,

0:27

as we always do. Yeah. The

0:29

fiber is swaying in the breeze. Oh

0:31

no. Have we talked about this before?

0:33

No. Is it hanging? Like, hold

0:36

on. Your window looks out on the street, right?

0:38

Oh no. This window looks out on the fire

0:40

escape. Oh. It looks out on

0:42

the other unit on this floor in this building,

0:45

which used to be my bedroom. I used to live in the other

0:47

unit on this floor. I lived in a one bedroom on this floor.

0:49

Really? Yes. And

0:52

then at some point the couple in this two bedroom that we're

0:54

in vacated. And you just moved down the hall? And

0:56

I had the easiest moving experience of my life and

0:59

about 15 feet down the hall. That's

1:03

incredible. I had no idea. Yes. But

1:06

now this office that I do all my work in here looks

1:08

out on my former bedroom window. Anyway,

1:10

it's a little like cut out fire escape. There's

1:12

a full on ladder and everything. The

1:14

fiber runs into this room. And it's just loose.

1:17

It's not like tacked up to the wall or anything.

1:19

It's always been loose. So when Sonic installed the

1:21

fiber here in 2016, like I was here, you

1:24

know, the guy was just like, Hey, what room do you want

1:26

the fiber to run into? I was like, Oh, the office where

1:28

I work when my computer is would be good. He

1:30

kind of just threw it over the top of the building. Okay.

1:34

And it just runs from

1:36

the roof down the side of two

1:38

floors from the roof. Yeah. Unattached

1:41

to anything. How does that make you feel? Until it terminates

1:43

in the place where he like ran it through, he drilled

1:45

a hole in the wall and ran it through. Wow.

1:48

Anyway, how are you feeling about that? Does that make you

1:50

feel good or not so much? A couple of times I've

1:52

gone out there. I have these little stick on, um, these

1:54

little stick on like plastic things I got from Ono Price

1:56

that are meant to run a zip tie through. It's

1:59

just a. adhesive backing. Yeah, it's like a

2:01

square with a little notch through the top.

2:04

Yeah. I've zip tied it to the

2:06

side of the building with those a couple of times, but every time

2:08

it rains a lot, as it has again recently, those things come right

2:10

off. Oh boy. You know what you need.

2:12

What's that? Nano tape. Well,

2:14

it's a weird shingled, it's like a

2:16

dappled or ridged kind of shingle siding

2:19

on the building, so I don't think nano tape would

2:21

stick very well. But like all your

2:24

common household adhesive things are also not going to stand

2:26

up to the elements. You need one of those 3M

2:28

goos, like one of the ones that's like a permanent,

2:30

it's like a gray kind of squishy thing that as

2:32

soon as it gets exposed to the air, it starts

2:34

to harden up and you just jam it on there.

2:36

It's going to glue on forever. Do you think I

2:38

need to ask the landlord before I do that? He's

2:40

going to tell you no. I know. So

2:43

I shouldn't ask. This is a forgiveness situation, not

2:45

a permission, I think. I think you're right. Or,

2:47

you know, maybe that's, maybe that cable

2:49

is very resilient. It's very thin. It's

2:52

a very thin, skinny, seemingly brittle cable.

2:54

I mean, it's fiber. Of course it's

2:56

brittle. Like, I don't

2:58

like my internet flying around in the wind. I

3:00

mean, the scary thing about this seems to like

3:02

if there's roofers or something that come in to

3:04

work on something else unrelated to the fiber, they

3:07

could very easily be the modern

3:09

equivalent of a backhoe on the May

3:11

West pipeline, May's pipeline shutting

3:13

down internet for two thirds of the country.

3:15

Yeah. Yeah. This is

3:18

a precarious situation. Actually, thinking back to the install the day he

3:20

got here, I had to wait like, I

3:22

don't know, a month or something. Those were fiber. Fiber was new

3:24

in the city at that time. Yeah, you were the first person

3:26

I knew with fiber, I think. You remember when I got fiber,

3:28

how excited was it? It was a,

3:30

look, I was like, hey, can I come

3:32

over to your house and try it? And you were like, no, you

3:35

can't come over. I'm sorry. That did not

3:37

happen for the record. Yeah, no, that did that.

3:39

No, I didn't ask. You didn't deny. That conversation

3:41

never took place. The

3:44

day he installed it, I had to wait like a month. And

3:46

I was actually about to go out of the country then

3:48

for like two days later. Is this a giant bomb already

3:50

or was this before that? Yeah, it was 2016. Okay.

3:53

Not only was I about to leave the country for

3:55

a week, also as the detail left out, they were

3:58

repaving the street that day. And

4:00

they were literally at the end of the block, like we're about

4:02

to start working their way up. And he was like, Hey, by

4:04

the way, I just talked to these road guys. Like they say,

4:06

I have to be out of here in like 90 minutes. Oh,

4:09

wow. And if I don't get this done before they tell

4:11

me I have to leave, you have

4:14

to go to the back of the queue for installs.

4:16

What? He couldn't just park on another block, which is

4:18

like several months long at this point. I think he

4:20

is. I think his truck had one of those like

4:22

the extending arm, you know, bucket or something. Yeah, because

4:24

he had to go to the pole. Got it. He

4:26

wasn't just using ladders. Oh, wow. He needed his truck

4:28

to do the install. And essentially

4:31

he was like, Hey, if I don't get this installed

4:33

in the next hour or so, you're going to have

4:35

to wait another three or four months for this. So

4:37

like, please do what you can to get this done.

4:39

It goes fast as you can, man. So I'll stay

4:41

out of your way and tipped him heavily. But

4:44

I think that may be why the line is

4:46

unsecured on top of the building. I feel like

4:48

I feel like having a real time pressure on

4:50

that is going to add some tension and stress.

4:52

It is not necessary, you know, but it was

4:54

very exciting when he said he came in and

4:56

he was like, All right, you're good to go.

4:58

But now I have the situation. You could just tack

5:00

it on there. Wait until the landlord's out. Stick a

5:03

thumbtack through. Get some nails. No, don't stick it. I'm

5:05

joking. Get the little loops. You get the half loops

5:07

with the nail on the side. Just bop, bop, bop,

5:09

three of them in. You're good to go. That's how

5:11

they did my fiber to the wall. I don't know

5:13

what the siding is made out of, but sticking nails in

5:15

it seems like a bad idea. Yeah,

5:18

it's probably fine. Like, like what's the worst

5:20

thing that could happen? Some Looney Tunes nonsense

5:22

where like the coyote gets the fiber hooked

5:24

up to him and he runs away and

5:26

like pulls everybody on the streets fiber out

5:28

of their houses. Oh, my God. I'll I'll

5:30

be right back. Welcome

6:02

to Brad and Will Made a Tech Pod. I'm Will.

6:05

I'm Brad. It is a rainy

6:08

Saturday here in the Bay

6:10

Area. It is. Man, it's been pouring.

6:12

I think I'm done... I think I... Like,

6:15

having spent some time in the desert in the summer and

6:17

seeing what real storms are like there, I don't really like

6:19

that. But I do like being warm, it turns out. Warm

6:21

is nice. I don't... I

6:23

don't think... I think I'm tired of living

6:26

someplace where it's this kind of miserable and wet all

6:28

the time. Well, not all the time. Let's

6:31

be specific here. Only eight months a year.

6:34

We're in the rainy season right now. You know what, frankly,

6:36

I'm happy the rainy season is back because we didn't have

6:38

it for some years. Look, this

6:41

is a hot take. I don't know that I'm as big

6:43

a fan of the rainy season as you are, Brad. Well,

6:45

I don't know if I'd say I'm a fan of it,

6:47

but it's a necessary evil. Look, I

6:49

like the state not burning and I like having

6:51

water. Not so much a

6:53

fan of the 90 days of

6:55

rain or whatever it is we're in right now.

6:57

I mean, you know what I like though is

6:59

all the parks around here are absolutely gorgeous after

7:02

this much rain because they're usually brown and dead

7:04

and desiccated. Have you gone walking in

7:06

one though? Because my... Oh yeah. They're

7:08

like my backyard and front yard. They're like two inches deep

7:10

of mud right now. Well, I mean once it dries, you

7:12

know, I get a few days. But like I just... I

7:15

walked by Golden Gate Park the other day and it is

7:17

an explosion of green right now. I

7:19

had to wheel my trash cans back in last night after

7:21

another inch of rain and I was like,

7:24

I think I have to lift up the cans

7:26

so they don't dig ruts in my yard. And

7:29

like anyway... This might be one

7:31

of those cases where not being a homeowner makes this

7:33

a little more palatable. Yeah. Like

7:37

there's a lot of stress that comes with... Just

7:39

to be clear, we're not talking about like

7:42

baby rain sprinkles or anything like that. We're

7:45

getting like one or two inches of rain per storm,

7:47

which sometimes is like a day. Sometimes it's like three.

7:49

Sometimes it's up to four inches. I think this next

7:51

one is supposed to be worse than the last one.

7:54

So it's like... It's gonna rain like all week. I

7:56

think I just saw. Isn't it here? Yeah. And

7:59

most of the dirt here does not absorb. water at that like

8:01

that kind of water like we have we

8:03

have like my house is on basically clay

8:05

so it it absorbs for a little bit

8:07

and then it just gets saturated and everything

8:10

turns into a swamp so it's

8:12

it's it's a it's a lot yeah it

8:15

is but at least at least theoretically maybe

8:17

the state won't burn again this fall so

8:19

that's good yes let's uh hang

8:22

on I got a segue you know what

8:24

else is a lot PC

8:26

game settings huh yeah

8:28

we had a discord user OS to

8:31

X or OS 10 you think it's

8:33

it's capital O lowercase sx is that

8:35

Oz X so yeah you could you

8:37

could actually be known to pronounce that

8:39

medically I'd certainly I miss

8:41

the days when we got to say OS 10

8:44

I hate I hate this transition

8:46

to back to Mac OS you just bring

8:48

that bring that energy over to 10 the

8:50

successor to Twitter god

8:52

damn it um OS X

8:54

said I would

8:56

love an episode about tuning PC games advanced graphic

8:58

settings some sample topics we come to mind our

9:01

DLSS 2 & 3 FSR 1 2 & 3

9:05

X ESS what does the sharpness lighter

9:07

do under these options do I need

9:09

to manually lower the in-game resolution for

9:11

them to upscale correctly what is SSAO

9:13

how do various variable refresh rate technologies

9:15

interact with v-sync and do I need

9:17

to turn it off v-sync I assume

9:19

in each game how can I verify

9:21

it's working what if I have multiple

9:23

monitors this is a good question

9:25

it turns out well OS X I'm glad you

9:27

asked we we have

9:29

been digging in and I've done

9:31

we've done a lot of reading I'm gonna say

9:33

and watching videos and and a lot of stuff

9:36

turns out PC game settings

9:38

kind of complicated most of the time

9:40

yeah yeah but but knowable I would

9:42

say they're not they're not they're not

9:44

so arcane that we can't unpack them

9:46

with full research yeah and I mean

9:48

the thing to remember and

9:50

like we like to peel the onion here and start

9:52

with the easy part first and go to the complicated

9:54

stuff later is that if you don't really care all

9:57

that much you can just load up the GeForce experience

9:59

or radio on settings app on your computer

10:01

and hit the button that says, make this pretty good.

10:04

Like hit the optimize button. And

10:06

it'll give you like, it'll usually

10:08

what they do is they use

10:11

the profiling data they have from testing games

10:14

to look at what the hardware is in your

10:16

system and they find a spec, a comp, the

10:19

closest comp in their database. And

10:21

then they put you settings that'll hit run

10:23

you at 60 frames a second. It is

10:26

the general guideline. So

10:28

I do that for

10:30

almost everything, except

10:33

for competitive shooters and

10:35

stuff with lots of ray tracing settings. See, this is

10:37

the difference between you and me. Look

10:40

I have a kid and like two and a half jobs.

10:43

I don't have time to think around. If I want to

10:45

ever play games, I have to just get into it. I

10:47

get it. Yes. I

10:49

mean, I also have like two and a half jobs. I don't have a kid though, to be fair.

10:53

I don't install Geforce experience. Why

10:55

not Brad? I, I just

10:57

don't feel like I need it. Well,

10:59

actually to be more specific, there are too many other overlays

11:01

already and I know you can disable parts. I just turn

11:03

off the overlay. Yeah. Like I know you can

11:05

turn parts of that off, but like there's

11:08

already the steam overlay. There's already game bar. I

11:10

wish you could turn off game bar. Game bar is

11:12

way better than it used to be. Have you used it lately? Yeah.

11:15

The little, I've used when we were doing profiling stuff on

11:17

the end increases, I used it because that overlay with the

11:20

frame rate data was pretty useful. Yeah. The

11:22

thing, honestly, the thing I like about Geforce experience and

11:24

sometimes if I'm troubleshooting perf stuff, I will

11:26

turn on the overlay because they have the 1%

11:28

frame rate. They

11:31

met, they report the frames

11:34

per second of the slowest 1% of

11:36

frames in real time. They try real time. The

11:38

1% lows. That's pretty useful, which

11:40

is maybe the most useful like real time

11:42

feature ever. It lets you know when you're

11:44

having a bad starter. Anyway, the

11:46

point is if you don't

11:49

care about this stuff, you can just skip

11:51

this entire episode basically and mash the optimized

11:53

button in Radeon settings or Geforce experience and

11:55

get something that's pretty good. Intel

11:58

has been leaning in this direction more too. I'm

12:00

not sure what they've got for their art cards

12:02

for their for their GPU stuff, but they rolled

12:04

out a tool for their CPUs What

12:06

is it? The name is escaping

12:08

me. It's more for

12:10

tuning CPU stuff. I want to say

12:12

I don't want to have to tune CPU

12:14

stuff to Brad. I just want to like start the game and

12:17

play the game and not think About this stuff. That's

12:19

my dream Application optimizer. I believe is the

12:21

name of it. I was trying to figure out what the

12:24

P stood for but it's one of those things where the

12:26

a and the P came from the same word. Oh Anyway,

12:30

this is getting a little tangential to the actual

12:32

topic of this episode, but like CPU

12:34

tweaking has gotten really interesting I don't know if

12:36

you've paid much attention to what Intel named you're both

12:38

doing where They both have

12:40

ways of identifying like which cores are the good

12:43

cores on your specific chip Yeah,

12:45

you know because just straight up variants

12:47

and manufacturing like silicon lottery style Everybody's cores

12:49

are going to be of different quality. So

12:51

it'll like actually tell you like, okay These

12:54

are the cores on your CPU that actually run the

12:56

fastest and it will will buy us toward those type

12:58

stuff Maybe I need to do run that on my

13:00

computer because I have to clock down to 52 X

13:02

on my 13 900 K To

13:06

get fortnight to not crash a lot

13:08

basically because of the memory corruption stuff

13:11

Anyway, so I'll say real quick before we

13:13

move back to graphics at launch APO was

13:15

just for 14th gen Okay,

13:17

but this headline I have just pulled from two weeks

13:19

ago. It makes it sound like they are going to

13:23

Roll it back roll it back at least at least to some

13:25

extent to 12 and 13 Same

13:28

socket anyway, yes So

13:30

yeah, the but from here on we're gonna go

13:32

down like we're gonna answer these questions from Oz

13:35

X or OS X. Yeah we

13:38

also went through and pulled some screenshots

13:40

from Kelly current

13:42

games cyberpunk Alan Wake fortnight The

13:46

Prince of Persia ones were comically thin I think I

13:48

think there's a couple others in there But and we'll

13:50

kind of go down the settings and explain what is

13:52

the perf? like what what stuff is

13:54

gonna really hit your perp what stuff doesn't matter as much and

13:58

Explain what some of the stuff means Yes. Yeah,

14:00

I threw Baldur's Gate 3 in there. It was

14:02

nice, I didn't see that. And the finals and

14:05

stuff, there was a bunch in there. But we

14:07

will, I will, people seem to like it a

14:09

couple weeks ago when we published the Google

14:12

Doc that we used for the show. Yeah.

14:14

So I will do that again. We have all those

14:16

images embedded in there. I hate the way the Google

14:18

Docs handles image embeds, which

14:20

is to say that it's

14:23

kind of hard. It's not easy to get

14:25

the full-size version of the images out. Hold

14:27

down Control. And then mouse

14:30

forward, mouse wheel forward. And then take a screenshot

14:32

of the document. That's how you get it out

14:34

of there. Yes. Well, so actually, have you ever

14:36

used this Google Docs Publish to Web feature? It's

14:38

really nice. That's what I used when we did

14:41

this a couple weeks ago. And

14:43

it just makes an HTML, like an HTML with an

14:45

image holder or something? Yes. When you go to Share

14:47

now, there's an option. Instead of sharing a link to

14:49

the Doc, you can hit Publish to Web. And it'll

14:51

just, yeah, like you said, it'll make a foot it.

14:54

It generates a link. Although, I should

14:56

probably make a short URL What's your

14:58

favorite URL shortener these days? Content.town.

15:02

Oh. I haven't actually

15:04

done it yet. But I've been looking at things I can

15:06

put on there that will generate short URLs.

15:08

That would be cool. We should do that. That

15:10

would be a good one. Anyway, so I mean,

15:12

you can scale that HTML-based version that we will

15:15

link in the show notes as well to see

15:17

the screenshots. It's a little hard to re-align the

15:19

settings up to default image size. But yeah.

15:22

Or you can zoom in also if you don't care. That's

15:24

what I mean. The Control mouse wheel up and

15:27

down tip applies there as well. So

15:29

let's get into it. The first thing, the

15:32

big one these days, this is a hot

15:34

button topic that everybody cares about because the

15:36

graphics vendors are in a

15:38

full ass war over super sampling

15:40

these days. Super

15:43

sampling is the idea that you take,

15:46

it used to be the idea that you render, take

15:48

the resolution that you're running and then

15:50

you render it up at like,

15:52

so say you want to run at 1080p, you

15:55

used to run it at 4K and

15:58

then you'd take each four. Every

16:00

four pixels on on the

16:02

on the the forte version

16:04

would get combine like mathematically

16:06

averaged. And you'd get. A.

16:08

Teddy P. image that as it has like smoother

16:10

edges and all the kind of the benefits of

16:12

super sampling. Zola we that's not what it does

16:14

anymore. It turns out that he would do that.

16:16

And games are You have a lot of have

16:18

room to spare like does me too For example

16:20

you could ease you do that on like a

16:23

ten even when the game as models that you

16:25

know it's it's a bit older will have a

16:27

as one lightweight enough with the are you and

16:29

and that obvious of that results in a much

16:31

cleaner image. but also yeah. Now that

16:33

you mentioned that is what Super Sampling

16:35

used to be stopped moving undiluted. Technically,

16:37

what we're doing now which is the

16:39

reverse direction still counts as super sampling,

16:41

but whatever. So that the thing that

16:43

the reason they call super sampling is

16:45

because what they're doing is they're generating

16:48

the four pixels with machine learning now

16:50

and damn despair and then dividing the

16:52

than combining them without doing their job.

16:54

Upsize render anyway. I'm dissing. That

16:56

doesn't mean I used to sell for deal us

16:58

as these at this is deal of a deal

17:00

Ss and Fs are and X E S s

17:02

on. And D M D

17:04

and Intel respectively. And and the thing

17:06

that they do as they're using these

17:09

two upscale in a way that doesn't

17:11

that adds more information. New get was

17:13

just straight up scaling. I'm.

17:16

As he sees him basically render the game

17:18

at a lower resolution, upscale it using these

17:20

technologies, and then get the benefit of the

17:22

higher resolution. Most of the benefit the higher

17:24

resolution at the frame rate of the lower

17:26

resolution. Guess and this is is me. This

17:28

this is the prototypical example of the good

17:30

use of Ai her machine where again I've

17:32

talked about recently where it's this is a

17:34

signal processing kind of a ivory taking an

17:37

input. And doing something to with

17:39

and you can do as well without the I

17:41

like this is this is the good type of

17:43

a I use a jaunt the mop up a

17:45

gentleman's agreement both centered of but it's not pal

17:47

hundred of yeah well time guess is not in

17:50

your eyes he generate yes you're you're right. it

17:52

is technically interpreted as generating new image rebirth of

17:54

it's it's free. North of the dataset a trade

17:56

off of is just footage of Games organ data

17:58

on games yeah to running. Different

18:00

think I'm so. The

18:03

upshot is that that.

18:06

Is three different vet There's three vendors

18:08

posts there's that is three vendor supported

18:10

options. couldn't deal, assesses in video is

18:12

actually assesses Intel and end our Ssr

18:15

as a D is generally speaking you're

18:17

better off using the one that goes

18:19

with your card. If that's an option

18:21

of ya need video card use deal

18:23

us as if you have an A

18:26

D card Usaf has are if you

18:28

have an Intel card use use Fcs

18:30

ask the The reason for that is

18:32

that on the ones that are Ml

18:34

accelerated. That. Use did dead at

18:36

that are the deal as as an

18:39

ex Css use the dedicated machine learning

18:41

acceleration hardware on those cards to that

18:43

means they don't take resources away from

18:45

the graphics rendering as much as said

18:48

say like running our computer chair would

18:50

rate. It

18:53

doesn't seem like that impacts Sci efforts are as

18:55

much because it just uses computers on all hardware.

18:57

Yeah, Was a D tends to like you're like

18:59

platform agnostic or spur features, which I I appreciate,

19:02

but there is a negative. The negative is that

19:04

you're taking resources away from the game as the

19:06

games Friday's sentences history. On the other yeah that

19:08

that's why they've been out there. Him a ballyhoo

19:11

in like a of Us are works on all

19:13

hardware is because it runs in a generic enough

19:15

way that the Us when pretty much where you

19:17

just said and sound like as or Three which

19:20

is not that old. Might.

19:22

Have some very the like light hooks and some

19:24

seen wearing. More from from

19:26

what I've what round out but but by the with your

19:28

point is that way game These. Machine

19:30

whirring hardware efforts are just kind of behind

19:32

it day. They have not done a bunch

19:35

of platform or vendor specific stuff. It seems

19:37

like they're the I think I like their.

19:39

I think they're acquisition of Xilinx several years

19:41

ago is what was intended to address their.

19:45

Oh they're They're being behind the May I race so

19:47

and in fact I think they've made a bunch of

19:49

announcements about like the next raisins and the next. Everything

19:51

they're doing is gonna have a bunch of. A

19:54

bunch of Ai hardware in it, but there are quite

19:56

there yet it's it's own. Seems like they're pretty well

19:58

on the past. the. Yeah. So

20:00

so anyway the upshot is used the when

20:03

you have if you don't have one. I

20:05

think it's. Efforts or Two

20:07

and three Year or two is a

20:09

subset of three. So too does the

20:12

that the physically for acting of resolutions

20:14

in the same way that deal as

20:16

as does I'm one of us or

20:19

one it has terrible image quality problems

20:21

that I don't think anybody's do so

20:23

many circumstances they did some weird stuff,

20:26

it just didn't work particularly well. I'm

20:28

too much better. It's odd it with

20:30

all of almost all of these things

20:33

when you look at like the digital

20:35

foundry like a. Comparison in games

20:37

his support both. They. Both

20:39

fail in different ways of vote. Both

20:41

fail in interesting ways that you probably

20:44

not going to notice in an actual

20:46

game. So for example, deal as as

20:48

is better at polling overlay. stuff like

20:50

text rendered on the on the on

20:52

the game world. For

20:55

fire like they're kind of does to

20:57

d textures that make up most fire

20:59

Vfx stuff like that I'm it does

21:01

will be better with that. Efforts

21:03

are does better with most of so it like. Now.

21:07

Generally speaking use the when the your hardware support

21:09

see if it video to advance a deal as

21:11

as as feel as as only works on into

21:13

the hardware now so. I'm and we've

21:15

gotten very good and also if we should say

21:18

as avoids the always as three point five and

21:20

is extremely capable of yeah they they do they

21:22

do me a to I don't I can't see

21:24

as much named he succeeded Sport of us are

21:26

on and a crisis that with deal us as

21:28

they look at what you implement and they're like

21:30

hey we noticed some ghosting year or something like

21:33

that and they'll say you know if you if

21:35

your hair is ghosting if your hair is rendered

21:37

in this way you may want to use this

21:39

version of the model and said this version the

21:41

model and they'll even bake stuff our special for

21:43

you if if you did if he. Kneeled Vector

21:45

help. Now. I'm I'm not gonna point

21:48

as I have agreed with my mother's for reporting

21:50

white. I've really much always turn it on, even

21:52

if it's a game like. I'm quite

21:54

sure I could run a neat rows response. Yet

21:57

so I have a three hundred sixty Hertz displays. I always have

21:59

a turn. Like definitely have to

22:01

smoke A But also I often don't run

22:03

games at the full frame rate because it's

22:05

It's nice to have it running slower Said:

22:08

doesn't spend up all the fans and the

22:10

computer. Sure, I'm. Anyway, So.

22:13

Did the two things that happened recently.

22:15

Yards with Deal as a Three Point

22:17

Five and Diaz and efforts are three

22:19

different generations. so three of us are

22:21

three is a super set of to.

22:23

It's basically to plus frame generation feel.

22:25

As for as Three Point Five is

22:27

is three with frame generation I think

22:29

the Deal Us has three south only

22:31

worth three and new are only works

22:33

on the Ada. Four. Thousand Series

22:35

cards think that's right. Yes, I believe

22:38

Frame Generation we runs on. For.

22:40

That it frames and reasons part of three

22:42

point know. If. I'm honestly I

22:44

really was better than Three Point Five subheads

22:46

the path, tracing all that spirit, retracing re

22:48

reconstruction amory, every reconstruction of a so. This.

22:52

Is where he gets complicated because if you

22:54

have a to series card or three series

22:56

card from in video you might want to

22:58

framed generation that comes with Apis Arts, Three.

23:01

Which is not supported on your card. on on

23:03

the deal as a side. I'm.

23:06

The the weird thing about. Frame.

23:09

Generation is you want to

23:11

have on both sides. A.

23:14

Sixty. Frames per second Render.

23:17

After. the of scaling so

23:19

after the deal us as to an

23:21

end of us are to stuff happens

23:23

before you start doing the frame generations

23:25

otherwise you get some weird aren't affecting

23:27

that. That is pretty noticeable in my

23:29

experience. so this isn't gonna have a

23:31

bull. It can effectively

23:33

double your frame rate in some

23:35

cases. But. You

23:37

have to already be running at sixty frames

23:40

per second in order for that to work.

23:42

So years this is. This is basically for

23:44

getting high refresh rates on I read monitors

23:46

on lower specs cards Now. I'm

23:48

due due to do you use fringe a reason

23:50

as a matter of course. very much so as

23:53

to say the problem with frame generation is that

23:55

increases input latency. Yeah, that's that's my question to

23:57

buy. I've only uses enough to I try it.

24:00

This five minutes of time into for

24:02

games like I don't, I don't, I

24:04

haven't you cycling? I seriously play a

24:06

game and so wonder what the perceptual

24:08

the fact is over time as you

24:10

get used to it or not. and

24:12

also what? What's I got How babbling

24:14

the actually is. So digital Digital Foundry

24:16

day. I didn't do that. Messing about

24:18

this earlier last year. I guess I

24:20

made last year when the suffers new

24:22

and and they're they're that. the the

24:24

way that both aimed in and video.

24:26

I try to counter the though latency though

24:28

it though the basically it's input to pixel

24:31

the lag as or talk about the times

24:33

when you make make your input choice and

24:35

know though change registries on screen so for

24:37

example pulling the trigger and the gun shooting.

24:40

I'm. It they they use

24:42

and buddies reflex i am d's

24:44

thing called anti lag that are

24:46

basically the same thing that that

24:48

and that basically they weren't they

24:50

to. Frames. So

24:52

that you don't render too many frames in

24:55

advance, which helps minimize that lag. Now.

24:58

With those technologies are with reflects on

25:00

and with anti lag on this it's

25:03

pretty middle was like seventeen set, seventy

25:05

millisecond, sorts of almost negligible. a pretty

25:07

much impossible with sex for most people.

25:10

I. Don't use for a i wouldn't you

25:12

stream generation anything. That's that's competitive. I'm.

25:16

I probably also wouldn't use frame generation of

25:18

i play on mouse and keyboard in my

25:20

experience to like when I was playing cyber

25:22

punk on mouse and keyboard didn't feel great

25:24

when I space ever buchan the game had

25:26

didn't notice a difference so he had kind

25:29

of silliness of the the the direct. My

25:32

expectation with a mouse is that there's.

25:35

Be a pixel. The pixel a

25:37

precision on a mouse movements. And

25:39

on the game pad I'm used to feeling a

25:42

little flowed year which didn't. Didn't

25:44

really impact my experience much. I'm

25:46

to the so that's that's the deal that's

25:48

the super sampling. Like. That's it. Super

25:51

sampling going to make your performance better. It.

25:53

Is it's with it's a weird choice for name's

25:55

super simple used to make your games or and

25:58

slow as l. So. I

26:01

generally, if I'm playing at 1080p and I'm playing a

26:04

multiplayer game, I set it to the performance mode, we

26:06

should explain what this means. So

26:08

I assume, that's a question from OSX here, I

26:10

assume by slider they're referring to the, and actually

26:12

I think some games do depict it as a

26:15

slider, the DLSS presets, the kind of

26:18

named presets, like quality, balance,

26:20

performance, ultra performance, what do

26:22

those mean? And those presets are

26:24

actually defined by Nvidia, they ask you to,

26:26

when you set up your settings

26:28

in your game, they ask you to make sure that they align

26:31

with what their, what players

26:33

expectations are, so that they're the same across

26:35

the board. Yeah, so those basically have specific

26:37

meanings for what size of image is being

26:39

rendered under the hood before it is then

26:41

upsampled to your output resolution, for example 4K?

26:44

Yeah, so on ultra performance you're upscaling from 7,

26:47

assuming you're running at 4K, ultra performance

26:50

you're upscaling from 720p to 4K,

26:52

which means you're rendering 11% of the pixels

26:54

on the screen for real, and then making the

26:56

other 90, 89% with DLSS. Yeah,

27:00

and in my experience if you've tried ultra performance you

27:03

can see that. Yeah, ultra

27:05

performance usually feels pretty muddy.

27:07

As you can imagine with

27:09

barely over a tenth of the actual pixel data to

27:12

work from, like you're, there's only so much you can

27:14

do. Yeah, a performance mode is upscaling

27:16

from 1080p to 4K, which means you're rendering 25% of

27:18

the pixels and generating the other 75%, and balanced is

27:20

from 1253p, which I didn't actually

27:27

put in the spreadsheet, so I don't know what the percentage is. 33%

27:30

I believe? That

27:32

seems right, split the difference. And

27:35

then quality is 44% of

27:37

pixels, it's rendering from 1440p up to 4K. Yes.

27:42

Or, you know, maybe an easier way

27:44

to think about that is, again, this

27:46

is all assuming you're outputting a 4K,

27:48

this is all trickled down from what

27:50

output resolution you're talking about, but at

27:52

4K, for example, quality means

27:54

you're doing a 1440p under the performance So

28:00

I believe performance is the one that would be

28:02

told to 1080p and I don't I don't have

28:04

the settings for 1080p or 1440p

28:06

in front of you, but I believe that the resolutions I

28:09

Don't I don't believe it's the same percentages across

28:11

the board. Oh, I thought

28:13

the percentages were fixed. I Think

28:16

that's right. Yeah, cuz like the whole Alan way to

28:18

controversy is the thing you saw people screaming Was

28:21

that they said at? at

28:24

1080p Performance deal says performance

28:26

you would get all the law 30 frames or whatever

28:28

it was But I swear that's where you saw the

28:30

Jillian people running around screaming like oh, so you can

28:32

only render at 540p. What the hell? So

28:36

that would make sense. Yeah, so I think I think I think

28:39

The percentage percentages are fixed on these

28:41

presets and that's all just the LSS

28:43

obviously Until an AMD are going

28:46

to have different different different

28:48

things going on. Yeah, the thing I'll

28:50

say is as a 1080p player Ultra

28:53

performance always looks real bad. I assume it's

28:55

because you're starting with like like 1990s resolutions

28:59

And and going up from there. Yeah Quality

29:02

on 1080p is great.

29:05

I run quality a lot. Anyway, um,

29:08

Let's see What other questions from OS X was

29:10

do you do I need to manually lower the

29:12

in-game resolution for them to upscale? No, you set

29:14

the resolution that you want generally you should set

29:16

the resolution for what your monitor is hard hard

29:18

Like the resolution of your monitor. Yeah native native

29:23

And as you said a minute ago Brad a

29:25

lot of games will actually show you what the

29:28

internal rendering res is I Think

29:32

X CSS lets you set the internal rendering

29:34

resolution rather than a quality profile or percentage

29:36

interesting That might be something that is

29:38

a per game Setting though cuz I've

29:40

seen it both ways. I think on X CSS. Okay,

29:43

so even for DLSS I'll just mention Alan wait to

29:45

is as an outlier here where they don't use the

29:47

names Of those precepts.

29:49

Well, just they just show they let you

29:51

pick the internal render resolution Okay,

29:53

but without the game running in front of me I can't

29:56

say for sure but I think I think all you're doing

29:58

is picking those resolutions that would correspond to those presets

30:00

to begin with. So

30:02

that is a case where the answer in

30:05

this one specific case, the answer is yes, you do

30:07

have to pick the render resolution in Alan Wake 2.

30:09

But again, that's I think it's just, it's

30:12

just mapping to those presets without them

30:14

using the names of the presets. Anyway,

30:17

also, sorry, I forgot to mention a

30:19

lot of at least with the LSS, a lot of the

30:21

preset selection now just lets you just set it to auto,

30:24

which I assume is just so is just

30:26

targeting some performance target and it maybe is adjusting

30:28

it on the fly to try to maintain frame

30:30

rate. That

30:34

would make sense, I think. Yeah.

30:36

Okay, so the

30:39

next one is screen space is SSAO.

30:41

What is SSAO? It stands for screen

30:43

space ambient occlusion. Yes. And

30:45

it's a post, it's a post process effect. So it

30:47

happens after the scene has been rendered. And

30:50

it adds things like reflections

30:52

and shadows,

30:55

fuzzy shadows around the edges of things. Yeah.

30:57

So yeah, ambient occlusion basically is just a

30:59

bunch of fine shadowing and the cancranis anywhere

31:02

that like the ambient light

31:04

would be occluded by objects,

31:06

you know, so you got like a bench or whatever, there

31:09

should be little bits of shadow right behind where the

31:12

the seat and the railing and stuff would be occluding

31:14

the light. Screen

31:16

space effects, like you mentioned, like ambient occlusion

31:18

and reflection, reflections, rely

31:21

on data that is being generated as a

31:24

part of rendering what's already in frame.

31:26

Yeah. So if you've ever played a game where

31:29

you like sweep the camera around and you've got something

31:31

like a big pond or like glass that's reflecting stuff.

31:34

And as the thing that's being reflected moves out

31:36

of frame, you see those reflections vanishing at the

31:39

same time. But that's a screen space

31:41

effect that that is, that is essentially like the

31:43

engine no longer has the data it needs to

31:45

render those reflections because it's not rendering the source

31:47

objects in first place. It's all just like a

31:50

shortcut, like the screen space effects are just a

31:52

shortcut to use data that already exists per

31:54

frame to create additional effects without

31:56

because otherwise it would be like way,

31:59

Way, way too. Computationally expensive to

32:01

calculate those from scratch, so

32:03

sometimes you over render the

32:05

the asshole. I'm buggy over

32:07

under the actual screen space suit. Actual thing

32:10

that's been rendered is larger than sexually been

32:12

shown. Source of the you avoid that kind

32:14

of stuff. Yeah, Ah, But you

32:16

see it more on councils were they don't have as much

32:18

performance. Had room to work with. Get

32:20

And and I'm. The

32:22

needy about ambient occlusion. Sweet. cheaper than doing

32:25

it for real. Like. This is this is

32:27

a lot of the. Most. What we're

32:29

talking about say place to rest are games

32:31

are games that use a traditional the K

32:33

We're gonna have to draw a bunch of

32:35

texture is mapped over wire frames and then

32:37

we're going to put some other textures that

32:39

are serving like maps that indicate the light

32:41

in the dark areas and was rather textures

32:44

that include indicate like whether the surfaces bumpy

32:46

are smooth and and all the stuff we

32:48

were written however a tracy a little but

32:50

that changes. Like

32:52

right now we're using mostly kind of combined

32:54

rastering rate racing games. It's

32:57

is still the rate. Racing is changing. Basically

32:59

everything they were talking about here Kislyak as

33:01

a sale doesn't apply. For. Ray Tracing.

33:03

yeah yeah yeah, maybe seclusion am dumb. Reflections

33:05

are to places where are you don't need

33:08

the screen space versions of those effects anymore

33:10

because retracing just does those inherently. yeah and

33:12

edits also would say as a seo his

33:14

memory intense and expensive but also it's like

33:16

if you're having per problem is it's a

33:19

it's not a terrible thing to turn them

33:21

down although I actually only one ear abu

33:23

I don't know what I decided that that's

33:25

and like a full past racing kind of

33:27

scenario. You get all that stuff like there

33:30

are games that use race racing and more

33:32

winded ways. That would not cover those use

33:34

cases. Of Okay so

33:36

next he is very of a variable refresh rate.

33:38

This is confusing. This was confusing even for us

33:40

as we were researching this episode. Yeah I learn

33:42

stuff today. Yes when I when I built as

33:44

machine last year I have figured this out years

33:46

ago in the promptly forgot it as you often

33:48

do when you figure out some obscure settings friends

33:50

at them once, rape or into thats good. Good

33:52

for your brain to does let us have fall

33:54

away and it'll It'll change before the next time

33:56

you need a problem life with. of The one

33:58

thing I can say is definitely. Never document

34:00

anything to do. Know that guy. documentation for

34:02

Trump's is my understanding. What would that accomplish

34:04

by the way. Other so this is relatively

34:06

fresh and mind because I had a singer

34:08

the stuff out again. Last. Year This

34:11

is only role inferred from video is why. Been

34:13

using for years. I can speak to. Him

34:15

the an Intel here but the that question is.

34:18

Ah, I'm. Okay so it had a

34:20

variable refresh rate technologies and rak of the

34:22

think the first question do I need to

34:24

turn it off in each game saying sorry

34:26

I'm just a just generally speaking Blurb a

34:28

series is a great resource for this because

34:30

they do a ton of like hands on

34:32

research and aim high speed cameras. it monitors

34:34

and do all sorts of crazy stuff with

34:36

all the available hardware. So if you if

34:38

you have questions about a setup that we

34:40

are going talk about here today. Blur busters

34:43

is the is the is the number one

34:45

resource I think death. So the

34:47

the combo settings I have found on a

34:49

video that will consistently good variable refresh to

34:51

work is in the Nvidia control panel. Are

34:54

you wants you to rongi seek silicon? So there's

34:56

a setting under three settings called monitor technology. Which.

34:59

Will say D C Orgy scene compatible as

35:01

you'll have an actual dissing monitors. it's like

35:03

a freezing monorails. I compatible the at any

35:05

rate, You. Wanna turn on the adaptable

35:07

adaptive. Refreshed, Technology under

35:09

Your technology. And. Then you are

35:11

force to seen to on in the driver control

35:13

panel. And then thereafter you

35:15

always want to turn this thing off. In.

35:18

Every game that you're playing. Yeah because been

35:20

when you're having these the driver uses club.

35:22

he lets his overriding the games behavior with

35:25

the to her her global. Driver:

35:27

V sync and as the only way I

35:29

found to very consistently good variable refresh to

35:31

work and everything. This is one of the

35:33

benefits of using G Force Experience for the

35:35

south as it does. Does this email from

35:37

worry about it? ah in the in games

35:39

going wrong why we had worry about until

35:41

the third asked oh you do not buy

35:43

me need I? You don't have to go

35:45

into each game set and chaser so rice

35:47

or officer and yeah to turn off you

35:49

think I'm the the blurb us His explanation

35:51

of this is it the T Sinks ability

35:53

to fall back on us discs Refresh rate

35:55

V Sync behavior changed in. twenty fifteen when

35:58

they expose the option to to do

36:00

this. Right

36:02

now, because you see

36:04

both, people think that G-Sync and V-Sync are

36:06

separate options. But when G-Sync is enabled in

36:08

the control panel, the vertical sync option in

36:10

the control panel no longer acts as V-Sync

36:13

and actually dictates whether the G-Sync

36:16

module compensates for frame rate variances

36:18

by the system, which prevents

36:20

tearing. And

36:22

that's if G-Sync and V-Sync are off. If

36:25

G-Sync is on and V-Sync is off, this

36:27

is really confusing. This is a lot. I

36:29

think what you're saying, this is just arbitrary

36:31

Nvidia driver behavior. Like I said, you need

36:33

V-Sync on, but then once you enable G-Sync,

36:35

then it's going to use G-Sync as V-Sync

36:37

is what you're saying. Well, yeah, with V-Sync

36:39

on, in order, it

36:42

has to also have to do with the fallback for when you

36:45

get outside the range where V-Sync batteries. So once your G-Sync is

36:47

really for is

36:51

when you're running at less than the refresh rate

36:53

of the monitor. And

36:56

so basically,

36:58

the upshot is if you turn G-Sync on

37:01

and you turn V-Sync on, the driver just

37:03

handles and you don't have to worry about

37:05

all the other nonsense. The

37:07

other part of this is that if your game

37:10

has a frame limit and your game consistently runs

37:12

above the refresh rate of your monitor, you want

37:14

to set the frame limit to just below the

37:17

actual frame rate of the

37:20

refresh rate of the monitor to

37:22

reduce tearing, to make sure tearing never happens.

37:24

And it's like three to five frames a

37:26

second, usually. Yeah, for example, I have

37:28

a 144 Hertz monitor and I

37:30

set a 140 FPS cap in my drivers. Yeah,

37:33

well, you can also do that in the game as well, if

37:36

that's an option. The

37:38

other thing is that NVIDIA added a new

37:40

fast V-Sync option between the time the BlurBusters

37:42

article about G-Sync and V-Sync was written and

37:45

now. And you only

37:47

want to use that if your frame rate is

37:49

2x your refresh rate or more, and

37:52

you're probably not using G-Sync monitors. So

37:55

generally speaking, if you have a G-Sync monitor, you do

37:57

not turn on fast V-Sync. If you don't have a

37:59

V-Sync monitor, monitor, a G-Sync monitor, and

38:02

you're running. So if you have a 60

38:04

Hertz non-variable refresh rate monitor, and your

38:06

games are running at more than 2x, your

38:09

refresh rate of your monitor turning on the fast

38:11

one will give you benefits. I

38:13

can't imagine that there's a lot of people in

38:15

that situation that don't have G-Sync monitors or a

38:17

free-sync monitor at this point. That's almost certainly the

38:19

case. Brad, how can you tell if this is

38:21

working? Yeah, next question. How can I verify it's

38:23

working? Your monitor's on-screen display should

38:26

show you in real time the sync rate

38:28

fluctuating up and down. I can't imagine there

38:30

is a monitor that has adaptive

38:32

refresh that doesn't show it there, right? Mine

38:35

does. Yeah, my ASUS

38:37

monitors do. My TV does. Like, I think

38:39

just about everything that does some form of

38:41

variable refresh will just show you in the

38:44

display settings what it is at any given moment. I even

38:46

have a little thing that I can put the number in

38:48

the top right corner if I really want

38:51

to feel inadequate or excited about my

38:53

frame rate, depending on how things are

38:55

going. Ooh, I might have to look for that. One

38:57

of the other questions is, what if you have multiple monitors? So

39:01

from a frame rate perspective, Windows 11 and I

39:04

think 10, both can

39:06

run multiple monitors at independent resolutions

39:08

and refresh rates with no problem.

39:11

So like, my main monitor is running at 360 hertz. My

39:14

top monitor is running at 60 hertz. My right monitor

39:16

is running at 60 hertz. No problem. If

39:19

you're, say, using a two-PC streaming setup,

39:21

you can even mirror as long as

39:23

you mirror them in the right order and have the

39:26

refresh rate set before you do the mirroring. You can

39:28

set your main monitor, run it the video game resolution,

39:30

like the 144, 360 hertz, whatever. And

39:33

the capture card will still run at 60, and it will

39:36

just take every x frames to send

39:38

to the capture card. So that stuff works great. The

39:41

weird thing is that multiple monitors in

39:43

the past has really impacted performance. In

39:46

the pre-Windows 8 days, like Windows 7

39:49

is like the last generation of

39:51

desktop window manager and the early,

39:53

early stuff that started with Vista

39:56

and got better gradually over time. There

39:58

was a major performance. hit when you're running

40:00

full screen games and you

40:03

had stuff running on other monitors. It's

40:06

less of an issue with Windows 10 and 11. Linus Tech

40:09

Tips did a

40:12

video where they tested up to I think

40:15

four monitors with videos running full screen on the

40:17

other monitors while you played a game, which

40:20

showed that there was some impact, but I think

40:22

realistically nobody's going to be watching four videos at

40:24

once. You're much more likely to just have one

40:28

video and then some text like a guide

40:30

or something like that, you know, like just

40:32

static text, basically maybe images.

40:35

If you were concerned about it and if you're in perf

40:37

issues, you can use Win Plus P to

40:40

turn off your secondary monitors when you're

40:42

playing games really easily. Or

40:45

just hit the power switch. Also,

40:47

we should talk about game mode. On Windows 11,

40:49

there is a game mode setting that is

40:52

on by default. Most people don't ever have to change

40:54

it. It's worth checking. You can open

40:56

the settings panel and just type game mode

40:59

and it'll pop up. Basically what it does is

41:02

it does some pain in the ass

41:04

stuff like turns off notifications while you're playing

41:06

games, but the big one is that it

41:09

lets the window manager know that you play

41:11

games on this computer. So it handles full

41:13

screen games a little bit differently sometimes. Another

41:16

thing to know about here is full screen

41:18

versus borderless window or windowed if

41:20

you really want to play in a window, but only

41:22

weirdos play in windows unless it's

41:24

like vampire survivors or something maybe. I'm

41:27

judging them too. As the name

41:29

implies, borderless windowed means you're technically actually still

41:31

playing in a window. It just

41:34

is displaying the contents of the

41:36

window at full screen without scroll bars or title

41:38

bars or anything like that. So it looks like

41:40

it's full screen. You're still getting the benefit of

41:43

it being a windowed application, meaning you can all

41:45

tab in and out very cleanly. In

41:49

a lot of games, if you have a UI

41:51

menu up that gives you a mouse cursor, you

41:53

can just move it out of the window like

41:55

a normal ass application. The

41:58

problem... Is

42:00

that there again, this is a place where there used

42:02

to be a performance hit for doing the borderless window

42:04

thing Yeah, because in the exclusive full screen Windows

42:07

use in the old days windows would just say

42:10

hey, if this is all you

42:12

buddy, you got everything You got the GPU got the

42:14

CPU got we're turning off the window manager We're just

42:16

gonna let you do your stuff and then if you

42:18

wanted to say all tab It would minimize

42:20

the game the game would go away It

42:23

works. Definitely now and and back in the day This

42:25

is less of a thing now that people game out

42:27

there native monitor resolution more But if you're running in

42:29

a different video mode It would require like a whole

42:32

monitor video mode switch when you tab down and out

42:34

of the game Which was slow and clunky it still

42:36

does sometimes if you have a weird refresh rate monitor

42:38

like my a lot of games don't know to do

42:40

with my monitor sure and So

42:43

then like if you're if your games max refresh

42:45

rate setting is 144 Hertz It'll

42:48

just bump you down to 144 Hertz

42:50

on that monitor if you go in the exclusive full

42:52

screen mode Yeah, so yeah, like like like

42:54

we said borderless window avoids all those

42:56

annoying issues But but yes, there there

42:59

is or was a performance Delta at

43:01

some point So from so from a

43:03

survey I went through and searched reddit

43:06

for people complaining about borderless window and

43:08

nixtria an exclusive full-screen performance problems and

43:11

it seems like Pub G over

43:13

the years has been a game. You should probably play

43:15

an exclusive full-screen GTA 5

43:17

is a game that has a significant performance

43:19

boost running in borderless Huh

43:22

It's it's it seemed all

43:25

over the place on what's faster and what's slower

43:27

in full screen or Borderless

43:29

and of course it's important to note like

43:31

the convenience of not having to like mode

43:34

shift the monitor when you switch in and

43:36

out Of the game is substantial. So yes,

43:38

I Would just test

43:40

on game by game basis for the game. So

43:42

you're playing. Yeah. Yeah, I will throw in I've

43:44

had a lot better luck getting variable refresh to

43:46

work reliably in full-screen modes than with borderless like

43:48

the video control panel Does let you turn on?

43:51

Variable refresh for for windowed applications,

43:54

but it seems a lot less reliable

43:56

in my experience Are you so

43:58

I've had the it's funny. I've had opposite experience because

44:00

usually it works pretty well. My

44:03

desktop is always working in the variable refresh mode and

44:05

it just, if the game works then

44:07

the desktop works when you're in borderless. I mean if

44:09

the desktop works then the game works when you're in

44:11

borderless and makes it interesting. Interesting. Yeah.

44:14

One other tidbit I threw in here and this actually I got this

44:16

from the Discord a year or two ago. This

44:18

was fascinating I thought. Yeah. So this happened I

44:20

believe sometime after Windows 10 came out. Microsoft

44:23

rolled out a feature and this is a very under

44:25

the hood thing. They call it full

44:27

screen optimizations which actually means, and

44:29

I think there is a way to disable this, but it's

44:31

on by default, it means that if you tell a game

44:34

to run in full screen it thinks

44:36

it's running in full screen but Windows actually

44:38

is kind of intercepting it and there's some

44:40

kind of best of both worlds like abstraction

44:42

layer happening where it's like the game is

44:44

doing what it would be doing in full

44:46

screen but they are still like Windows

44:48

has still got some kind of compatibility layer where

44:50

it's also making it easier to all tab in

44:52

and out and stuff like that. You can read

44:55

this blog post I linked if you want

44:57

to know more about the technical details but

44:59

the idea there is to try to kind

45:01

of capture the best benefits of both modes.

45:03

Yeah. The desktop window manager is the thing

45:05

that handles rendering everything on the desktop generally

45:07

except for exclusive full screen apps. So the

45:09

fact that they pulled this in, you can

45:11

disable this in the launch options in the

45:13

compatibility tab for pretty much every executable on

45:17

your computer. I think

45:19

it is literally called full screen disable full

45:21

screen optimizations. Okay. There we go. I've had

45:23

really good luck with it though. I run

45:25

most all games in quote unquote full screen

45:28

but with this under the hood of course

45:30

and it's all worked pretty well for me.

45:33

So do they still minimize when you all tab out? Some

45:36

of them do. Actually that behavior is kind

45:38

of unpredictable like some decent number games

45:41

still do minimize but some of them do not.

45:44

Because I'm often streaming and I don't necessarily want

45:46

to show my desktop to everybody when I all

45:48

tab I do tend to

45:51

use borderless window. Yeah. That makes sense.

45:53

I mean you're doing a video output

45:55

into a stream machine, right? I'm not

45:57

doing a like I'm not I'm not

45:59

I don't don't use OBS to capture a window

46:01

and then broadcast it out someplace anymore.

46:03

I used to do that, but it was removing

46:07

layers of hassle. It turns

46:09

out it's important for me.

46:11

The next big place where there's stuff

46:13

that's confusing is post-processing.

46:16

Usually, people group these together

46:18

in their display options often. Here's

46:22

an interesting question for you, Brad. When

46:24

you're looking at display options

46:26

or graphics options in a

46:29

game, typically, people put them in the same

46:31

order. It's always resolution at the top. It seems like the

46:34

preferred way or maybe the scheme

46:36

that people use is to put the stuff that's

46:38

the highest impact toward the top and

46:40

the lowest impact toward the bottom in terms of performance.

46:42

Yeah. A lot of things, I think that is intuitive.

46:45

Except for now that we have ray tracing stuff and that's always at the

46:47

bottom. We've completely upended

46:49

the entire hierarchy. Post-processing stuff is

46:52

all the stuff that happens after

46:54

the scene is rendered. You've drawn

46:56

your polygons, you've applied textures to

46:59

them, you've applied your lightmaps, you've

47:01

done all the highlights and the

47:03

reflectivity and all that stuff. Then,

47:06

you have this time to work inside the

47:09

rendered scene. Stuff like reflections and

47:11

ambient occlusion, which we've talked about a little bit.

47:13

Depth of field usually comes in at this point,

47:15

so you blur the stuff that's far away or

47:17

too close that's outside of the focal range of

47:19

the camera. Bloom,

47:23

where you brighten

47:25

up areas that have light sources. Anybody

47:28

who played video games from, let's

47:31

say 2002 to 2006 or 2007, remember more than that,

47:36

wasn't it? It definitely kept going

47:38

from there. There was a period

47:40

like mid-original Xbox with Fable and

47:42

some stuff like that. Like blooms

47:44

were so overused. It was everywhere.

47:47

I mean, it was the

47:49

color lighting of its time.

47:51

Let's see, bloom, chromatic aberrations.

47:55

It turns out- Very polarizing. Well,

47:58

chromatic aberrations are one of the things to make. people

48:00

motion check really anything else. Yeah,

48:02

really? Yeah, we we we

48:05

had some early on we had some chromatic aberrations

48:07

on some VFX in our game and we had

48:09

to pull them because it was making people feel

48:11

real bad. Wild. I wouldn't even associate that with

48:13

motion because it's like a color separation thing. It's

48:16

it's it's mimicking a lensing effect right like it's

48:18

something if you have glasses of a certain or

48:20

the high enough prescription you'll yeah

48:22

like what you get red and blue depending on

48:24

which side of the lens you're pointing out. You

48:26

see the colors separating out a little bit. That's

48:29

what that is. Some some people just hate the

48:31

way it looks. Also, I find yeah,

48:33

it's I'm trying to think I think control used a lot

48:35

of chromatic aberrations. Oh, hold on. I was going to put

48:37

something in there. The

48:40

color balance. So this is this is

48:42

there's also post processing effects that are

48:44

like almost like film production stuff where

48:46

you where you can change the tone mapping. You

48:48

can change the color balance of a scene. So you can make it go black

48:51

and white. You can make it be full color.

48:53

You can only have certain colors be

48:55

visible stuff like that. Vignetting

48:57

which is just kind of putting a frame around the edge of the

48:59

screen like you would if you were on on a too

49:01

narrow a lens for your sensor size on

49:04

your camera. TAA

49:06

temporal anti aliasing which is

49:09

does anti aliasing between frames basically.

49:12

FXAA which is fast approximate. This is another one where

49:15

the X is the letter in the acronym is in

49:17

the middle of the word which makes me crazy. Smooth

49:20

edges often that's used for

49:23

things like fences and narrow skinny

49:25

stuff. I thought I never knew what the X

49:27

stood for and that I'm glad I know now

49:29

even though even though it's even though it's an

49:32

illegal abbreviation. Yeah, this

49:34

is true crimes here.

49:36

And then super sampling which we talked

49:38

about before and film grain which everybody should

49:40

turn off immediately. I think motion

49:42

blur fits in this category too but I'm not

49:44

100% about that. It wasn't on

49:46

the list here. Yeah. I'm realizing

49:49

that now. Look man, film grain has its

49:51

place. It can absolutely be overused. It has

49:53

its place in film but when you're watching

49:55

a digital medium that's recorded digitally and rendered

49:57

digitally like it's film grain. is a, is

50:00

a, what was the bad word from the

50:02

early days of the iPhone? Uh,

50:06

you know, when you, when you made your, your movie

50:08

collection look like a shelf. Oh,

50:11

skeuomorphic? Skeuomorphic. Yeah. Most

50:14

skeuomorphic. I guess. Look, it's an aesthetic

50:16

choice, okay? Our direction. I

50:19

get it, etc. Whatever. I liked it.

50:21

I liked it in Mass Effect. It looked good with those

50:23

70s lens flares. I turned it off immediately. Give me the

50:25

lens flare though. I love a lens flare. Lens

50:27

flare is also a post, a post-processing.

50:30

So a lot of this stuff

50:32

is, is low impact. Uh, like film grain, uh,

50:35

um, depth of field, um, tone

50:38

mapping, color balance, all this. So a lot of these things

50:40

you don't get access to. Dramatic aberrations,

50:42

people let you turn off because some people hate them and some

50:44

people make barf when you use them. Uh, bloom

50:46

often you can turn off. None of this

50:48

stuff is super high impact except for

50:50

SSAO and the anti-aliasing stuff. And

50:52

all of those are basically supplanted by

50:55

the ray tracing and or DLSS, uh,

50:57

the, the suit, the, the MR, ML

50:59

super sampling stuff. Yeah.

51:02

Um, HDR is another weird one. We, we did realize

51:04

we forgot to put this on the main list. I

51:06

kind of wish we had left it off. HDR

51:09

is a mess in PC games. It's really

51:11

all over the place. Like everybody implements it

51:13

differently. Some people don't support it right. That's,

51:17

that's kind of what I wanted to say is it sounds

51:19

like I'd often just turn it off. Some

51:21

give you too many options. Um, it's

51:23

in the no stock. If people want to look it up, cyberpunk, literally,

51:26

well, I can't see all the options right now,

51:28

but cyberpunk literally offers like two different HDR modes

51:30

that use like terminology, like sRGB,

51:32

something, something. And then I forgot the other, you

51:34

know what I mean? It's like, unless you really,

51:36

they let you pick the spec that you use,

51:38

right? Like unless you already professional working in like

51:40

color timing or something, you have no idea what

51:43

those things mean. And I couldn't tell the difference.

51:45

Oh, uh, speaking of color timing, that

51:47

is another thing. If you use it, if you actually take

51:49

the time to color calibrate your monitor and

51:51

you let games use exclusive full screen,

51:54

they won't use the color timing, the

51:57

color calibration profiles that you set up

51:59

through windows. So worth

52:01

worth knowing if you've bothered to color

52:04

calibrate your monitors. Um, I

52:06

so I usually What

52:08

I'm gonna say is don't use the auto HDR

52:10

setting that's in Windows because it blows does

52:13

it I think it stinks Some

52:15

people I haven't I've not spent a lot of time with it

52:17

personally But some people things seem to think it works pretty well.

52:19

I mean I have had almost

52:22

exclusively bad experiences with it On

52:26

games that do support HDR and you

52:28

take the time to do the calibration like don't

52:30

skip the calibration steps if you're doing HDR Don't

52:33

cheese the calibration steps Where

52:36

you you know, cuz look we've all done that thing

52:39

where it's like hey adjust the gamma until you can

52:41

barely see this Darkest bit of the scene right and

52:43

then we're like, you know what? It's pretty cool when

52:45

you're playing a multiplayer game to be able to see

52:47

people hiding. Oh shadows Yeah, you can't do that with

52:50

HDR because your stuff gets really washed out and janky.

52:52

It looks terrible It's just gonna make everything would blown

52:54

out Um, it's if you're gonna a lot of games

52:56

will actually ask you the max brightness of your monitor

52:58

Which is a kind of insane thing to ask. Yeah,

53:00

well, I've yes I've noticed that trend of like they

53:03

they literally like asked you how many knits can your

53:05

display do or something? Like those people are not going

53:07

to know that Well, I mean you're gonna

53:09

want to go look up your monitor on the internet

53:11

to see what it does Yeah, because if you get

53:13

that wrong, it'll make things look real bad, too Like

53:15

this is this is why it's messy. It's cuz there's

53:17

no As far as I

53:20

can tell there's no good way for the monitor to

53:22

say to Windows Hey, here's what I support and Windows

53:24

doesn't care enough to report that back to games So

53:26

it's up to us as players to know this stuff

53:28

on a game by game basis, which is janky and

53:31

bad Another thing I should mention. I think

53:33

I've got this right again. The stuff is pretty hard to figure

53:35

out I think you have to you have to have a shower

53:37

enabled in Windows at the desktop level before you then go into

53:39

a game And turn it on there, right? It has to be

53:41

on on your monitor It has to be on on Windows and

53:43

it has to be on in the game Like

53:46

you can't point is you can't leave HDR

53:48

off at like say the Windows desktop and then

53:50

just make a game force it on it Already

53:52

has to be on in Windows A lot of

53:54

people don't like the way that the desktop looks

53:56

with HDR on me included Especially if you have

53:58

like two non HDR monitors in one HDR monitor.

54:01

There is a Windows shortcut for HDR

54:04

to enable HDR. It's Windows key plus

54:06

alt plus B turns on HDR and

54:08

turns on alt B. If you're gonna

54:10

use HDR on

54:14

your Windows PC that one's a banger. Yeah

54:16

that's pretty good. I will keep that in mind.

54:19

That's pretty close to the reset the graphics

54:22

driver entirely keyboard shortcut. What

54:24

is that? Control alt B?

54:27

Is B for graphics? I guess. It's Windows

54:29

control shift B for that one which completely

54:31

restarts the graphics driver if you're having display

54:33

issues. There you go. That's a good one

54:35

to know. Last thing I

54:37

should mention here I think we talked about this a few months ago

54:39

in some other contexts. Maybe it was when I

54:41

bought my TV or something. Yeah we talked about HDR. Windows

54:44

11 has an HDR calibration

54:46

tool now. Oh really? You do have to go

54:48

to the Microsoft Store and download it. It doesn't

54:51

just stop built into Windows 11

54:53

for whatever reason. Literally just called

54:55

Windows HDR calibration. Just go to the

54:57

Microsoft Store and search for that. It's

54:59

actually basically identical to the Xbox like

55:02

the Series X HDR calibration. So Series

55:05

X and PS5 pretty much have this solved. I

55:07

mean of course they do because they're consoles. Yeah. You

55:10

know it's a 6 platform but it mimics that

55:12

where it's literally just three quick screens of like

55:15

adjust the brightness so you can barely see this

55:17

symbol now adjust it the other direction. Same thing

55:19

like it takes two minutes. That's really good. So

55:21

definitely do that in Windows 11 if you're going

55:23

to try using HDR in games. I wonder if

55:25

it passes it forward to games or if that

55:27

just sets the baseline for the monitor. So Alan

55:30

Way 2 recommends that you do this. Okay. So

55:32

I think it is working with games in some

55:34

direct way. Good to know. Hey so

55:36

the good news is we've done all the complicated

55:38

stuff and all we have left is ray tracing.

55:40

Oh yeah that's I mean it's just a fire

55:42

and forget right. Yeah ray tracing is really complicated

55:44

it turns out. Yeah. Well for now I guess

55:47

like theoretically eventually it's just gonna be the

55:49

thing and it's gonna be a well I

55:52

I mean at some point

55:54

in the future yeah but like right now

55:56

the way people are implementing in games it

55:58

seems like it's Like

56:00

everybody picks and chooses the stuff that they want

56:03

to apply retracing to. Yes. They're

56:05

pretty tactical about it. Like with the

56:07

exception of like cyberpunk where everything's traced to

56:09

hell. You

56:12

have to play it game by game basically is the upshot

56:14

here. There's no common

56:17

rules. The hardware

56:19

vendors are doing it typically a pretty

56:21

good job of highlighting the work for

56:23

flagship games. Yeah. Like I found this

56:25

great blog post about control on Nvidia's

56:27

blog where they highlighted the five things

56:29

that control used ray tracing for. Which

56:32

is like floor like reflections, environmental

56:35

reflections like furniture reflecting onto the

56:38

shiny concrete floors, white

56:41

walls reflecting lights differently, stuff

56:43

like that. Transparent

56:45

reflections where you can see your character in glass windows

56:47

or if like there's a fight happening in a room

56:49

with a bunch of glass, the

56:52

particles are reflected in the windows which

56:54

normally doesn't happen on glass reflections when

56:56

they're done by Q maps or other

56:58

pre-computed volumes. Indirect

57:01

diffuse lighting which is where the color light bounces to

57:03

nearby surfaces so you get more than one bounce on

57:05

the light maps including dynamic. So

57:08

typically with a light map you only

57:10

see lights on non-dynamic objects. So like

57:12

if there's an enemy in

57:14

a shadowy hallway the light won't necessarily

57:16

like the light won't

57:18

reflect onto them and in this case the wood. Right.

57:21

Like in that case like the game developer on the

57:23

rasterized situation is just using a different lighting technique to

57:26

simulate that the light from the light map is hitting

57:29

the dynamic object even

57:31

though there actually is different technique under the hood. They're

57:34

saying hey you're standing inside a light map therefore we

57:36

should reduce your lighting by this much. Right.

57:39

Right. But also it means that like

57:41

a projectiles light won't necessarily impact on

57:43

anyway. Yeah. So that's why

57:45

like if you've watched like me flip out of

57:48

Q2 RTX on a stream a couple of years

57:50

ago or whatever and specifically you're like oh my

57:52

god all the power-ups are affected by the same

57:54

lighting as everything else stuff like that. Yeah.

57:57

That's why like that's that's that's where having

57:59

a fully general. unified

58:01

lighting model makes makes everything look really consistent

58:03

and cool. And the thing

58:05

to know is that you're not replacing

58:07

like they're not the raytrays stuff isn't

58:10

replacing isn't

58:12

replacing nothing. It's replacing hacks that

58:14

were less good, right? It's it's

58:17

things that we did to make things look okay

58:19

But and it's kind of like

58:22

I use this example up It's kind of like

58:24

video compression right where after you watch each impact

58:26

to impact to video for a long time You

58:28

don't see the blockyness of it anymore after you

58:30

watch h264 compression for long enough You don't notice

58:32

the kind of glitches around the edges of things

58:35

anymore And and this is all stuff that is

58:37

people who've been playing games for a long time

58:39

You probably don't even notice anymore until it's not

58:41

there and it's working the way it does in

58:43

the real world, right? Contact

58:46

shadows was another good one so if you think

58:48

about how shadows on a thing like a chain

58:50

link fence or like a thin metal rod or

58:53

like a Like like really

58:55

fences are the classic example because there's a lot

58:57

of vertical things aligned with each other and they

58:59

often get stairs Kind of stair stepped out of

59:01

existence because the light maps are lower

59:03

resolution than the world and and the shadows don't look

59:05

good as a result These

59:08

these give fine detail shadows That

59:11

without having to do the kind of fine-tuning

59:14

Like basically you put you put a higher resolution

59:16

The way to fix is to put higher resolution

59:18

light maps in areas where these types of things

59:20

are But that is incredibly time intensive and requires

59:23

an artist going back over the levels again and

59:25

again and again as things change So

59:28

so this just it kind of happens automatically. It looks

59:30

really nice It also means

59:32

that like things are shadowed like the inside of

59:34

pipes and stuff like that are shadowed in a

59:37

way That's appropriate even when lighting in the world

59:39

is changing and then the last one is debris

59:41

so there's a couple of scenes in control where

59:43

like monsters come in and just destroy the whole

59:45

world around you and It

59:48

what they're doing is essentially letting the things

59:50

that are typically visual effects and Particles

59:52

that are that don't maybe don't look like particles

59:55

Cast shadows and reflect light and do

59:57

things and it looks bonkers when you

59:59

you see it in person. Yeah.

1:00:03

So, so, but the upshot on

1:00:05

that is that the settings for

1:00:08

control and the settings for Alan Wake and

1:00:10

the settings for cyberpunk and the settings for

1:00:12

Fortnite are all wildly different because

1:00:14

their implementations of ray tracing are very highly specific

1:00:16

to their games. Yeah. And London in

1:00:18

some cases, like you said, they kind of take and choose which

1:00:21

parts of the render pipeline are handled by

1:00:23

reasoning and which are traditional like call of

1:00:25

duty has ray tracing, but they ray traced

1:00:28

some shadows and I think a couple

1:00:30

of reflections occasionally, but

1:00:32

not like, not like world reflections.

1:00:34

It's basically just you see the players in in

1:00:37

puddles and glass and stuff like that. Yeah. Yeah.

1:00:39

And I mean, to be clear, the reason for

1:00:41

that is that so far ray tracing has had

1:00:43

an enormous performance cost. Like it's, it's not

1:00:46

what a lot of people would consider comfortably playable

1:00:48

depending on the kind of hardware you're on. Like

1:00:50

we're just getting to the point now where hardware

1:00:52

is fast enough and also the software is improving

1:00:54

enough. Like that's why it was such a big deal

1:00:56

like marketing wise, like they made such a big push

1:00:58

behind the cyberpunk path tracing mode.

1:01:01

Yeah. Path racing in this context is a much

1:01:03

more generalized, Hey, now we're using ray tracing for

1:01:05

everything in the game. Yeah. So ray

1:01:08

tracing traces out from the camera viewpoint to

1:01:10

all the pixels in the scene and path

1:01:12

tracing is the other way. It bounces from

1:01:14

light sources until it intersects with the camera,

1:01:16

right? Um, I can't remember which direction. I

1:01:19

think they're still going from surface back. I

1:01:21

think they're still working backwards, but

1:01:23

we're, we're definitely not experts on that. I

1:01:26

think it's tracing rays back from surface to camera instead of

1:01:28

the other way around. Right. But

1:01:31

the race. Yeah. Yeah. Anyway, but,

1:01:33

but the upshot is there are way

1:01:35

better algorithms coming out for like ray casting and resampling

1:01:37

and all that stuff. Like if you've seen some of

1:01:39

the, I know this is a very Nvidia

1:01:41

heavy episode. I mean, they've been pushing

1:01:44

ray tracing artists. Um, RTX DI,

1:01:46

I believe is the name of their package that

1:01:48

they put out along with cyberpunk. Um,

1:01:51

it is incorporating a lot faster types

1:01:53

of ray tracing processing. Um, so

1:01:56

I think we're going to start seeing more games that use ray

1:01:58

tracing and a much more general. Global way. Well, I

1:02:00

think the other thing that's gonna happen is we're gonna start

1:02:02

seeing Current gen console exclusives that

1:02:05

utilize the ray tracing stuff more more Yeah,

1:02:07

we're actually start like spider-man 2 was a

1:02:09

surprising example of that where there is no

1:02:11

mode on the PS even on the ps5

1:02:13

There is no mode that doesn't use ray

1:02:15

tracing. Yeah, I agree And

1:02:17

that carries over to console ports, of course So

1:02:22

path tracing according to digital trends Path

1:02:26

tracing is ray tracing is more

1:02:28

accurate path tracing is faster can

1:02:30

have some problems But you

1:02:32

can kind of work around the problems by by

1:02:34

applying some other clever solutions to

1:02:36

them And that's why like

1:02:39

way to RTX that that's why the games that

1:02:41

have done full lighting model replacements have done path

1:02:43

tracing for the most part The

1:02:48

let's see last thing I guess here

1:02:50

I I'm

1:02:53

trying to think I feel motion blur. We didn't talk about

1:02:55

motion blur Brad People

1:02:57

love motion blur. Do they well?

1:03:00

Do you remember a couple years ago when when digital foundry

1:03:02

did the thing that was like, hey You know what

1:03:04

forts of horizon looks as good at 30 frames

1:03:06

a second with motion blur on is it does it

1:03:08

60 frames a second? Mm-hmm people

1:03:11

reacted pretty negatively to that as I

1:03:13

understand. Yes. Yes, that is unsurprising Yeah,

1:03:16

I tried it. Did you try it? Did you turn

1:03:18

it on? I did not I was

1:03:20

really shocked It adds what how

1:03:22

effective it was I couldn't really tell

1:03:24

the difference. Oh Interesting.

1:03:27

Yeah. Um, so anyway, let's see

1:03:29

last couple things the latency reduction

1:03:31

stuff like reflex and and Anti-lag

1:03:35

on the AMD side are

1:03:37

both really good. And if you have those supported in

1:03:39

your games It's worth turning them on especially if you're

1:03:41

using mouse and keyboard because they do reduce the input

1:03:43

to pixel latency quite a bit Yeah, they're actually I

1:03:46

was gonna ask earlier Is there any reason not to

1:03:48

ever enable those is it just like a if you

1:03:50

run into problems turn them off? Maybe

1:03:52

the boost setting can increase your power

1:03:54

consumption on video side for the for

1:03:56

the reflex make

1:04:00

the GPU run it. We can keep

1:04:02

this at the TDP if you have to, to

1:04:04

hit the frame rate calls.

1:04:08

My general advice is like turn off

1:04:11

the frame generation for most stuff probably

1:04:13

at this point, unless it's like heavy

1:04:15

ray tracing type content. Yeah, again,

1:04:17

Cyberpunk there was the example where like, even

1:04:20

a 4090 could only run Cyberpunk on the path tracing

1:04:22

mode at 4K, like basically about

1:04:25

30 FPS native. You

1:04:27

can see in my Allen

1:04:30

Wake settings, I have DLSS,

1:04:32

I'm using DLAA for the render

1:04:34

resolution at my native NAP and

1:04:37

I have frame generation on there. That

1:04:40

was nice. It performed well. Yeah, Allen Wake

1:04:42

is another one of the few games that's very heavy

1:04:44

on ray tracing. But again, I think we're going to

1:04:46

see a lot more of it in the next two,

1:04:48

three years. Well, and also like both

1:04:50

of those, I mean, Cyberpunk a little bit less so,

1:04:52

but Allen Wake is a game where like your input

1:04:55

latency doesn't really matter, right? It's

1:04:57

not like there's fast shooting sequences

1:04:59

or you're basically walking around being

1:05:01

a kind of grumpy old guy and doing cop

1:05:03

stuff. So yeah, that's it. That's

1:05:05

settings. Do we miss anything else important? I

1:05:07

don't think so. I mean, let's

1:05:10

miss something else a little bit. That's shocking, yeah. But please

1:05:12

let us know if you thought this was useful or helpful.

1:05:15

Oh, the other one is Unreal 5 stuff. Unreal 5

1:05:17

has a completely different pipeline for ray tracing than anybody

1:05:19

else. So as we start to see Unreal 5 games,

1:05:21

they will probably be a little bit more in common

1:05:23

if they're using Nanite and Lumen and all that stuff,

1:05:26

right? But it's a kind of

1:05:28

hybrid CPU GPU solution that's

1:05:33

not super well documented from what this means

1:05:35

for people. Standpoint, if

1:05:37

you're playing say Fortnite, which is a competitive

1:05:39

game, probably I turn all

1:05:41

that stuff off. If

1:05:43

you're playing something that is not competitive and you want

1:05:46

to look nice, turn it all on, get crazy. Going

1:05:48

right back to the days of Quake 1 when I

1:05:50

had the three effects card and thought it was cool.

1:05:52

And then I went back to the software mode because

1:05:54

I got 200 frames a second. The

1:05:57

guy who ran the Linux lab that I've talked about, I think...

1:06:00

in the past had had settings

1:06:03

for Quake one with 3D acceleration that

1:06:05

basically turned the MIP levels all the way down.

1:06:08

Oh yeah. Oh, that's a classic. It was like

1:06:10

you'd like it like the health box was just

1:06:12

like a brown box with a

1:06:14

red circle on it. Basically a fuzzy

1:06:16

red circle. Yeah, that's the classic. I've had good

1:06:18

times. Good times. I don't

1:06:21

think it really mattered all that much, Brad. Probably

1:06:24

not. No, but but yeah, let us know. Let us

1:06:26

know if there's other settings. If there's other things you

1:06:29

don't understand, you'd like to know more about, we

1:06:32

will track them down post posted. Send us

1:06:34

an email at techpod at content

1:06:36

down or in the question seeking answers channel

1:06:38

in the discord and we'll we'll we will

1:06:41

we will drill down and figure it out. And

1:06:43

if we don't know the answer, we'll call Nvidia and AMD

1:06:45

and ask them to please help us answer

1:06:48

your questions. But

1:06:50

I think that's it for us this week, Brad. Yeah.

1:06:53

As always, Brad and Will Made a TechPod is a 100%

1:06:56

listener supported show. So

1:06:59

without you all, we wouldn't be here. So thank you to all of

1:07:01

our patrons for supporting the show. Yes, thank you. If

1:07:03

you would like to find out how to support Brad Will Made

1:07:05

a TechPod, you can go to tech patreon.com/tech

1:07:07

pod. Again, you can go

1:07:10

to patreon.com/tech pod. And

1:07:12

for five bucks a month to get access to

1:07:14

the fabulous tech pod discord, our monthly patron exclusive

1:07:16

episodes and and you get to know

1:07:18

that like we get to continue making a podcast. So

1:07:21

that's that's really good, too. Yes, that helps. That's

1:07:23

that's cool. Yeah, we I like right now in between. I'm

1:07:25

going to go and tell you having a little bit of

1:07:27

income while I'm in between and kind of

1:07:29

talking to people about jobs. It's comforting to know that

1:07:31

this is there. It helps helps a lot. So we

1:07:33

really appreciate it. I've also had that thought about this

1:07:35

podcast. Thank you to everyone who keeps it going. Also,

1:07:38

also, you get to be on the discord.

1:07:40

And for example, when you're

1:07:43

trying to continuity tests and HDMI cable and want

1:07:45

to get some advice about a better way to

1:07:47

do that, that's a good place to ask questions

1:07:49

like that. Oh, it seems like you're ripped from

1:07:51

the headlines kind of example there,

1:07:53

Brad. Yeah. Guaranteed,

1:07:56

if it's if it's technical in nature, somebody is going to

1:07:58

have some idea about it there. I've

1:08:00

enjoyed the conversations about the last

1:08:02

patron episode where we talked about

1:08:05

vertical monitors and everybody's been sharing pictures

1:08:07

of their vertical monitors, their ultra-wide vertical monitors.

1:08:10

Oh no, I hadn't seen that. I'll

1:08:12

have to see that. And we also got

1:08:14

into the B-Links, B-Link versus how to wake

1:08:16

on LAN a B-Link and how to do

1:08:19

a remote start on a B-Link, which

1:08:21

I've been curious about because it's a situation of

1:08:25

concern to me. Does that thing have an

1:08:27

accessible BIOS or UEFI? UEFI?

1:08:29

It's a little thin, Brad. UEFI.

1:08:32

It's a bad... Who the hell thought... BIOS.

1:08:35

UEFI. BIOS. Listen

1:08:37

to that. Just listen to it. BIOS? UEFI.

1:08:40

BIOS. UEFI. UEFI?

1:08:44

UEFI. UEFI. UEFI. UEFI.

1:08:48

UEFI. I can't even say it three times. UEFI.

1:08:50

UEFI? UEFI. UEFI?

1:08:54

Say UEFI. BIOS. I'm just going to keep saying BIOS.

1:08:56

Thank you all so much for watching. We

1:08:59

really appreciate the support. And we especially appreciate our

1:09:01

executive producer tier patrons, including Paddle Creek Games,

1:09:03

makers of Fractured Veil, Andrew Slosky,

1:09:06

Jordan Lippett, Bunnyfiend, duh, Just Wedge,

1:09:08

Joel Krauska, Twinkle Twinkie, David Allen,

1:09:10

James Kammack, and Pantheon,

1:09:12

makers of the HS3 high-speed 3D printer. Thank you

1:09:15

all so much. We appreciate you. Yes, we do.

1:09:17

And with that, we will be back next week with

1:09:20

another edition of the TechPod. Thanks everybody for listening. We

1:09:22

hope you have a productive and lovely week. Bye.

1:09:25

Bye. Bye. Bye.

Unlock more with Podchaser Pro

  • Audience Insights
  • Contact Information
  • Demographics
  • Charts
  • Sponsor History
  • and More!
Pro Features