gn fps sucks and we are reading it wrong...

[thrasher_565]

so im just watching this video now i no im late... but im kinda geting it. even if you have say 300+ fps in cs each frame can be good or bad. and when its bad there a spike and that make a micro studder. and that is what you feel makes the game feel bad even thow your getting 300+ fps.. means nothing.

 

consistent timing is better then fast timings but we know that all ready...

 

anyway skip to say 11:00 the meat and potatos imo.

 

yes with the 1% lows and .1% lows what he's saying is we need to calculate how many errors were in that test and get a % from that.

 

errors per second

 

Edited March 14 by thrasher_565

[Sharkyx1]

Yep I watched this 

 

 

Feels like they are talking about developing tools to measure standard deviation in presentation of smooth animation. Makes sense to move to more defined measurements as we are able to more closely measure things.

[Blazepoint5]

2 hours ago, thrasher_565 said:

so im just watching this video now i no im late... but im kinda geting it. even if you have say 300+ fps in cs each frame can be good or bad. and when its bad there a spike and that make a micro studder. and that is what you feel makes the game feel bad even thow your getting 300+ fps.. means nothing.

 

consistent timing is better then fast timings but we know that all ready...

 

anyway skip to say 11:00 the meat and potatos imo.

 

yes with the 1% lows and .1% lows what he's saying is we need to calculate how many errors were in that test and get a % from that.

 

errors per second

 

1%lows and .1%lows diffrence? I know 1%lows but idk the .1

 

So do you want error per second to be a thing added on softwares like MSI AB

[thrasher_565]

8 minutes ago, Blazepoint5 said:

1%lows and .1%lows diffrence? I know 1%lows but idk the .1

 

So do you want error per second to be a thing added on softwares like MSI AB

well it be nice to no that just because xgpu can do xfps dose not mean its "faster" if the micro stutters give you a worse experience. and yes there alot that can give problems not just the gpu. there problems all over like keyboard lag, monitor lag, ram problems os problems heck even different motherboards can give different out comes. then thow silacon lotty in to the mix. thow they bin them so that less off a problem these days faster cpu they just sell it for more...

 

40fps with low stutters is nicer then 300 fps with stutters

if it plays smooth at 40 fps probably wont care how much fps its doing

Edited March 14 by thrasher_565

[Sharkyx1]

29 minutes ago, Blazepoint5 said:

1%lows and .1%lows diffrence? I know 1%lows but idk the .1

 

So do you want error per second to be a thing added on softwares like MSI AB

.1% lows are just a more precise show of worst case scenario frame drops. It isn't uncommon to see them in benchmarks.

 

The video is talking about making a new measurement for msi AB yeah, afterburner uses Intel's open source Presentmon function internally and Intel is adding new features that will basically tell you "average smoothness error" which I'm sure will take time to educate the audience in but it will give users more ability to diagnose issues with their games.

[Blazepoint5]

26 minutes ago, thrasher_565 said:

well it be nice to no that just because xgpu can do xfps dose not mean its "faster" if the micro stutters give you a worse experience. and yes there alot that can give problems not just the gpu. there problems all over like keyboard lag, monitor lag, ram problems os problems heck even different motherboards can give different out comes. then thow silacon lotty in to the mix. thow they bin them so that less off a problem these days faster cpu they just sell it for more...

 

40fps with low stutters is nicer then 300 fps with stutters

if it plays smooth at 40 fps probably wont care how much fps its doing

I agree, stuttering and micro stutters breaks emersion and it is annoying

[Blazepoint5]

3 minutes ago, Sharkyx1 said:

.1% lows are just a more precise show of worst case scenario frame drops. It isn't uncommon to see them in benchmarks.

 

The video is talking about making a new measurement for msi AB yeah, afterburner uses Intel's open source Presentmon function internally and Intel is adding new features that will basically tell you "average smoothness error" which I'm sure will take time to educate the audience in but it will give users more ability to diagnose issues with their games.

Ohhh makes sense. 

Let me watch the video then

[thrasher_565]

30 minutes ago, Sharkyx1 said:

.1% lows are just a more precise show of worst case scenario frame drops. It isn't uncommon to see them in benchmarks.

 

The video is talking about making a new measurement for msi AB yeah, afterburner uses Intel's open source Presentmon function internally and Intel is adding new features that will basically tell you "average smoothness error" which I'm sure will take time to educate the audience in but it will give users more ability to diagnose issues with their games.

ya it tells you that frame there was a problem with it when it spikes but it dosent tell you why it spiked and how many of thows spikes you have. its not the hiness of the spike its the inconsistencies of them. and he saying 1 frame is not on screen or processed the same time some are faster and some longer. 3 longer frames is a big difference from 3 fast frames thats what people dont no about fps. i dont no we will see what this data gives us. could just be a red haring

 

we were just toght that 1 fps a second.

Edited March 14 by thrasher_565

[Motifator]

FPS is a meaningless metric without full context.

Say you're sailing on a boat, you want navigation / sonars up. You're sailing globally, you want a curved screen not a flat. The mKart system is your friend. You can hook it up, and use it as a global chartpilotter.

Put that on gaming context, you still want an ultrawide monitor just because you get to see more to the edges.

[Doobeedoo]

Yeah some of us knew this already and in some games can be a big problem, like online games. Especially those with scuffed engines.

For example I play Quake Champions and game is notorious for having issues sadly. The time I tried optimizing the game though. I have locked fps and game at times can feel horrible like it's at running much lower at intervals or micro stutters etc. Sometimes it runs fine, it's also netcode issues on top. I've bench tested it and 1% and .1% lows getting logged but still not good enough as it's not telling enough in this case.

[Sharkyx1]

On 3/14/2024 at 1:10 AM, Motifator said:

FPS is a meaningless metric without full context.

Say you're sailing on a boat, you want navigation / sonars up. You're sailing globally, you want a curved screen not a flat. The mKart system is your friend. You can hook it up, and use it as a global chartpilotter.

Put that on gaming context, you still want an ultrawide monitor just because you get to see more to the edges.

Weird to use ultrawide as the metaphor, and not just higher FOV, I use a taller than standard monitor for the most part, it doesn't limit how much I can see 

[Mark Kaine]

On 3/14/2024 at 3:21 AM, thrasher_565 said:

so im just watching this video now i no im late... but im kinda geting it. even if you have say 300+ fps in cs each frame can be good or bad. and when its bad there a spike and that make a micro studder. and that is what you feel makes the game feel bad even thow your getting 300+ fps.. means nothing.

 

consistent timing is better then fast timings but we know that all ready...

 

anyway skip to say 11:00 the meat and potatos imo.

 

yes with the 1% lows and .1% lows what he's saying is we need to calculate how many errors were in that test and get a % from that.

 

errors per second

 

i keep saying this since years... people don't care all they see is higher (or lower) numbers,  placebo effect is incredibly strong with people these days...

 

 

 

[porina]

I missed this earlier. Interesting stuff from Tom. Be interesting to see where this goes. The basic problems were known for a long time, but getting something meaningful out to move beyond that is the tricky part.

 

Also, anyone know what case this is in the background? Or just the character?

[Poinkachu]

11 hours ago, porina said:

 

 

Also, anyone know what case this is in the background? Or just the character?

https://www.aliexpress.com/i/4000920082580.html

 

 

Didn't look into it much, but considering the character has a Ningmei logo on it's chest armor, probably an original character made by either Phanteks or Ningmei.

[porina]

12 hours ago, Poinkachu said:

Didn't look into it much, but considering the character has a Ningmei logo on it's chest armor, probably an original character made by either Phanteks or Ningmei.

Thanks. Have gone through the video, which is 3 years old now. This specific one isn't for me, but I hope to see more accessible collab designs in the west. Been watching Hyte for example who have done vtuber/game collabs. Having the right waifu on your case will increase fps more than blue LEDs or airflow ever will!

[Poinkachu]

1 minute ago, porina said:

Thanks. Have gone through the video, which is 3 years old now. This specific one isn't for me, but I hope to see more accessible collab designs in the west. Been watching Hyte for example who have done vtuber/game collabs. Having the right waifu on your case will increase fps more than blue LEDs or airflow ever will!

I think GN showed a ningmei x inwin waifu case in that video as well, which has an LCD screen inside the case showing a dancing waifu.

I wonder how much FPS will that increase

[Motifator]

On 3/16/2024 at 2:22 AM, Sharkyx1 said:

Weird to use ultrawide as the metaphor, and not just higher FOV, I use a taller than standard monitor for the most part, it doesn't limit how much I can see 

Not really since UW makes its thing:

[Sharkyx1]

1 hour ago, Motifator said:

Not really since UW makes its thing:

On a standard monitor,if you increase FOV to 100 you'll probably have the same wide view. 

 

Ultrawides are not bad, but I have not seen anything to make me think they are the Superior monitor format. I just think it's weird to say 

Quote

Put that on gaming context, you still want an ultrawide monitor

 

By best friend has a 34x14 screen and another friend has a 52x14 screen and I will take my 16:10 and 16:9 monitors over them every day. 

 

Edit: changed one word 

[Motifator]

8 hours ago, Sharkyx1 said:

On a standard monitor,if you increase FOV to 100 you'll probably have the same wide view. 

 

Ultrawides are not bad, but I have not seen anything to make me think they are the Superior monitor format. I just think it's weird to say 

 

By best friend has a 34x14 screen and another friend has a 52x14 screen and I will take my 16:10 and 16:9 monitors over them every day. 

 

Edit: changed one word 

You won't, you'll have fisheye effect. Normal monitors suck when rolled into ultrawide FOV's.

Also, you want an aggressive curve in certain games like Siege, they give you a different level of immersion.

All first hand experience.

[Sharkyx1]

14 minutes ago, Motifator said:

You won't, you'll have fisheye effect. Normal monitors suck when rolled into ultrawide FOV's.

Also, you want an aggressive curve in certain games like Siege, they give you a different level of immersion.

All first hand experience.

I'm not trying to say that you are wrong about your first hand experience and your opinions, you're allowed to believe what you want ,and it can be true to the way you your games.

 

When I was talking about my friends monitors I wasn't talking about 2nd hand information, I've used both 21:9 32:9 and 48:9 eyefinity/surround. I don't see them as worth it. I occasionally use my oled tv to do an ultrawide in the centre, haven't really felt any need to buy one.

 

My only point is I think it's a incorrect generalization saying that all users want ultrawide, it's nice but it's not objectively superior in my eyes.

 

On the other hand I do not really believe in immersion, that's not the way I play games