The Framerate Analysis Thread part 2

'VBL' = Vertical Blank Line (at least it did back in the Amiga days). The etymology comes from refresh on a CRT, and denotes one screen refresh = 1/60th second on an NTSC screen. Thus 1 VBL = 1 frame on a 60 FPS display = 60 fps, and 2 VBL = 2 frames = 30 fps.

Nebula said:
So care to share the video with framerate counter like the regular Digital Foundry videos or as PS360 analysis videos?
Click to expand...
You don't need high accuracy for a rough empirical estimation. eg. If I say "the doctor measured me and found I was ten feet tall", it wouldn't need someone else to measure me with a tape measure to prove that false. Seeing me standing in a normal sized room with my head nowhere near the ceiling is conclusive enough.

If the footage seen is clearly above 30 fps based on prior experience with viewing moving images, this empirical observation is good enough reason to doubt claims of much lower framerates. From that estimate of accuracy, we can look into reasons why (wrong footage measured, fault in equipment, wrong measurements taken, incorrect interpretation of data, etc.) and how methods can be improved so future measurements are trustworthy. Though numbers look so confident and trustworthy, it can be dangerous to accept them always at face value without questioning!
 
Shifty Geezer said:
'VBL' = Vertical Blank Line (at least it did back in the Amiga days). The etymology comes from refresh on a CRT, and denotes one screen refresh = 1/60th second on an NTSC screen. Thus 1 VBL = 1 frame on a 60 FPS display = 60 fps, and 2 VBL = 2 frames = 30 fps.
Click to expand...

Thank you for your reply.
 
Polite reminder : Remember this is the analysis thread, not the discussion of findings thread. There's been some serious lateral shift in the conversation but all related in the end to measurement of framerate and screen-tear. Please refrain from posting personal opinions on results, and if any posts go missing, it'll be a clean-up job.
 
ps360 said:
Tearing Judgment
Brief description
/http://gamess.web.fc2.com/image/coj/tearing.PNG
It is actually complex a little more.

http://www3.telus.net/public/dhwag/COJ.zip
framerate-001.jpg XOR framerate-002.jpg
/http://gamess.web.fc2.com/image/coj/comp_framerate-002.JPG

My ScreenShot
X0002.PNG XOR X0003.PNG
/http://gamess.web.fc2.com/image/coj/comp_x0003.PNG

Tearing Frame --> No Frame Count
Click to expand...


I've manually sewn the torn frames together to recover the complete images.

http://www3.telus.net/public/dhwag/COJ_complete.zip

48 unique images out of 60 as I have previously estimated :cool:
 
MazingerDUDE said:
I've manually sewn the torn frames together to recover the complete images.

http://www3.telus.net/public/dhwag/COJ_complete.zip

48 unique images out of 60 as I have previously estimated :cool:
Click to expand...

Thanks for taking the time MazingerDUDE. :)

I prefer the Quaz/Digital Foundry "old school" way of calculating FPS where each unique frame regardless of tearing is considered a frame.

Keeping track of torn frames already allows users who are interested in that facet adjust the FPS numbers to account for torn frames if they wish.

I don't see the point of penalizing a system for torn frames AND keeping track of torn frames still.

For example saying X game which has 40 unique frames but 50% torn frames is actually 20 FPS with 50% torn frames gives the impression that there is only 20 unique frames being rendered per second and out of those only 10 of them are untorn.

Regards,
SB
 
ps360 said:
frame_count.gif


COJ is 2VBL
Click to expand...

There are two consecutive torn frames throughout my gameplay captures. And there are MANY of them in the cut-scenes.

I really don't think it is 2VBL. At best, it switches between 1VBL and 2VBL in a similar way to GTA IV on 360.


MazingerDUDE said:
I've manually sewn the torn frames together to recover the complete images.

http://www3.telus.net/public/dhwag/COJ_complete.zip

48 unique images out of 60 as I have previously estimated :cool:
Click to expand...

I salute your Photoshop skills!!
 
To the best of my knowledge GTA IV doesn't have two consecutive tearing frames though. I might be wrong. CoJ definitely does though, both in gameplay and cut-scenes.
 
i don't think it's fair to not count the torn frame, part of it is a new frame and should add to the total fps.

But on the other hand you shouldn't count it as a whole new frame because it isn't. Why not calculate the percent of the frame that is new and add it to the FPS. So you don't get any redundant information in your FPS count.

Hey just an late night idea.

edit. ps360 already doing this in post#387.

Why doesn't the tearing count on every torn frame in "60fps tearing ver.2" ?
 
Last edited by a moderator:
habbe said:
i don't think it's fair to not count the torn frame, part of it is a new frame and should add to the total fps.

But on the other hand you shouldn't count it as a whole new frame because it isn't. Why not calculate the percent of the frame that is new and add it to the FPS. So you don't get any redundant information in your FPS count.

Hey just an late night idea.

edit. ps360 already doing this in post#387.

Why doesn't the tearing count on every torn frame in "60fps tearing ver.2" ?
Click to expand...

examine
 
You must log in or register to reply here.

Similar threads

Shifty Geezer
Digital Foundry Article Technical Discussion [2024]
138 139 140
Replies
3K
Views
246K
GhostofWar
G
L
ARM based Game Consoles on the way?
2 3 4
Replies
73
Views
8K
Lurkmass
L
R
State of GDeflate *spawn
2
Replies
25
Views
2K
Remij
R
BRiT
  • Locked
Digital Foundry Article Technical Discussion [2023]
252 253 254
Replies
5K
Views
534K
chris1515
C
G
Nvidia G-Assist
Replies
2
Views
1K
Davros
D
Back
Top