Announcement

Collapse
No announcement yet.

Stuttering Rage - the new analysis method from Rage3D

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Stuttering Rage - the new analysis method from Rage3D

    In my latest review I'm using a new benchmark analysis method which I dubbed Stuttering Rage. The point of it is to show more details of performance which can be used on a comparitive basis and get away from misleading or simple min, avg, max FPS.

    http://www.rage3d.com/reviews/video/.../index.php?p=3

    Take a look and please offer your criticisms. This is a work in progress and I'd love to hear your thoughts about the analysis, assumptions, presentation etc.

    If your post starts with 'I haven't read your complete article' then my reply will likely start the same, too

    #2
    can you add the similar metric that TR uses.. no. of frames above 50ms,40ms,30ms (distrib graph)

    GIGABYTE G1.Sniper Z87 - i7-4770K | 2x4Gb G.Skill PC16800 DDR3| XFX R9 290 (Cat14.12) | Asus PB278Q | 128Gb OCZ Agility3 SSD | 240Gb Samsung EVO SSD| Seagate Hybrid Drive 2TB | Fractal Design R4 | PC Power & Cooling Silencer 750 | Samsung SH-B123L BD/DVD/RW | W8.1

    GIGABYTE GA-Z68XP-UD3 - i7-2600K | 4x4Gb G.Skill PC16800 DDR3| Sapphire HD 7970 (Cat14.12) | BenQ XL2720T | 2x 128Gb OCZ Agility3 SSD | 2TB WD Caviar | Antec Lanboy | Corsair H60 w/S-Flex 120x25 F fans | PC Power & Cooling Silencer 950 | Creative X-Fi Titanium | Samsung SH-B123L BD/DVD/RW | W7 Ult

    Gigabyte EP45-UD3 - Q9550 oc 4.0 | 2x2Gb OCZ Reaper 8500 DDR2 | Sapphire 6970 (Cat13.6b2)| Dell 2407WFP | 74Gb WD Raptor | 500Gb WD SE16 | 1Tb Seagate | Thermalright XP-90C w/Delta 92x25 3-blade fan | Corsair 650TW | CM 690 case| Creative X-Fi Platinum | Liteon 165H6s DVDRW | WinXP Pro SP3

    Comment


      #3
      why is that useful to you?

      Comment


        #4
        I suppose there is some cut off where any frames with latency above x ms has a greater impact on performance.

        I would think that cut off would be game engine specific however. Some may be more sensitive than others.

        Comment


          #5
          because the frequency of frame performance above these time marks is useful to represent how much frame latency exists.

          GIGABYTE G1.Sniper Z87 - i7-4770K | 2x4Gb G.Skill PC16800 DDR3| XFX R9 290 (Cat14.12) | Asus PB278Q | 128Gb OCZ Agility3 SSD | 240Gb Samsung EVO SSD| Seagate Hybrid Drive 2TB | Fractal Design R4 | PC Power & Cooling Silencer 750 | Samsung SH-B123L BD/DVD/RW | W8.1

          GIGABYTE GA-Z68XP-UD3 - i7-2600K | 4x4Gb G.Skill PC16800 DDR3| Sapphire HD 7970 (Cat14.12) | BenQ XL2720T | 2x 128Gb OCZ Agility3 SSD | 2TB WD Caviar | Antec Lanboy | Corsair H60 w/S-Flex 120x25 F fans | PC Power & Cooling Silencer 950 | Creative X-Fi Titanium | Samsung SH-B123L BD/DVD/RW | W7 Ult

          Gigabyte EP45-UD3 - Q9550 oc 4.0 | 2x2Gb OCZ Reaper 8500 DDR2 | Sapphire 6970 (Cat13.6b2)| Dell 2407WFP | 74Gb WD Raptor | 500Gb WD SE16 | 1Tb Seagate | Thermalright XP-90C w/Delta 92x25 3-blade fan | Corsair 650TW | CM 690 case| Creative X-Fi Platinum | Liteon 165H6s DVDRW | WinXP Pro SP3

          Comment


            #6
            Originally posted by SpecTP View Post
            because the frequency of frame performance above these time marks is useful to represent how much frame latency exists.
            The amount of time spent within 20% of the mode isn't useful to you?

            Comment


              #7
              Originally posted by caveman-jim View Post
              The amount of time spent within 20% of the mode isn't useful to you?
              it is useful, but I'd like to also see the frequency and duration of frames at their 'outliers' as well.

              GIGABYTE G1.Sniper Z87 - i7-4770K | 2x4Gb G.Skill PC16800 DDR3| XFX R9 290 (Cat14.12) | Asus PB278Q | 128Gb OCZ Agility3 SSD | 240Gb Samsung EVO SSD| Seagate Hybrid Drive 2TB | Fractal Design R4 | PC Power & Cooling Silencer 750 | Samsung SH-B123L BD/DVD/RW | W8.1

              GIGABYTE GA-Z68XP-UD3 - i7-2600K | 4x4Gb G.Skill PC16800 DDR3| Sapphire HD 7970 (Cat14.12) | BenQ XL2720T | 2x 128Gb OCZ Agility3 SSD | 2TB WD Caviar | Antec Lanboy | Corsair H60 w/S-Flex 120x25 F fans | PC Power & Cooling Silencer 950 | Creative X-Fi Titanium | Samsung SH-B123L BD/DVD/RW | W7 Ult

              Gigabyte EP45-UD3 - Q9550 oc 4.0 | 2x2Gb OCZ Reaper 8500 DDR2 | Sapphire 6970 (Cat13.6b2)| Dell 2407WFP | 74Gb WD Raptor | 500Gb WD SE16 | 1Tb Seagate | Thermalright XP-90C w/Delta 92x25 3-blade fan | Corsair 650TW | CM 690 case| Creative X-Fi Platinum | Liteon 165H6s DVDRW | WinXP Pro SP3

              Comment


                #8
                Slightly off topic but the "smoothness graphs" for Supreme Commander 2 are the graphs for Hitman Absolution.
                Edit: so are the Shogan2 graphs.
                Last edited by spike700; Jan 19, 2013, 09:26 PM.
                ------------------------------------------------------
                Aoccdrnig to rscheearch at Cmabrigde Uinervtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht the frist and lsat ltteer be at the rghit pclae. The rset can be a total mses and you can sitll raed it wouthit porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe.
                ****nig amzanig huh?
                ------------------------------------------------------

                Comment


                  #9
                  Thanks, I'll get those fixed.

                  Comment


                    #10
                    The data is very useful and does give a good representation of game smoothness. The analysis part of the report may be hard for some to understand that are not so tech savy though. I had to look at the data and reread this section a few times to make sure I understood what I was looking at. Though it was kind of early in the morning.

                    Do you have any plans to add in any comparisons? For me the only time I read reports are when shopping for a new card or just out of curiosity how a competitors card fairs against my current card.
                    Last edited by Craz; Jan 22, 2013, 07:31 AM.
                    BF3 Clan: Soldiers of War

                    Cheaters will be killed on site!

                    Comment


                      #11
                      The intent is to simplify presentation a bit (more on that soon) and to provide a better way to contrast cards, so yes. I am limited in which products I have access to, unfortunately.

                      Comment


                        #12
                        Originally posted by caveman-jim View Post
                        The intent is to simplify presentation a bit (more on that soon) and to provide a better way to contrast cards, so yes. I am limited in which products I have access to, unfortunately.
                        Cool I am looking forward to seeing the finished report format. These detailed reviews will be a big help when it comes time make a decision on my next purchase. Also a thing I would like to see tested is stability test of some sort. I was plagued with system hangs running 7970's in trifinity on my x79 system and none of the reviews I read ever mentioned any issues. I am sure it probably had to do with the amount of time the reviewers had with the cards. Do you think a few hrs of heaven and 3dmark for each card tested would be of any help to look for issues that I experienced?
                        Last edited by Craz; Jan 22, 2013, 11:03 AM.
                        BF3 Clan: Soldiers of War

                        Cheaters will be killed on site!

                        Comment


                          #13
                          Did your system have the hangs when running that test in that configuration?

                          Review time is certainly limited, the more time spent looping through a burn in test the fewer titles get tested overall.

                          Comment


                            #14
                            Originally posted by caveman-jim View Post
                            Did your system have the hangs when running that test in that configuration?

                            Review time is certainly limited, the more time spent looping through a burn in test the fewer titles get tested overall.
                            Yes I did have system hangs running that configuration and with 2 cards. Single screen was about impossible to recreate the hang though it did happen on certain titles.

                            With time being limited, I can see where it would be hard to fit in burn in test with multiple configurations. Though once a burn in test on a card is complete I wouldn't think it would need to be redone for future test unless other components were changed.
                            BF3 Clan: Soldiers of War

                            Cheaters will be killed on site!

                            Comment

                            Working...
                            X