Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
10.8 GM - lag improvements?
- Thread starter rikbrown
- Start date
- Sort by reaction score
Are some of the people posting their personal disdain for this product also illiterate? The point of this thread is for those who have ML installed on the MBPr they own. Let me know what part of this eluded you, and I'll break it down further.
Here are my findings on my Retina MacBook Pro with ML GM.
-With discrete graphics and native resolution, everything is butter smooth.
-With discrete graphics and scaled, OS animations are smooth and some sites may slightly stutter in Safari. In Firefox beta, however, everything is butter smooth. So, Apple has some catching up to do. It really looks bad that they are beaten by firefox in this department. On the iPad, the situation is reversed.
-With both integrated and discrete, Firefox scrolling is MUCH smoother on web sites Safari has problems with (such as Google+, Facebook)
-With scaling via integrated graphics, animations and transitions are sometimes smooth and sometimes stutter depending on what you have going on.
I have no doubt that scroll lag will be addressed. What concerns me more is how well they can address interface animation stutter when running scaled via integrated graphics.
Also, I have a 17" MBP early 2011. Honestly, scrolling isn't really all that smooth with integrated graphics either. In other words, the focus being placed on this topic is magnifying the importance of butter smooth scrolling on large web pages.
After looking at it some more, it seems to me that fixed elements on webpages break hardware acceleration at least on Chrome under Lion. I have no idea if Safari on ML is better. But if you hide the fixed elements on FB or The Verge, scrolling is buttery smooth in Chrome (with HW acceleration enabled) under all circumstances. I knew this was a software flaw ever since I enabled HW acceleration in Chrome and witnessed heavy websites scroll perfectly even when a flash video was running, Engadget for example.
That is the most extreme thing I have ever heard. That also has to be the dumbest thing I have heard since the rmbp was released.
Yes it's totally valid because scrolling in Facebook dictates how fast your computer can covert video, process images, play games, so on...
It's quite obvious that facebook scrolling is only one example of lag. Clearly people have been experiencing lag all over the place which would affect how efficient it can actually do work.
Technically it's the browser that was half baked, and the hardware in other laptops was essentially serving as a bandaid. Now that a display that can actually stress a single core out has arrived, Apple had to code the software a bit better.
So will a cmbp have an advantage? Perhaps. But with 4 cores + hyper threading, that advantage will become pretty small if the apps are coded well
So does this mean that essentially with SwitchResX (non hidpi) at 1440 res or 1680 will run just like the cMBP? If so, why is everyone complaining and sending back their rMBP? They could just run switchresX and wait for problems to be sorted and in the meantime get the UI performance they want in a nicer, lighter machine...
I'd imagine so, but don't have a 2012 cmbp to do any comparisons with or I'd try.
RDM is a free alternative instead of buying SwitchResX...
but anyways, some people would have no issues with that, but many here will complain that you should never run any screen at less than its full resolution, because they claim it looks too bad... so they want an actual lower res screen. I find that a bit absurd, but maybe these people have eyes of an eagle.
It will run as fast, but it won't look quite as nice. The font smoothing in Mac OS X uses subpixel rendering to get extra clarity when drawing text.
For instance, if you need to draw a vertical bar 1 1/3 pixels thick, you could draw it as one black pixel, and to the right of it, a slightly yellow (G + B) pixel. So you'd have: R (off), G (off), B (off), R (dim), G (on), B (on). The dim R is next to three dark subpixels, so it just makes it look like the line is a little thicker.
But, if you do that on a retina display scaled up at 2x, you actually get two black pixels, and then two yellow pixels. The dim R subpixel in the second yellow pixel is separated from the other dark subpixels, and actually makes the text a little blurrier.
So, I wouldn't recommend buying a rMBP and running it without HiDPI, especially not to someone so picky as not to buy a beautiful rMBP because of a slightly lower frame rate bringing up Expose!
If you install Windows on the rMBP and install Chrome with the Smooth Scrolling extension, you get scrolling similar to OS X Safari scrolling but with none of the lag.
It's perfectly smooth on every site I tested. Facebook, The Verge, PhoneArena, GameSpot, etc.
It's perfectly smooth on every site I tested. Facebook, The Verge, PhoneArena, GameSpot, etc.
I second that. The UI is smoother than a few Intel Macs I've owned in the past. I remember upgrading Tiger to Leopard on a Macbook Pro made it go from smooth to choppy/laggy. Nobody seemed to care then.
Very cool find. Thanks for input! I tried the firefox beta as you mentioned and I too can confirmed its as smooth as butter. So it seems Safari does have some work to do. It will be interesting to see if how the next release of Chrome and Firefox compares with Safari.
I also noticed when my batter on my rMBP drops to 10% the integrated graphics kick in or so I think because that's I can notice the lag in safari the most.
Everything else seems to be pretty good. We are basically all beta testers sort of speak as we are dealing with a brand new macbook pro design and brand new OS release
