Quantcast

Maximum PC

It is currently Sun Sep 21, 2014 3:23 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 8 posts ] 
Author Message
 Post subject: Refresh Rate vs Response time: Is there a conflict here?
PostPosted: Sun Feb 05, 2012 10:50 pm 
8086
8086

Joined: Tue Jan 03, 2012 2:18 pm
Posts: 61
I'm looking at display specs for monitors and there is something that doesn't make sense to me.

Response time is defined as the amount of time it takes for a pixel to change its value. So a monitor with a response time of 5ms can change the value of a pixel in 5ms of time.

Refresh rate is the rate at which the screen is draws the data it is given. A typical monitor has a refresh rate of 60hz or 60 times per second or roughly one time every 16.6 ms.

So my question is this. If the fastest a monitor draws a screen is 16.6ms then what exactly does a response time of 5ms mean? Does that mean that if the instruction to redraw that pixel is given 5ms before the refresh occurs then the pixel will be redrawn in 5ms but if the instruction is given 4m before a refresh it wont actually be drawn until 20.5ms after the instruction is given? Isn't a 5ms response time therefore very misleading? Doesn't that mean that in most cases there wont really be any difference at all between a monitor with 10ms response and one with 5ms response on a 60hz monitor?


Top
  Profile  
 
 Post subject: Re: Refresh Rate vs Response time: Is there a conflict here
PostPosted: Tue Feb 07, 2012 5:49 am 
[Team Member]
[Team Member]

Joined: Sat Jun 26, 2004 4:31 am
Posts: 11092
Location: Home Sweet Home
Took a while but I found a site that explains it in plain English:

http://www.ehow.com/facts_7378341_lcd-5ms-vs_-8ms.html

Nasty


Top
  Profile  
 
 Post subject: Re: Refresh Rate vs Response time: Is there a conflict here
PostPosted: Tue Feb 07, 2012 9:49 am 
8086
8086

Joined: Tue Jan 03, 2012 2:18 pm
Posts: 61
Thank you but my issue isn't with the definition of response time, I think that I understand (although your article did clarify a few things for me so thank you). My issue is that the whole idea of a response time measured in a few milliseconds doesn't really make sense if the monitor is refreshing at a rate of 60 Hz. After all 60 Hz is roughly once every 16ms so the response time being 5ms is somewhat meaningless or perhaps just an absolute best-case scenario if the instructions line up perfectly with the refresh rate.


Top
  Profile  
 
 Post subject: Re: Refresh Rate vs Response time: Is there a conflict here
PostPosted: Tue Feb 07, 2012 11:27 am 
8086
8086

Joined: Tue Jan 03, 2012 2:18 pm
Posts: 61
Here let me just lay out how I'm picturing this and you all can tell me if I am wrong or not. The comparison will be between a monitor with 60Hz refresh 4ms response and one with 60Hz refresh and 8ms response.

Time zero: monitor refreshes
2ms: Instruction given to the monitor to change a pixel from black to white
17ms: Monitor refreshes, applying instructions
19ms: 4ms response monitor makes pixel white (2ms to go from black to white 17ms after instruction)
21ms: 8ms response monitor makes pixel white (4ms to go from black to white 19ms after instruction)

So if true a 4ms response time monitor isn't twice as fast as a 8ms response time monitor its only the difference between something like 17ms to carry out an instruction and 19ms to carry out an instruction (in this example). You would get much much MUCH more benefit from getting a 120hz monitor than you would from seeking better response time.


Top
  Profile  
 
 Post subject: Re: Refresh Rate vs Response time: Is there a conflict here
PostPosted: Tue Feb 07, 2012 12:46 pm 
Million Club - 2 Plus
Million Club - 2 Plus

Joined: Thu Aug 02, 2007 11:07 pm
Posts: 2622
Location: NC
Remember, the article barely touches on the idea of how the industry defines the pixel "response time".

    Rise time (tR): Time it takes to go from 0 (black) to 255 (white).

    Fall time (tF): Time it takes to go from 255 to 0.

    Black-to-Black(tR + tF): This is standard definition of a response time as defined by the ISO 13406-2 method. This is the time it takes to do a complete cycle from black to white and back to black.

    Gray-to-gray (G2G or GtG): This is the most common value that's associated with reported response times from the manufacturers, since its usually the quickest and looks better for marketing people to tout. The biggest issue with G2G times is that there is no standardization on test procedures so you never know what you are actually getting. It can mean going from 128 to 255, but it can also be going the opposite way or with much small value changes; or just the complete average. They never say.


Realistically, G2G is a bit more practical since most of the time, your videos would rarely strobe between 0 and 255 values. If there was a standardized way of displaying the response times to their customers, then this value would have more meaning, but this requires lots of time and money to educate their customers when they really don't need to know too much about this information.

The biggest thing to note is that Rise time will tend to be very very slow relative to Fall time. Say if you have a tR of 10ms from 0 to 255, 7ms of that time could be spent just by going into an "on" state; from 0 to 32.

Fall time tends to be very fast since its just a matter of increasing voltage. Going from 255 to 0 in most cases can be 2ms or faster. Some panels may have problems when dropping to dark gray values, but typically, they should still be quite a bit faster than similar Rise times.


Overdrive/Response Time Compensation (RTC) is a technique used in TN panels to apply over-voltage to the Liquid Crystals (LC) to accelerate or motive them into orientation faster. In a TN panel, the LC's are used to block the light coming from the backlight; in this case a black or "off" (0) is actually in an active state where voltage is being applied to the LC. In a fully white state or "on" (255), no voltage is being applied to the LC allowing for full transmission of the backlight.

RTC forces the LC to go into a full white (inactive state) and then to a black (active state) transition first before the LC is allowed to drop to the required grey level. This speeds up the change for G2G values, but does nothing for the actual black to white (tR) response time.

However, when using an RTC mode, a few issues can arise. Color reproduction error/overshoot (aka RTC error %), video noise, and colour trailing resulting in dark halo's on moving objects. RTC error % is the chance of a transition will overshoot the value during the RTC cycle, thus needing to be corrected after the fact. This results in longer response times and can result with artifacts and video noise.

On quality panels, RTC error's are usually very minimal and only occur when going between small values (like 64 to 96); It's occurs more frequently from dark gray to lighter gray transitions. Video noise can become an issue in areas of soft gradation, specially when a panel uses dithering and the RTC error is high for that particular grey value. Colour trailing tends to be a hard issue to see since the black halo's only occur in moving objects, in high contrast scenes and the halo's only last for 2-3 frames; plus, it can only be scene for less than 10-15ms in most modern panels and hardly noticeable.



So now that you just read all of that, lets put that to a practical meaning with an example.
Image
Above is the Samsung SyncMaster SA950 TN panel monitor response time 3d bar graph; It has a stated 2ms response time rating on the box. As you can see, in normal mode, it doesn't even come close. It has an average of 8.2ms G2G with a max of 16ms response time. You can also see that going from any "on" (32-255) state to "off" (0) state is 1ms or less, but requires a little over 9ms to go from 0 to 32.

Image
Above is the same panel, but using an RTC overdrive setting. Clearly it lives up to its expectation for the most part with an average of 2.4ms G2G response time with no higher than 4ms; although it does have some outlier issues with transitions with dark-grays and almost white values. Most panels tend to have similar reproduction problems so its generally acceptable.

Image
Now above is the RTC error % 3d bar chart when running the panel in the RTC mode. This is a bit high for a quality panel but it does highlight some issues with RTC control abilities in trying to hit the right color value the first time. Going from values 32 to 128 will result in an 82% failure rate, requiring the need for a correction for that particular transition. In some monitors, it could take 35ms or longer to correct the color, but generally the rule of thumb is 3 frames to clear and correct.


If you want more information or complete detailed reviews on monitors, visit xbitlabs.com; they have more than enough information to throw a stick at. Other sources is tft central and tom's hardware.


Top
  Profile  
 
 Post subject: Re: Refresh Rate vs Response time: Is there a conflict here
PostPosted: Tue Feb 07, 2012 12:53 pm 
8086
8086

Joined: Tue Jan 03, 2012 2:18 pm
Posts: 61
Wow well that was certainly a lot of good info Jbaz and I learned a lot there. Don't take this the wrong way but I don't feel like any of that addressed my actual question which is that isn't the importance of response time dwarfed by the amount of time it takes for a refresh.

Unless I'm mistaken the pixels dont even start to shift color until the screen refreshes and at 60hz thats going to be ~17ms between refreshes. So even if response time was perfect and exactly as advertised the difference between a 2ms response time monitor and a 4ms response time monitor wouldnt be 2x as fast but rather just slightly faster as its the 17ms refresh rate thats going to cause the majority of the lag time.

I guess my bottom line is should I really care that much about response time. I see gamers typically use that as the main stat for buying but I'm not sure I believe it, I think they might be fooling themselves. Seems like the only downside to IPS is a slgihtly slower response time than TN but honestly I don't see why I should care.


On a sidenote I assume a refresh rate of 60hz means that the most the monitor could ever display is 60 fps and anything past that gets lost in the shuffle or worse causes screen tearing.


Top
  Profile  
 
 Post subject: Re: Refresh Rate vs Response time: Is there a conflict here
PostPosted: Tue Feb 07, 2012 1:43 pm 
Million Club - 2 Plus
Million Club - 2 Plus

Joined: Thu Aug 02, 2007 11:07 pm
Posts: 2622
Location: NC
Yes, you are correct, the pixel's only change when the screen refreshes unlike old CRT's. Once a pixel hits its value, it will stay that color until it needs to change again.

But in my post, I'm highlighting the idea that the G2G response times of what the manufacture claims can be grossly exaggerated in most cases. You might get some shitty ones that require 16ms or slower response times if the panels have issues with a particular transition, even with RTC overdrive enabled. Plus whatever RTC error time on top of that if the RTC overshoots, which can add another 20-50ms depending on the model. Once you start hitting 25ms, you'll visually see ghosting effects in fluid videos.

The bottom line is that in most cases, like games and movies, it's not going to matter; specially considering most of the PC panels are still in the 60hz range. TV? they got 240hz+ panels already and we are finally getting more mainstream 120hz PC monitors. Most people can't see past 60hz anyhow.

Honestly, the more important spec you should be looking at is input lag... and its not a spec they tend to list. You have to find reviews for your monitor to find this out; generally IPS monitors tend to have a high input lag rate compared to TN's.


Top
  Profile  
 
 Post subject: Re: Refresh Rate vs Response time: Is there a conflict here
PostPosted: Sun Feb 26, 2012 10:16 am 
8086
8086

Joined: Mon May 02, 2011 6:50 am
Posts: 51
A refresh rate of 60 Hz means every 16ms, monitor's hardware draws a new picture on the screen. A response time of 5ms means that 5ms out of those 16ms the pixel is changing, and the rest 11ms the pixel stays that color, so that your eyes can pick that color and identify it.

Response time is the measure of the LCD matrix itself, while the refresh rate is the measure of the monitor's internal logic hardware.

Theoretically, a matrix with 5ms response time can have a MAXIMUM refresh rate of 200Hz (1 / 0.005), but it is usually limited by the logic hardware.


Top
  Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 8 posts ] 

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 4 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group