1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Bits Understanding Colour Depth

Discussion in 'Article Discussion' started by Tim S, 9 Nov 2007.

  1. OleJ

    OleJ Me!

    Joined:
    1 Jul 2007
    Posts:
    2,024
    Likes Received:
    10
    That was a very good article! Educative, informational and interesting. Thank you :)

    I must say I agree with the editor concerning defining absolutes of our vision. One of my BIG questions regarding vision is refresh rates. Usually we define motion as being fluid and indiscernible at approx. 24 frames per second. BUT this does not match the time it takes the synapses to fire. Neither does it match what our brains interpret.
    Studies also show that we do not and can not comprehend the amount of data that our eyes receive as our brains and nervous system simply don't have the bandwidth to deal with the amount of data captured by our eyes.
    Psychological studies have also shown that we do not percept visual entireties but only fragments and then our brains complete the imagery.

    All my blabbering hopefully helping to argument that our visions absolutes are indeed very limited (in terms of bandwidth and the eyes perception) but the percepted does not equal this as the final image is created as a separate whole.

    I owe to say I sought all this information because I did not understand why turning wheels sometimes look like they are turning backwards when you look at the spokes (be that in sunlight as electric light has an obvious explanation) :)
     
  2. spazmochad

    spazmochad Minimodder

    Joined:
    21 Aug 2006
    Posts:
    572
    Likes Received:
    42
    Wow some of the best writing I've ever seen on bit tech!

    Makes me want a 3 chip DLP projector so bad.
     
  3. The_Gimpy

    The_Gimpy What's a Dremel?

    Joined:
    15 Jul 2004
    Posts:
    196
    Likes Received:
    0
    Does the response time in LCD televisions work on the same priciple? Or would the response time those mean anything different?
     
  4. Flyingsheep

    Flyingsheep What's a Dremel?

    Joined:
    26 Jun 2006
    Posts:
    25
    Likes Received:
    0
    I don't believe <=16ms completely eliminates ghosting. For LCDs, I don't even have to think about it and I notice ghosting. I also don't believe in the common belief that the human eye can't perceive anything faster than 60fps. It's nonsense. On my CRT at 85hz I often times still notice the flickering. I'm not making myself think I can see it. I never think about it or try to think if I can notice it or not. I'll just realize sometimes that my monitor is flickering even though it's at 85hz. When I switch it down to 75hz I notice a big difference, and of course 60hz flickers way too much for me. I don't believe there is a limit to how many frames per second we can perceive. There is a nice article here that explains:

    http://amo.net/NT/02-21-01FPS.html
     
    Last edited: 10 Nov 2007
  5. Arkanrais

    Arkanrais What's a Dremel?

    Joined:
    14 Jul 2007
    Posts:
    406
    Likes Received:
    5
    Great article allthough bit depth and image quality is somewhat lost on me (cataracts = sh*t vision)
    one thing I have noticed with flickering screens, is that when my cataracts were somewhat worse, that I would notice more flickering on my screens. my guess is that the less quality I can see, the more movement I can see. Not sure though.
     
  6. Ramble

    Ramble Ginger Nut

    Joined:
    5 Dec 2005
    Posts:
    5,595
    Likes Received:
    41
    I tried it and saw nothing. Perhaps it's something to do with colour depth.
     
  7. Firehed

    Firehed Why not? I own a domain to match.

    Joined:
    15 Feb 2004
    Posts:
    12,574
    Likes Received:
    16
    No, that's ghosting. You just have a decent monitor. I see it only on one of my displays at work - if I move my cursor around the blue sidebar over there ------> on one monitor, it's totally white the whole time. On the other monitor, part of it will be slightly blue until the movement stops. It's the LCD taking more than a single frame to go from #004080 to #FFFFFF.

    TBH, I think that explaining ghosting properly (and conclusively proving that anything below 16.667ms response time is a waste) would be quite difficult without a high-speed camera, so you could look at the exact - rather than theoretical - changes occurring every sixtieth of a second. At the end of the day, the monitor is only receiving 60 frames per second from the video card provided you've got vsync on (and you should). Either a pixel can make the complete transition from one color to another within that 16.667ms window or it can't. If it can, the monitor won't ghost. Of course, don't mix it up with motion blur, which is more of a mental function than anything else - look at a wheel on a car spinning on the highway, and notice how it's all blended together or looks somewhat like a semitransparent solid where there's some open space.
     
  8. ufk

    ufk Licenced Fool

    Joined:
    3 Jan 2004
    Posts:
    760
    Likes Received:
    10
    ..And know my brain hurts.

    Interesting article, informative and concise, ah well I learned something new today

    *goes to look for information on setting a TFT up properly
     
  9. Cobalt

    Cobalt What's a Dremel?

    Joined:
    24 Feb 2006
    Posts:
    309
    Likes Received:
    2
    The difference here isn't that we can percieve more than 60fps but that the manner in which this 60fps is displayed allows us to see artifacts on the image anyway. For the record the 60fps limit is colour preception, our monochromatic vision is a lot more acute and so the flickering on a CRT is very obvious at 60Hz. Some people have sensitive enough vision that even at very high refresh rates they can still see flickering. I have to use 100Hz CRTs to not get headaches and nearly everybody can see the flickering if they loook at a CRT out of the corners of their eyes (as the rods have higher concentrations around the outside of the retina).

    Of course all of this only applies to the CRTs with their black time. LCDs don't suffer from it at all. The reason ghosting can be seen on those <16ms monitors is due to the grey to grey measurements the article was talking about. The black-white measurement is probably still above 16ms so you can percieve it. My current monitor has a listed response time of 16ms and the only ghosting I can ever see is in extreme black/white transitions and even then it is only minor. I've used panels with 8ms response times and 6bpc where the image quality was noticably worse with lots of ghosting even in low contrast situations.

    The millisecond myth really is a myth but because of the inconsistant ways of meauring response time people still get the impression that a 2ms panel will be better, if only because the black-white transition may creep in under 16ms if the grey-grey gets low enough.
     
  10. Tomm

    Tomm I also ride trials :¬)

    Joined:
    12 Apr 2004
    Posts:
    2,249
    Likes Received:
    0
    You got your dorsal and ventral streams the wrong way around :D
     
  11. phonon45

    phonon45 What's a Dremel?

    Joined:
    28 Sep 2003
    Posts:
    32
    Likes Received:
    0
    Just so we are clear Tim, you are claiming that the eye has variable gain correct? The arguement only works if the eye can dynamically adjust the gain and not saturate. Do either of you have a reference for this point or the counter point?
     
  12. dognosh

    dognosh What's a Dremel?

    Joined:
    22 May 2006
    Posts:
    138
    Likes Received:
    0
    Excellent article:clap:

    You get shearing ? I did with my last setup but not with the new one, must be a hardware thing ?

    also I like >60 fps , because when the action gets intense and the Gcard becomes very very busy I still get over or around 30 fps:hip:
     
  13. legoman666

    legoman666 Beat to fit, paint to match.

    Joined:
    24 Aug 2003
    Posts:
    927
    Likes Received:
    19
    Hmm, I wasn't saying that films have motion blur, I'm saying that they could benefit from being >24fps. You stated that >30fps is inpercievable, which is, in my opinion, simply wrong. When you have images as big as a theatre screen, you can notice the fact that something is moving a measureable distance between 2 frames. This is one of the reasons that I wish the framerate would increase along with the resolution in HDTV, it's simply more noticeable when its 1080p and 60".

    Anyway, I don't think I got across what I was trying to in regards to motion blur. I'm agree with you that <=16ms response time will get rid of ghosting (for a 60hz display) but you'll need less than 16ms to get rid of motion blur. Another post after yours talked about the difference in the way CRT's and LCD's display an image, which goes a long way in explaining the percieved motion blur of an LCD. On a CRT, the screen is blackened between each frame, so when the next frame is drawn, there are no "artifacts" left on the display from the previous frame. On an LCD, it goes directly from 1 frame to another, so the previous frame has to "fade" into the next.

    Is that correct? Or am I just spouting more BS?
     
  14. completemadness

    completemadness What's a Dremel?

    Joined:
    11 May 2007
    Posts:
    887
    Likes Received:
    0
    No I'm pretty sure that's wrong

    Firstly, a CRT has no mechanism to "blacken" a pixel
    What happens with a CRT, an electron hits a phosphor, which then glows, but starts to dull immediately
    Because the CRT refreshes 60hz or more, you don't (well normally) notice

    The LCD on the other hand, changes the colour by moving the crystal, so the time taken to change a pixel is determined by how fast the crystal can change (assuming your only controlling that 1 pixel directly)

    This also means, that every frame the entire screen is "refreshed" on an LCD, whereas every other frame on the CRT's are refreshed
     
Tags: Add Tags

Share This Page