The Naked Scientists

The Naked Scientists Forum

Author Topic: How many pixels can we actually perceive?  (Read 5183 times)

Offline chris

  • Neilep Level Member
  • ******
  • Posts: 5339
  • Thanked: 65 times
  • The Naked Scientist
    • View Profile
    • The Naked Scientists
How many pixels can we actually perceive?
« on: 03/07/2009 22:56:24 »
I was asked by someone today at what point we should stop adding smaler and smaller pixels to screens on the grounds that the eye won't discriminate them.

Can anyone enlighten me?

Chris


 

Offline Don_1

  • Neilep Level Member
  • ******
  • Posts: 6890
  • Thanked: 7 times
  • A stupid comment for every occasion.
    • View Profile
    • Knight Light Haulage
How many pixels can we actually perceive?
« Reply #1 on: 04/07/2009 15:11:36 »
At normal magnification I think we have already passed the point of perceivable difference. I don't think it's the screen/play-back medium, so much as the record/capture medium which now determines clarity and definition of the image.
 

Offline chris

  • Neilep Level Member
  • ******
  • Posts: 5339
  • Thanked: 65 times
  • The Naked Scientist
    • View Profile
    • The Naked Scientists
How many pixels can we actually perceive?
« Reply #2 on: 04/07/2009 15:16:37 »
Sure, so what is that resolution please?

Chris
 

Offline LeeE

  • Neilep Level Member
  • ******
  • Posts: 3382
    • View Profile
    • Spatial
How many pixels can we actually perceive?
« Reply #3 on: 04/07/2009 19:04:52 »
It would vary from person to person.

I'm using a 20" crt monitor running at 1600x1200, with a vertical screen height of almost 12".  This works out to about 100 dpi (dots per inch).  Without anti-aliasing I can clearly see stepping artifacts, so quadruple the resolution i.e. 400 dpi, would probably be sufficient for me.
 

lyner

  • Guest
How many pixels can we actually perceive?
« Reply #4 on: 04/07/2009 20:14:41 »
A factor which no one has mentioned is viewing distance or angle subtended by the screen. If you want to sit with your nose on the screen, you'll need nearly 300 dots per inch. If the screen is to cover a whole wall with that resolution, you couldn't aford it ( or the bandwidth).
If you watch normal TV, with about 600 lines of verticall resolution, the recommended viewing distance has always been five times picture height. For HDTV you sit half as far away and it looks fine. Any nearer and you start to see lines / pixels.
 

Offline RD

  • Neilep Level Member
  • ******
  • Posts: 8132
  • Thanked: 53 times
    • View Profile
How many pixels can we actually perceive?
« Reply #5 on: 04/07/2009 21:07:24 »
Quote
In photography, the circle of confusion diameter limit (“CoC”) is sometimes defined as the largest blur circle that will still be perceived by the human eye as a point when viewed at a distance of 25 cm ... a person with good vision can usually distinguish an image resolution of 5 line pairs per millimeter (lp/mm), equivalent to a CoC of 0.2 mm in the final image.
http://en.wikipedia.org/wiki/Circle_of_confusion

So that's 0.2mm at a viewing distance of 25cm, or an angular measurement of about 1/22nd of a degree.

This figure will only relate to the spacing of cells on the human retina: your pet eagle will think the resolution of HDTV is crap  :).
« Last Edit: 05/07/2009 15:45:08 by RD »
 

lyner

  • Guest
How many pixels can we actually perceive?
« Reply #6 on: 05/07/2009 00:35:39 »
It's like the length of a piece of string.
Are we talking about simple angular resolution? In which case, the field of vision is relevant if you want to determine how many pixels  are necessary.

Most people choose to have their TV displays far too big for impressive viewing. A big glubby picture is nothing like as nice as a smaller picture which you could believe would be sharp even if you looked closer. People still insist on screens taking up half of their living room wall, though. Even if it's HD ready the majority of programme material which is likely to be available  in the life of a display of the present generation will mean that most pictures will be big and fuzzy.
Never mind the quality, feel the width.
 

Offline techmind

  • Hero Member
  • *****
  • Posts: 934
  • Un-obfuscated
    • View Profile
    • techmind.org
How many pixels can we actually perceive?
« Reply #7 on: 07/07/2009 20:42:46 »
The resolution of the eye should be expressed as an angular resolution, as sophiecentaur has already pointed out. The visual acuity is highest in the central few degrees of vision than in the periphery (but I can't remember a figure of the top of my head).

However, I do remember in the past both calculating, and doing some practical experiments with stepping backwards from pieces of paper printed with fine parallel lines and measuring where they went 'wuzzy'/flickery/whatever. The conclusion was that in good sunlight, at a viewing/reading distance of about 14inches (35cm) the effective resolution of the eye is approximately 250-300dpi. i.e. a 0.1mm dot-pitch is sufficient. This resolution will decrease in low light as the pupil enlarges.

Since the limiting factor is angular resolution, the above figure scales with distance, so at 18cm (7inches) it's more like 500-600dpi. You might therefore specify quite a high pixel density for a handheld electronic widget.

For a TV 3-4 metres away, you would ideally want about 1mm dot pitch.
 

The Naked Scientists Forum

How many pixels can we actually perceive?
« Reply #7 on: 07/07/2009 20:42:46 »

 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums