Minimal resolution for retina resolution

[Rikard]

Hi, I was away and my posting here might remain less frequent than last year, but I thought I should share this with you:

I have been planning on getting a 27" or 30" 120 Hz screen once they become available for us mortals, but I also want to have better resolution than my current 1080p. According to Steve Jobs, humans cannot resolve more than 53.5 pixels per degree. This you might agree with or not, but lets go with that definition of retina resolution, it is as good as any.

Using the angular resolution as a constant, I then calculated the minimum horizontal resolution required for obtaining retina resolution as a function of the distance between the viewers eyes and the screen. You see the result below for a few popular screen sizes. I also indicate a few common resolutions with flat, thin, lines

So, this tells me that I need to sit 85 cm from the screen using my current 24" 1080p screen. Unfortunately I sit about 57 cm from screen while typing this, so I would need better resolution and in fact at this distance I would need a 4K display... However a 4K display would still be fine at this distance even if I would go for a 30" screen.

Meanwhile, it looks like most 50" and 55" TV's are already at retina resolution at typical viewing distances.

I hope this is of use to someone more than me.

(A 30" 4K 120 Hz screen and the graphics card to give that beast 120 FPS, I am not going to get close to that anytime soon!)

 

[KingFatty]

The problem with the chart is we don't know what number of pixels you are using for your inch sizes.

For example, you show a 27" 16:9 display. But how many pixels is that? There is a 27" display with 1920x1080 pixels, which has big fat pixels so you must sit far away. But there is also a 27" display with 2560x1440, which has very tiny pixels so you can sit much closer and still have 'retina' experience.

Furthermore, two different sized displays, such as 27" 1440p and 30" 1600p, can have identical sized pixels, so you can sit exactly the same distance to achieve retina experience. However, your chart suggests that this would be impossible.

So, maybe create the chart based on pixel size/pitch, and then show some common pixel size pitches used for various types of displays?

 

[Rikard]

The problem with the chart is we don't know what number of pixels you are using for your inch sizes.

For example, you show a 27" 16:9 display. But how many pixels is that? There is a 27" display with 1920x1080 pixels, which has big fat pixels so you must sit far away. But there is also a 27" display with 2560x1440, which has very tiny pixels so you can sit much closer and still have 'retina' experience.

That is what the graph above is telling you. For a given viewing distance and a given size of the screen, how many horizontal pixels do you need to achieve retina effect. Get it now? (The reason for keeping those two constant instead of the resolution is that it is much easier to change hardware than to change location in most scenarios.)

Furthermore, two different sized displays, such as 27" 1440p and 30" 1600p, can have identical sized pixels, so you can sit exactly the same distance to achieve retina experience. However, your chart suggests that this would be impossible.

You mean 2560x1440 and 2560x1600? Only the first one is 16:9 and is thus included, but it is not difficult for you to get the corresponding screen size for a different aspect ratio, in fact you already did that. So there is nothing impossible about the graph. You just did not read it right.

 

[blackened23]

edit: nevermind

 

[Phynaz]

FYI, human visual acuity is one arc second.

 

[KingFatty]

Ah thanks for clarification, I see now what you mean, how to read the chart.

So for "minimum horizontal pixels" of 2560, and a display of 16:10 aspect ratio, can I use your chart? That would be hard to figure out when you have two monitors that have the same horizontal number of pixels, but different vertical.

Such as 1920x1200 vs 1920x1080, or 2560x1600 vs 2560x1440. So I guess there is information we need to know how to assume and impose upon the chart, as in bring our own information to interpret result, because it's impossible to see the data?

It just hurts my brain to think of the number of pixels as a variable that changes. I mean I can't exactly change the number of pixels on my display. But I can change the viewing distance. Hmm I guess it's hard to make an intuitive chart that accounts for all of this.

But in the end, is it correct that we must know beforehand that the chart is only good for 16:9 aspect ratio displays? I mean, if I have a 30" 16:10 display, where does it fall on your chart?

 

[KingFatty]

Also, the lines on the chart are "exactly" retina resolution. However, what is beyond retina resolution or below retina resolution for a given display? I guess we know that as you move further away from the monitor (increasing viewing distance), the display becomes more and more retina. So, on your chart, would the entire right side of the chart be fully retina? It's just when you fall to the left of a given line that you are no longer retina?

Charts are nice for being able to tell a story with barely using any words. But I think I need a PhD to read your chart, and also I would need additional information that is not available anywhere in the chart, in order to interpret it?

 

[Rikard]

FYI, human visual acuity is one arc minute.

FIFY

 

[Rikard]

Ah thanks for clarification, I see now what you mean, how to read the chart.

You are welcome. And sorry for the late reply, as I said I am not able to check the forum as often as I used to.

So for "minimum horizontal pixels" of 2560, and a display of 16:10 aspect ratio, can I use your chart? That would be hard to figure out when you have two monitors that have the same horizontal number of pixels, but different vertical.

Such as 1920x1200 vs 1920x1080, or 2560x1600 vs 2560x1440. So I guess there is information we need to know how to assume and impose upon the chart, as in bring our own information to interpret result, because it's impossible to see the data?

...

But in the end, is it correct that we must know beforehand that the chart is only good for 16:9 aspect ratio displays? I mean, if I have a 30" 16:10 display, where does it fall on your chart?

Right, what really matters is the width of your screen. (Or, if I would have given it as minimum vertical pixels it would have been the height of course.) Since that is not something people normally know I plugged in the diagonal, which together with the aspect ratio gives you the width. So for a 16:10 display you would need to find the diagonal of a corresponding 16:9 display that have the same horizontal width as the 16:10 screen and use that in the table above. If you find it useful I could add some lines to the plot above instead.

It just hurts my brain to think of the number of pixels as a variable that changes. I mean I can't exactly change the number of pixels on my display. But I can change the viewing distance. Hmm I guess it's hard to make an intuitive chart that accounts for all of this.

Well, I could have turned it around like that, but usually I sit at a certain distance for a reason such as access to keyboard, size of living room etc, and these things tend to be harder to modify than buying a screen with a different resolution

Also, the lines on the chart are "exactly" retina resolution. However, what is beyond retina resolution or below retina resolution for a given display? I guess we know that as you move further away from the monitor (increasing viewing distance), the display becomes more and more retina. So, on your chart, would the entire right side of the chart be fully retina? It's just when you fall to the left of a given line that you are no longer retina?

Charts are nice for being able to tell a story with barely using any words. But I think I need a PhD to read your chart, and also I would need additional information that is not available anywhere in the chart, in order to interpret it?

Oh well, I have a PhD, that might be the problem right there! Let me try to walk you through it:

1. Decide what distance from the screen you want to sit. That tells you where on the horizontal axis you should look.
2. Decide which size of screen you want. That tells you what colored curve to look at.
3. The intersection of the colored curve (from size of screen) and the vertical dotted line (from viewing distance) tells you the minimum horizontal pixels you need by looking at the vertical axis for that point.
4. The flat horizontal lines are just there to help your eyes. They represent common horizontal resolutions.

So for example you view from 90 cm, you want a 24" 16:9 screen. Then you see from the blue curve that you would need approximately 1800 horizontal pixels. A comparison with the thin purple line, which corresponds to 1920 pixels, shows that if you would get a 1920x1080 24" screen you would have already obtain retina resolution at that distance (since the 1920 line is above the intersection between 90 cm and 24" curve).

Did it help? I am so used to look at graphs so I tend to be a bit blind to potential misconceptions others might have. A table might be easier for reading the values from, but it is not as compact and thus less forum friendly.

 

[Braxos]

It is for 16:9 monitor since I don't have a phd. So if you are seeting x cm away from monitor a inches you need a y horizontal resolution to get retina.

So if y is 16 then how much is z on 9. Z=(9÷16)*y => z= 0.5625*y

So try it z=0.5625*2560 => z=1440

Congrats

 

[Phynaz]

FIFY

Derp, yes you did!

 

[crisium]

I still find it odd we have 5" phones at 1920x1080, but you cannot get anything bigger than 1080 below 27" for PC monitors. We have 1080 monitors commonly available at 21.5", 23", 24", and even 27". But 1440 is only 27".

But I guess they are waiting for 3840x2160 to become standard. Once the TV manufacturers start churning these out in all sorts of sizes, I'm sure we'll get 2160 monitors from at least 24"-32" range.