Looking for an article about resolution

[aaronchan]

Dear all,

I'm looking for an old article which talks about resolution.
Part of it mention the resolving power of human vision, and talks about very accurate number of how many "dpi" we can see by different distance.
Has anyone know which article I'm talking about?

Thank you very much.

Regard,
Aaron

[Bart_van_der_Wolf]

Dear all,

I'm looking for an old article which talks about resolution.
Part of it mention the resolving power of human vision, and talks about very accurate number of how many "dpi" we can see by different distance.
Has anyone know which article I'm talking about?

Thank you very much.

Hi Aaron,

There have been a number of such articles, but perhaps my attached summary is enough for your requirements?

The 1 arc minute values are used for average visual acuity, and the 0.4 arc minute values are used for very high visual acuity (e.g. young persons or very well corrected eyesight).

Cheers,
Bart

[Mark D Segal]

Very useful Bart. So here in Canada where we can be both metrically and "imperially" challenged, and as I grew up in the "imperial system", let's use feet; assuming my viewing distance of a 13*19 sheet is arm's length, or about two feet (depending on your arms), and my vision is well-corrected, I can divide the 716 by 2 and I'm close to  360, which is one of the native resolutions (perhaps the most often applicable) of the Epson print head. But I could print lower and still be well within range of average human visual acuity - in fact according to this data at 140 PPI. This range makes sense to me, as I've seen from my own testing that from about 180 downward one begins to see the appearance of sharpness deteriorating, and below 140 is "no go" territory.

[Bart_van_der_Wolf]

Very useful Bart. So here in Canada where we can be both metrically and "imperially" challenged, and as I grew up in the "imperial system", let's use feet; assuming my viewing distance of a 13*19 sheet is arm's length, or about two feet (depending on your arms), and my vision is well-corrected, I can divide the 716 by 2 and I'm close to  360, which is one of the native resolutions (perhaps the most often applicable) of the Epson print head.

Yes, that's what the output file's PPI should be to meet the required visual acuity when viewed from that distance. Viewing it from a larger distance will not resolve more if the eye cannot resolve finer detail, the details will just become smaller to the eye and the finest detail can no longer be resolved. The contrast by which the eye resolves detail will become too low to separate details as one approaches the limiting resolution of one's eyes. That contrast varies with detail size. This also allows to visually improve resolution by boosting the contrast of the finest detail, with output sharpening.

But I could print lower and still be well within range of average human visual acuity - in fact according to this data at 140 PPI. This range makes sense to me, as I've seen from my own testing that from about 180 downward one begins to see the appearance of sharpness deteriorating, and below 140 is "no go" territory.

That's how that works. Of course, we photographers are not normal people so we tend to look at image detail from very short distances. So to somewhat satisfy that requirement, one could use a normal reading distance of 1 foot (or instead, the length of one's nose if a forensic analysis is required) for minimum required PPI. One can also put an obstacle, like a couch or table in front of the wall ...

This also assumes viewing at reasonable illumination levels, because at dark conditions our eyes lose resolution to spherical aberrations, and at high illumination levels we'll get into diffraction limited eye resolution.

Cheers,
Bart

[NAwlins_Contrarian]

From both reports and my own experience, similar figures apply. I haven't read the primary research, and what I did see many years ago was more about distance vision, but it was that 'perfect' human vision could see about 0.5 minute-of-angle (MOA), at least with fairly high-contrast subjects in black and white or certain colors; and that more typically achieved or mostly-corrected vision might be down to more like 1 MOA.

In the early 2000s, I was considering what it took to get 8x10 inch prints from digital files that looked good. For me, viewing distance on 8x10 inch prints is usually about 18 inches. At 18 inches, 0.5 MOA comes to 382 ppi,* and 1 MOA is 191 ppi. Once we got to the better 4 MP compact digital cameras, they could deliver an 8x10 inch print at 216 ppi with a real resolution more typically around 162 ppi**--and IMO those could look good, but were a little lacking in detail. Improvements in achievable 8x10s were visible but increasingly subtle but up to about 12 MP DSLRs, which gave 356 ppi with real resolutions of about 267 ppi, after which some or all among a bunch of factors*** made visible improvements very iffy and hard to come by.

Can you / I ever see more? Yes, but only under extreme conditions. When B&W laser printers went from 300 ppi to 600 ppi, the improvement in things like the smoothness of a lower-case italic i was visible to me. But that's an extreme case.
 

* A minute of angle is 1/60th of a degree. The height (h) occupied by that minute is calculated relative to the distance (d) from the viewer to the subject, h = d * TAN(1/60th degree). If we use d in inches, we get h in inches, and to convert to ppi, we use ppi / h (expressed in inches).

** 4 MP on 4:3 is 2304 x 1728 pixels; 1728 / 8 = 216. Effective linear resolution on the better ones was somewhere around 75% of what the pixel count implied.

*** Was the visible performance limited by my eyes, printers that were in various cases typically 250 or 300 ppi, lens performance, limited depth of field, sub-optimal processing / scaling? Yes, to all, I suspect.

[Schewe]

Part of it mention the resolving power of human vision, and talks about very accurate number of how many "dpi" we can see by different distance.
Has anyone know which article I'm talking about?

Was it The Right Print Resolution?

[aaronchan]

Was it The Right Print Resolution?

Thank you very much Jeff
This is exactly what I'm looking for!
You are amazing!

Regard,
Aaron

[Schewe]

You are amazing!

You are very astute!