The fact is we are able to distinguish details down to 8 arcminutes at 16 feet. And just so you know, that is 40% of the size of a human cone photoreceptor.
How does 8 arcminutes at 16 feet compare to the other claim:
"Maybe you'd like to source some research that says the retina can pick out detail higher than 326 ppi at 10 inches away before making nonsensical arguments."
I assume that the paper shows greater acuity is possible than 326 PPI at 10 inches, but I can't convert the two.
The brain also plays a huge roll in perceivable resolution. Our eyes jitter and then the brain reassembles the input into coherent imagery. The jittering is actually advantageous though. It allows perception of detail smaller than the size of the receptor. For instance, lines thinner than a single receptor can still be seen.
(Or at least that is my recollection of something I might have read somewhere years ago maybe.)
The brain also plays a huge roll in perceivable resolution. Our eyes jitter and then the brain reassembles the input into coherent imagery. The jittering is actually advantageous though. It allows perception of detail smaller than the size of the receptor. For instance, lines thinner than a single receptor can still be seen.
(Or at least that is my recollection of something I might have read somewhere years ago maybe.)
Yes, that's correct. But the size of the optical blur on the retina is about 4microns, which is about 2 to 4 photoreceptors on the fovea. In this sense the jitter helps enormoulsy to improve resolution. In the special case of Vernier acuity, it does wonders. But the 300 dpi is the optimal for everyday tasks.
Vernier acuity means aligning to vertical lines with each other. So if you are doing somer vernier readings (callipers) then that is really good. Apart from that, perhaps when passing thread in a needle eye... Not helpful on small screen, I think.
. I would also recommend as starters, "Sensation and Perception" 5th. ed. by Harvey Richard Schiffman, "Vision Science, photons to phenomenology" by Stephen Palmer, and "Foundations of Vision" by Brian A. Wandell, and "Visual Perception, a clinical orientation" by Steven H. Schwartz.
A million years ago when I took Introduction to Psychology in college, we read "Eye and Brain" by an author I can't remember.
Is that still a standard text? Has it been updated? Is much of what (everything?) I learned there wrong? Or are the basics still agreed upon, with new research refining the 1970's conclusions?
If you could recommend a single layman's/college kid's text for me to read with current research on the subject, I'd appreciate it. I remember being fascinated by the subject of eye/nerve physiology vs. brain interpretation.
Considering that as contrast goes down, we lose acuity quickly, even the paper by Levi shows that, we have to know what contrast levels we are talking about. This is well known photographically, which is why lenses are characterized at differing contrast levels.
How are " lenses ... characterized at differing contrast levels"?
What units of characterization are used? All I am familiar with is the maximum f-stop rating, which does not use contrast as a factor.
Yes, that's correct. But the size of the optical blur on the retina is about 4microns, which is about 2 to 4 photoreceptors on the fovea. In this sense the jitter helps enormoulsy to improve resolution. In the special case of Vernier acuity, it does wonders. But the 300 dpi is the optimal for everyday tasks.
Vernier acuity means aligning to vertical lines with each other. So if you are doing somer vernier readings (callipers) then that is really good. Apart from that, perhaps when passing thread in a needle eye... Not helpful on small screen, I think.
Thank you iStud, and others for correcting Mr. 'vernier acuity' Bowser. I was annoyed at his 'patently false' claim, particularly since as a long time screen user in late 40's i can only just read most small text as it is. Already the iphone (3), images looks incredible, so increasing resolution 4 fold, must surely take resolution to a point where few (o.1% of the population could discern any pixellation.
I tried basic trig, based on 8 arc seconds, for a handheld device at 20 inches - seems like the claims are not patently false.
All of a sudden I'm taking a closer look at all the 300dpi printers scattered around my office. We use various zebra direct thermal printers to print labels for use in inventory and manufacturing.
While I can definitely see jaggies on printed text at 300dpi, this is mostly because these thermal transfer printer are not grayscale. For an individual pixel, they can only do 100% black or clear. Also, printed pixels are more distinct than lcd pixels because there is no bleed through between adjacent pixels.
This isn't meant as criticism or to say that a 300dpi display is insufficient. Rather, just a few random thoughts on dpi. Personally, I can't wait to take delivery of the new iPhone.
All of a sudden I'm taking a closer look at all the 300dpi printers scattered around my office. We use various zebra direct thermal printers to print labels for use in inventory and manufacturing.
While I can definitely see jaggies on printed text at 300dpi, this is mostly because these thermal transfer printer are not grayscale. For an individual pixel, they can only do 100% black or clear. Also, printed pixels are more distinct than lcd pixels because there is no bleed through between adjacent pixels.
This isn't meant as criticism or to say that a 300dpi display is insufficient. Rather, just a few random thoughts on dpi. Personally, I can't wait to take delivery of the new iPhone.
A 300dpi printer, has alternate 150dpi rows half a point displaced from each other. This corrects from jitter during printing. So the actual resolution you get 150dpi, not 300dpi as you would get from a display. That's why you see lower quality in "equivalent" prints
Better late than never, a reality since 2006 for some Nokia users...
Wow, that's late. My NEC had it in 2005, and I think that model originally came out in late '04. Though i daresay the experience on Nokias was equally as unsatisfactory - small screen, poor picture and stream quality, massive data costs.
Seems like it might actually be a usable feature on the iPhone.
But your understanding of optics, and geometry don't help you. As I have pointed out in my previous posts, there are a number of factual mistakes in your arguments that undermine your credibility.
While you are correct in your assertions about the vernier acuity, the numbers you provide are wrong, the figure of 8minarc is baseless and wrong. Photoreceptors are much smaller than that!
Vernier acuity is a very interesting topic, and it has its applications. Reading and playing games is not one of them. But correct me please.
The limits of normal human vision (every day tasks) is set by the arrangement of retinal cones, but most importantly the size of the PSF on the retina. The combination of these 2 factors gives you the standard 300dpi. If a PhD in visual perception did not give you the background needed to understand this very basic concept, and being able to differentiate between normal every day activities, and super-resolution in specialised tasks like vernier acuity, you PhD training left many holes behind, that will have to be filled in over the years.
How about we stop the ad hominems and arguments about vernier acuity and talk about the IPHONE? Hmm?
Thank you iStud, and others for correcting Mr. 'vernier acuity' Bowser. I was annoyed at his 'patently false' claim, particularly since as a long time screen user in late 40's i can only just read most small text as it is. Already the iphone (3), images looks incredible, so increasing resolution 4 fold, must surely take resolution to a point where few (o.1% of the population could discern any pixellation.
I tried basic trig, based on 8 arc seconds, for a handheld device at 20 inches - seems like the claims are not patently false.
Of course is not patently false.. Try it with 1 arcminute and 12 inches away.
How about we stop the ad hominems and arguments about vernier acuity and talk about the IPHONE? Hmm?
Looks great. I'm upgrading.
Good point. the argument arose because this guy said that the 300dpi being the optimal resolution of display was "patently false" due to his vernier acuity PhD...
the point was to show the statement about resolution during the keynote was correct.
It is about the iPhone after all. At least an aspect of it. Not just its specs.
A 300dpi printer, has alternate 150dpi rows half a point displaced from each other. This corrects from jitter during printing. So the actual resolution you get 150dpi, not 300dpi as you would get from a display. That's why you see lower quality in "equivalent" prints
Good point, but in my mind that is still 300dpi. Perhaps what you're referring to is the fact that pixels are slightly over-sized such that there is no screen door effect, especially when accounting for slop in the media feed speed.
I see this as reinforcing the notion that 300dpi is a sufficiently high resolution for almost all uses of an LCD. It should definitely look better than a 300dpi 1-bit printer. One of the main reasons for ludicrously high dpi printers is so that they can do imperceptible dithering.
Good point, but in my mind that is still 300dpi. Perhaps what you're referring to is the fact that pixels are slightly over-sized such that there is no screen door effect, especially when accounting for slop in the media feed speed.
I see this as reinforcing the notion that 300dpi is a sufficiently high resolution for almost all uses of an LCD. It should definitely look better than a 300dpi 1-bit printer. One of the main reasons for ludicrously high dpi printers is so that they can do imperceptible dithering.
Wow and this is why I love Mac forums--learning random crap like how the eye works. Thanks iStud, dfiler, Bowser for starting the discussion. Very relevant to the iPhone! Great stuff! Makes me even more excited to get the iPhone 4, sounds like the screen really is the best out there and the best an eye could possibly see. Sweet!
No argument there. It's just that between the poor sensitivity of these tiny sensors, and the really slow lenses they're coupled with, a flash will help. They're really just intended for a portrait shot at close distance. Anything further away that about three feet won't get much benefit anyway. I'm assuming that the flash is automatic, or that the software will compensate for the distance. Two LED's will give maybe 4.5 feet of effective light. It's better than nothing with these things.
That's an interesting point when you look at the backlit sensor. It is much more sensitive in low light environments, and it appears to have a better signal to noise ratio. I wonder if the flash is variable, or just a full on/full off sort of situation? it would be ideal if the phone could sample ambient light and regulate the brightness of the flash.
Comments
The fact is we are able to distinguish details down to 8 arcminutes at 16 feet. And just so you know, that is 40% of the size of a human cone photoreceptor.
How does 8 arcminutes at 16 feet compare to the other claim:
"Maybe you'd like to source some research that says the retina can pick out detail higher than 326 ppi at 10 inches away before making nonsensical arguments."
I assume that the paper shows greater acuity is possible than 326 PPI at 10 inches, but I can't convert the two.
(Or at least that is my recollection of something I might have read somewhere years ago maybe.)
The brain also plays a huge roll in perceivable resolution. Our eyes jitter and then the brain reassembles the input into coherent imagery. The jittering is actually advantageous though. It allows perception of detail smaller than the size of the receptor. For instance, lines thinner than a single receptor can still be seen.
(Or at least that is my recollection of something I might have read somewhere years ago maybe.)
Yes, that's correct. But the size of the optical blur on the retina is about 4microns, which is about 2 to 4 photoreceptors on the fovea. In this sense the jitter helps enormoulsy to improve resolution. In the special case of Vernier acuity, it does wonders. But the 300 dpi is the optimal for everyday tasks.
Vernier acuity means aligning to vertical lines with each other. So if you are doing somer vernier readings (callipers) then that is really good. Apart from that, perhaps when passing thread in a needle eye... Not helpful on small screen, I think.
. I would also recommend as starters, "Sensation and Perception" 5th. ed. by Harvey Richard Schiffman, "Vision Science, photons to phenomenology" by Stephen Palmer, and "Foundations of Vision" by Brian A. Wandell, and "Visual Perception, a clinical orientation" by Steven H. Schwartz.
A million years ago when I took Introduction to Psychology in college, we read "Eye and Brain" by an author I can't remember.
Is that still a standard text? Has it been updated? Is much of what (everything?) I learned there wrong? Or are the basics still agreed upon, with new research refining the 1970's conclusions?
If you could recommend a single layman's/college kid's text for me to read with current research on the subject, I'd appreciate it. I remember being fascinated by the subject of eye/nerve physiology vs. brain interpretation.
Video calling now a reality...wow FaceTime
Better late than never, a reality since 2006 for some Nokia users...
Considering that as contrast goes down, we lose acuity quickly, even the paper by Levi shows that, we have to know what contrast levels we are talking about. This is well known photographically, which is why lenses are characterized at differing contrast levels.
How are " lenses ... characterized at differing contrast levels"?
What units of characterization are used? All I am familiar with is the maximum f-stop rating, which does not use contrast as a factor.
Yes, that's correct. But the size of the optical blur on the retina is about 4microns, which is about 2 to 4 photoreceptors on the fovea. In this sense the jitter helps enormoulsy to improve resolution. In the special case of Vernier acuity, it does wonders. But the 300 dpi is the optimal for everyday tasks.
Vernier acuity means aligning to vertical lines with each other. So if you are doing somer vernier readings (callipers) then that is really good. Apart from that, perhaps when passing thread in a needle eye... Not helpful on small screen, I think.
Thank you iStud, and others for correcting Mr. 'vernier acuity' Bowser. I was annoyed at his 'patently false' claim, particularly since as a long time screen user in late 40's i can only just read most small text as it is. Already the iphone (3), images looks incredible, so increasing resolution 4 fold, must surely take resolution to a point where few (o.1% of the population could discern any pixellation.
I tried basic trig, based on 8 arc seconds, for a handheld device at 20 inches - seems like the claims are not patently false.
I think they use Windows NT don't they?
The space shuttle use IBM's AP-101s avionics computers.
While I can definitely see jaggies on printed text at 300dpi, this is mostly because these thermal transfer printer are not grayscale. For an individual pixel, they can only do 100% black or clear. Also, printed pixels are more distinct than lcd pixels because there is no bleed through between adjacent pixels.
This isn't meant as criticism or to say that a 300dpi display is insufficient. Rather, just a few random thoughts on dpi. Personally, I can't wait to take delivery of the new iPhone.
All of a sudden I'm taking a closer look at all the 300dpi printers scattered around my office. We use various zebra direct thermal printers to print labels for use in inventory and manufacturing.
While I can definitely see jaggies on printed text at 300dpi, this is mostly because these thermal transfer printer are not grayscale. For an individual pixel, they can only do 100% black or clear. Also, printed pixels are more distinct than lcd pixels because there is no bleed through between adjacent pixels.
This isn't meant as criticism or to say that a 300dpi display is insufficient. Rather, just a few random thoughts on dpi. Personally, I can't wait to take delivery of the new iPhone.
A 300dpi printer, has alternate 150dpi rows half a point displaced from each other. This corrects from jitter during printing. So the actual resolution you get 150dpi, not 300dpi as you would get from a display. That's why you see lower quality in "equivalent" prints
No one makes a 64GB phone. You actually going to spend $300 on a 32GB micro SD card?
Huh, you can find them around $90, and their price will keep falling... Dunno how you invented that $300 price...
With this in mind the $100 difference of price between the iPhone 4 16GB and its 32GB version, i.e. $100 for an addition of 16GB, is really a scam.
Better late than never, a reality since 2006 for some Nokia users...
Wow, that's late. My NEC had it in 2005, and I think that model originally came out in late '04. Though i daresay the experience on Nokias was equally as unsatisfactory - small screen, poor picture and stream quality, massive data costs.
Seems like it might actually be a usable feature on the iPhone.
But your understanding of optics, and geometry don't help you. As I have pointed out in my previous posts, there are a number of factual mistakes in your arguments that undermine your credibility.
While you are correct in your assertions about the vernier acuity, the numbers you provide are wrong, the figure of 8minarc is baseless and wrong. Photoreceptors are much smaller than that!
Vernier acuity is a very interesting topic, and it has its applications. Reading and playing games is not one of them. But correct me please.
The limits of normal human vision (every day tasks) is set by the arrangement of retinal cones, but most importantly the size of the PSF on the retina. The combination of these 2 factors gives you the standard 300dpi. If a PhD in visual perception did not give you the background needed to understand this very basic concept, and being able to differentiate between normal every day activities, and super-resolution in specialised tasks like vernier acuity, you PhD training left many holes behind, that will have to be filled in over the years.
How about we stop the ad hominems and arguments about vernier acuity and talk about the IPHONE? Hmm?
Looks great. I'm upgrading.
Thank you iStud, and others for correcting Mr. 'vernier acuity' Bowser. I was annoyed at his 'patently false' claim, particularly since as a long time screen user in late 40's i can only just read most small text as it is. Already the iphone (3), images looks incredible, so increasing resolution 4 fold, must surely take resolution to a point where few (o.1% of the population could discern any pixellation.
I tried basic trig, based on 8 arc seconds, for a handheld device at 20 inches - seems like the claims are not patently false.
Of course is not patently false.. Try it with 1 arcminute and 12 inches away.
How about we stop the ad hominems and arguments about vernier acuity and talk about the IPHONE? Hmm?
Looks great. I'm upgrading.
Good point. the argument arose because this guy said that the 300dpi being the optimal resolution of display was "patently false" due to his vernier acuity PhD...
the point was to show the statement about resolution during the keynote was correct.
It is about the iPhone after all. At least an aspect of it. Not just its specs.
A 300dpi printer, has alternate 150dpi rows half a point displaced from each other. This corrects from jitter during printing. So the actual resolution you get 150dpi, not 300dpi as you would get from a display. That's why you see lower quality in "equivalent" prints
Good point, but in my mind that is still 300dpi. Perhaps what you're referring to is the fact that pixels are slightly over-sized such that there is no screen door effect, especially when accounting for slop in the media feed speed.
I see this as reinforcing the notion that 300dpi is a sufficiently high resolution for almost all uses of an LCD. It should definitely look better than a 300dpi 1-bit printer. One of the main reasons for ludicrously high dpi printers is so that they can do imperceptible dithering.
Good point, but in my mind that is still 300dpi. Perhaps what you're referring to is the fact that pixels are slightly over-sized such that there is no screen door effect, especially when accounting for slop in the media feed speed.
I see this as reinforcing the notion that 300dpi is a sufficiently high resolution for almost all uses of an LCD. It should definitely look better than a 300dpi 1-bit printer. One of the main reasons for ludicrously high dpi printers is so that they can do imperceptible dithering.
Agreed.
Agreed.
Wow and this is why I love Mac forums--learning random crap like how the eye works. Thanks iStud, dfiler, Bowser for starting the discussion. Very relevant to the iPhone! Great stuff! Makes me even more excited to get the iPhone 4, sounds like the screen really is the best out there and the best an eye could possibly see. Sweet!
No argument there. It's just that between the poor sensitivity of these tiny sensors, and the really slow lenses they're coupled with, a flash will help. They're really just intended for a portrait shot at close distance. Anything further away that about three feet won't get much benefit anyway. I'm assuming that the flash is automatic, or that the software will compensate for the distance. Two LED's will give maybe 4.5 feet of effective light. It's better than nothing with these things.
That's an interesting point when you look at the backlit sensor. It is much more sensitive in low light environments, and it appears to have a better signal to noise ratio. I wonder if the flash is variable, or just a full on/full off sort of situation? it would be ideal if the phone could sample ambient light and regulate the brightness of the flash.