Hello I came to this website http://6mpixel.org/en/ Talking about how 6 mp is the best resolution for image quality and was wondering what you all think about it? Is there some truth to it? If there is, is the difference effectively noticeable?
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Thoughts on "Best picture quality with 6 megapixels!" ?
- Thread starter MindBrain
- Start date
- Sort by reaction score
I think he was referring more to the smaller sensors in p/s cameras, and perhaps APS-C size sensors, and a bit of a criticism of the megapixel race in comsumer cameras - which I tend to agree with. It seems the issue is more about sensor photosite density, but I haven't read too much of what he says - yet.
Does the Canon G9 fall under p/s category as you say, or will it have a larger sensor then the more compact cameras? Wondering if when buying a camera I should be looking at the sensor size and be going for a larger one?
The Canon G9 has a 1/1.7" sensor, which is on the larger side for point and shoot cameras.
1/1.7" translates to a sensor that's 7.6mm x 5.7mm.
The typical 1/2.5" sensor in most point and shoots is 5.76 x 4.29.
APS-C in most DSLRs is around 23 x 15.
"full frame" is 36 x 24.
http://www.dpreview.com/learn/?/Glossary/Camera_System/sensor_sizes_01.htm
What Miloblithe said...! Great answer, great information.
One thing to remember is that technology never stands still, and I would hesitate to assume that what holds up today will not change in a few years.
One thing to remember is that technology never stands still, and I would hesitate to assume that what holds up today will not change in a few years.
I have only anecdotal evidence for this, and clearly there are other factors at play here, but looking the quality of shots taken with a "modern" Olympus 12MP P&S as compared to my old Olympus Stylus P&S (3.2MP) or Olympus C5060 (5.1MP), there's no comparison. The C5060, especially, has far less noise at ISO 400 than the 12MP model I tried last week. What's worse, the new model goes up to ISO 6400 (!), at which point the images are basically useless.
What is ironic is that these pixels are being wasted anyway; very few people make prints that would benefit from these huge megapixel counts, and those that do are mainly using dSLRs anyway. Most people would be far better off with a 5-6MP P&S, from which they could easily make nice, reasonably low noise 8x10 prints for their granny. But people fall for the marketing crap.
This is one of my major complaints about digital P&S vs. old film P&S. Yes, there is the immediacy factor and convenience of a digital P&S vs. film. However, these tiny sensors are so bad that IQ really suffers. I would wager that even grocery store 400 ISO Fuji Superia is going to look far better than a typical 400 ISO shot from a modern 12MP digital P&S. Moreover, I could stick Velvia or Kodachrome or Provia or anything in a film P&S and get all of the benefits of that particular film. But with a digital, I'm stuck with whatever crap sensor I start with; no way to improve or tailor things to my liking.
dSLRs are, of course, immune to this, since the sensor is big enough (and pixel density low enough) to make this a non issue. Frankly, the 6.3MP from my 10D is more than I've ever needed; I could use a faster buffer, a bigger screen, or slightly better high ISO performance, but for my typical print size, this sensor is fine.
What is ironic is that these pixels are being wasted anyway; very few people make prints that would benefit from these huge megapixel counts, and those that do are mainly using dSLRs anyway. Most people would be far better off with a 5-6MP P&S, from which they could easily make nice, reasonably low noise 8x10 prints for their granny. But people fall for the marketing crap.
This is one of my major complaints about digital P&S vs. old film P&S. Yes, there is the immediacy factor and convenience of a digital P&S vs. film. However, these tiny sensors are so bad that IQ really suffers. I would wager that even grocery store 400 ISO Fuji Superia is going to look far better than a typical 400 ISO shot from a modern 12MP digital P&S. Moreover, I could stick Velvia or Kodachrome or Provia or anything in a film P&S and get all of the benefits of that particular film. But with a digital, I'm stuck with whatever crap sensor I start with; no way to improve or tailor things to my liking.
dSLRs are, of course, immune to this, since the sensor is big enough (and pixel density low enough) to make this a non issue. Frankly, the 6.3MP from my 10D is more than I've ever needed; I could use a faster buffer, a bigger screen, or slightly better high ISO performance, but for my typical print size, this sensor is fine.
Technology changes quickly. Physics never changes.
Much of what is said about small sensors is based on physics. No mather how advanced the camera if the lens has a given f-stop and the pixels are a given size the same number of photons will hit the pixel. Over time technology will allow those photons to be put to better use but "noise" is determined by the fact that light is "packaged" into photons and when the number of photons is low you at at the mercy of the statistics of small numbers. In other words the root cause of noise in high ISO images is due to the nature of light. It will ALWAYS be the case that small sensors are noisier than larger sensors. Same goes for resolving power. It is a law of nature that the resolving power of a lens is limited by it's actual diameter. Small lenses can never be as sharp as large ones. The BEST technology can hope for is to approach the limits imposed by physical laws. With optics we are already very close.
What happed was that many years ago afordable cameras did not have enough pixels and everyone knw that digital picure looked "blocky" and more pixles ment a better image. Way back when this was true.
We see the same thing with LCD monitors. Years ago they could not display fast moving images and even curcur movement left "ghost trails" on the screen so they started using "responce time" as an indicator of quality. Well now they are in the single digit times and it absolutly does not mater. with vertical refresh rates fater then responce time.
The thing is that casual users simply do not understand the technology and the people trying to sell cameras are in a bad way. How to you promote your camera is being better to a buyer who knows absolutly nothing? Your only choise is to latch onto the one number that he might understand and say "my number is bigger"
The other thing that is driving this is that consummers are very non-critical of poor image quality. 99% of them just don't care and can't tell crap when they see it.
I agree the physics won't change, but perhaps our understanding of it might... after all at some point men came to realize that the sun was not revolving around the earth. Nothing changed, but we got smarter.
Improvements in how we deal with physical limits is what technology seems to bring. New optical formulas and improved manufacturing methods and coatings all tailor new optics for digital sensors, as opposed to film. Lenses have more difficult jobs to do, so they have to get better, make the light path straighter out at the edges of the sensors and improve convergence of the various wavelengths of colors where they meet the perfectly flat, reflective surface of the sensors. These are issues older lenses didn't have to contend with. Sensor photobuckets can become more efficient at collecting light, requiring less "gain" which causes noise and software can learn how to remove the noise more efficiently by artificially rebuilding/enhancing the image more realistically.
I don't know what modern technology will manage to overcome next, but there have been too many leaps just in my lifetime for me to say, "That's it! Done! It can't be topped or overcome..." I agree, basic fundamentals of physics still apply, it's just how they are confronted that keeps me interested in the story.
Speaking as someone with a degree in physics from back in the stone age... I wouldn't hold my breath on this, because it's a fatally flawed analogy. You're trying to draw a parallel between a belief that was demonstrably wrong - a philosophical belief that only held up while it couldn't be measured - with a quantum phenomenon that's demonstrably right.
Physics occasionally finds a way to "cheat" and nibble away at the edges, but the fundamental physical principle can't be violated.
I think the P&S photo companies, though, have a very good idea about the types of photos people are taking, and realize that the physical limitations of the newer sensors will have very little (to no) impact on the poor quality photos the average person takes. Who cares about poor low-light sensor performance when a photo is already blurred due to camera movement? The market isn't driven by the small number of people that know what they're doing and want a technically good P&S - that'd just be a blip on their overall sales figures. So they do what makes sense, and market higher-density sensors (and maybe pay some lip-service to the few that have heard about the "noise problem" of these sensors by including some crap "feature" like pixel binning).
Frankly, when I look at the random photos available on Flickr I rarely see anything that couldn't have been taken (at likely higher quality, no less) with my old 3 megapixel Canon S30.
Okay, in a strict sense... you got me!
But, let's not forget that what eventually became a demonstrably wrong belief, was not exactly demonstrably wrong at the time... and we mere mortals did get smarter... and proved that belief wrong. Who's to say there aren't going to be other surprises down the pike in the world of physics, at least as we know it to be now?If you're only interested in web photos then a 1 megapixel camera is more than adequate.