Izu
Leading Member
Well, that's how film actually works... if a grain is hitted by a number of photons it wil became black during development... the trick is: use many, many, many little, little, little grains (or pixels).
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Hopefully the Canon 7D Review Puts to Rest the Too Many Megapixels Issue
- Thread starter Michael Fusco
- Start date
Gary Martin
Veteran Member
It is interesting that with 50% more pixels than the D300, the D7 only has 12% more resolution. So there seems to be a decreasing effectiveness in added more and more pixels to any sensor.
--
http://www.pixelstatic.com
http://www.flickr.com/photos/pixelstatic/
--
http://www.pixelstatic.com
http://www.flickr.com/photos/pixelstatic/
John Sheehy
Forum Pro
The 7D has 22% more linear pixel resolution. You need a lens and technique that is mostly limited by the 7D pixel pitch to get 22% more resolution at high contrast (it is guaranteed at low contrast).
That depends on the lens and technique. There are many lenses with high-contrast resolution well beyond what 4.3 micron pixels can record.
--
John
Gary Martin
Veteran Member
John Sheehy
Forum Pro
Nope. They are the same thing; you're comparing apples to oranges (length to area).
The number of pixels does not represent relative resolution; the square root of the number of pixels does. Measured resolution typically measures the number of lines resolvable across an image, not the number of coins recordable with legible dates.
--
John
John W Peterson
Senior Member
You raise a valid point. While Canon may be inclined to introduce new lower end digital cameras when it feels the market is ready for or "wants" them, it seems capable of holding back and introducing new professional series cameras when it feels technically ready to do so.
As example, marketing would have demanded a 1DIV much earlier than canon introduced it, as a way to try to shed the difficult reputation that the 1DIII had accumulated early on in it's debut.
So to to the issue - given the notion that between the introduction of the 5DII and the 7D, you suspect that Canon has made significant technical advance in photosite noise, the next reasonable question becomes, are we looking at the sensor technology that will be harnessed (albeit in FF form) to make the 1DsIV.
I have used, and am quite fond of the 1DsIII since it was intoduced. It is a magnficent camera under the right circumstances, which is to say eneough light to allow low iso, and controlled contrast / lighting ratios so as to avoid exceeding the range of the sensor. Nonetheless, such things are not always possible for the landscape shooter, who, with the setting sun or other challenging lighting may need to push that camera beyond it's preferred "sweet spot".
While many might look to a 1DsIV as yet another generation of "more pixels", it would seem to me that a sensor with the other improvements offered by the 7D technology, might be more to the point of what we really "need".
I put "need" in quotes, conceding the point that the abilities of the camera, far exceeded by own skill set, back in the 1970's and I have been struggling to catch up ever since !
As example, marketing would have demanded a 1DIV much earlier than canon introduced it, as a way to try to shed the difficult reputation that the 1DIII had accumulated early on in it's debut.
So to to the issue - given the notion that between the introduction of the 5DII and the 7D, you suspect that Canon has made significant technical advance in photosite noise, the next reasonable question becomes, are we looking at the sensor technology that will be harnessed (albeit in FF form) to make the 1DsIV.
I have used, and am quite fond of the 1DsIII since it was intoduced. It is a magnficent camera under the right circumstances, which is to say eneough light to allow low iso, and controlled contrast / lighting ratios so as to avoid exceeding the range of the sensor. Nonetheless, such things are not always possible for the landscape shooter, who, with the setting sun or other challenging lighting may need to push that camera beyond it's preferred "sweet spot".
While many might look to a 1DsIV as yet another generation of "more pixels", it would seem to me that a sensor with the other improvements offered by the 7D technology, might be more to the point of what we really "need".
I put "need" in quotes, conceding the point that the abilities of the camera, far exceeded by own skill set, back in the 1970's and I have been struggling to catch up ever since !
John,
Have I seen a compact shot without noise reduction? Yes, if you count RAW.
So you believe that high mp count high density chip compacts don't suffer image degradation problems at high ISO? Jesus. I'm tempted to ask you if you actually use cameras in the real world or just study the spec and test shots under ideal conditions but me and you will never see eye to eye on this John so it's pointless continuing this.
You buy and use what gear you want and I'll buy and use what I want.
Have I seen a compact shot without noise reduction? Yes, if you count RAW.
So you believe that high mp count high density chip compacts don't suffer image degradation problems at high ISO? Jesus. I'm tempted to ask you if you actually use cameras in the real world or just study the spec and test shots under ideal conditions but me and you will never see eye to eye on this John so it's pointless continuing this.
You buy and use what gear you want and I'll buy and use what I want.
John Sheehy
Forum Pro
Converted with something like dcraw, or with something that destroys detail with NR?
I manually convert high exposure index image from my G9, apply minimal noise reduction (filtering of high-frequency chroma), and at small sizes you can't even tell it's from high ISO. Forget all the fools who look at high-ISO high-MP compact images at sizes about 4 feet across or greater on their mon
No, they don't. Do you know the difference between an image and a pixel?
I take about a thousand "real photos" for every test shot.
There's no difference between real world and test shots, except that the test shots can be controlled for accurate comparison. The PQ and IQ does not change for a camera depending on your intentions. the fact that you think this would make a difference for the discussion at hand shows that you have no idea what you are talking about. You are just one of many people lost in a world of illusion.
--
John
Clint Sanders
Senior Member
But pixel counts increase at an exponential rate, so it evens out.
Clint Sanders
Senior Member
Like I say, if your theoretical results don't align with real-world results, your theory is wrong.
John Sheehy
Forum Pro
My real world results are that my older digicams with less pixels look like cr*p compared to my G9 at high ISOs, and my G9 looks a little sad compared to some G10 samples I have seen. There has been no general deterioration of RAW IQ in digicams with increased pixel density; that is a myth and an illusion.
--
John
igb
Senior Member
Test shot are not theories but real world shots under controlled conditions .
--
-------------------------------------------------------
My Galleries: http://webs.ono.com/igonzalezbordes/index.html
--
-------------------------------------------------------
My Galleries: http://webs.ono.com/igonzalezbordes/index.html
John W Peterson
Senior Member
To Gary's point -
This has been pointed out - when camera pixel count "rocketed" from 6 to 8 megapixels, and then 8 to 10 MP; assuming that the performance of the system (including lens) is limited by the resolution of the sensor only, one would need to at least double pixel count to produce "notable" improvements, and a going up by "double-double" i.e. four fold should produce a very substantiative step up.
Hence, going from a 6 to a 24 MP camera would produce discernible improvement in resolution when looking at real world images. This is making several assumptions as to what those real world circumstances are, but most of us would agree that going from 12 to 18 MP, which is less than a "doubling" of pixel count produces improvements that are technically measurable, but modest at best, and of equivocal likelihood to effect most printed images.
One reassuring note - we do not see (can you just imagine it) floods of recently married couples coming back in every year or two demanding to have their wedding rephotographed with "more pixels". That would be quite a racket of we could get that to happen, but customers that had their wedding recorded by a professional armed with an 8MP camera several years ago, still seem quite happy today.
This has been pointed out - when camera pixel count "rocketed" from 6 to 8 megapixels, and then 8 to 10 MP; assuming that the performance of the system (including lens) is limited by the resolution of the sensor only, one would need to at least double pixel count to produce "notable" improvements, and a going up by "double-double" i.e. four fold should produce a very substantiative step up.
Hence, going from a 6 to a 24 MP camera would produce discernible improvement in resolution when looking at real world images. This is making several assumptions as to what those real world circumstances are, but most of us would agree that going from 12 to 18 MP, which is less than a "doubling" of pixel count produces improvements that are technically measurable, but modest at best, and of equivocal likelihood to effect most printed images.
One reassuring note - we do not see (can you just imagine it) floods of recently married couples coming back in every year or two demanding to have their wedding rephotographed with "more pixels". That would be quite a racket of we could get that to happen, but customers that had their wedding recorded by a professional armed with an 8MP camera several years ago, still seem quite happy today.
John Sheehy
Forum Pro
With images printed at small sizes, and the diffusion of the printers, how much difference would you expect to jump off the print? Crop from both cameras, however, and bring the pixel sizes well above the diffusion radius and the threshold of obvious, course visibility, and even the difference between 12 and 18MP will be obvious (if the subject was in focus and stable, of course).
--
John
John W Peterson
Senior Member
Given a lens, as has been described above (perhaps the new Canon L macro which is by the way a thing of beauty) I would imagine that the 7D will shine, even with aggressive cropping and subsequent use of up-sizing by e.g. the genuine fractals plug-in.
But, just as 36 exposure rolls of film taught us to be careful with burning up exposures, digital cameras seem to want to teach us to crop before rather than after pressing the shutter.
Still, a few extra pixels, as you say, are very nice to have...
But, just as 36 exposure rolls of film taught us to be careful with burning up exposures, digital cameras seem to want to teach us to crop before rather than after pressing the shutter.
Still, a few extra pixels, as you say, are very nice to have...
Michael, do you think that you are representative of the "vast majority" of 7D buyers?
I don't.
You may like using the ultra-high megapixel bodies for your work but most don't need that much resolution. Canon can make more profits by selling them, and can get buyers to accept them because of idiotic postings like yours. I, for one, don't like having to pay for your supposed needs. Why don't you just buy a 1Ds or a medium format back, and let the lesser photographers get by with a more reasonable amount of megapixels on out mid-range cameras?
I don't.
You may like using the ultra-high megapixel bodies for your work but most don't need that much resolution. Canon can make more profits by selling them, and can get buyers to accept them because of idiotic postings like yours. I, for one, don't like having to pay for your supposed needs. Why don't you just buy a 1Ds or a medium format back, and let the lesser photographers get by with a more reasonable amount of megapixels on out mid-range cameras?
It will not end the argument when other reviews have found a problem with the 7D due to the increase in MP. Not at high ISO but at the lower end where they said it was clearly visible on plain surfaces.
You're kind of giving a blanket reason for your argument that people always say that newer technology is not needed. In some cases I think it's true but with mega pixels on a DSLR sensor there's a negative to the increase as opposed to say an increase in RAM for a PC. Personally I can't see a downside to more PC memory but I can see a downside to more mega pixels in a DSLR sensor, i.e. NOISE!
They keep bumping up the mega pixels without upgrading the image chip or increasing the size of the sensor. The 5D MK II has roughly HALF the pixel density of the 7D, 60D & 550D, just look at the difference in noise performance.
TrojMacReady
Veteran Member
Size of sensor makes the difference here, not much else. Every new APS-C sensor from Canon has more pixels and shows better S/N figures yet, so the above statement is factually wrong.
hjulenissen
Senior Member
Mr. Sheehy:
If Canon and friends are having a "significant" number of photons read noise, do you think that it would be practical for others to reduce that number, possibly by increasing pizel pitch?
What do you think about the measurements on R.N. Clarkes webpage?
-h
If Canon and friends are having a "significant" number of photons read noise, do you think that it would be practical for others to reduce that number, possibly by increasing pizel pitch?
What do you think about the measurements on R.N. Clarkes webpage?
-h
Similar threads
Keyboard shortcuts
- f
- Forum