AOKH
Senior Member
So a 5D 12mp can print larger than an 18mp 7D before pixelation occurs? I am not talking IQ, just max native-resolution prints. I find that hard to believe, although I am no expert.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
5D 12mp = @ 24mp on 1.6 crop sensored cameras
- Thread starter Peter Nelson
- Start date
seanfoulkes
Member
Peter, I'm starting to think, both by your oddly worded posts and by your spelling, that you are in over your head. Let me clarify, because, while everyone here has been more or less correct with explaining this concept, you still seem to be confused.
These people are trying to explain to you the difference in sharpness between sensors, as it applies to your cloudy question. Larger sensors will have bigger, crisper pixels in the output image, due in part to lens resolution, actual pixel size and ability to collect photons, and also the fact that bigger pixels means sharper received image through the low pass (anti-alias) filter placed in front of the sensor.
A smaller 12 megapixel image from a full-frame 5D, enlarged to print at the size of an 18 megapixel printed image from a 7D will have much better resolution than a 12 megapixel APS-C image enlarged to the size of the 7D's image. This is because full-frame images are inherently sharper at a pixel level (100% zoom in photoshop). You will notice that, side-by-side, APS-C cameras, including the 7D, will be slightly softer in most cases at the pixel level, which means that a sharp 100% enlargement of a 5D image might, and in most cases will, be very sharp, whereas a 100% enlargement of a crop sensor image will be slightly less sharp.
So to answer your original question, an enlarged 12mp 5D image could and would be at least as sharp as a 7D image printed natively (at 300, 250, 200dpi, take your pick). This won't always be the case, I've made impressive 24" wide prints with the 7D, which were @ 250dpi, and I have also made very impressive 18-24" wide prints from a 5D, that were upsampled (digitally enlarged), and sharpened, post processed from there.
There is no mathmatical equation that universally describes detail comparing crop and full-frame sensors. Each camera and lens combination will give you different resolving capabilities.
But, generally, full-frame sensors, even smaller 12mp sensors, give sharper images than their current higher megapixel crop-sensor cousins. Sharp enough, that many people can argue that a 12mp 5D will give you equal or even better enlargements on print than an 18mp 7D.
Hope that clears things up.
These people are trying to explain to you the difference in sharpness between sensors, as it applies to your cloudy question. Larger sensors will have bigger, crisper pixels in the output image, due in part to lens resolution, actual pixel size and ability to collect photons, and also the fact that bigger pixels means sharper received image through the low pass (anti-alias) filter placed in front of the sensor.
A smaller 12 megapixel image from a full-frame 5D, enlarged to print at the size of an 18 megapixel printed image from a 7D will have much better resolution than a 12 megapixel APS-C image enlarged to the size of the 7D's image. This is because full-frame images are inherently sharper at a pixel level (100% zoom in photoshop). You will notice that, side-by-side, APS-C cameras, including the 7D, will be slightly softer in most cases at the pixel level, which means that a sharp 100% enlargement of a 5D image might, and in most cases will, be very sharp, whereas a 100% enlargement of a crop sensor image will be slightly less sharp.
So to answer your original question, an enlarged 12mp 5D image could and would be at least as sharp as a 7D image printed natively (at 300, 250, 200dpi, take your pick). This won't always be the case, I've made impressive 24" wide prints with the 7D, which were @ 250dpi, and I have also made very impressive 18-24" wide prints from a 5D, that were upsampled (digitally enlarged), and sharpened, post processed from there.
There is no mathmatical equation that universally describes detail comparing crop and full-frame sensors. Each camera and lens combination will give you different resolving capabilities.
But, generally, full-frame sensors, even smaller 12mp sensors, give sharper images than their current higher megapixel crop-sensor cousins. Sharp enough, that many people can argue that a 12mp 5D will give you equal or even better enlargements on print than an 18mp 7D.
Hope that clears things up.
Maybe this helps:
A digital picture has no actual size by itself, only a certain amount of pixels. It is only a file with information.
The size of the digital picture is determined by the output medium you choose, be that a print, a screen or a projection. Now every single pixel will cover a certain space (it is NOT "enlarged", for it has no size), depending on the size of the total output - yet the size of the area the pixel covers has nothing to do anymore with the size of the sensor the picture was taken with.
So you definitely cannot compare sensor sizes like film sizes.
But of course there will be a difference depending on the quality of a pixel - i.e. the amount of information it carries. And of course this will get more obvious the bigger the area one pixel has to cover (in relation of the resolution of the human eye).
That's were all the mentioned qualities like better "per pixel resolution", sharpness, micro-contrast, less noise etc come into play. A picture made up of less pixels may carry more total information than one made up of more pixels. But as others pointed out, the math is not as simple as your OP because it depends on many different factors.
To me too it seems obvious that with todays sensors the size of the individual sensor-pixel is one of the key factors to its information-quality. So much so, that a FF-sensor like the 5Ds might even outperform a three generations newer APS-C-sensor with too many MP in certain areas and under certain circumstances.
Hope this helps.
Greetings
A digital picture has no actual size by itself, only a certain amount of pixels. It is only a file with information.
The size of the digital picture is determined by the output medium you choose, be that a print, a screen or a projection. Now every single pixel will cover a certain space (it is NOT "enlarged", for it has no size), depending on the size of the total output - yet the size of the area the pixel covers has nothing to do anymore with the size of the sensor the picture was taken with.
So you definitely cannot compare sensor sizes like film sizes.
But of course there will be a difference depending on the quality of a pixel - i.e. the amount of information it carries. And of course this will get more obvious the bigger the area one pixel has to cover (in relation of the resolution of the human eye).
That's were all the mentioned qualities like better "per pixel resolution", sharpness, micro-contrast, less noise etc come into play. A picture made up of less pixels may carry more total information than one made up of more pixels. But as others pointed out, the math is not as simple as your OP because it depends on many different factors.
To me too it seems obvious that with todays sensors the size of the individual sensor-pixel is one of the key factors to its information-quality. So much so, that a FF-sensor like the 5Ds might even outperform a three generations newer APS-C-sensor with too many MP in certain areas and under certain circumstances.
Hope this helps.
Greetings
I'm no expert either, but I would say: the 5Ds print will "pixelate" earlier. But the real question is, does the 7Ds print show blotches of "empty" pixels before the 5D pixelates?
18MP with less total resolution than 13MP (according to repeatedly mentioned DxO-results) must mean, some of the 7Ds pixels carry no information. Makes sense, that the effect is smaller with very good glass.
(And who will buy that very good glass with say, a 550D? - but that's another discussion...)
I prefer more information stored in less MP anyway
18MP with less total resolution than 13MP (according to repeatedly mentioned DxO-results) must mean, some of the 7Ds pixels carry no information. Makes sense, that the effect is smaller with very good glass.
(And who will buy that very good glass with say, a 550D? - but that's another discussion...)
I prefer more information stored in less MP anyway
AOKH
Senior Member
Thanks, that sounds like a plausible explanation. Interesting thread. Regards
Regards. Anders
--
Regards. Anders
Perhaps. Perhaps Canon made an unannounced slight tweak in one area which had a slight (yet still measurable) negative impact on the high ISO performance. Or maybe just sample variation. Or maybe all the above or none of the above ... who knows???
Yep.
--
Good Day,
Roonal
Faintandfuzzy
Senior Member
- Messages
- 3,328
- Reaction score
- 6
Check the resolution test here at DPreview. The 7D outresolves the olkd 5D by a good margin. I've confirmed this myself.
Ah the Let's-Jump-To-Conclusions game
Otherwise you would have asked do I thus think the test results are not accurate, and if so, why.
And so since you know / have concluded I don't think it is accurate, you want to know my reasons.
Well I thought I was giving a simple reply to a simple question. Silly me I didn't consider that that simply reply would be viewed well beyond what it was.
Oops almost forgot
--
Good Day,
Roonal
I list a couple of test results from DxO and simple act of doing so was apparently (in your mind) the equivalent of me saying the DxO tests are not accurate.
Otherwise you would have asked do I thus think the test results are not accurate, and if so, why.
And so since you know / have concluded I don't think it is accurate, you want to know my reasons.
Well I thought I was giving a simple reply to a simple question. Silly me I didn't consider that that simply reply would be viewed well beyond what it was.
So not only do I think the 1DMkII & 1DMkIIn DxO high ISO performance test results are inaccurate, now I choose not to trust DxO period - wowIn every measurement there is a margin of error ... and since I assume you don't know what either of the margins of error is, you choose to not trust the DxO....
Oops almost forgot
--
Good Day,
Roonal
bronxbombers
Forum Pro
of course the actual results become tricky, what lens, what sensor size and MP count, at most levels the factor will have nothing remotely to do with the factors derived from your math although under the perfect set of circumstances it might get a bit closer to taht but it's all wrong how you are going about it and it leads to all sorts of crazy results as I said below
Daniel Clune
Senior Member
It sure does and its easy to see. http://www.dpreview.com/reviews/canoneos7d/page28.asp
http://www.dpreview.com/reviews/canoneos5d/page31.asp
Another option would be to ask this over in the nikon forum. As the d300 and D3/D700 are both 12 meg and D300 is a 1.5 crop body. I have read it over there that nikon owners who own both cant see any real world differance in low iso prints they made between the two bodies even though one is FF and one is a crop body.. But if you ask the question when the iso is raised above base then its a differant story and they say its a big differance because of higher noise on the crop body.
There is always a little better crispness on a FF body compared to a crop at the pixel level mainly do the less stress on the lens. But the differance isnt even close to the crap the OP was saying. He then goes on to say he read it some were and throws names like Thom Hogans about but cant produce a link to what "he thinks" he read. Thom Hogan does post over on the nikon forum once in a while so i suggest the OP ask his question to see if Thom Hogam might respond.
http://www.pbase.com/dc9mm
http://www.dpreview.com/reviews/canoneos5d/page31.asp
Another option would be to ask this over in the nikon forum. As the d300 and D3/D700 are both 12 meg and D300 is a 1.5 crop body. I have read it over there that nikon owners who own both cant see any real world differance in low iso prints they made between the two bodies even though one is FF and one is a crop body.. But if you ask the question when the iso is raised above base then its a differant story and they say its a big differance because of higher noise on the crop body.
There is always a little better crispness on a FF body compared to a crop at the pixel level mainly do the less stress on the lens. But the differance isnt even close to the crap the OP was saying. He then goes on to say he read it some were and throws names like Thom Hogans about but cant produce a link to what "he thinks" he read. Thom Hogan does post over on the nikon forum once in a while so i suggest the OP ask his question to see if Thom Hogam might respond.
--
http://www.pbase.com/dc9mm
Fraxinus excelsior
Forum Enthusiast
Wrong! Its because you choose show as print which is locked at a given print size! Try and show as on screen at 100% and the result is the opposite! I have Sony A900 FF myself and Sony APS-C.
Fraxinus excelsior
Forum Enthusiast
Fraxinus excelsior
Forum Enthusiast
Hi Hans you are wrong.
if you print a 12MP file at 360 dpi and a 18MP, or 24MP file at 360dpi the 12MP file is smaller! If you want a 12 MP file to just print as big as a 24MP file you have to interpolate the 12MP file regardless of the sensor size!
Kind regards
David
Hans Kruse
Senior Member
When you print on e.g. an Epson printer you will always interpolate anyway to the native PPI og 360 of the printer. However if there is more resolution in one file even with less pixels you will see the difference on the print if it is big enough. The interpolation can be done in Lightroom (which I always do depending on the size of the print) to 360 PPI.
--
Kind regards,
Hans Kruse
Home Page -- http://www.hanskrusephotography.com , http://www.hanskruse.com
Workshops -- http://www.hanskrusephotography.com/workshops
Facebook Photography http://www.facebook.com/pages/Hans-Kruse-Photography/271477435625
Workshop Newsletter signup http://eepurl.com/bA0Pj
Keyboard shortcuts
- f
- Forum