Velocity Reviews > question about relationship between sensor size and print size.

# question about relationship between sensor size and print size.

ftran999
Guest
Posts: n/a

 02-22-2007
This is something that I'm just courious about. I understand that to
determine what print size a digital camera (or scanned image) is capable
of, one divides the desired dpi (e.g. 200 or 300) into the pixel dimensions
produced by the camera. Now my question is if two cameras produce nearly
identical resolutions but have different sensor sizes how does this affect
print size? For example take two different DSLRs with nearly identical
resolutions, the Nikon DX2 and the Canon 5D. If one was to determine that
they wanted to print at 300dpi by dividing that number into the pixel
dimensions of both cameras, you would get a print size of approx 9.5x14.5
give or take a few fractions of an inch. Now the thing is that the Canon
uses a full size sensor (24x36) whereas the Nikon uses a DX sensor
(15.7x23.7). Now I'm not asking specifically about the superiority of one
camera over the other, just happened to pick them out to use as examples,
but THEORETICALLY if a FF camera has a sensor size that is 2.3 times the
area of DX camera then shouldn't it be able to produce a print that is
larger by the same factor ? Keep in mind everything will be equal, printer,
paper, viewing distance.

Charles Schuler
Guest
Posts: n/a

 02-22-2007

"ftran999" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed). ..
> This is something that I'm just courious about. I understand that to
> determine what print size a digital camera (or scanned image) is
> capable of, one divides the desired dpi (e.g. 200 or 300) into the pixel
> dimensions produced by the camera. Now my question is if two cameras
> produce nearly identical resolutions but have different sensor sizes how
> does this affect print size? For example take two different DSLRs with
> nearly identical resolutions, the Nikon DX2 and the Canon 5D. If one was
> to determine that they wanted to print at 300dpi by dividing that number
> into the pixel dimensions of both cameras, you would get a print size of
> approx 9.5x14.5 give or take a few fractions of an inch. Now the thing is
> that the Canon uses a full size sensor (24x36) whereas the Nikon uses a
> DX sensor (15.7x23.7). Now I'm not asking specifically about the
> superiority of one camera over the other, just happened to pick them out
> to use as examples, but THEORETICALLY if a FF camera has a sensor size
> that is 2.3 times the area of DX camera then shouldn't it be able to
> produce a print that is larger by the same factor ? Keep in mind
> everything will be equal, printer, paper, viewing distance.

In a nutshell, the bigger the sensor (physically speaking) the better the
signal to noise ratio and dynamic range.

ray
Guest
Posts: n/a

 02-22-2007
On Wed, 21 Feb 2007 19:01:16 -0500, ftran999 wrote:

> This is something that I'm just courious about. I understand that to
> determine what print size a digital camera (or scanned image) is capable
> of, one divides the desired dpi (e.g. 200 or 300) into the pixel dimensions
> produced by the camera. Now my question is if two cameras produce nearly
> identical resolutions but have different sensor sizes how does this affect
> print size?

It doesn't.

> For example take two different DSLRs with nearly identical
> resolutions, the Nikon DX2 and the Canon 5D. If one was to determine that
> they wanted to print at 300dpi by dividing that number into the pixel
> dimensions of both cameras, you would get a print size of approx 9.5x14.5
> give or take a few fractions of an inch. Now the thing is that the Canon
> uses a full size sensor (24x36) whereas the Nikon uses a DX sensor
> (15.7x23.7). Now I'm not asking specifically about the superiority of one
> camera over the other, just happened to pick them out to use as examples,
> but THEORETICALLY if a FF camera has a sensor size that is 2.3 times the
> area of DX camera then shouldn't it be able to produce a print that is
> larger by the same factor ? Keep in mind everything will be equal, printer,
> paper, viewing distance.

A pixel is a pixel. The biggest factor is going to be the compression
involved in making the image. If they are both producing lossless images
then the print quality will be equal - the one with the larger sensor
should have less 'noise' at a given iso setting.

MarkČ
Guest
Posts: n/a

 02-22-2007
ray wrote:
> On Wed, 21 Feb 2007 19:01:16 -0500, ftran999 wrote:
>
>> This is something that I'm just courious about. I understand that
>> to determine what print size a digital camera (or scanned image)
>> is capable of, one divides the desired dpi (e.g. 200 or 300) into
>> the pixel dimensions produced by the camera. Now my question is if
>> two cameras produce nearly identical resolutions but have different
>> sensor sizes how does this affect print size?

>
> It doesn't.

Not directly (since yu can print ANY size [200'x300' if you're so
inclined...even with 1MP]...so long as quality isn't factored in), but the
SHAPE of the sensor can make a difference to the extent that you may have to
crop based on the print dimension ratio.
If you shoot with a 3:2 sensor (most DSLRs), then you can print 12x18, or
the same ratio at other sizes and take full advantage of the resolution of
the sensor. Once you go to 8x10, or other ratio, you have to crop--throwing
away a portion of your image.

--
Images (Plus Snaps & Grabs) by MarkČ at:
www.pbase.com/markuson

erqua
Guest
Posts: n/a

 02-22-2007
Charles Schuler wrote:
> In a nutshell, the bigger the sensor (physically speaking) the better
> the signal to noise ratio and dynamic range.

Would this be more obvious in low light conditions?
Would this be offset by the camera having a bigger lens (DSLR vs
compact)?

Paul Rubin
Guest
Posts: n/a

 02-22-2007
"erqua" <(E-Mail Removed)> writes:
> > In a nutshell, the bigger the sensor (physically speaking) the better
> > the signal to noise ratio and dynamic range.

>
> Would this be more obvious in low light conditions?

Yes.

> Would this be offset by the camera having a bigger lens (DSLR vs
> compact)?

DSLR's have bigger lenses for the precise reason that they have bigger
sensors.

David J Taylor
Guest
Posts: n/a

 02-22-2007
Paul Rubin wrote:
> "erqua" <(E-Mail Removed)> writes:
>>> In a nutshell, the bigger the sensor (physically speaking) the
>>> better the signal to noise ratio and dynamic range.

>>
>> Would this be more obvious in low light conditions?

> Yes.
>
>> Would this be offset by the camera having a bigger lens (DSLR vs
>> compact)?

>
> DSLR's have bigger lenses for the precise reason that they have bigger
> sensors.

.... and because of cost or size considerations, many of the DSLR lenses
have a smaller aperture (f/5.6) and the long end of the zoom compared to
the f/2.8 offered by some small-sensor cameras, thus throwing away quite a
lot of the advantage of the large sensor....

David

Randy Berbaum
Guest
Posts: n/a

 02-22-2007
ftran999 <(E-Mail Removed)> wrote:
: This is something that I'm just courious about. I understand that to
: determine what print size a digital camera (or scanned image) is capable
: of, one divides the desired dpi (e.g. 200 or 300) into the pixel dimensions
: produced by the camera. Now my question is if two cameras produce nearly
: identical resolutions but have different sensor sizes how does this affect
: print size? For example take two different DSLRs with nearly identical
: resolutions, the Nikon DX2 and the Canon 5D. If one was to determine that
: they wanted to print at 300dpi by dividing that number into the pixel
: dimensions of both cameras, you would get a print size of approx 9.5x14.5
: give or take a few fractions of an inch. Now the thing is that the Canon
: uses a full size sensor (24x36) whereas the Nikon uses a DX sensor
: (15.7x23.7). Now I'm not asking specifically about the superiority of one
: camera over the other, just happened to pick them out to use as examples,
: but THEORETICALLY if a FF camera has a sensor size that is 2.3 times the
: area of DX camera then shouldn't it be able to produce a print that is
: larger by the same factor ? Keep in mind everything will be equal, printer,
: paper, viewing distance.

No. Since you are not asking about the theories and limitations of the
image capturing and storage systems, the size of the sensor is not
directly helping or hindering the print size. If you have two sensors of
different sizes with the same exact dimensions in pixels, each pixel will
be representing the exact same proportion of the total image. The
different size of the respective pixel sensing element may effect the
ability to capture the light qualities for that single pixel, but the
stored pixel is still representing the same portion of the finished image.

The 300 dpi benchmark is generally a reflection of the current state of
the art in printing technology to accurately produce a pixel on the page.
Too high dpi, and the printing device may begin to blur the pixels due to
the limits of ink to be deposited without difusing into the adjacent
pixel. And too low a dpi and the individual pixels become visible. Of
course these values are for a standard viewing distance of 3-6ft or so. If
you plan on viewing from the length of a football field the "best" dpi
will change. A Jumbotron image that looks nice and clear at a distance
looks silly up close (a single color element is made up from 3 seperate
color elements, each being about 4" square). Each "pixel" is about 8" by
8". On the other hand if you can find a printer that can deposit fine
enough "drops" of ink on a surface, and the ink does not spread into the
next pixel, you could print an image with a very high dpi that would
satisfy our readers who MUST examine every print with a microscope.

But using 300dpi is a good benchmark, no matter what the sensors physical
dimensions. Now if the physical size change is reflected in a change in
pixel count, the change in size would have an effect in the print.

Randy

==========
Randy Berbaum
Champaign, IL

Don Stauffer in Minnesota
Guest
Posts: n/a

 02-22-2007
On Feb 21, 6:01 pm, "ftran999" <(E-Mail Removed)> wrote:
> This is something that I'm just courious about. I understand that to
> determine what print size a digital camera (or scanned image) is capable
> of, one divides the desired dpi (e.g. 200 or 300) into the pixel dimensions
> produced by the camera. Now my question is if two cameras produce nearly
> identical resolutions but have different sensor sizes how does this affect
> print size? For example take two different DSLRs with nearly identical
> resolutions, the Nikon DX2 and the Canon 5D. If one was to determine that
> they wanted to print at 300dpi by dividing that number into the pixel
> dimensions of both cameras, you would get a print size of approx 9.5x14.5
> give or take a few fractions of an inch. Now the thing is that the Canon
> uses a full size sensor (24x36) whereas the Nikon uses a DX sensor
> (15.7x23.7). Now I'm not asking specifically about the superiority of one
> camera over the other, just happened to pick them out to use as examples,
> but THEORETICALLY if a FF camera has a sensor size that is 2.3 times the
> area of DX camera then shouldn't it be able to produce a print that is
> larger by the same factor ? Keep in mind everything will be equal, printer,
> paper, viewing distance.

The final resolution of the camera is a chain of the resolution of
each of several limiting factors. The larger the sensor, the less
likely that IT is the primary limitation to resolution. i.e, a chain
is as strong as its weakest link.

Now, by resolution I mean the true, measured resolution based on bar
chart or mtf tests, NOT simply the number of pixels in the sensor.
The actual resolution of a camera is ultimately limited by the number
of pixels, but a camera CANNOT resolve a bar frequency equal to the
number of pixels in a given direction. That is, a 3000 pixel wide
camera will NOT, on average, be able to resolve a chart with 3000
lines of alternating black and white bars.