LOG IN    OR   REGISTER TO FORUMS


18 vs 36 mp in print...

FORUMS Canon Cameras, Lenses & Accessories Canon EOS Digital Cameras
Thread started 20 Feb 2012 (Monday) 23:29   
LIST NEARBY THREADS
 
Lowner
"I'm the original idiot"
Lowner's Avatar
Joined Jul 2007
12,888 posts
Salisbury, UK.
[MORE/SHARE]

Moppie wrote in post #13934621external link
Don't confuse print resolution with camera resolution, don't confuse print quality with mega pixels and dont assume more mega pixels means a better quality image.

The 5d2s ability to capture better tonal ranges and more accurately produce colour changes than a 30d has nothing to do with it having more mega pixels.

I don't confuse resolution with the tonal quality I'm referring to. But to an extent the extra MP of the 5D2 DOES have a bearing, along with other atributes, on the tonal range. Accuracy or otherwise I cannot comment about, a "smoothness" of tonal changes is the only way I can explain it.

Post #16, Feb 21, 2012 07:09:49


Richard

http://rcb4344.zenfoli​o.comexternal link

LOG IN TO REPLY
sponsored links
(this ad will go away when you log in as registered member)
boerewors
Goldmember
boerewors's Avatar
Joined Sep 2009
1,942 posts
South African living in Indonesia
[MORE/SHARE]

ISO performance has a big role to play on the megapixels as well. I would assume an ISO 12800 large print from a 1Dx would look better than an ISO12800 print from a 36mp camera. I also wonder if ISO performance would equal out if the 36mp were downsized to 18mp??

Post #17, Feb 21, 2012 09:21:56


The most important piece of gear you own, resides in your head and its called your brain.

LOG IN TO REPLY
tzalman
Fatal attraction.
tzalman's Avatar
Joined Apr 2005
12,598 posts
Gesher Haziv, Israel
[MORE/SHARE]

Lowner wrote in post #13934893external link
I don't confuse resolution with the tonal quality I'm referring to. But to an extent the extra MP of the 5D2 DOES have a bearing, along with other atributes, on the tonal range. Accuracy or otherwise I cannot comment about, a "smoothness" of tonal changes is the only way I can explain it.

The smoothness of tonal change of the 5D2 is not because of resolution - it has exactly the same resolution as the 30D. It is because of sensor size - to get an 8x12 print the optical image recorded on the FF sensor is enlarged x8, from an APS-C it is enlarged x13. A 66 MP MF sensor would also have the same resolution, but would produce even greater IQ. When I compare 20 x 30 prints from my 40D and my 5D2, despite the slightly higher resolution of the 40D the difference is obvious.

But the above is OT because the OP asked about comparing prints from same sized sensors and to answer this the technology of printing can't be ignored. Any image sent to a printer will be resampled to the printer's input requirements, unless it is already at that ppi. Resampling reduces IQ; if the amount is small the loss will be difficult to discern, if a lot of resampling is done the loss will be greater. The 1DX will be able to supply an Epson printer with an image that will not need resampling if the print size is 10.2 x 15.3 inches. For the 36 MP sensor it will be 14.28 x 21.42. For a 20 x 30 print the difference between the 2X upres and the 1.4X upres would probably need a careful side-by-side comparison to be seen. For a 40 x 50 the difference between a 4X and a 2.8X resampling would be easier to see.

BTW, the argument that 36 MP files will be too large to be processed seems to me to be ludicrous. We live in our DSLR bubble while outside it increasing numbers of fine art photographers are using 80 MP MF backs.

Post #18, Feb 21, 2012 09:58:55


Elie / אלי
I look at the world and I notice it's turning, while my guitar gently weeps.
With every mistake we must surely be learning, still my guitar gently weeps.
Paul Harrison
Siteexternal link

LOG IN TO REPLY
AJSJones
Goldmember
AJSJones's Avatar
Joined Dec 2001
2,267 posts
California
[MORE/SHARE]

Moppie wrote in post #13934435external link
But only at very large print sizes and close viewing distances.


At the correct viewing distance for a given print resolution you can not tell any difference if more pixels are added at the time the photo was taken.


I shoot with a 5D2 and a 30D, and I'll be damned if I can tell them apart sometimes.
Even had a second shooter with a 1000D the other day, and the prints look exactly the same as they do from my 5D2.

Some photographers (quite a few judging by the popularity of large fine art prints) strive for large prints. A large fraction of people who view these prints view them from a distance to take in the whole image and also move closer to examine the details as they are drawn in to the image (at least the good ones). If you can see the pixel structure or printer dithering, you are probably too close and will step back till it is no longer visible/distracting. In other words, the viewer determines the correct distance to view the print.
I have some very nice 8x10's from my D30 (3MP) They were uprezzed a bit to get there but those huge pixels allowed it. They are almost as good as the 8x10s from my 5D2. At a small enough size, adding more MP doesn't do very much. For anyone who only prints small and thinks 12x18 or 14x20 is a big print, then 12-18 MP is probably sufficient. For those striving for large detailed prints, 36MP is welcome. How big you can print depends on your individual quality standards and technique in taking, processing and printing. There is no "one size fits all" - the best succinct answer so far is "300ppi" - not because 300 is necessarily the "correct" number, but because there is usually a ppi that you will feel meets your own individual quality standards. Then it's arithmetic to calculate the size you can print to :D

Post #19, Feb 21, 2012 10:19:44


My picture galleriesexternal link

LOG IN TO REPLY
wimg
Cream of the Crop
wimg's Avatar
Joined Jan 2007
6,590 posts
Netherlands, EU
[MORE/SHARE]

boerewors wrote in post #13935526external link
ISO performance has a big role to play on the megapixels as well. I would assume an ISO 12800 large print from a 1Dx would look better than an ISO12800 print from a 36mp camera. I also wonder if ISO performance would equal out if the 36mp were downsized to 18mp??

Actually, no. Based on the same technology, a 36 MP image and an 18 MP image printed or viewed at the same size will show the same amount of noise.

Kind regards, Wim

Post #20, Feb 21, 2012 11:23:02


5D Mk II, 1D Mk III, with a solitary zoom and a gaggle of primes, GF-2 with 3 primes and 3 zooms, G10, an accessory plague, and some analog stuff

LOG IN TO REPLY
sponsored links
(this ad will go away when you log in as registered member)
wimg
Cream of the Crop
wimg's Avatar
Joined Jan 2007
6,590 posts
Netherlands, EU
[MORE/SHARE]

As to prints from images of differently sized sensors, MP-wise: if the prints are the same size, the higher resolution images will look sharper, even if looked from distances where one can't see the individual pixels.

One can try this exercise, by printing the same image, downscaled to print at 300 dpi, and 600 dpi, where the 600 dpi image has 4X the number of pixels of the 300 dpi image. The printer used has to be capable of both, of course.

Even with my bad eyes I can see the difference.

Kind regards, Wim

Post #21, Feb 21, 2012 11:28:48


5D Mk II, 1D Mk III, with a solitary zoom and a gaggle of primes, GF-2 with 3 primes and 3 zooms, G10, an accessory plague, and some analog stuff

LOG IN TO REPLY
mmahoney
Goldmember
mmahoney's Avatar
Joined Jan 2007
2,784 posts
[MORE/SHARE]

mafoo wrote in post #13933810external link
So, I am debating someone, who says more pixels is always better.

I asume when printing, there is a point where you run out of pixels, and more is better....

For the 1Dx, and an 18 mp sensor, if you did not crop the image, how large of a print could you make before the pixel density started to negatively impact the print?

How about with a 36 mp sensor?

"Well Johnny, I've got a 6 foot by 4 foot print of my old dog Shep hanging on the barn wall and it's so sharp we shear the sheep with it" :rolleyes:

This discussion is about as old as the internet forums that host them .. my thoughts are that most photographers opinions on the subject are formed largely by the cameras they own, and their own eyesight (or lack of).

90% of people offering opinions have never worked with really high resolution files from MF backs or drum scans of LF film nor have they ever seen the absolute magic quality that a large print from these files produces.

As a yardstick most print houses suggest 300dpi, and most magazines print at 360dpi .. you can do the math from there but as a general observation more pixels are better.

There is some good reading in the link earlier in this thread (Browning?)

Post #22, Feb 21, 2012 12:56:19


Newfoundland Wedding Photographerexternal link

LOG IN TO REPLY
h4ppydaze
Goldmember
Joined Oct 2011
1,324 posts
[MORE/SHARE]

hollis_f wrote in post #13934624external link
The understated utility of smaller pixels - by Daniel Browning, a man whose brain can be quite astounding.

wow, great read. Almost makes me regret purchasing my 1Diii over a 5Dii. Almost.

Focus has a larger effect on image quality than megapixels, so I went 1D :lol:. Though I won't deny that more pixels are useful. That's why the 50D is still around. In tightly controlled conditions it'll be slightly easier to coax big prints out of it.

The major advantage of the 5Dii was sensor size. Upping sensor size is the best way to increase 'true' resolution. You just reach a point where the number of pixels isn't what's limiting your image quality the most. Right now, I bet my sensors matches the resolving power of my glass in most situations, like when shooting wide open or past f4-5.6, which I do for most of my shooting.

For truly large prints, I'll stick with 4x5 and up to maybe 8x10 if I'm feeling rich at the moment.

Post #23, Feb 21, 2012 13:26:12




LOG IN TO REPLY
lungdoc
Goldmember
lungdoc's Avatar
Joined May 2006
2,101 posts
St. Catharines, Ontario Canada
[MORE/SHARE]

noisejammer wrote in post #13934520external link
Per Moppie's comment -

Under absolutely perfect conditions, the exeptional eye can resolve about half an arc-minute (that's about 150 micro-radians.) This corresponds to 0.15 mm at a metre, or 7 pixels per mm (175 pixels per inch.) This suggests that a print with 5000 pixels along the long axis could be 28 inches wide and even people with perfect vision could not resolve the pixels. Printing at higher resolution effectively blurs the pixels so that you can't see blocks (even if there is no increase in detail.)

This brings me to the value of 36Mp. If the lens isn't good enough, then there's no information to print. If focus is not perfect (twice as good as currently achievable) then the benefit is lost. If the lens is stopped down past f/4, then diffraction kicks in and the benefit is (progressively) lost. If the camera shakes by a tiny amount, then the benefit is lost...

I suppose that it may be possible to obtain somewhat better resolution but it seems that even my Zeiss glass will struggle to produce any contrast at the native resolution of the sensor (200 pixels per mm.) It can be argued that the true resolution of a Bayer sensor is closer to half the pixel density - so say 100 pixels with full information per mm... but even that is beyond the resolving capability of my best glass (in this case the 50/2 MP and 100/2 MP, the Canon 100/2.8L is similar.) I assume Nikon is also ruled by the economics of optics and their lenses will be just as limiting.

So here's the rub - even if you record 36 Mp in the image, there's probably no additional information present beyond what might be found on a 21 Mp frame.... in fact assuming the lens can reasonably resolve 60 lp/mm - few can - there's probably very little information present that wouldn't be present in a 12 Mp frame. There will be bigger file sizes, more noise and lots of gee-whiz factor... but not a lot of real difference.

I don't know if this is true - the general ideas sound reasonable but I'm doubting if they apply at the resolutions you describe. If that were the case there should be no advantage of any crop sensor camera with a resolution over about 8 MP, after all the cropped portion of a 5D II image equates to an 8.2 MP APS-C sensor. I think most do find real resolution improvement to APS-C images, albeit under best conditions, from say a 7D sensor which is equivalent to a FF 46 MP sensor, versus say a 30D sensor.

Post #24, Feb 21, 2012 14:06:24


Mark
My Smugmugexternal link Eos 7D, Canon G1X II, Canon 15-85 IS, Canon 17-85 IS, Sigma 100-300 EX IF HSM, Canon 50mm 1.8, Canon 85mm 1.8, Canon 100mm 2.8 Macro, Sigma 50-150 2.8, Sigma 1.4 EX DG , Sigma 24-70 F2.8 DG Macro, Canon EF-S 10-22, Canon 430EX,

LOG IN TO REPLY
mafoo
THREAD ­ STARTER
Goldmember
mafoo's Avatar
Joined Dec 2011
1,503 posts
[MORE/SHARE]

Thanks everyone for your responses...

I think I am on the wrong side of this argument. I think more pixels might be better. Here is an example with a D3s (12mp), and a D3x(24mp), where they down sampled the higher mp camera to match the lower one. This is a good example, because the higher MP sensor is older, so there is no chance of newer technology influencing the results.

http://heylong.com/d80​0/d3x_d3x_12mp_compare​.JPGexternal link

The down sampled image, is clearly better.

Post #25, Feb 21, 2012 14:19:23


-Jeremy
5D Mk II | SL1 | 24-105 f4.0L IS | 70-200 f2.8L IS | S35 1.4 | 40 2.8 Pancake | Samyang 14 2.8 | 430EX II

LOG IN TO REPLY
h4ppydaze
Goldmember
Joined Oct 2011
1,324 posts
[MORE/SHARE]

mafoo wrote in post #13937590external link
Thanks everyone for your responses...

I think I am on the wrong side of this argument. I think more pixels might be better. Here is an example with a D3s (12mp), and a D3x(24mp), where they down sampled the higher mp camera to match the lower one. This is a good example, because the higher MP sensor is older, so there is no chance of newer technology influencing the results.

http://heylong.com/d80​0/d3x_d3x_12mp_compare​.JPGexternal link

The down sampled image, is clearly better.

wow! Indeed it is. By a good margin, too. I'd imagine at 100% crop the D3s is suffering from artifacts that are a fact of life when it comes to digital sensors. It's robbing detail and contrast with anti-aliasing. The extra pixels in the D3X must reduce the effect of these artifacts. Now I'm even closer to regretting choosing the 1Diii over the 5Dii :confused:

Post #26, Feb 21, 2012 14:27:49




LOG IN TO REPLY
sponsored links
(this ad will go away when you log in as registered member)
AJSJones
Goldmember
AJSJones's Avatar
Joined Dec 2001
2,267 posts
California
[MORE/SHARE]

mafoo wrote in post #13937590external link
Thanks everyone for your responses...

I think I am on the wrong side of this argument. I think more pixels might be better. .

No, that's the right side. The point is that once you print small enough, you can't see the benefit of the extra pixels - viewing that comparison at 100% is like looking closely at a large print.

Post #27, Feb 21, 2012 17:12:35


My picture galleriesexternal link

LOG IN TO REPLY
Panoz
Senior Member
Joined Feb 2012
195 posts
[MORE/SHARE]

Pixels have a small factor in print quality as long as you exceed a minimum amount, say 8MP. I have 20x24 images from my old Canon 20D that are tack sharp when you look up close. That was an 8MP camera.

What everyone is leaving out is that the printer driver - the one that pushes data from your file to the printer - has more of an impact on the print quality than the quantity of pixels do. How does an 8MP camera produce a sharp 300dpi print? The driver dithers the missing data. A 20x24 300dpi print would require a 43MP source for 1:1 ratio of pixels to dots. My image only had 8MP but produced a razor sharp print. It's the printer driver's algorithm that expands the missing data for you.

Pixels and long lenses sell hardware. But you don't need them to get a great print.

Post #28, Feb 21, 2012 17:27:34 as a reply to AJSJones's post 14 minutes earlier.


Canon G12, Canon 5D MkII, Canon 24-105 L, Sigma 16mm fisheye

LOG IN TO REPLY
Old ­ Baldy
Senior Member
Joined Sep 2010
708 posts
South Lyon, MI
[MORE/SHARE]

Reach, reach, reach, reach!!!

Everyone seems to forget the HUGE advantage of greater MP for so MANY types of photography not being discussed here.

Dammit!

:)

Post #29, Feb 21, 2012 18:43:06 as a reply to Panoz's post 1 hour earlier.


OB
NEX-5N vs NEX-7external link
NEX-5N compared....external link
Gear list
Flickr Galleryexternal link

LOG IN TO REPLY
Ricardo222
Cream of the Crop
Ricardo222's Avatar
Joined Mar 2010
15,984 posts
Nelson, New Zealand
[MORE/SHARE]

FlyingPhotog wrote in post #13934461external link
SOOC @ 10Mp from a 1DMkIII

Final image generously massaged to about 7.5 Mp via Crop, Up-Res and Sharpening...
I'm looking at a very nice 16x24 of this above my desk printed by Mpix. ;)

(It also won 3rd Place in a world-wide aviation photo contest and was published back in December)

Ahhhhh! In the hands of a master so much is possible. Funny how so many photographers, including me, fuss about what resolution we can achieve and a myriad of other "problems", when we have yet to learn to get the very best out of the gear we have right now.

With the new range of sensors starting to be announced the Can/ikon argument goes on apace yet few of us are TRULY affected because we seldom really challenge our equipment. (To those to whom it DOES make a difference I defer...you are the elite.);)

Post #30, Feb 21, 2012 19:09:36


Growing old disgracefully!

LOG IN TO REPLY


LIST NEARBY THREADS
4,641 views & 0 likes for this thread
18 vs 36 mp in print...
FORUMS Canon Cameras, Lenses & Accessories Canon EOS Digital Cameras



NOT A MEMBER YET? CLICK HERE TO REGISTER TO FORUMS

CHANGE BODY TEXT SIZE FOR ALL THREAD PAGES
POWERED BY AMASS 1.0version 1.0
made in Finland
by Pekka Saarinen
for photography-on-the.net


SEND FEEDBACK TO STAFF  |  JUMP TO FORUM...  |  FORUM RULES


Spent 0.00191 for 4 database queries.
PAGE COMPLETED IN 0.02s
826 guests, 620 members online
Simultaneous users record so far is 3341, that happened on Dec 11, 2014
Latest registered member is DrewB777

COOKIES DISCLAIMER: By using this site you agree that some cookies will be stored on your browser. For unlogged users we store one session id cookie. For registered members we store (in addition to login session cookie) only cookies that are essential for required functionality, we do not store any personal tracking data in cookies or other browser's data storage methods.