Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Print sizes vs. megapixel count  (Read 9300 times)
lluis
Newbie
*
Offline Offline

Posts: 33


WWW
« on: November 14, 2009, 10:13:04 AM »
ReplyReply

Can anybody tell me a general relationship between megapixel count and optimal print sizes, bearing in mind we want to get good print quality?
So for example, to make good A3 print, which is the minimum pixels I need?

Thank you.
Logged
Dick Roadnight
Sr. Member
****
Offline Offline

Posts: 1730


« Reply #1 on: November 14, 2009, 10:45:55 AM »
ReplyReply

Quote from: lluis
Can anybody tell me a general relationship between megapixel count and optimal print sizes, bearing in mind we want to get good print quality?
So for example, to make good A3 print, which is the minimum pixels I need?

Thank you.
It depends on what you call "good print quality".

With film the guideline was 10 times film size, with digital, you will not get much visible improvement above 300 or 360 (Epson) Pixels per original print inch, depending what the default res of you printer is.

240 would be OK... for large prints to be viewed at a distance 180 would be OK.

A3@ 360 ppi  is (29 *42* 360^2)/(2.54^2*10^6) = 24.46 Megapixels.

A2@ 360 ppi  is (58 *42* 360^2)/(2.54^2*10^6) = 48.9 Megapixels

Many will tell you that they can produce good A2 prints with 6Mpx, but it depends what you call "good", and it depends on how fine the detail is in the picture... if you are using scans or film lenses, you might want to print at 720 ppi.
« Last Edit: November 14, 2009, 10:46:57 AM by Dick Roadnight » Logged

Hasselblad H4, Sinar P3 monorail view camera, Schneider Apo-digitar lenses
Dan Berg
Sr. Member
****
Offline Offline

Posts: 1532



WWW
« Reply #2 on: November 14, 2009, 11:57:25 AM »
ReplyReply

Heres what I use
8 X 10 - 11 x 17    - 360
 13 X 19 - 17 X 22 - 300
24 X 30 - 24 X 36 - 240
30 X 40 to 40 X 60 - 180
Run anything larger then 24 X 30 through Genuine Fractals 6. On ocassion I drop to 180 on all the 24" prints as well. Really no visible difference. This is all on canvas.
 This is with 12mp bodies D2Xs and D300. I just printed 3 images from Africa from a 2.8mp file. One is excellent the other 2 are fair to poor. The larger image is printed to 32 X 42 at 180 and from 2 feet it is super sharp. The others are good only at 6feet or further.
pretty marginal for the size they were printed at 24 X 32. Without Genuine Fractals 6 they were not printable at these sizes.
« Last Edit: November 14, 2009, 12:03:40 PM by Dan Berg » Logged

lluis
Newbie
*
Offline Offline

Posts: 33


WWW
« Reply #3 on: November 14, 2009, 01:38:58 PM »
ReplyReply

Quote from: Dan Berg
Heres what I use
8 X 10 - 11 x 17    - 360
 13 X 19 - 17 X 22 - 300
24 X 30 - 24 X 36 - 240
30 X 40 to 40 X 60 - 180
Run anything larger then 24 X 30 through Genuine Fractals 6. On ocassion I drop to 180 on all the 24" prints as well. Really no visible difference. This is all on canvas.
 This is with 12mp bodies D2Xs and D300. I just printed 3 images from Africa from a 2.8mp file. One is excellent the other 2 are fair to poor. The larger image is printed to 32 X 42 at 180 and from 2 feet it is super sharp. The others are good only at 6feet or further.
pretty marginal for the size they were printed at 24 X 32. Without Genuine Fractals 6 they were not printable at these sizes.

Thank you for your kind answers.
Regarding Genuine Fractals, I always have thought one is producing a "fake" image. If I don't have enough resolution to make a "good" print (say, 240ppi), is it worth trying Genuine Fractals?

Lluis
Logged
Dan Berg
Sr. Member
****
Offline Offline

Posts: 1532



WWW
« Reply #4 on: November 14, 2009, 02:21:58 PM »
ReplyReply

Nothing fake about it. The greatest thing since sliced bread! I have another right in front of me. Perfect example. A beautiful flower shot in med format jpeg,325 kb is the file size. Big enough for a 4 X 6 or 5 X 7.  I used to turn people away if they brought me a file this size. Today I sold an $80.00 print because of this software. I just finished printing this as a 24 X 30 on Exhibition Fibre. After Genuine Fractals it is a 116 mb tiff. You would never believe it,absolutly stunning.
Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7333


WWW
« Reply #5 on: November 14, 2009, 04:33:20 PM »
ReplyReply

Hi,

I have a small write up on the issue, needs some updates, tough.

http://83.177.178.241/ekr/index.php/photoa...xels-do-we-need

Best regards
Erik

Quote from: lluis
Can anybody tell me a general relationship between megapixel count and optimal print sizes, bearing in mind we want to get good print quality?
So for example, to make good A3 print, which is the minimum pixels I need?

Thank you.
Logged

Alan Goldhammer
Sr. Member
****
Offline Offline

Posts: 1632


WWW
« Reply #6 on: November 14, 2009, 08:04:08 PM »
ReplyReply

There is also a nice write up on Thom Hogan's Nikon site:  http://www.bythom.com/printsizes.htm

Remember that it also depends on the viewing distance.  Classic example would be Roy Lichtenstein paintings (the comic book interpretations).  From across the museum room the color looks continuous.  As you get closer you can see the dot pattern.  Steinmueller has a formula in his book - resolution = 300 / viewing distance (in feet)  Finally, we have to remember that the printer values are in dots per inch not ppi

Logged

Deepsouth
Jr. Member
**
Offline Offline

Posts: 81


« Reply #7 on: November 15, 2009, 06:46:54 AM »
ReplyReply

Quote from: ErikKaffehr
Hi,

I have a small write up on the issue, needs some updates, tough.

http://83.177.178.241/ekr/index.php/photoa...xels-do-we-need

Best regards
Erik


Both of these write-ups are very good. I might take issue with the quoted formula

print res=300/viewing distance in feet

since it does take into account the kind of image . People's expectations vary according to their knowledge of the image. A landscape you have never seen in person may look fine at a lower res. A portrait of something (or someone) you know well brings much more critical judgements into play. Commercial portrait photographers rarely (maybe never) use a high-res workflow since too much detail will disappoint the customer-nobody wants every wrinkle, acne scar and pore to to be seen in an expensive portrait.

I appreciate Erik's remarks about printer resolution. The "resolution race" still goes on with mfrs solemnly stating they have multi-thousands of dpi. Nonsense. If you look at the basic technology, the truth is that the vast majority of high end printers are hardware-limited to 600 dpi, period. Interpolation can inflate those numbers. One of the "old hands" who was a silver-based darkroom guy and made the transition to high-end commercial inkjet printing told me years ago that inkjet printing should be done at 300 dpi, period, regardless of size (if the original image and file would support that resolution). He based that on his experience and the fact that at the time RAM and processor speed were very dear. Printing a at 600 dpi would have slowed down his workflow, and, he asserted, would result in no visible difference to the customer. Time and expectations have moved on, but I still think his maxim has value.
Logged
neil snape
Sr. Member
****
Offline Offline

Posts: 1432


WWW
« Reply #8 on: November 15, 2009, 07:13:24 AM »
ReplyReply

While I have not seen substantial differences between GF and late Photoshop uprezzing, I certainly do see a lot better printing from Qimge from the same images printed through the drivers.

The 12 MPX full frame cameras easily go up to A3, but start loosing out if the view distance is close compared to the full frame >20mpx cameras.

Many say the MF captures up rez much better for the same pixel count than Dslr.

So it's not always uniquely pixel count that defines decent print quality.

Logged
Dan Berg
Sr. Member
****
Offline Offline

Posts: 1532



WWW
« Reply #9 on: November 15, 2009, 07:45:05 AM »
ReplyReply

In reference to my 24" X 30" print it was only acceptable from a viewing distance of 2 feet at 4 feet pretty darn good and at 6 feet excellent. So that viewing distance as we all know will make or break an image that has been pushed to a larger size then it was intended.
Logged

edwinb
Full Member
***
Offline Offline

Posts: 143


WWW
« Reply #10 on: November 15, 2009, 09:00:02 AM »
ReplyReply

Quote from: neil snape
...

Many say the MF captures up rez much better for the same pixel count than Dslr.

So it's not always uniquely pixel count that defines decent print quality.

I support what Neil says here,
better quality pixels print much better quality pictures
and users of Sinars 6MP cameraback are still using it for professional work because of that and can up-res the files and still look good.
It is related to the noise in the pixels - clean data can be extrapolated but noisy pixels (eg from scanned film) will look poor in comparison.
Edwin
Logged

Edwin Blenkinsopp
Technical Manager
image2output
m:  +44 (0) 7836 674749
e: edwin.blenkinsopp@image2output.com
w: Sinar | Foba | Inkjet
Dick Roadnight
Sr. Member
****
Offline Offline

Posts: 1730


« Reply #11 on: November 15, 2009, 11:45:16 AM »
ReplyReply

Quote from: edwinb
better quality pixels print much better quality pictures
and users of Sinars 6MP cameraback are still using it for professional work because of that and can up-res the files and still look good.
It is related to the noise in the pixels - clean data can be extrapolated but noisy pixels (eg from scanned film) will look poor in comparison.
Edwin
Better quality pixels does mean Medium Format...

My theory is that with three 60Mpx shot slide-and-stitch, using a H4D-60 back on a Sinar P3 with Apo-digitars, I should be able to produce some very good, very big prints?
Logged

Hasselblad H4, Sinar P3 monorail view camera, Schneider Apo-digitar lenses
lluis
Newbie
*
Offline Offline

Posts: 33


WWW
« Reply #12 on: November 15, 2009, 01:14:58 PM »
ReplyReply

Well, MF is clearly out of my reach, and if it were I coudn't justify that expense. I'm working now with just 6 MP, a KM Dynax 7D, great camera but limited for big prints. I'm afraid I'll be able to make only good A4 prints. I'm waiting for my Epson 3880 to arrive soon. I plan to buy a Sony a900 and go for A2 prints.
Let me say I find we are in a great moment regarding photography. We can control and participate in the full creative process, from capture to the final print. I've been doing photography for 25 years and I thought it couldn't be possible. Wonderful!

Lluis
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7333


WWW
« Reply #13 on: November 15, 2009, 03:19:15 PM »
ReplyReply

Hi,

In my experience:

- 12 MP is good for A2.
- Much depends on optics and processing pipe line.

With my processing pipeline it can be hard to see diference between 12 MP and 24 MP in A2-print even if the advantage is very obvious in Photoshop.

http://www.pbase.com/ekr/image/107619976/original  (photshop)
http://www.pbase.com/ekr/image/107823207/original  (scanned from A2-print)

Experienced observer could not see difference in print.



Best regards
Erik


Quote from: neil snape
While I have not seen substantial differences between GF and late Photoshop uprezzing, I certainly do see a lot better printing from Qimge from the same images printed through the drivers.

The 12 MPX full frame cameras easily go up to A3, but start loosing out if the view distance is close compared to the full frame >20mpx cameras.

Many say the MF captures up rez much better for the same pixel count than Dslr.

So it's not always uniquely pixel count that defines decent print quality.
« Last Edit: November 15, 2009, 03:49:29 PM by ErikKaffehr » Logged

Scott Martin
Sr. Member
****
Offline Offline

Posts: 1311


WWW
« Reply #14 on: November 15, 2009, 03:44:15 PM »
ReplyReply

As someone who's been making and exhibiting 5x7 foot prints for the last few years, I think the only rule of thumb here is to make a print as see how you like it. Localized contrast techniques, grain/noise adding, media, lens sharpness, image content, viewing distance and subjective preference all play a huge role and screw up formulas. I never would have thought I'd be happy with a 5x7 foot print from 12-21 megapixel camera files but seeing is believing, and the challenge of pushing these traditional limits can be a good teacher.
Logged

alain
Sr. Member
****
Offline Offline

Posts: 275


« Reply #15 on: November 15, 2009, 04:01:47 PM »
ReplyReply

Quote from: ErikKaffehr
Hi,

In my experience:

- 12 MP is good for A2.
- Much depends on optics and processing pipe line.

With my processing pipeline it can be hard to see diference between 12 MP and 24 MP in A2-print even if the advantage is very obvious in Photoshop.

http://www.pbase.com/ekr/image/107619976/original  (photshop)
http://www.pbase.com/ekr/image/107823207/original  (scanned from A2-print)

Experienced observer could not see difference in print.



Best regards
Erik

I've even gone to 82cm length (about 40% more length than A2) with 12 mp and it's hardly noticeable on the print (used Qimage).  With good light (aka a light box) I can see some artefacts if I go closer than 30-40cm and I now where to look.   It is important that the starting image is technically good.
Logged
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #16 on: November 17, 2009, 12:57:15 PM »
ReplyReply

Quote from: ErikKaffehr
Hi,

In my experience:

- 12 MP is good for A2.
- Much depends on optics and processing pipe line.
I would agree with this. 16x24" (roughly A2) has always been my cutoff for most 12mp images. This requires using good glass, shooting technique, and post-processing as well as avoiding cropping. I find that 24mp does make a noticeable difference over 12mp in 16x24" prints on close examination; but the real benefit to 24mp is being able to make 20x30" prints with slightly better quality than my previous 16x24" prints. 20x30" at 200ppi looks great even from fairly close viewing distances.

Quote
With my processing pipeline it can be hard to see diference between 12 MP and 24 MP in A2-print even if the advantage is very obvious in Photoshop.

http://www.pbase.com/ekr/image/107619976/original  (photshop)
http://www.pbase.com/ekr/image/107823207/original  (scanned from A2-print)

Experienced observer could not see difference in print.
Really? To me the difference is apparent even in the scans of the prints from your example. Of course that's not quite the same as comparing actual prints. Maybe the scans are equivelent to sticking your nose in the print. But my experience is that you don't need a loupe to see the difference between 12mp and 24mp at that print size.
Logged

Dick Roadnight
Sr. Member
****
Offline Offline

Posts: 1730


« Reply #17 on: November 17, 2009, 01:26:51 PM »
ReplyReply

Anyone printed A0 from 50 or 60 Mpx?

...or 44 * 60 (with 44 or 60 inch paper) with 4 shot 2 row slide and stitch?
Logged

Hasselblad H4, Sinar P3 monorail view camera, Schneider Apo-digitar lenses
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7333


WWW
« Reply #18 on: November 17, 2009, 03:24:49 PM »
ReplyReply

Hi,

The prints were actually crops corresponding to A2.

I consulted two of my friends, one is much younger and may be better eyseight, the other one is a guy who was working at one of the best professional labs in Sweden. This lab was employed by the famous photographer Lennart Nilsson, famous for pictures about life,  http://en.wikipedia.org/wiki/Lennart_Nilsson .

What the lab guy said was that the A-700 print had slightly higher contrast and therefore may look sharper, but he failed to see any difference.

The both images were done using the following pipeline:

- Raw conversion lin Lightroom with same sharpening presets (landscape?)
- cropping to 50% in linear scale (so we see 1/4-th of the area)
- printing on A4 paper with interpolation to 480 PPI, and standard output sharpening for glossy paper

So the prints were actually A4 but the crops corresponded to A2.

I was very surprized when I saw the result.

I have also scanned the prints at 600 PPI, but the files were to big to use.

I'm not really in the lense/camera/printer testing business, but want to share my experience. The idea that some readers may some money on my input.

Best regards
Erik





Quote from: JeffKohn
I would agree with this. 16x24" (roughly A2) has always been my cutoff for most 12mp images. This requires using good glass, shooting technique, and post-processing as well as avoiding cropping. I find that 24mp does make a noticeable difference over 12mp in 16x24" prints on close examination; but the real benefit to 24mp is being able to make 20x30" prints with slightly better quality than my previous 16x24" prints. 20x30" at 200ppi looks great even from fairly close viewing distances.


Really? To me the difference is apparent even in the scans of the prints from your example. Of course that's not quite the same as comparing actual prints. Maybe the scans are equivelent to sticking your nose in the print. But my experience is that you don't need a loupe to see the difference between 12mp and 24mp at that print size.
Logged

Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad