[TonyW]

No one has ever printed at 72 ppi (they may think they have!) on an inkjet because it will be upsampled to printers native resolution if you do not do it yourself. 72 dpi is a typical dot matrix printer resolution from many years ago

Upsampling yourself to 300 or 600 ppi is fine if you have Canon printer but is wrong for Epson as it will use 360 or 720 ppi. Doing it yourself to printers native resolution is better than letting print driver upsampled using less than stellar algorithms.

Placing no value on print resolution is absolutely fine. By the same token no value needs to be placed on lens quality/resolving power, diffraction or any other aberration as without treating image data correctly none of it matters as you will have discarded the benefits of good glass with poor printing

Happy with throwing away the benefits of quality equipment at the printing stage is ok by me if it suits the individual.

Take chances with your printing absolutely, and for that matter your photography, but do yourself a favour and at least understand where your printing may fall down

[normhead]

No one has ever printed at 72 ppi (they may think they have!) on an inkjet because it will be upsampled to printers native resolution if you do not do it yourself. 72 dpi is a typical dot matrix printer resolution from many years ago

Upsampling yourself to 300 or 600 ppi is fine if you have Canon printer but is wrong for Epson as it will use 360 or 720 ppi. Doing it yourself to printers native resolution is better than letting print driver upsampled using less than stellar algorithms.

Placing no value on print resolution is absolutely fine. By the same token no value needs to be placed on lens quality/resolving power, diffraction or any other aberration as without treating image data correctly none of it matters as you will have discarded the benefits of good glass with poor printing

Happy with throwing away the benefits of quality equipment at the printing stage is ok by me if it suits the individual.

Take chances with your printing absolutely, and for that matter your photography, but do yourself a favour and at least understand where your printing may fall down

I'd also recommend everyone do some printing at 300 DPI or 360 DPI, whatever is native to your printer, real DPI, not upsampled, there are images, I think of Pinholecam's cityscapes that will use every bit of the resolution provided. Even if on a print you decide to ignore printing guidelines, you still need to recognize which prints won't be affected by print resolution and which will be much better if you exploit it. There will always be images for which it matters little, but there will also be situations where it matters a lot.

If you haven't looked at some images that map your image to a printer 1:1, you won't be able to evaluate whether staying hi-res is appropriate. Everyone should do one or two at least.

But for most of us printing big it's low res or nothing. I've never been disappointed when exceeding my own guidelines, but usually the alternative would be buying a 100 MP camera so I had the required resolution, and that's not going to happen.

[bobbotron]

It's best not to optimize prints for people that will sit 12 inches away from your large print and nit pick. I've larger prints on walls, even if the resolution isn't great up close, it's incredibly pleasing to have a larger poster sized image to look at in a room.

[angerdan]

With 300dpi, you can print on A3 with 6MP with good quality.
So 12MP would suit A2, 24MP A1 and 48MP A0.
How Large Should You Print? - Photo Review

Here's a table about corresponding resolutions to formats with a fixed 300dpi print quality.

Format	MP (300dpi)	camera model
9x13	2	*istD
10x15	2	*istD
13x18	3	*istD
20x30	9	K-m, K200D
DIN A4	9	K-m, K200D
DIN A3	17	K-50, K500, K-5 II(s), K-01
30x45	19	K-S1, K-S2
DIN A2	35	K-1 (II)
50x75	54	645Z
DIN A1	70
DIN A0	131

This table shows the recommended dpi for each viewing distance:

Distance	min. Resolution
0.6m / 2ft	300 dpi
1m / 3.3ft	180 dpi
1.5m / 5ft	120 dpi
2m / 6.5ft	90 dpi
3m / 10ft	60 dpi
5m / 16ft	35 dpi

You can calculate everything in Excel or with this online tool:
Blog - The Relationship Between Viewing Distance and DPI

---

Last edited by angerdan; 04-24-2021 at 07:58 AM.

[angerdan]

I'll put the values into a table soon.

[TonyW]

I would be wary of taking figures from a website that confuses printers DPI with a printers required resolution which is measured in PPI and is really the most relevant measure of resolution. And further compounds confusion by stating that they can be viewed as being the same - they are not!

Calculation for the required PPI can be made by using this formula:
PPI = 1/((distance x 0.000291) / 2)

An easier to remember method is to take 3438 and divide by the viewing distance. Therefore at a viewing distance of 12” a minimum resolution equal to 286.5 PPI is required.

When considering PPI requirements based on these calculations there are some points to consider:
1. The figures quoted are based on your average print being viewed under average lighting conditions. These figures can easily be doubled for ideal viewing conditions (e.g. exhibition lighting) and images containing great detail.

2. Be aware of false precision. The above 286.5 PPI is an example, it should only be taken as a rough guide to what may be required as a minimum

3. For optimal results the PPI requirements of the printer should be used regardless of native PPI of your image at your required print size. Epson 360/720 ppi, Canon/HP/Fuji 300/600 ppi

[biz-engineer]

An easier to remember method is to take 3438 and divide by the viewing distance. Therefore at a viewing distance of 12” a minimum resolution equal to 286.5 PPI is required.

I guess this is kind of a general rule and it doesn't fully work for my eyes.

Let me explain a bit why I think the ppi proportion to viewing distance doesn't fully work for me. I have done an experiment.
1) View a crop of picture containing small details, say I can appreciate details of 300 ppi at 12" viewing distance
2) Blow up the same image crop 4x, and go back 12"x4 from the print, see if I can perceive and count the same details

Surprise (at least for my eyes, having correction glasses for close and far), I can see much better the 4x details at 4x the distance, than the same 1x details at 1x distance.
My visual acuity isn't a constant vs distance. So that means, if I print small, I need proportionally less resolution than if I print big, contrary to common belief. That could mean, for example, I'd print at 300 ppi for a small print viewed at 12", I would still need 250 ppi instead of 150 ppi for a print 2x larger viewed at twice the distance , 24" for instance. So, at least for all people like me, a larger print viewed from further away will still benefit from more pixels than the common 300 ppi at 12".

---

Last edited by biz-engineer; 04-23-2021 at 08:44 AM.

[house]

Different images will require different resolution prints. Any rule of thumb can be significantly broken.

[TonyW]

I guess this is kind of a general rule and it doesn't fully work for my eyes.

It is a guideline based on the maths of visual acuity based on someone with 20/20 vision. Obviously if your vision is better or worse than 20/20 the recommendation will change

...
My visual acuity isn't a constant vs distance. So that means, if I print small, I need proportionally less resolution than if I print big, contrary to common belief. That could mean, for example, I'd print at 300 ppi for a small print viewed at 12", I would still need 250 ppi instead of 150 ppi for a print 2x larger viewed at twice the distance , 24" for instance. So, at least for all people like me, a larger print viewed from further away will still benefit from more pixels than the common 300 ppi at 12".

The idea of PPI and print size needs to be understood. The required PPI is the figure that at a given viewing distance an image comprised of dots will be perceived as a continuous tone . In other words a viewer will not see the spaces between the dots.

Contrary to what you seem to have found is the fact that a smaller print will, if viewed at a reasonable distance require more pixels to resolve detail that may be contained in your image data than an enlarged version may require viewed at a similar ‘reasonable’ distance - taking into account that even if detail is present it may not be able to be seen with the unaided eye due to distance from print.

I prefer to think of what minimum pixel requirements for a certain viewing distance than the commonly quoted 1.5 - 2x print diagonal. So I use the 3438/estimated closest viewing distance for a print to be viewed in a particular setting.

So, as an example an 10x8 image viewed at 12” would require a minimum of 286 ppi whereas a 20x16 images viewed at 24” would require a minimum of 143 ppi when viewed at 24”. Note figures quoted can be considered to a degree moot, due to the way printers handle image data!

The commonly quoted 300 ppi is somewhat of a myth relating to inkjet printing and visual acuity. 300 ppi is the lower resolution required by Canon and HP printers and is one that some quality magazines require for offset printing. They will use lines per inch (LPI) and may quote 150 LPI as the requirement, which when translated to PPI will require a minimum of 300 ppi.
Epson as already covered require 360 ppi at the lowest resolution requirements.

Regardless of what you think you are sending to the printer in PPI your image if outside of the printers declared resolution will be resampled to the printers requirements prior to printing. So your examples of 150ppi and 250 ppi would be resampled to 300 ppi for a Canon printer or 360 ppi for Epson

---------- Post added 04-23-21 at 11:17 AM ----------

Different images will require different resolution prints. Any rule of thumb can be significantly broken.

Not really anything to do with rules of thumb, as has already been covered in depth

[biz-engineer]

So, as an example an 10x8 image viewed at 12” would require a minimum of 286 ppi whereas a 20x16 images viewed at 24” would require a minimum of 143 ppi when viewed at 24”.

That's a very good baseline. I was just mentioning that my eye don't have the same angular resolution at each distance. If I look at a print as close as 6 inches, my visual acuity drops so I can't tell the difference between a 200ppi print and 300ppi print at that distance. Now, if I look at a print at 30" distance, my eyes are very good at noticing small details, in that case I can appreciate the 3438/estimated closest viewing distance rule. Surprising to me was that small prints are easy to make because of looking at them closer and less total amount of pixels to cover the area at a given print PPI figure. 40" large photographic print is much more demanding, and not only because my eye get razor sharp at 40" viewing distance, but in addition I can easily look closer than 40" if I want to and do so effortlessly. Whereas it would be a big effort to look closer than 6" on a small print, since most people can focus their eyes that close anyway . So , my conclusion was that the 286ppi at 12" viewing distance my not be as valid at other distances. The real world vision is non-linear, visual acuity angle drop when looking very close and drop again when looking far away. Hence, 40"x 60" print is most demanding in terms of showing details.

[house]

Not really anything to do with rules of thumb, as has already been covered in depth

The OP is about "megapixels" ie camera file size. You can up sample some images without the print suffering while other photos may show distracting softness. Ie it's not just a technical question.

[biz-engineer]

The OP is about "megapixels" ie camera file size. You can up sample some images without the print suffering while other photos may show distracting softness. Ie it's not just a technical question.

I have a picture, taken with Pentax K1 and DFA28-105, ISO 200, shutter speed 1/400, the picture is printed 30"x 40".

In the foreground left and right bottom corners of the prints are two pine trees, the detail is amazing, looking close and far, everything is that area of the frame looks excellent :-).
In the background is a mountain with jagged peeks covered with snow; when processes the file, I didn't pay much attention to lens CA, not fully corrected. When looking at the print, I can immediately see the lens CA at the mountain edges, even looking at more than 1.5m away from the print. If I used a good 35mm prime lens instead of the D-FA28-105 at 35mm, the picture would be perfect. In that case, it's printed at 160 PPI (camera, before up-sampling for the print), the lens optical performance gets in the way of getting the print excellent corner to corner.

[TonyW]

That's a very good baseline. I was just mentioning that my eye don't have the same angular resolution at each distance. If I look at a print as close as 6 inches, my visual acuity drops so I can't tell the difference between a 200ppi print and 300ppi print at that distance. Now, if I look at a print at 30" distance, my eyes are very good at noticing small details, in that case I can appreciate the 3438/estimated closest viewing distance rule. Surprising to me was that small prints are easy to make because of looking at them closer and less total amount of pixels to cover the area at a given print PPI figure. 40" large photographic print is much more demanding, and not only because my eye get razor sharp at 40" viewing distance, but in addition I can easily look closer than 40" if I want to and do so effortlessly. Whereas it would be a big effort to look closer than 6" on a small print, since most people can focus their eyes that close anyway . So , my conclusion was that the 286ppi at 12" viewing distance my not be as valid at other distances. The real world vision is non-linear, visual acuity angle drop when looking very close and drop again when looking far away. Hence, 40"x 60" print is most demanding in terms of showing details.

Thanks, I appreciate that you are approaching this from your perspective relating to your own visual acuity.

Your conclusion that a viewing distance of 12" for 286 ppi may not be correct for other viewing distances is sound. The answer usually given to the often quoted 'how many pixels needed for a print?' is based on taking the diagonal of the print and multiplying by 1.5 - 2x to get to the viewing distance then applying the formula for the visual acuity of someone with 20/20 vision.

So for your 40" x 60" print you would be looking at a requirement of a native file size of around 32 ppi minimum to give the impression of continuos tone and a reasonably well resolved image at a viewing distance of 108" (9'). The closer a viewer gets the more MP required, so at a viewing distance of 36" we would be looking at a minumum of 95 ppi and at 12" we are back to 286 ppi. Based on someone with 20/20 visual acuity

---------- Post added 04-24-21 at 03:32 AM ----------

The OP is about "megapixels" ie camera file size. You can up sample some images without the print suffering while other photos may show distracting softness. Ie it's not just a technical question.

Interesting take, as according to the OP title Megapixels vs Maximum Printing size it could be taken that we are talking about how many pixels (megapixels if you wish) do we need to print at a certain size or how do megapixels translate to print size.

So in reality it is a technical question and prints showing 'distracting softness' are generally the result of either going to far with not enough pixels for the required print size and viewing distance or the image data is intentionally soft.

[house]

To clarify with an extreme example. I saw Anders Petersen's Stockholm exhibition a year or two ago. Some prints were 1.5 - 2m tall others were just a couple of inches. All photographs were taken on fast black and white 35mm film using an analog Contax compact camera. None were inappropriately sized for that show.

Of course Anders Petersen has a extreme gritty style and technical perfection is way down the list of priorities. The logic does apply to other styles of photography as well as individual photographs.

About the photographer: Anders Petersen | LensCulture
Video flick through the book version of the exhibition:

You Tube

You Tube

the photographer and the camera is on the left at
https://youtu.be/t8BdC_AWmps?t=366

[MrNPhoto]

I've printed a 24MP image at 5 ft x 9ft. It is an image of an oil painting, so there is already a graininess to the image. It is mounted on a high wall, so you can't get closer than 8' to it. Looks great and the owner was very happy with the result.