Where to set resolution during workflow?

Iliah Borg said:

jimhughes said:

Late to this party, but I recently wrote a blog post attempting to clear up the PPI/DPI question and maybe put a stake through its heart:

https://jimhphoto.com/index.php/2020/12/21/dpi-a-concept-you-should-forget/

Click to expand...

This isn't quite right:

"A photo from a camera has just pixels. That’s all – no physical size and therefore, no inches. It has pixel numbers, like 6000×4000, depending on its sensor size, but the image has no physical size. So it can’t have a DPI or PPI value. It just can’t."

Before demosaicking it's Bayer raster, thus one can say they are dots, like dots in CMYK raster. Also, one can argue that the image has the physical size, and that's the sensor image area size. This makes dpi metrics for a raw image valid, and ppi metrics - valid for demosaicked images.

Click to expand...

Iliah Borg said:

The reason why size makes sense is because, for example, the effect of diffraction, as well as artifacts that have fixed real dimensions, scale with the size. Imagine having a 0.1 mm dust spec on a m43 and FF sensors having the same pixel counts; if images are printed at the same size, magnification is different and the dust spec will look larger on the print from m43.

Click to expand...

Jestertheclown said:

Iliah Borg said:

jimhughes said:

Late to this party, but I recently wrote a blog post attempting to clear up the PPI/DPI question and maybe put a stake through its heart:

https://jimhphoto.com/index.php/2020/12/21/dpi-a-concept-you-should-forget/

Click to expand...

This isn't quite right:

"A photo from a camera has just pixels. That’s all – no physical size and therefore, no inches. It has pixel numbers, like 6000×4000, depending on its sensor size, but the image has no physical size. So it can’t have a DPI or PPI value. It just can’t."

Before demosaicking it's Bayer raster, thus one can say they are dots, like dots in CMYK raster. Also, one can argue that the image has the physical size, and that's the sensor image area size. This makes dpi metrics for a raw image valid, and ppi metrics - valid for demosaicked images.

Click to expand...

You're really clutching at straws with that scenario.

Click to expand...

davidedric said:

I don't know exactly who Mr Borg is, but he rarely if ever calls it wrong

Click to expand...

Jestertheclown said:

You're really clutching at straws with that scenario.

.

"It's good to be . . . . . . . . . Me!"

Click to expand...

Jestertheclown said:

davidedric said:

I don't know exactly who Mr Borg is, but he rarely if ever calls it wrong

Click to expand...

There's always a first time . . .

"It's good to be . . . . . . . . . Me!"

Click to expand...

Raul said:

Jestertheclown said:

davidedric said:

I don't know exactly who Mr Borg is, but he rarely if ever calls it wrong

Click to expand...

There's always a first time . . .

"It's good to be . . . . . . . . . Me!"

Click to expand...

Mr Borg is a sage.

Click to expand...

gscotten said:

the metrics of a sensor chip have nothing to do with the DPI tag in an image file which is, for the OP's purpose, absolutely meaningless.

Click to expand...

Iliah Borg said:

gscotten said:

the metrics of a sensor chip have nothing to do with the DPI tag in an image file which is, for the OP's purpose, absolutely meaningless.

Click to expand...

You've misunderstood me.

I was referring to "the image has no physical size".

My point is this: how do we apply vignetting, angular response, and other math / physics corrections not knowing the geometry of the sensor and a rule to calculate polar coordinates of a pixel? What do we need to calculate those coordinates without a notion (explicit or implicit) of pixel density / image physical size?

Click to expand...

gscotten said:

Iliah Borg said:

gscotten said:

the metrics of a sensor chip have nothing to do with the DPI tag in an image file which is, for the OP's purpose, absolutely meaningless.

Click to expand...

You've misunderstood me.

I was referring to "the image has no physical size".

My point is this: how do we apply vignetting, angular response, and other math / physics corrections not knowing the geometry of the sensor and a rule to calculate polar coordinates of a pixel? What do we need to calculate those coordinates without a notion (explicit or implicit) of pixel density / image physical size?

Click to expand...

Iliah, I agree with you 100%. It just has nothing to do with the OP's question about the DPI tag in an image, and diverts this thread to a completely different place than the OP was asking about. Forum threads do that a lot, and I am not complaining. I simply wanted to make that clear to the OP.

Click to expand...

Flycaster said:

Although my posting has lead to further discussion that is beyond my ken or needs, I have no problem as I got the responses that I was looking for. And, you guys got to discuss stuff at a higher level. So, all is well...

Click to expand...