Luminous Landscape Forum
Raw & Post Processing, Printing => Digital Image Processing => Topic started by: Jonathan Wienke on May 24, 2005, 08:12:06 pm
-
I already answered that. Read the article I linked.
-
Jan, sorry english isn't my mother tongue, with "All of these contribute to the illusion of resolution, so that the number of pixels doesn't even tell half the story" you mean that even we would able to find the resolution concept in the physical sensor of the camera, it probably only adds confusion because digital image doesn't have resolution (like px/mm) ¿?
It's not just because it's digital, it's because the different sensors of different cameras work in different ways.
In many ways, this can be considered to be similar to how different films work, Fuji Provia and Velvia, various Kodaks, if you just do b&w Ilford with red, green and blue filters, what your film's grain size is, etc.
-
Thank you for your reply, Jack
-
One of my doubth is if you can set or handle resolution in a digital camera, I think what dinarius says -My Canon 20D RAW conversions- is about some printing software that makes a raw file to be printed, but no handle in the camera. Can you confirm that¿?
-
Can anybody tell me if in the RAW conversion setting you can set resolution (in px/inch, px/cm or whatever it be)? Thanks in advance
I'm not sure about exactly which setting you're thinking of, but in the Camera Raw window, you can set the target print resolution, yes.
This is also possible to do in most other raw converters.
-
Jonathan Wienke, I have asked to Nikkon about the tagging of RAW files with a number of PPI, and they have said to me that there is no possibility of doing something like that or similar. Also somebody has told me through private message that he know nothing about this tagging. I need the information for third persons, not only for me, and because of it I would like the information was complete. At first you gave me very useful posts but I can’t reach the reason why you leave this doubth and the thread open, is it so difficult to specify the purpose on it like for example perhaps the default resolution (px/inch, px/cm,..) for printing after, or to say that you don’t know its purpose, or to say that it likely weren’t the correct words? Can you or anybody say something about that in the open forum? I would be gratefull for any information
-
> It has been answered by MarkDS, but thank you very much anyway
But I didn't agree with his answer ("it serves no purpose whatsoever") which is why I offered my opinion. It depends on your workflow whether early tagging is useful or not. If you do everything in PS then it doesn't matter except than tagging correctly for your workflow to start with saves an extra step in PS. For my workflow I often resize in the raw converter (C1) so I'd be lost without the ability to tag and show inch dimensions at that stage.
- DL
-
Hi, we have two different opinions about a so basic and at once vital conception like resolution. I hope a enquiry about it be appropriate in this forum. I put below the two versions and I would be grateful for any help in order to determine which is the right one. Thank you in advance.
Version 1:
1.a. The image size professionally is, in both analog and digital, measured in metric units (centimeters and its multiples or inches). And the resolution is always measured in pixels/metric unit. The result of this two variables (which is a specific amount of pixels) is, in general conceps in digital photography, only the resulting amount of pixels, only that. Finding a reason for what in some digital cameras is erroneously spoken about «image size» in pixels without that image being for the web, I find that really there is a fixed relation between size and pixels, independently of resolution, but it refers to the image file’s size, measured in MB, which certainly is always directly proportional to the amount of pixels.
1.b. When we do a photo with a digital camera, the size is always the same, those that lens is able to catch. And resolution is the amount of pixels in which that photo is going to turn into, in other words, the density of pixels that a size of photo is going to be: those that lens is able to catch and give to the «scanner» of the camera in order to transform it by the scanner in a whole spectrum of pixels, and which is going to determine the accuracy of the photo’s details, that is: its resolution.
Version 2:
2.a. In photography with digital camera, resolution (px/cm, px/inch) doesn’t exist, photography with digital camera exclusively works with image size in pixels, which is named resolution. The same occurs in digital image in screen (for example web, also screen proyection): photography with digital camera exclusively works with image size in pixels, which is named resolution.
2.b. In photography with digital camera, the size of the caught photo isn’t always the same: generally, the photosensor is a Area Array CCD, that consists in a reticulated matrix of hundred of thousands of microscopic photosensitive cells (photodiodes). Each photodiode matchs with one pixel, so the more photosensor the CCD have, the better the quality obtained with the camera will be. Those number of photosensors can be associated to a physical size (that which they themselves take up inside the camera, that isn’t the one of the image because it doesn’t exist as we physically know it), this physical size in any case would be variable by changing the resolution with which we take the photo (size in pixels x pixels).
-
I see, this whole link matchs perfectly with the question, clearly answered: "Digital images are nothing more than an array of pixels (numeric values sepcifying color values at specific points in a rectangular grid); they have no intrinsic size or weight. The resolution of a digital image is correctly designated by the number of pixels it contains horizontally and vertically. The more pixels the file has, the greater the resolution of the image".
It's very concrete and interesting. Thanks for all
-
Anyway, resolution (px/inch, px/cm) would be always linked to the physical support of the digital camera, while resolution (px/inch, px/cm) isn't inherent to the image itself, which only needs size in pixels. Right?
-
In the camera, the number of pixels does not change UNLESS you choose a "small" jpeg or use "digital zoom." In both cases you "throw away" pixels
When we post-process the image in a program like Photoshop AFTER it has been taken, we are also controling the pixels.
-
HOWEVER! You probably won't be able to see those gains in an A4-sized print.
Just what I wanted to hear! ;-)
90% of what I do rarely goes above A5, never mind A4, so I guess I can stick with my 20D then.
I put the question of size to a bunch of photographers recently, and they agree with me. Most of what they do stays A5 or smaller.
Who needs PhaseOne!? ;-)
D
-
Peter, going back to your question several posts above, if you are referring to the Adobe Camera Raw converter for raw files you have downloaded from the camera to the computer, YES, you can specify the amount of PPI in a dialog box in the Camera Raw window. However, Jonathan is absolutely correct - it serves no purpose whatsoever. For example, with my Canon 1Ds, this resolution is at 240 PPI. Because the pixel dimensions of the RAW image from a Canon 1Ds are 4064 by 2704, without changing anything the print dimensions would by 4064/240 by 2704/240, or 16.9 by 11.3 inches. I send the image from Camera Raw to Photoshop in this state, assuming no cropping at the Camera Raw stage. Then in Photoshop I have the choice of (a) without resampling the image increase the PPI, which correspondingly reduces the image size (calculation same as above because the number of pixels is fixed but if I change the pixels per inch, Photoshop shows the resulting image dimensions (in inches or cm whatever) of dividing the new PPI value into the number of pixels for each dimension), or (b) resampling the image in which case I can select whatever PPI and image dimensions I want.
Does this clarify it completely?
-
in the Camera Raw window, you can set the target print resolution, yes.
I understand you refer to printing the raw file from the camera, don't you? I meaned if you can set the image resolution (px/inch, px/cm,..) to be stored in the raw file of the image. Thanks again
-
dlashier, you have explained the basis of resolution concept (even through dimensional analysis), i know the utility of resolution but i meant the tagging in some raw conversions. It has been answered by MarkDS, but thank you very much anyway
-
Digital images are pixels and have no physical size whatsoever until they are printed. See http://www.visual-vacations.com/Photography/digital_dpi.htm (http://www.visual-vacations.com/Photography/digital_dpi.htm) for more details. The camera sensor has physical dimensions, but the bits in the image do not. How wide is the number 23464679?
-
This has been discussed a few times on these forums.
There is, however, something that's not mentioned in Jonathan's article (but it is in several posts here on this site), and that's the issue of the physical properties of the CCD or CMOS sensor.
If you have e.g. a 6 Mpx sensor laid out as 3000x2000 px
over a 30x20mm area, then you have 100 px/mm density.
But the actual photosites capturing a pixel may be smaller than what this implies; instead of being 10 micron wide, depending on manufacturing process, they could be e.g. 9 or 7 micron wide.
This makes a difference in how much light each photosite can capture.
You also have the issues of the antialiasing filter, the colour filter ("bayer matrix" or lack thereof), whether photosites are laid out in a regular matrix or a 45-degree rotated matrix, etc.
All of these contribute to the illusion of resolution, so that the number of pixels doesn't even tell half the story.
Search the forums for "foveon", "bayer", "sigma", "fuji S2 Pro", "photosite", and similar keywords for more information.
-
Thanks Jack, I understand that in a unique digital camera (obviously with a unique lens) resolution never change, and user never handle it, but user handle the image size (pixels). Could you only tell me if it's right¿?
-
Your P1 has more native pixels at A4, so by default you have more detail.
HOWEVER! You probably won't be able to see those gains in an A4-sized print.
-
Just been with a client looking at high res scans of 6x7cm trannies that I shot last year, side-by-side with shots taken yesterday on my 20D.
There is nothing between them. 20D images might even have the edge in sharpness. Colour in 20D images is undoubtedly better, though to be fair, the scans could be corrected later.
D.
-
I just had two of my best 6x7 trannies high-res scanned, and compared with the same shot taken with my 10D (not even a 20D). And sectional blowups to 16x20" show no more detail and no greater dynamic range in the 6x7s than from the 10D. ???
Anybody want to buy a great Pentax 67 II kit with lots of good lenses?
Eric
-
The camera may (or may not) select an arbitrary PPI number to tag the RAW with, but it's completely meaningless, and frankly not even worth wasting oxygen thinking about. You have total freedom to override this tag at any point in your workflow. It's much more useful to simply think of digital images in terms of pixels instead of physical dimensions like inches or centimeters, since a digital image file has no physical size whatsoever. Only prints have a physical size, and the pixels/inch or whatever are only determined when you send the print job to the printer. Any tags in the file are completely meaningless until then, and most of the time afterwards as well. There is no reason a print has to have any correlation to the Pixels/inch or pixels/cm tag in the image file. It's merely lucky happenstance if it does.
-
MarkDS, absolutely. I wonder why Nikon said that, I understand that not all professional cameras makers give this possibility. Anyway, the main doubth that my group discussion could have with Jonathan's post was if you can set the photograph resolution before you take it or before the photograph is transformed in a spectrum of pixels , the reply for me is clearly NOT but not for them and I wouldn't be able to solve them this doubth without your post. Thank you very much indeed
-
So you can also resize in the raw converter, but in the same way that you would do in any other software, and always after taking the photo. It's something more to learn. Thanks
-
Thanks for reply, Jonathan, and can you tell anything about first question? Does digital image itself have resolution as we know (px/inch or px/cm) or itself have only size (px)¿¿??
-
Jan, sorry english isn't my mother tongue, with "All of these contribute to the illusion of resolution, so that the number of pixels doesn't even tell half the story" you mean that even we would able to find the resolution concept in the physical sensor of the camera, it probably only adds confusion because digital image doesn't have resolution (like px/mm) ¿?
-
1000px x 1000px is always 1000px x 1000px, which will always be 1 million pixels or 1 MP...
1) It does not matter what the diameter of those pixels are for that comparison. However,
2) Now enter different diameters for pixels. If a set of pixels is twice the diameter of another, they will require a chip with 4x the area to hold the same number.
3) With the larger AREA of the chip and 2xs the diagonal, you will need a lens with 2x the focal length to get the SAME field of view with the two image systems.
(I believe concept #3 is the resolution hurdle where folks get confused.)
4) Soooo, both chips have the same resolution but different image-sensor sizes and hence require different focal lengths to get the same picture. IF you now try to "compare" the cameras using the SAME lens, the sensor with the smaller pixel diameters will be able to offer greater DETAIL, assuming the lens can accomodate it, but the resultant image will be as though you cropped the center from a chip with 4X the resolution.
Hope this helps,
Jack
-
My Canon 20D RAW conversions at 300 PPI end up as near as dammit to A4 in size.
Am I right in understanding that, if I *know* that my images are never going to be used larger than A4, I don't gain anything by shooting for the same size print on, say, a PhaseOne?
Or is it the case that, because the P1 native size is greater, if it is reduced to A4, the resolution would be better?
Thanks.
D.
-
90% of what I do rarely goes above A5, never mind A4, so I guess I can stick with my 20D then.
I put the question of size to a bunch of photographers recently, and they agree with me. Most of what they do stays A5 or smaller.
Who needs PhaseOne!? ;-)
I see the smiley.
I don't know about a PhaseOne back, I have nothing to attach it to (unless it magically fits that No. 2 Folding Cartridge Hawkeye), but I'd really love to do prints as wide as A1 or A0 that I could stand a foot away from and still not see all the details in.
And I'd like to do those kinds of prints with as little stitching as possible.
Well, I can dream, can't I?
-
Can anybody tell me if in the RAW conversion setting you can set resolution (in px/inch, px/cm or whatever it be)? Thanks in advance
-
ut I would like to know the purpose of this taging because this point has been cause of discord in same recent discussion. In other words, I would like to see as detailed as possible. because taging the raw is meaningless and you can override this tag at any point in your workflow, is there any purpose on it like for example the default resolution (px/inch, px/cm,..) for printing after?
The only reason for tagging is to offer an option of showing out size in real dimensions - eg. inches or centimeters. You must specify resolution (pixels/physical-measurement) in order to mathematically convert from pixels to measurement (inches etc), and carrying along a tag can simplify the workflow assuming it is appropriately set.
Think back to HS math - dimensional analysis. "Pixels" is dimensionless. In order to convert to a dimensioned value (inches) a conversion factor is required, in this case pixels/inch. Size(inches) = pixels / resolution, or dimensionally analyzed inches = pixels / (pixels/inch) = pixels * inches/pixel, and the pixels cancels leaving you with inches.
- DL
-
I totally agree with you Jonathan, but I would like to know the purpose of this taging because this point has been cause of discord in same recent discussion. In other words, I would like to see as detailed as possible. because taging the raw is meaningless and you can override this tag at any point in your workflow, is there any purpose on it like for example the default resolution (px/inch, px/cm,..) for printing after? Thanks indeed
-
Don, the reason I said that is because whether you dimension in RAW or dimension in PS, you are still at one stage or another deriving inches from pixels as a function of resolution, and as you say - where you do it is a matter of personal preference. Your arithmetic is of course correct and, as I am sure you would acknowledge, complements what I was explaining to the gentleman about how these relationships work, differentiated by whether or not one resamples the image.