Okay, I think I usually write too much, so I’ll try to make this really short. Hopefully that makes it simple too, rather than just more confusing. 🙂 (Sorry…I think I need a good editor. This came out just as long as other messages I’ve written 🙁 ).
Here’s an actual short version that I have inserted here, after having written another essay. Feel free to ignore the essay and pay attention only to this actual short version:
* With respect to the image file, the resolution of the file is unimportant. What is important is the number of pixels across the width and height of the image. The scans you are getting back have 1536×1028 pixels (which, by the way, I don’t really consider "professional quality"). If you change the image size without changing the resolution, you remove pixels (and thus important data). In the example you gave, you are removing 90% of the image data, if I read your message right.
* With respect to your printer, the resolutions of the image and the printer are only comparable if you actually print the image at the size that corresponds to the resolution of the image. In most cases, you will be choosing a specific size for your printout, and there is a new resolution that can be calculated from that size along with the number of pixels in the image. That new resolution IS comparable to the resolution of your printer; getting close to that resolution is useful, going past that resolution is not.
* Modern printers almost always are going to be at least as high resolution as the images printed to them (after adjusting for actual printed size), and often much higher resolution. So generally speaking, it’s best to just get as many pixels as you can for the original image. Because of that, "resolution" as stored in the image file is pretty much not useful information for many people. Just get scans of your images that have as many pixels in them as you can get, and let the software on your computer worry about the resolution stuff.
Okay, the long version…
When dealing with the digital data on your computer:
* Resolution basically just doesn’t matter.
* The number of pixels across the width and height of the image matters a LOT
The resolution is simply a way of describing how large the image is "intended" to be displayed, whether that be on the monitor or on a piece of paper. But it’s somewhat arbitrary; you can make the picture larger just by reducing the resolution, and you can increase the resolution just by saying the picture is smaller. All without actually changing what image information is in the file.
When dealing with digital output to a printer:
* Resolution basically just doesn’t matter.
* The number of pixels across the width and height of the image matters a LOT
Surprised that it’s the same for a printout as for working with it on the computer? Well, again…all that really matters is how much "information" is stored in the file, and that depends entirely on the width and height of the image in *pixels* (along with compression used, if any).
You are getting scans back on a CD that are 1536 pixels wide by 1028 pixels high. That’s the only information that really matters. When you print them to a printer, you decide at that time how large you want the print to be (using Elements, you set this in the Preview dialog…other programs have a variety of ways of deciding how large the printed image is). Then, the "final" resolution of the printed image is simply the pixel count in each direction divided by the actual printed size in each direction.
As far as the specifics of your situation goes:
* You don’t say how you’re changing the image size, but it sounds like you are actually removing "information" from the image, since you say you’ve reduced the image from 21×14 down to 7×5 without changing the resolution. This is a bad idea, since you want to retain as much information as possible, at least for the purpose of printing the image (sending in email is a whole ‘nothing deal, and often reducing the pixel count is useful). It sounds to me like you’ve altered the image from 1536×1028 pixels to 500×360 pixels, thus decreasing the total number of pixels by about 90%. That’s a huge reduction in information, and it will show up in your final prints.
* The resolution of your printer is only nominally related to the resolution in the image. That’s because the resolution indicated as part of the image file is simply a "suggested" size. What’s really important is the number of pixels and how large you wind up printing the image ("final" resolution, as described above).
* As far as that latter pair of factors go (number of pixels and how large it’s printed), you can calculate the final resolution as I mentioned above and compare that to your printer’s resolution. With an ink jet printer (or any dithering device, for that matter) you lose some detail because the printer cannot reproduce the same number of colors per pixel as is stored in the image file, and so mixes a fewer number of colors in a larger area to achieve the same range of colors.
Even so, you will find that you get the best results when the image’s final resolution as printed is near to or equal to the printer’s resolution. Printer’s with non-square pixels (such as the printer you have) complicate things, since digital images almost always have square pixels, but the same general rules apply. Try to get the final resolution in the ballpark of one or the other resolution numbers for the printer.
* With respect to you finding that the "pro" and "consumer" scans result in the "same size" image, what do you mean by "same size"? If you really mean that you are getting the same number of pixels (that is, both get you 21×14 images at 72 dpi) with both types of scans, I am with you wondering what the difference is. Certainly there’s no difference in the scan resolution. Maybe the shop does some kind of color balancing or the like as part of the "pro" service? Or maybe they are just ripping you off. I don’t really know.
* As far as what format to work in, I would use the .PSD format, simply because that’s the native format for Elements. Definitely do not use JPEG for anything other than final output, and even then you will probably want to limit that to situations where file size is more important to you than quality (for example, emailing pictures or putting them on a web site). You could use TIFF, and in fact in some situations it will be required (for example, giving the results to a printing shop that only accepts digital images in TIFF format). But TIFF will not store 100% of the characteristics of the file (certain kinds of layers, vector objects, that sort of thing), so again…it’s more useful when you are sure you’re done working on the image.
Sorry in advance if none of this is helpful. 🙂
Pete
Emeril, I’ll only add one point to Pete’s comprehensive explanation: if you’re paying extra for a ‘pro level’ scan, you should be getting images on the CD that are larger than 1500×1000, which is what the math says you’re seeing when you open the file in Elements – that’s the equivalent of what you get from a 1.5 megapixel camera. An image of 3000×2000 is what you’d really like, and it’s one of the options on Kodak’s Professional CD. To make large prints from the image, the larger scans are highly desirable.
Chuck
Chuck
To add one more point to your post: If Emeril cannot get the 3000X2000 on a CD it would probably be better, (albeit not easier), to scan your own prints and choosing your own settings.
PeteD
Thanks for the info it helps, one of these days I will understand this program and the digital darkroom.(can’t wait till I can get my cannon 10D and I won’t have to worry about film processing and scanning).
I have batch saved some of the jpegs I recieved form the film processor I use. I saved them as psd so I can work on them later at 4200 X 3000 pixels at 600 dpi this gives me a 7 X 5 inch pic according to what elements tells me this is about a 36 meg file. The pics look good on screen and are a good size to work with, at 200 magnifacation on screen they still seem good and clear.
Now if my understanding is correct, if I bump up my pixel setting after editing I should be able to print them out in as a larger print, say larger then 8 X 10 and still get a good clear print.
Emeril
Emeril, now I’m a little confused: did the original images from the processor come in at 4200×3000 or did you bump them up? If the former, that’s great; if the latter, and they really came in at 1500×1000 or something like that, increasing them to 4200×3000 wasn’t a good idea. To the extent possible, you want to start with the maximum pixel dimensions (forget about ppi and inch dimensions for the moment) that the camera, scanner or photo processor serves up. Then, in Image<Resize<Image Size, you want to UNCHECK the resample box and change the document size (in inches) to your desired print size. When you do that, the ppi will automatically recalculate; if you wind up with a ppi of 200 or more, you should have a very good print.
However, when you bump up the pixel dimensions (which you can only do with the resample box CHECKED), you are adding calculated pixels to what was contained in the original image. A little bit of that may be okay, but adding a lot will seriously degrade your image in sharpness and even in color rendition. Even if you do decide you need to add pixels in this manner, the approach that has been successfully demonstrated by Leen Koper, a professional photographer and forum regular, involves increasing the pixels in small steps. If that’s needed for what you want to do, please come back and we’ll add some details.
Bottom line: whatever pixel dimensions you start with when opening the image, stick with that if at all possible.
Chuck
Now if my understanding is correct, if I bump up my pixel setting after
editing I should be able to print them out in as a larger print, say larger then 8 X 10 and still get a good clear print.
Emeril,
What you are suggesting here is referred to as "upsampling". It will add those pixels but it is very questionable if it will do it well. While this function is useful sometimes, most of the time (for going to 8X10) the quality of the print will e degraded. The photo itself has a lot to do with the success of upsampling. If the image is not complicated it may work well.
The thought is to "scan" at a size and resolution that will be your final print output. (or ask for that when they burn Cd’s).
(You could scan one of the images you have setting scanner to 8X10 and resize the same picture from the CD and print, then compare your results to see what effect you get).
PeteD
Chuck,
I did not notice you over there or I would have just let you answer this one 🙂 (was I really typing for 14 minutes? … time between posts?)
PeteD
Hey, Pete, as long as we’re telling the same story, more is better!
🙂
Chuck
"Chuck Snyder" wrote in message
Emeril, I’ll only add one point to Pete’s comprehensive explanation: if you’re paying extra for a ‘pro level’ scan, you should be getting images
on
the CD that are larger than 1500×1000
Heh…in my defense, I did actually say that too. It’s just that I wrote so much other crap, it was probably hard to find that little tidbit in all that other verbiage. 🙂
Pete, I’m sorry – missed it on the first read of your post. But it’s a point worth reiterating, right?!
🙂
"Chuck Snyder" wrote in message
Pete, I’m sorry – missed it on the first read of your post. But it’s a point worth reiterating, right?!
Indeed. Probably THE most important point, since that has the biggest effect on the quality of Emeril’s printouts.
Pete
Peter,
I read your long, informative tutorial and I understand most of it. There are a couple of points that confuse me. And so, at the risk of making a fool of myself, I’ll just charge ahead….
With regard to display of images on the monitor, resolution is irrelevant, as you say. Monitors display pixels, and the resolution information in the image file is generally ignored. I say generally, because some slideshow software will downsample to fit the screen if the image is too large. (too many pixels)
The first advice I would give to Emeril is to UNcheck the Resampling box when adjusting the size or resolution of the image. That box should only be checked as a last resort. Resampling will always destroy image data, whether you are upsampling or downsampling. Of the two, downsampling is less destructive usually. The Constrain Proportions box should also always be checked unless you are trying for special effects, because changing one dimension of an image without changing the other will distort the picture. Generally not desirable.
Now let’s talk about printing. Printer dpi has absolutely nothing to do with image resolution! The printer uses multiple ink dots to produce a pixel of the downloaded image. Each pixel is NOT converted to an ink dot. If you think about it, it is clear that the printer is using blobs of C, M, Y and K ink to create a pixel that started out with three 8-bit values for R, G and B. The printer software does the conversion from RGB to CMYK, but it still takes a bunch of ink dots to reproduce the exact color of each pixel. In fact, sometimes it is unable to, and uses a procedure called "dithering" to get as close as possible. That’s another whole subject. There is a lot of confusion about ppi vs. dpi. The way Wayne Fulton describes it in has book "Scanning Tips" is as follows: Ppi and dpi are the SAME thing if you’re talking about monitor display of images. PPi and dpi are NOT the same if you’re talking about printing for the reasons described above.
A general rule is that 300 ppi is an optimum value for most printers. Higher resolution is at best a waste, at worst can actually degrade image quality. I know Peter disagrees with me on the last statement. Maybe it depends on the printer software. If you go much below 200 dpi, you will begin to see a lack of detail in the print. Even lower, and you will see pixelation.
So, given all that, let’s take one of your 1536 x 1024 images. If you follow the above instructions and set the resolution to 300 the print size will be 5.12 x 3.41 inches. If you go down to 200 ppi, the print size will be 7.64 x 5.12. In other words, you can’t get a good 8×10 from this image! If you want an 8×10 (or larger) you have two choices: Upsample or go down to 150ppi resolution.
Neither one will give you perfect results, and the best procedure can very from image to image. I often do it both ways, using the Unsharp Mask to sharpen it up as much as possible.
Anyway…another long, boring read…sorry.
Bert
Chuck,
The resolution I get from the processing lab on cd is 1536 X 1028 pixels in a 21.333 X 14.270 inch jpeg at a 72 dpi resolution. The is according to elements image info at the bottom left of my screen.
I load the jpegs into elements and then batch process them to my hard disk. The settings I use in the batch process is convert open to PSD, convert image size to 4200 X 3000 pixels resolution set to 600 DPI, rename files filename to newfilename and then save them to my HD.
I am reducing the picture size to 7 X 5 inches, a standard print size so I can edit and fix them in elements when I get a chance to. The origenal jpeg is not touched (so if I scr–ed up I can fix it).
I will try what you had suggested, turn off the resample and then change the image size and see how that works. One more thing does elements resample automaticly when you open a jpeg from cd and can you turn the resample off if it does ?
One of these days I will get the hang of the digital darkroom, it’s a far cry from what I use to do in my B/W darkroom 20 years ago. It’s been a long time since I have had the chance to get involved in photography like I did back then.
Emeril
Bert,
Not a boring read for me, every little bit helps. To everyone else once again I thank you for the info, I try to read each posting in here on a daily bases. It all helps out be it a long posting or a short posting, I do get around to reading every word in them.
But now a new question, I have just got back into photography after 20 years (marriage, kids and divorce, long story). If one wanted to get back into things, where and how does one get his photos displayed, sold or get a job/assignment. My usual line of work is drying up, I may not be as good as some of the other photographers on these forms, but I would like a chance to get involved full time.
Any ideas drop an e-mail my address should be listed.
Emeril.
Emeril, thanks for explaining your work flow. Going from an original of 1536 x 1028 to a final of 4200×3000 is really introducing a lot of calculated pixels via resampling, which is not helping your image quality. Here’s what I would do to get that 5 inch by 7 inch picture:
1. Open up original.
2. Go to Image<Resize<Image Size and uncheck resample. You should see the 1536×1028 in Pixel Dimensions, and your 21.33 inches, 14.27 inches and 72 ppi in the Document Size Segment.
3. Then replace the 14.27 inches with 5 inches; that should change the long side to around 7.48 inches; that’s as close as you can get to 5×7 without cropping a little off that long side. You could actually do that with a paper cutter or pair of scissors after you make the print, or you could crop in Elements; if you want to crop, we can give some suggestions in that area, too.
4. You’ll note that the ppi has changed from 72 to around 205 ppi; that’s really pretty close to the minimum for a decent print, but go ahead and try it to see if you’re okay with it. We may have to go back later and add a few pixels in via stepwise resampling to get the best print possible.
Bottom line on your photo processor; he’s not providing you with enough pixels at 1536×1028 to print much more than a 4×6; if you’re paying a lot extra for that service, it’s not cost-effective. As someone earlier in the thread suggested, you might be better off scanning a print rather than settling for 1536×1028…
To answer your other question, Elements doesn’t do any resampling upon opening; what you’re seeing is simply a calculation based on your image’s pixel dimensions (1528×1036) and a default resolution of 72 ppi. HOWEVER….once you start resizing or cropping, you can introduce resampling even if you don’t want it. Keeping that resample box unchecked until you really want to resample is very important.
Chuck
"Bert Bigelow" wrote in message
[…] PPi and dpi are NOT the same if you’re talking about printing for the reasons described above.
They are not the same, but provided one understands the relationship between the two, they are certainly comparable.
A general rule is that 300 ppi is an optimum value for most printers.
That general rule is simply wrong. The "optimum value" depends entirely on the actual printer and the technology the printer uses. But even for a consumer-level ink jet printer one can buy at CompUSA today for $150, printing at a resolution higher than 300 dpi will produce better results than at 300 dpi.
Pete
Ok Chuck, I will try what you suggest and see how it works out. As for paying the extra for the scanning at time of processing, it almost triples my cost. I have an epson scanner, it takes forever to scan the prints in (flat bed).
Epson has this little quirk,it doesn’t like you to use any other printer except epson, as for scanning directly into elements it wouldn’t even allow me to assosate elements as one of the application to scan directly to.
I will be getting a new scanner and printer soon, after I get the rest of my new photo outfit (just got back into photography after 20 years). I have to make due with what I have at the moment, new camera and lenses do take a bite out of ones budget. I’ve been using those little throw away point and shoot cameras for snap shots and family pics for years. Now I’m getting back into photography like I use to before I was married (not any more which is why I can get back to it). So any help, suggestions on equipment would be greatly appeceated.
Emeril
Emeril, re the scanner – can you just scan and save as a JPG or TIF file? Many (most?) scanners will let you do that. Then you can open the saved file in Elements and go from there.
I checked on the Kodak website; they do have an ISO (!) 400 speed black&white film that uses C-41 processing; I’ve never used it, but it may be worth a shot (no pun intended…)
Chuck
The "optimum value" depends entirely on the actual printer and the technology the printer uses
Pete,
You are correct, it does depend on the printer technology. I should have said that. I was referring to the consumer-level inkjet printers that most people in this forum use. I have read that number in several books, and it is often quoted by knowledgeable people on this forum. It is obviously not true for commercial printing.
I am curious, though. What $150 printer are you referring to? Bert
"Bert Bigelow" wrote in message
I am curious, though. What $150 printer are you referring to?
Take your pick. Just as a "for example", Epson’s Stylus Photo 925 uses 6 colors and has a resolution 5760×720, which is plenty high enough to take advantage of a higher image resolution than 300 dpi. They have a less-expensive model, the Stylus C84 with 5760×1440 resolution, that would do even better.
It would be hard today to buy a printer that could NOT take advantage of more than 300 dpi of resolution. You don’t even need to spend $150…I just picked that number because there’s a number of great $150 printers on the market today. It’s sort of a "sweet spot" for price/performance.
Pete
Chuck, Kodak even has two films that have to be processed in C-41. One is the usual amateur emulsion , the other one is part of the professional Portra film range and can be printed at the same settings as the CN Portra films. Both are rated at ISO 400, but generally the best exposure is a setting at ISO 200. Both films scan very well.
Emeril, if you want to be a professional photographer, join the Professional Photographers of America. These people can tell you how to get started again. There is a lot of knowledge in this organisation and joining their print competitions might provide you with a sound knowledge of todays demands. (This summer in England I happened to meet PPoA past president Mrs Helen Yancy and I was quite impressed by what she told me about the judging of the print competitions in the USA)
If you ‘ve got any questions, please do mail me; I ‘ll be glad to help you wherever I can. (You can find my e-mail adress by clicking my name.)
BTW, many pros will tell you you cannot do without the full version of Photoshop; I try to prove just only Elements will do. 😉
Leen
Leen
Leen, thanks very much for the info on the Portra films!
Chuck
Pete,
Thanks for the info. My old Epson Stylus Photo 780 has a max resolution of 2880 x 720. I have tried printing at higher resolutions than 300 ppi, and if there is any significant difference, my old eyes can’t see it.
I’m sure somebody must have done some tests on the new ones. I’d like to know what they recommend…both for the best printer and for optimum resolution.
Bert
Chuck, you are welcome. I’m glad finally I know something you obviously didnot know. Usually you impress me so much with your knowledge and a lot of the things I know now I owe to you and other people on the forum.
About both Kodak C-41 B&W films: extremely fine grain, wonderful enormous exposure latitude and they scan and print very well. It is pretty hard to see any difference between the T-400CN and the Portra 400 B&W in a finished print. Especially in MF cameras these films -provided they have been processed in a professional way- show qualities that almost equal LF prints.
Hi res scanning these films, enhancing -if needed- and printing on fine art papers provides you with images of true exhibition quality.
Ansel Adams -my all time favourite photographer!- would have been extremely jealous! It took him many years and three wonderful books on the technique of the zone system to reach standards so "easily" achieved today.
Leen
"Bert Bigelow" wrote in message
[…] I have tried printing at higher resolutions than 300 ppi, and if there is any significant difference, my old eyes can’t see it.
Well, you raise an interesting (but very different) point here.
Obviously, even if you could print an image at 4 billion dpi (to pick a random number that is absurdly large) with a full 24-bit color resolution, there may or may not be any real reason to do so. At some point, the human eye is incapable of distinguishing the difference.
However, I have not noticed anyone making that argument when it comes to printer resolution. The question I have been addressing in my posts is simply whether the printer is capable of reproducing the resolution of the image it’s being asked to print. With respect to my comments, the ability of the human eye to tell the difference is irrelevant.
That does not mean it’s an irrelevant aspect of the bigger picture though. For printed reproductions intended *only* for visual presentation, there probably IS a maximum resolution beyond which there’s no added benefit. It’s just that that’s a question of physiology, not of technical specifications of printers.
Pete
Leen, I really don’t know much…just where to find it. Or I try it out for myself. Keeps the old brain exercised (now if I could just do the same for the rest of me…)
Chuck
Emeril
Epson has this little quirk,it doesn’t like you to use any other printer
except epson, as for scanning directly into elements it wouldn’t even allow me to assosate elements as one of the application to scan directly to.
I have the same problem with an HP scanner but In Photoshop Elements I can go to File / Import / and there I find my scanner in the list (provided I have powered it up). Yours may be also.
PeteD
Leen,
I also admire Ansel Adams. For me, he is the greatest nature photographer of the twentieth century. I have studied his images, and dreamed of producing such works of art.
His pictures take my breath away.
Bert
With respect to my comments, the ability of the human eye to tell the difference is irrelevant
Peter,
I respectfully disagree with you. If an improvement in image quality is not discernable by the human eye, then it IS irrelevant. What is the purpose of producing/displaying/printing an image? It isn’t an abstract artifact to be admired for it’s mathematical purity. It is something MORE than that…something to stimulate an emotional response. If an improvement in image quality does not enhance that emotional response, then it is indeed irrelevant, in my humble opinion.
bert
EDIT: Having said all that, I must admit I am still interested in the new printers you mentioned. I’m going to start a new thread on the subject of printer technology, and invite all the "old pros" who have taught me so much on this forum. I want their response to your comments, Peter, so please join in. It should be interesting.
"Bert Bigelow" wrote in message
I respectfully disagree with you. If an improvement in image quality is not discernable by the human eye, then it IS irrelevant.
That may well be true for your purpose. However, it has no relevance in a purely technical discussion, and there are plenty of applications for which being able to print something smaller than the human eye can discern is useful.
Consider:
* Micro-print security features on checks
* Data compression along the lines of microfiche
* Transparencies meant for projection on a large screen
The number of such examples is limited only by one’s imagination. Just because a printer’s ability is not useful to you, that does not mean it’s not useful at all.
Beyond that, it’s simply incorrect to state that image resolutions above 300 dpi don’t produce higher quality prints, and it’s especially incorrect to state that image resolutions above 300 dpi produce *inferior* quality prints.
It’s fine if you want to qualify your statement to say that there’s no point in going above 300 dpi for the purpose of an art print, but to claim that the printer generally either doesn’t perform better with better data, or that it actually performs worse, is simply unfair to every reader that takes such a statement as fact.
Sorry if I sound harsh, but I really think it’s time for this misinformation to stop. I don’t like rumor and mythology in any area of my life, but it surely has no place in such a deterministic environment as digital imaging.
Pete
Peter,
Your point is well taken. There are certainly special applications where higher resolution in a print would be useful.
I was speaking from the point of view of an amateur/hobbyist which is the point of view that would be of interest to most of the readers of this forum. Most, but not all, and you are right to point out my error.
I would like to invite you to join the discussion in the new thread on this subject that I just started called "Printer Resolution…all you ‘old pros’ gather ’round."
Bert
EDIT: I must add one comment on your statement about higher resolution images (more pixels) always providing higher quality prints. That is true up to some point, but after that, without a complete understanding of the interpolation and dithering algorithms in the printer software it is not possible to say deterministically that a higher quality print will result.
but to claim that the printer generally either doesn’t perform better with better data, or that it actually performs worse, is simply unfair to every reader that takes such a statement as fact.
There is no evidence to suggest that the data is better simply because there is more of it, and Bert never claimed it to be worse data. More does not equal better data.
If you want to talk about dispelling myths, then clarify the point.
I really think it’s time for this misinformation to stop. I don’t like rumor and mythology in any area of my life, but it surely has no place in such a deterministic environment as digital imaging.
That was uncalled for – Bert posed a perfectly decent question.
"YrbkMgr" wrote in message
That was uncalled for – Bert posed a perfectly decent question.
My comments were not in response to a question.
Bert has claimed not only that 300 dpi is the maximum useful resolution for any image printed to a printer, but that presenting an image of higher resolution will actually result in a *worse* printout.
That’s not a question, it’s a statement of fact, and a false one at that.
Pete
I’m sorry to see the outcroppings of self-righteousness, even anger and denigration of "adversarial positions" expressed on this and the other thread on printers and resolutions.
This is super-interesting and useful subject matter, and I really appreciate the input of those of you who have a much higher technical knowledge of the subject matter than I. I think many of us of lesser understanding can benefit from the information on material we are struggling to learn.
As far as the Elements forum is concerned, the reason I come here is the combination of opportunity to learn, the congeniality and helpfulness of others, and in effect a social relationship which is very relaxed and enjoyable.
I hope we can continue to emphasize the congenial, yet keep on getting the help from people with an extremely broad range of knowledge and skill who really extend themselves to be helpful.
My 2 cents.
Al
Please move back to the other thread Bert started. My missive is now posted. I need somebody to yell at me.
