test print small - final print big?

[digitalshiver]

Does anyone have experience with printing small (8x10) until the image is right - or in this case, a series of images - and then printing bigger for a final version.

I'd like to print on 8x10 paper, and once I've edited, masked, color & tone corrected, then print on 13 x 19 paper.

Will the tonal relationships hold up? Will the larger print be brighter or darker? Will I inherit a slew of headaches? Should I just stick with my final output size as I make editing decisions?

I understand that ALL sharpening should be done at final output size and resolution. That's no problem. I'm wondering about achieving repeatable tonal precision in my calibrated and soft-proofed work flow.

I would work on the print at my final (large print) resolution and size, then I'd flatten a copy and drop it down in size before sending to RIP. (ImagePrint  RIP & Epson 2200 printer)

Thanks in advance for any experience that you may have.

Charles

[David Good]

Charles,

I would only work on the image once (in this case 13x19) and reduce the image size to test on smaller paper. The two different sized images, 8x10 and 13x19 would contain different information depending on cropping so that may make a difference if that approach is taken.

Dave

[digitalshiver]

Thanks Dave. I may have been unclear. I will work on a single image - an image that anticipates being printed 13 x 19. Once all edits are made, I'd save a flattened copy, and reduce the size of that copy so that it fits entirely on 8x10 paper. Then I'd print the entire image 8x10. If it looks good, I'd move on to another image.

Later, I would return to all these images and print them big, simply by skipping the size reduction step.

I intend to test this, but I'm about to make a paper buying decision, and I'm considering buying lots of 8x10 paper for a project that will ultimately wind up on 13 x 19.

The question is do tonal values shift or fall apart at different sizes? Can I "proof" my images on 8 x 10 paper, even though they will eventually be printed larger?

charles

[Ken]

I always make a "test strip" from the final size image. When you look at anything from a distance (smaller) tonalities and details compress. As you get closer to it (larger) you discover a lot more visual information. Another analogy would be a contact sheet... even magnification doesn't reveal what you see in an enlargement.

Try this: Size your image to 8X10. Just like making a good darkroom test strip, find a section of the image that has a variety of details, colors, contrasts and tonalities. Create a new blank 8X10 image with the same ppi and profile as your print. Call it your "Test Sheet." Copy a section (6x4 or so) from your 8X10 image and paste it onto the Test Sheet. Then go back to your image and enlarge it to the final print size. Look for the same test area as before, and select a 6x4 piece of it. (You will be able to use only a portion of the same area, of course.) Copy that and paste it onto your Test Sheet. Print the Test Sheet. If you judge the difference between the two is or isn't worth the trouble... there's your workflow.

[seangirard]

I have to wonder in this case if the difference in size between your proofs and your final prints really makes it worth all the effort. I guess it depends how many rounds of proofing you intend to do.

[Bruce Watson]

Does anyone have experience with printing small (8x10) until the image is right - or in this case, a series of images - and then printing bigger for a final version.

I'd like to print on 8x10 paper, and once I've edited, masked, color & tone corrected, then print on 13 x 19 paper.

Will the tonal relationships hold up? Will the larger print be brighter or darker? Will I inherit a slew of headaches? Should I just stick with my final output size as I make editing decisions?
[a href=\"index.php?act=findpost&pid=226339\"][{POST_SNAPBACK}][/a]

Yes, I do this all the time. Just finishing one up as we speak. I use a sheet of 13x19 paper and make two or three proofs on it (how many depends on the aspect ratio of the image in question). Then the final print at whatever size fits on the 13x19 paper which again depends on the aspect ratio of the image in question.

The final print will measure the same as the smaller proofs with a spectrophotometer. That is, the colors and tonal relationships hold up. However, your perception isn't a spectro. The final print may "feel" somewhat lighter or darker depending on the image. Often the more the enlargement the more the perceived difference. I seldom see the colors or relationships between colors change, but often the overall tone does, especially with B&W prints.

This is why many (most?) professional print makers will insist or at least strongly advise that your final proof be at full size. Because size does in fact make a difference.

You'll have a tougher time perhaps if you are proofing on a different paper. Then you can easily see color and tonal relationship changes. Sadly ICC profiles aren't perfect.

[digitalshiver]

Thank you everyone for your feedback. Good to have a number of opinions to sample. I will do some testing and I'll be happy to share my experiences here!

Meanwhile, if others have relevant experience, I'd love to hear it.

Thanks again everyone.

Charles

[thompsonkirk]

Larger prints always look a bit less contrasty & less saturated than small ones.  

You have to adjust your expectations to this tendency - rather like in the darkroom, when you had to allow for the print to dry down.

Kirk

[Geoff Wittig]

There's a big shift in color perception & tonality with scale; the same image, even the exact same file, will look quite different at 8x10" compared to (say) 24x30". Therefore if you're going to proof an image on a smaller piece of paper, you really need to proof a small section of the image at the intended final size to get an accurate sense of its appearance.

I actually got some insight into this from my childhood hobby of building and painting plastic models. There is a surprisingly large shift in tonal perception when you change the scale of an object; when you go larger, you see deeper into shadows and the perceived tonality is lighter. When you scale down and go smaller, the perceived tonality looks darker, even if a spectrophotometer measures exactly the same numbers.

This is not something you can correct mathematically or automatically; you need to eyeball it and correct it perceptually. There is a brief but good discussion of this on John Paul Caponigro's website, under the 'proofing' section of free PDF's.

[digitalshiver]

Larger prints always look a bit less contrasty & less saturated than small ones. 

You have to adjust your expectations to this tendency - rather like in the darkroom, when you had to allow for the print to dry down.

[a href=\"index.php?act=findpost&pid=226423\"][{POST_SNAPBACK}][/a]

Can I compensate for this with saturation and contrast control?

[ScottWald]

My experience has been that you need more gamma for smaller prints and less for larger ones.  I keep all my files at 16 bits so that I can make the requisite curves adjustments just before I print without any worry of posterization.  I haven't noticed a difference in perceived saturation.

But, really, if you're going to show the images to people who know images well, you'll always have to make your final test print on the final paper at full size, place it under glass in a room with lighting very similar to what will be in the exhibition space (i.e., similar brightness, same type of light source, etc.) and then look at it again and again over the next few days.  I've found that only then will you consistently get things right.  For me, there are few things worse than having a print on the wall for three months at a major gallery that makes you cringe whenever you see it.

Scott

Can I compensate for this with saturation and contrast control?
[a href=\"index.php?act=findpost&pid=226433\"][{POST_SNAPBACK}][/a]

[Ernst Dinkla]

Thanks in advance for any experience that you may have.

Charles
[{POST_SNAPBACK}][/a]

There are differences but the sizes you quote will show them less in my opinion.

It's quite complex in practice when you go from a good large print to a small one with the same data. Aliasing can play a role then too, contrast increases. In reproducing art you have to print samples at the same size to get the colors fixed, editing the image data based on a print at a quarter of the original size doesn't work. I'm not so sure whether saturation decreases from small to larger prints, I have seen it going the other way or be good for every size, depends on the color texture of the art in my opinion. Must be related to how downsampling, ink mixing on the print copes with the color texture and how the eye interprets the result of that. The same for contrast and brightness, lots of small detail in the image that benefits of enlargement works different between small and large than hard edge compositions. There's a difference between film grain (aliasing) and no film grain in the image data so rules that existed for analogue prints are no longer usable for all (semi)digital work.

It may be a good idea to print a small 1/4 one with the data of a big perfect print and then say what has been changed between the two. It is still difficult to display them both in an equal way to the eye then. Viewer distance, white margin, surrounding space, illumination play a role next to size. The quarter print superimposed on a large one and on display next to a large one ?  Several images with different contents to check. Nice project for RIT research I think. I doubt it will ever deliver an algorithm that can be put in Qimage to cope with print size differences.

This issue has an implication for soft proofing too. I always had reservations on soft proofing. The additive RGB display with transmitted light isn't a subtractive CMYK++++ print in reflected light and all the profile + CM compensations for the softproof do not overcome that.  The size difference adds to that.

It's a bit woolly what I write here but that's also how I face it and cope with it. Every time the approach to fix it is different but no proof equals print proofs at 1:1 and a final proof at that size.


Ernst Dinkla

Try: [a href=\"http://groups.yahoo.com/group/Wide_Inkjet_Printers/]http://groups.yahoo.com/group/Wide_Inkjet_Printers/[/url]

[hsmeets]

Hi folks,

Then I must belong to the group of ignorant colourblind b*st*rds that softproof and directly print the final image and think that they have a marvellous print in their hands.

:-)


When I left the wet darkroom earlier this year and stepped into the light I was sooooo pleased that I could see on screen what dodge/burn would do to the print.

Yes, there is a difference between screen and print, but I quickly learned to factor that into what I see on screen versus how the print will look&feel.

Am I ill, must I 'go see the doctor'?

Cheers,

Huib
www.huibsmeets.com

[digitalshiver]

Thank you for your thoughts, Geoff & Scott & Ernst, and everyone else.
I think I am likely to continue proofing at my final size, but I will test the use of smaller paper as well.

Thanks again, all!

charles

[jule]

Some more thoughts and experience on this one I found on John Paul Caponigro's site. http://www.johnpaulcaponigro.com/downloads...f_for_scale.pdf

Julie

[langier]

My master prints are genrally set to go for my 9800 at 19x13 on a 24x16 sheet. Many times, I won't need a print quite this large, so I simply shrink it down using Bicubic Sharper in PS or simple reduce to fit in the printer dialog. It seems to work just fine comparing the the smaller prints to larger when I get to them. Some images I master larger, but that is fairly rare. In any case, shrinking down either way the prints are identical for all practical purposes.

I tend to work once and repurpose smaller with my master images, though I've found quick and easy workarounds if I screw up and master a smaller image that I should have started larger. In most cases, I start at the native resolution of the original capture and work it to near perfection. Work larger then print smaller, if you want to save paper or only need a smaller image for your project. Going the other way you will generally loose quality.

When I'm proofing a bunch of my work or master images, I'll run Contact Sheet in the Automate section of PS and run a gang sheet of 3x5 proofs. I can look at a glance at quite a few images and be quite certain that what I see there in color and tonality will be the same in the final and larger images.

Last year, I sold 140-lot group of 20x30, 24x36 and 16x48 and larger panos by simply taking the master files and running Contact Sheet to make 36-up, 24x30 gang sheets. When I compared the 3x5 inch prints with the larger and finished big prints, they had substantially the same color and tonality and the client and gallery were both pleased. This year, it was a bunch of 3x5 foot canvas wraps for the same gallery and another client. We ran small samples on the canvas and sold the job.

The gang sheets allowed both the gallery and client to see an actual print on the same paper. No muss or fuss and fairly cheap and easy to pull off. The final prints once matted, framed and hanged are simply stunning and look even better than the gang sheets. The gang sheets set the stage and sell themselves. Couple a sheet with a large framed and matted print from your selection and the client will be blown away.