Print Problems / analysis needed

[GaryTorello]

Hi All!

I'm currently working with my better half (KarinaB1970) to resolve some severe printing issues (since I tend to give her advice and help on calibration and printing issues, etc.)

My experience is extensive in producing accurate color results for print work (web press, etc.) but not specifically with photo processing houses. Therefore, I REALLY could use an independent analysis of what we're being told by our printing source...

IN BRIEF: Karina is using an on-line hosting solution for her galleries, and them to print as well.. 3 print orders, ALL had problems. She then went to a local processor I recommended, and prints came out great! However, when she reported these results to the on-line processor, they had a different response which leaves us wondering if they could possibly be correct and that we have some extreme calibration problems. BOTTOM LINE is that we would like to furnish a file to be printed and specify NO adjustments by the printer, and get what we see.

Here's some samples (scanned) illustrating the problem:


http://www.qgm.com/temp/printproblems.jpg

Here's the RATHER LENGTHY (sorry) EMAIL EXCHANGE between Karina and her on-line processor:


First, (significant portions of) her COMPLAINT to them:





Karina Brach wrote:



>First order (# 17698): I placed this order for a just few prints, only one included a full range of colors, the remainder were sepia tones or muted colors. The color appeared somewhat saturated but it did not ruin the print so I thought nothing of it.

>Second order (# 18502): the majority of the prints were all heavily magenta saturated. Many of them not even usable (except for those files that did not have a significant magenta component to begin with. I assumed that it was a problem on my end with my monitor. I purchased a new monitor & calibrated it; file appearance on-screen did not change from the old monitor. Back to square one, as still I did not know what the cause was. [ref: seagull shot]

>I finally placed a third order with a wide variety of prints (color, b&w, sepia), along with reprints from the previous order, untouched and unchanged. (order # 20705): [ref: train wheel and axe factory]

>The images containing no color (black and white, and the sepia toned) came out within acceptable margins of error. However the color photographs are EXTREMELY saturated with color. Not only are the images incredibly saturated, one of the reprints from the previous order is totally different in color (no changes were made to this image)!

In EACH of my orders my instructions have been to make NO enhancements/adjustments to my images (with the exception of one image which looked terrific on screen, but printed horribly saturated with magenta).. print them AS you receive them.

>Totally frustrated I had all of the same images reprinted by a local processor. Success! They were printed exactly as they were captured and match my monitor perfectly.


THEN (significant portion of) THIER REPLY:



On 21 Dec 2004 at 22:01, ****** wrote:> So...why would your prints look great from another lab and have saturation values too high or magenta shifts from (del)? Can't say for sure without knowing how the lab is set up, but the most common thing is they run your photo through I2E, as 90% of labs do. I2E adjusts to memory colors like sky and also reduces saturation when the values are very high, as they are on the side of the old ax factory (I'm measuring 89% magenta, 81% yellow) . (del) has an I2E option and we've been considering offering it. It certainly makes grass green, sky blue, and and adjusts saturation and exposure pretty well. It's just that we know from experience it can also wreck some good prints and then we are left saying d'oh to our customers.

> We've actually fulfilled from 3 labs in the past but we've only been able to hit the level of consistency we run at now from (del) and (del), both of whom use Fuji paper, chemicals, and printers, and offer a true color option. We ended up going (del) because they offer more selection (26 paper sizes and 3 finishes).

> If you can point me to specific photos that are not printing correctly, I can investigate more.


PHEW.. end of the long emails...

Bottom lines is, does what they are telling us have ANY merit? Is it possible that the images Karina is sending off for print are IN FACT oversaturated, but APPEAR ok on screen, AND the online printer is the only one printing the TRUE result even though we can't see it?

ANY technical analysis of this would be appreciated. For reference/examination, larger images can be found here:


Seagull (http://galleries.smugmug.com/gallery/289368/1/11715314/Large) | Axe Factory (http://galleries.smugmug.com/gallery/287066/1/11418423/Large) | Trainwheel (http://galleries.smugmug.com/gallery/287976/1/11460795/Large)

THANKS IN ADVANCE! :)


Gary

[Scottes]

It seems to me that they are making adjustments to the images - increasing saturation, adjusting contrast, etc. They should not be doing this if you specifically told them not to, and you definitely should be telling them. If you want control of your images you should tell ANY lab not to make corrections or color adjustments. If there are any problems then it's your fault, and after the first batch you'll know how to adjust to make the pictures come out how you want.

Some though, appear to be double-profiled. The Magenta seagull looks like both you and the printer profiled the image.

For a new lab I like to print 10 or 20 test images in the first batch. They'll be saved as sRGB, since most photo printers seem to expect this. These will have varying degrees of saturation, contrast, and sharpness along with crop marks and varying white & black points. When I get these images back I'll know how to print to that lab. I'll also know if they're listening to me.

Most photo-printers (Fuji Frontier, Agfa dlab, etc) will print some numbers on the back of an image. These numbers indicate how the printer adjusted the image in regards to saturation, contrast, etc, and they're used if a customer wants reprints or less magenta, etc. Generally you will want to see all Zeroes if you told them not to make adjustments. (Of course this depends on the exact printer model.) If it's not a photo printer then you might be out of luck since they probably won't print anything on the back.


If you go to a pro lab and they know what they're doing then it's a whole different ball game. But I don't consider anyplace to be a pro digital lab unless they can give me very specific instructions on how to prepare the images. And they should provide a profile for their printer. Without those two items it's luck that will dictate the final results, so why bother paying the price of a pro lab?

[GaryTorello]

...saved as sRGB, since most photo printers seem to expect this. These will have varying degrees of saturation, contrast, and sharpness along with crop marks and varying white & black points. When I get these images back I'll know how to print to that lab. I'll also know if they're listening to me.

Most photo-printers (Fuji Frontier, Agfa dlab, etc) will print some numbers on the back of an image. These numbers indicate how the printer adjusted the image... First off, THANK YOU for weighing in on this! I've started to bang my head on things. :(


although I'm not sure WHERE this leaves us then; All of the images were saved using an sRGB profile. Is there some way to tell conclusively if the image was saved/printed with a particular profile?

I looked at the prints from the on-line processor, and unfortunately there's NO data on the back except a file reference number and date.

The prints from the local printer however DO have some numbers on them. Here's what's on the back of the train wheel image:

http://www.qgm.com/temp/back-train.jpg

It doesn't LOOK like zeroes :( Does this mean that even though instructed not to adjust, they did? What do these mean? Other images have similar (although different) sets of numbers.

Thanks so much!

Gary
___________________________
with a brain that's about to -=POP!=-

[Scottes]

The N's are "No Changes" basically. Same as 0's. That is **probably** Contrast, Red, Green, Blue - so it looks like they din't do anything. However, I don't like the "AS+01" at all, but I don't know *what* it is so I'm just paranoid. Find out the model of printer they use and do a little research on it and you may find a bit more.

But the N's and the fact that they look good tells me that they didn't do anything extra to your prints. This is a good thing. The other numbers could be anything, but look innocuous (though the "AS+01" still has me wondering.


Alas, there is no conclusive way to tell what profile was used. But the magenta cast is pretty conclusive to tell that it was double-profiled. So it might actually be your fault - if you profiled it with your printer and they didn't know then they probably would have profiled with their printer. Bingo - you get garbage.

But this seems doubtful since the local printer got it right.


My conclusion: go local, and pay more for useful stuff rather than paying less for stuff you're going to throw away. :-)

[F1_Fan]

But the N's and the fact that they look good tells me that they didn't do anything extra to your prints. This is a good thing. The other numbers could be anything, but look innocuous (though the "AS+01" still has me wondering.
http://www.drycreekphoto.com/Frontier/using_printer_profiles.htm

scroll down... AS+01 is Auto Sharpening. The other codes are explained as well.

The link seems to have a good method for dealing with profiles and labs.

[Scottes]

F1, are you sure those are Frontier codes? They don't look like what I remember - but that could be understandable....

[F1_Fan]

F1, are you sure those are Frontier codes? They don't look like what I remember - but that could be understandable....Ths sample posted above is most likely Noritsu (since it's Kodak paper... Fuji Frontier uses Fuji paper)

They are also a match for the Noritsu codes on a photo that I'm looking at right now. The Frontier codes on another print are different. This one infront of me is:

9.28 P0 0ANX0N NNN- 1 0253

I'm guessing the NNN- is similar to the Noritsu codes

[GaryTorello]

ok, PHEW, I think I've learned a TON

Based on all this It's apparent that:

The local lab did NOT color manage at all, but did sneak in a sharpness tweak (AS+1) - will have to talk to them about that next time.

I checked the embedded profiles (by simply reopening them in PS) to see if the were sRGB or not. They were. It's the only way I know to check this.. if the profile is NOT my working space profile (sRGB) I'd get a warning.

I also think some good advice in the drycreekphoto.com link is to CONVERT the image to the sRGB profile rather than embedding it to elliminate any problems with printers that choke on profiles. (image/mode/convert to profile...)

The online lab is *apparently* giving me some fast-talk, but I now have enough info to be dangerous and discuss it with them.

I think the best advice I've gotten here is to stay local, at least it's easier to discuss these issues as they arise.. Too bad though, the online service is a hosting service with the nice benefit of being able to assign your own prices either globally or individually and automate the purchase of reprints by customers - they do all the work, you get the profit. This was one of the deciding factors in going with smugmug.com -rats-

I'm still curious however about their reference to "I2E" - is this another name for color correction, or something in addition to it? I would *think* seeing no color corrections in the codes on the back mean exactly that and "I2E" is a red herring.

Thanks again!

Gary
______________________________________________
with 23,231,312 brain cells now alloted to printer profile info

[Scottes]

I2E: http://www.express-imaging.com/products/i2e.htm

Seems like a very smart way for a computer to automatically "enhance" an image. Great for P&S and disposables cameras and people who don't want to do all the work that you and I love to do in order to retain control.

Since both you and the I2E stuff were trying to correct the images you got junk. Too many cooks in the kitchen, basically.


The local lab did NOT color manage at all, but did sneak in a sharpness tweak (AS+1) - will have to talk to them about that next time. Bring the image and show them the back and ask them to make sure. Explain to them that *you* want to do everything to prepare the image, and that if they do something then it will mess things up. Don't complain about the extra sharpening - it may have been soft in their eyes and they might have truly been trying to do you a favor. But explain that you want to know the required sharpening so that they don't ever have to worry about it. Eventually they should understand that they don't have to do any work for your photos and that you're making their job easier.

I went through all this with my local printer, and actually taught them quite a bit. Now I just drop off disks and they run them and I'm happy with the results and they're very happy. They've got 2 of my images on the wall at 20x30 because they've never seen anything come out of their printer that looked that good. After a couple hundred 4x6s and 20+ large prints I know exactly what that printer is going to do, so I was able to produce a file that would look excellent to my eyes. While there's no doubt that they care about their output they just can't afford to put in the time and effort that I can. They print hundreds and hundreds of images every day, while I print a couple hundred a year, so I can afford the time to really tweak the image.