PDA

View Full Version : ImagePrint Experts Question...



jim kitchen
5-Sep-2006, 21:00
I am about to send a few of my latest image proofs to a local lab for processing. They have a new 9800, compared to a 9600, and the latest version of Imageprint, which reportedly allows the use of 16bit files. I am curious to know if someone could answer a question, regarding this software?

I researched the Yahoo Imageprint Group for an answer, but to no avail...

It is my understanding that an Epson 9800 is capable of receiving files with a native resolution of 720 pixels per inch, since this can be set within the Epson printer.

If I send a file to print that is 16bit and set to 720dpi, will the latest version of Imageprint knock my image file back to 360, before it is sent to the 9800?

Yesterday I saw a print window within Imageprint, that seems to regulate the dpi to the printer with two settings, which are 180 and 360. My local printer struggles with these questions, since they must ask Imageprint, and they do not have great success in that area.

Just curious...

Thank you in advance,

jim k

Kirk Gittings
5-Sep-2006, 22:12
Jim,

The 720 native resolution claim is argued by many people. The Imageprint manual is obtuse on this subject, but I believe Image print interpolates to 360 if that setting is used. I send everything to it at 360. IPs interpolation is excellent by the way.

The real question is what you can see in a print. I did the 360/720 input test thru the Epson Rip test and could not see the difference with a 16x loupe on mat paper. I can however see the real difference between 1440 4-pass, 1440 8-pass, and 2880 output.

Bruce Watson
6-Sep-2006, 04:21
It is my understanding that an Epson 9800 is capable of receiving files with a native resolution of 720 pixels per inch, since this can be set within the Epson printer.

If I send a file to print that is 16bit and set to 720dpi, will the latest version of Imageprint knock my image file back to 360, before it is sent to the 9800?
The 720/360 ppi question is really one for the Epson driver, not the printer itself. The first thing the Epson driver does is resize the image to 360 ppi, and drop 16 bit images down to 8 bit. This makes the math easier for creating a dither pattern. The results are quite nice; it's not a bad way to make a driver.

My understanding is that Imageprint drives the printer directly, bypassing the Epson driver. If it's anything like StudioPrint, which I use, it will use all of image information you send it and create its dither pattern without resizing the image data. IOW, ImagePrint is not sensitive to the file's output resolution, and doesn't need or want a specific output resolution in the same way that the Epson driver does.

So, send it all the information you have. Don't upsize or downsize your file. Let the RIP do its job.

jim kitchen
6-Sep-2006, 06:56
Thank you gentlemen for your feedback...

I just supplied the print house with three files, all set to 16bit, where the files are set to 360, 720 and 1440dpi. I will have the 16X20 proofs this afternoon. According to the print house, their latest Imageprint software handles 16bit files, but they did not know if Imageprint simply lowers the file to 8bit automatically.

That said, I will see if there is a difference under a loupe later today. I suspect that the files will be equalized, and that there will not be a difference at the 1440 4-pass setting.

The next test will be the 8-pass setting.

Thank you again,

jim k

Kirk Gittings
6-Sep-2006, 07:18
With 4 pass there is oftentimes subtle banding in smooth areas of midtone. I don't find it good for anything but proofing. I find 8 pass a necessity.

Bruce Watson
6-Sep-2006, 07:38
I just supplied the print house with three files, all set to 16bit, where the files are set to 360, 720 and 1440dpi.
In re-reading this thread, it seems somewhat confusing to me. Probably has something to do with my caffeine-deprived state when I wrote my first post ;-) Just in case, I'm making this post just to be sure that we are all on the same page. If it's not necessary, please forgive my intrusion.

I think it's possible that we might be confusing the output resolution of the file (in ppi) with the print resolution from the printer (in dpi). The two are completely separate things.

The file's output resolution is typically set in an image editor like Photoshop. In Photoshop, it's set in the Image...Image Size dialog box. This is the actual density of pixels, and is expressed in ppi.

The print resolution is typically set by the printer driver (or RIP in this case). This is the actual density of ink droplets used to form the image on the paper, and is expressed in dpi.

If you send a 360 ppi file to the printer, you can have the printer print it at a range of settings - for example, from 360 dpi to 2880 dpi. What the printer/driver is doing is deciding how many ink dots to use, and where to place them, to create your pixels (this is the "dither pattern"). In the case of 2880 dpi printing, the printer is using 8 ink dots to make each pixel (2880/360 = 8). Note that these ink dots can be, and usually are, different colors.

For fine art printing, most people set the printer for 1440 dpi or higher. I think you'll find prints run at 360 dpi and 720 dpi less than optimal for fine art purposes.

Now, off in search of even more caffeine! I have a feeling I'm going to need it...

jim kitchen
6-Sep-2006, 10:30
Yes, my mistake...

I sent files to the printer with 360, 720, and 1440 ppi.

You did not require a coffee to see that typing error of mine.


jim k

jim kitchen
6-Sep-2006, 19:55
Well I received the proofs...

The three images all look the same under a 6X lupe. I can not tell if one image is better than the other. Tomorrow I will see if there is a difference with the 8-pass setting, as Kirk suggests, and the 4-pass images printed today.

jim k

Kirk Gittings
6-Sep-2006, 20:18
Jim, Which suggests that ImagePrint interpolates everything to 360 or the difference is too fine for us to see. I suspect the former. I suspect if you had sent a 180 file to the printer you would see a difference because it it would be upresing rather than downresing. I don't remember. Are these on mat paper? What paper? I think you will find the difference on toothy papers is subtle but there. On smooth papers it is more pronounced.

jim kitchen
6-Sep-2006, 22:29
Kirk,

I think you might be correct with the 360 number, and hopefully Imageprint will confirm this, but for the moment I do not think they will. As I mentioned earlier, and since I only had a glimpse at the screen, there seems to be only two settings 180 and, or 360 in Imageprint's printer dialog window. Imageprint will probably think I can't read...

At the moment, I am experimenting with Silver Rag, and it looks quite nice actually, but I have Phil Bard of Cirrus Digital Printing do all of my major printing with carbon inks. I always send him files that are 360ppi, and he does not use Imageprint as a RIP. His work is rather outstanding.

His site is here: http://www.cirrus-digital.com/home.html

K3 inks and Imageprint seem to do a brilliant job on Silver Rag, and the finished product is quite different compared to Hahnemuhle 308 and Cone inks. I just do not put them in the same room, since I want to compare forever. I love my carbon prints and they sure do hold their own in a gallery, but it sure is nice to see a broader range of tones, and deeper blacks with the K3's, once in a while. Unfortunately, this paper and ink combination still produces minute metamerism, even with Imageprint as a RIP.

My local print house can not produce carbon prints.

That said, I will finish my testing tomorrow afternoon, and see what happens.

Thank you again,

jim k

jim kitchen
7-Sep-2006, 15:05
Today I received a print on Silver Rag, printed at 1440 8-pass. Under the 6X lupe, I could not see the difference between this print and the 4-pass print.

The 2880 print will not be ready till tomorrow. So, I wait again. My expectations, at seeing a major difference between the two prints, are probably too high. Then again, I wonder if they even changed the setting to 8-pass?

I will try one more time...

jim k

Kirk Gittings
7-Sep-2006, 19:55
That's great. Maybe the 4 pass banding issue is not a problem with the 800 series printers (I use a 4000). Or maybe, it is sometimes simply the different papers. It will make your lab happy as 4 pass prints much faster. The 2880 setting has a different "look" to it as it does not use variable droplets but a single size very small. It gives a kind of creamy look to the midtones, which I like for some images.

jim kitchen
8-Sep-2006, 17:15
Dear Kirk,

Today I received another proof, and this one set at 2880dpi single pass. For the life of me I can not see any difference between it and 1440 4-pass, under a 6X loupe. I am surprised, but then again I might have a really crappy loupe, or a really bad eye.

These incremental settings must be a placebos...

jim k

Bruce Watson
8-Sep-2006, 19:27
Today I received another proof, and this one set at 2880dpi single pass. For the life of me I can not see any difference between it and 1440 4-pass, under a 6X loupe. I am surprised, but then again I might have a really crappy loupe, or a really bad eye.

These incremental settings must be a placebos...

So you sent a 360 ppi file out, and got back prints made at 720dpi, 1440 dpi (one 4 pass, one 8 pass), and one at 2880dpi, all made from this same 360 ppi file, and you can't tell them apart? Normal viewing distance or under 6x loupe?

Very interesting. My inclination is to say that somebody must be lying to somebody, but I don't have nearly enough information for that. Could be that whoever is doing the printing is playing with you. It could be that ImagePrint's dither patterns are magically good to the point were there's no noticable difference. Could be a lot of things.

I can say that with my 7600 using PiezoTone inks and the StudioPrint RIP for B&W output, that 1440 blows 720 out of the water, and that 2880 is sufficiently smoother than 1440 that I always print at 2880 eventhough it takes considerably longer to print. And I can also see the differences with a loupe, although my print loupe is 10x.

Interesting indeed.

Ted Harris
8-Sep-2006, 20:00
I'll add that, using the Colorburst RIP (the full version not the baddly crippled version supplied in Epson's pro packages) on a 4800 I see significant differences in 720, 1440 and 2880 settings.

Another possibility is that the machine operator really is not expert. You would be amazed at the number of folks working in "Pro Labs" including some of the best, are not experts at their craft. You might want to sit down with your local guy while he is setting up the print and talk through what he is doing ... you may both learn.

jim kitchen
9-Sep-2006, 18:32
Gentlemen,

You are probably correct with your assumption that the operator might not know what the software does at the moment. Your software is different, and your support is probably a bit more excellent, compared to what I experience at the moment. Imageprint, for some unknown reason, does not divulge information, and I am at a loss as to why the images look so identical, under a 6X loupe. The operator can not answer my questions, as I mentioned earler, and the printhouse finds little comfort with Imageprint returning phone calls.

Cruising the Imageprint newsgroup, indicates that everyone has an opinion on how the software works, but I can not find a definitive answer from Imageprint themselves. I believe the different settings should give some incremental changes, but alas, these images look identical.

It seems from the Imageprint newsgroup, a few other people share my experience...

That said, I envy the fact that you can see a difference. I will sit down with the print house operator on Monday, and give this issue one last try.

jim k

jim kitchen
11-Sep-2006, 23:21
Dear Group,

I received my final test prints today from the print house, where I sat and watched them work the software and the printer...

We printed three prints from the same file at 1440 4-pass, 1440 8-pass, and 2880, where all were single passes and not the high speed bidirectional passes. The images look identical. It should be noted that I used Crane's Silver Rag with the Imageprint grey curves seleceted for each setting, where there is a single curve for the 1440 prints and a curve for the 2880.

That said, we all reviewed the images, and we could not see any discernable difference between the images. They all looked great, but not different. This issue might be specific to the curve for this paper, but for the moment I am tired of testing, and I will send a few of my images off to Phil in Portland to print a carbon ink image or two for the gallery.

Anyway, I am done with my test, and I will go forward with this paper at the lower dpi setting, until the paper supply is exhausted.

Thank you for your replies...

jim k

Brian K
12-Sep-2006, 04:45
I used the test version of Imageprint, having recently been badly burned by StudioPrint. I found that using 4 pass gave me noticeable banding, as did 2880 however the banding was nearly invisible (you'd have to seriously look for it) with 8-pass.

The B&W prints with Image Print were neutral right out of the box, dead neutral no metamerism. The print I've been getting with StudioPrint were olive in color with noticeable metamerism. This using a gretag produced linearization and StudioPrint profiles. The biggest problem I had with StudioPrint was that when I had a recent deadline, due the tuesday after Labor day, StudioPrint had bugs and refused to print. I kept getting error messages and images would not open in the program. Needless to say this happened on the Saturday of the 3 day weekend and I had to rely on the native Epson drivers to finish my project. To be fair the native Epson drivers were about as good as StudioPrint. So for me StudioPrint is a waste of $895. plus I bought a PC, (I'm a mac) just to use it.

I'm still not thrilled with the results(banding) I got with ImagePrint, although it is clearly better than StudioPrint, but I am hesitant to throw another $800 away on these boutique softwares that have little reliability and little customer support. They all seem more like a work in progress than a finished product.

jim kitchen
12-Sep-2006, 21:29
Dear Brian, et al...

Some notes for the archives.

My troubles and frustration with Imageprint, stem from the fact that I am not in control of my printing, other than setting up a file to print within Photoshop, and believing what I see on my calibrated monitor would be printed properly. Imageprint is a black box, and for the moment, I consider it a Black Hole. I am a part-time Apple Developer, and I can empathize with Imageprint to a certain degree, regarding their absolute silent position with their proprietary software. I also agree with their stance about paying for information, since that is part of their revenue stream, but not quite to the degree they enforce this issue. They are notorious for not returning phone calls, and the lack of excellent knowledge among the users astounds me. This lack of knowledge, which is purely speculation on my part, could simply be a global position taken by many users that own the software, that do not wish to divulge any information, since they paid for it. I might do the same if I owned this software, but I am for the moment, a user at a print house that owns and distributes the software.

That is why I call it a Black Hole...

That said, I must admit though, this software and using Silver Rag looks quite astonishing, where the image is absolutely beautifully neutral with only a slight bit of metamersim, even if I don't know how it was printed and I can not see the difference between the image settings. Go figure...

As a side note, one kind soul contacted me directly recently that seems to know Imageprint quite well. He offered some intuitive advice, and for some unknown reason he decide to pass it on to me. He states the following, and again this is a quote from this gentleman and not Imageprint. This gentleman emphatically suggests that I should understand the following:

1. Choose the correct grey profile for the printer output, such as 1440 for 1440, and 2880 for 2880. Selecting the correct profile will automatically select the desired printer dpi;

2. Do not use the bi-directional printing option for high quality images, since it might compromise image quality;

3. Set your files to 360ppi only, with all of your attributes set, such as sharpening, etc. This effectively turns off the interpolation engine within Imageprint;

4. Do not let Imageprint upsize or downsize your images, since it is a black box, where you will lose all control of your attributes set in the file;

5. Imageprint has a radically different screening algorithm, and you must be aware when downsizing or upsizing your image. Users of heavy sharpening attributes might find that Imageprint makes it too sharp, and your image will appear lighter than expected;

6. The 2880 setting is there because too many people asked for it. The 2880 setting supposedly produces single fine dots. Whether it works is another matter;

7. Some people believe that 2880 produces better prints, but that is hard to decern on certain images.

8. Use the 2880 setting for small prints only, otherwise use the 1440 4-pass;

9. The 1440 versus 2880 discussion is image specific, where the results vary just as the image does;

10. The 1440 8-pass setting is for use with specific papers that require additional ink drying time. It seems that the 8-pass supposedly offers a higher level of micro weaving as the ink is laid down, where the ink has a chance to dry properly before the next layer is laid down. This setting has no incremental visual imapct with the print (Such an interesting comment...);

11. The 1440 4-pass setting is the preferred setting;

12. The issue with 16bit files was not spoken too, since he could not answer the issue as to whether Imageprint dropped the image back to 8bit internally or not, since this feature could be just a convenience for the larger print houses, that regularily receive 16bit files;

13. A Black Hole is God's answer to division by Zero.

I hope this information helps someone down the road, although these items are not the Gospel, according to Imageprint.

jim k

Brian K
13-Sep-2006, 04:16
Jim thanks for the detailed response. As I have read in other posts, there are banding problems when using 1440 4 pass, I have had this problem. The only correction for it has been to work in 1440 8 pass. To be honest and as the prints I sell are silver and the inkjet prints are only for gallery catalogs and presskits, I may just use the native Photoshop/Epson print driver. I do so because all of these RIP software programs seem to be works in progress, are all black holes and I'm tired of spending $800 a pop for products this unfinished.

Good luck with your efforts.

Ted Harris
13-Sep-2006, 06:40
Jim, Brian, Bruce ..... and anyone else that wants to chime in. Kirk and I have talked about doing an article comparing the positive and negative aspects of several of the major RIPs, including but not necessarily limited to QTR, ImagePrint, ColorBurst, PowerX and StudioPrint. I have gathered a lot of the preliminary information and talked with several of the RIP companies but we are still in early stages.

We will concentrate on using RIPs with the wide formt Epson printers with some mention of the 2400 and the new Canon printers (if ny RIPs are available for them when we are ready to go to press) and will discuss RIPs v. native drivers. Test prints will be produced on several diferent papers. Sine I am still in the planning and organizing stages there is not much more to say right now but this thread has underscored for me the usefulness of this article.

Thanks in advance for any and all comments. Please email me at tedharris@mac.com

Brian K
13-Sep-2006, 06:42
Ted, I look forward to your results. I think what you're doing will be beneficial to many.

tim atherton
13-Sep-2006, 09:56
Jim, Brian, Bruce ..... and anyone else that wants to chime in. Kirk and I have talked about doing an article comparing the positive and negative aspects of several of the major RIPs, including but not necessarily limited to QTR, ImagePrint, ColorBurst, PowerX and StudioPrint. I have gathered a lot of the preliminary information and talked with several of the RIP companies but we are still in early stages.



What's that one for greyscale that Helen Bach on here uses? Is that on the list

tim atherton
13-Sep-2006, 10:02
What's that one for greyscale that Helen Bach on here uses? Is that on the list


okay - found it IJC/OPM from Bowhaus (always liked the look of it, never tried it....)

Ted Harris
13-Sep-2006, 20:34
Tim, thanks for the lead. I contacted them and will test it.