Mike Chaney's Tech Corner

This is my first post here. I have been using Qimage some time now.Basicly I am happy with the product. However a couple of times I have torn my grey hairs. My knowledge of Qimage is still far from good, so some of the problems have been caused by this.

Here is one problem which I encountered the other day and I have not found a solution to it. The problem will probably pop up for me some day again and maybe somebody else has had too.

The other day I printed some photos with my Epson 3800. I planned to print four images to A4 sheet and then use my trimmer paper cutter to cut one sheet to four photos.

Because I had quite many photos I planned to make it easy. So I planned to use the paper cutter only three times per A4 sheet (one cut to halve the A4 and then same again with the two halves) ending with 4 photos from one A4.

I like to have white borders in my photos. I knew that Qimage has option to include white borders "automatically". This is where the troubles came. No matter what I tried I did not get even borders to my photos, so I ended up using the paper cutter a lot.

Afterwards I examined the task more closely, and I found out that it was not easy at all. There are a lot of variables involved (basically in two pieces of software: Qimage and the Epson driver).

Here is what I wanted to do:

(http://www.kivikukkula.com/misc/target2.gif)

The size of A4 is 297x210 mm. That makes 148.5x105 mm per image. Using 10 mm border in each side leaves 128,5x85 mm (ratio 1.511765) for the image itself. The originals were 1.5 in ratio, so I had to crop them a bit to get ratio 1.511765.
I used Qimage Professional Edition (version 2009.268) with :
- Custom size 2 col x 2 rows
- Auto Cropping: Off
- Print Orientation: Unlock
- Borders: B
- Border 1: 5mm
- Border 2: 5mm
In the Epson driver I had:
- Source: Sheet
- Borderless: Yes
- Size: A4
- Page Layout: Layout: Output Paper: Same as Paper Size

This is the best I could produce. The borders are not the same in each side, however.

(http://www.kivikukkula.com/misc/qimage10.gif)

The borders shown in the output from Epson Print Preview are exactly the same as in the actual print (the colours are not).

Here I used different colours for the borders to show the problem. In the final print the border colors are all white.

As you can see the borders are not the same:
1) In the top there is a small white border (0.5mm).
2) In the center there is a white border (1.5mm).
3) The bottom border is cropped (1mm).
4) In the left there is very small whiter border (less than 0.5 mm, hardly visible in the screen)

There is an option in Qimage I also tried. Page Formatting -> Borderless Overspray/Expansion -> Disable. This seemed to cancel the borderless option in the Epson driver and result was worse than above.

It is possible to change to the Amount Enlargement for the borderless print (Main -> Paper Settings -> Expansion). Changing the Expansion from Max (factory setting) to Mid or Min, gave also worse result in my case.

So that leaves the question: Is it possible to have exactly the same borders in 4 images in one sheet?

(I would not like to start calculating and including the borders in actual images itself).

The day Qimage gives correct feedback on mm sizes you can create equal borders. As long as it rounds off actual sizes from say 1,4 to 1 mm in some menus you will get issues like you reported. It is meanwhile a five year old request: let all feedback be precise to 0,1 mm. We need that in the metric world.


met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

I don't think the problem is anything to do with that, it's related to the fact that Oldfox is using borderless and the expansion affects the sizes.
He's also got cropping off so the images will not fill a quadrant (4 images per page), hence the extra white space.

Sorry Oldfox, I've not time at present to look in detail and simulate here but will try later on today. There may be some clues in what I said above but using borderless is always dodgy for accurate sizing, even with expansion disabled in Qimage.
A screen shot from Qimage would be helpful too, right hand side of main screen and the full page editor.
Terry

Hi Oldfox,
I've now checked.
I was a little puzzled that you first mentioned 10mm and then 1.5mm B1 and 2.5mm B2 borders and the screen shot appeared to show something like 5mm for each ???
However I simulated with  each at 5mm on an R800.
First:
There was no need to do this with the image, just use print cropping (scissors icon) and it's done for you precisely.

I set up A4 borderless with Custom 2x2 = 4 per page and border as above. Expansion was disabled in Qimage.
LIke you, from print preview the borders looked unequal with some white spaces in strange places, ie. at the top in my case.
See attached screen shot below.
I then did the same but without borderless set, ie. normal page margins.
See attached below, although the image black outline does not show so well, the borders are equal and the only white space are the page margins as expected.
Conclusion: the borderless setting creates some distortion so for accurate dimensions, don't use it!  :o
Terry

B1 = B2 = 5mm as you figured. (there should be colons in my text, maybe they show up different in your screen).

Anyway, thx for quick answer(s), I'll have to study them more later this week.

/old Fox

ps. the same post has been in Steve's Digicams' Qimage Support Forum two weeks now, no replies, 184 views (now)...

You are right, I did not look carefully enough on my hi res screen ::)

That web site is dead duck now as far as Qimage is concerned it's all here now  ;D
Terry

Hi

I think you are asking a lot of the 3800 to expect millimetre perfect paper feed. It's not the best printer I've used in this respect, although in most other ways it's fantastic.
Given that you are using white borders I don't think a mm would be noticed, and you also have the problem of trimming to a 1mm tolerance.

Brian

Brian, do you use the metric system or "imperial"?


met vriendelijke groeten, Ernst Dinkla

spectral plots of +100 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

Ernst

I can work in either without too much trouble. I suppose I think in inches but measure in millimetres. In the 'old' days when we were totally Imperial we didn't use tenths of inches, just 1/8ths, 1/16ths, 1/64ths, etc.

Brian

Brian,

It is not the computing metric<>imperial I'm thinking off, I have to do it often too. I ask this because a 1mm difference on a 10mm border is quite visible. That's the kind of borders the thread started with. And cutting that kind of borders with a 1mm error would show in my opinion. Cutting with 0.25mm precision is doable. One way or another I get the impression that the 1mm looks smaller in the imperial eye than in the metric eye. It is less than 1/25th of an inch so to speak.


met vriendelijke groeten, Ernst Dinkla

spectral plots of +100 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

Ernst et al,
Here are some real facts.
I printed the white images with different set borders shown attached below, the numbers are the set mm border sizes in QU.
I measured each one with a hand held measuring microscope  and, (within the limitations of a hand held device resting on the print), each ones measurement matched the number on the print.
I can even see the differences on the print - that comes from practice at looking at small differences in the engineering industry.
Conclusion, Qimage in metric mode is VERY precise, is at least 0.10 mm, 2 hairs' width, good enough?
End of story
Terry
(edited)

Terry,

The thread took a different turn with Brian's: "Given that you are using white borders I don't think a mm would be noticed, and you also have the problem of trimming to a 1mm tolerance". That was what I commented on.

You are entirely right on Qimage's precision in metrics and inches and I did not argue that fact, I know it is like that. My first comment in this thread wasn't correct, I hadn't noticed that Oldfox used a borderless print setting in both trials, I thought he did it only in the second attempt and used mm fractions in the first attempt.

Remains the fact that in several menus when you use an x,5mm choice or even more precise the backfeed in the menu isn't showing the actual size in mm's. It rounds it off to whole mm's. Which can lead to related issues as I notice in practice.


met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

Hi,

You are right about the distortion with borderless. Thinking the way borderless is achieved, this is obvious.

I tried your suggestions (no borderless, print cropping on) and got the same result as you. However, there are the Epson physical margins which should be cut away.

I found out that there are also Page Margins within Qimage. So I changed the right margin to -8m giving total right margin -4.966mm. The result was promising (attach 1, qimage11.gif).
Next I changed all four Page Margins to -8mm. The preview in Qimage looked still promising. See the white lines indicating the boundary of "image" to be printed (attach 2, qimage12.gif).
However the Epson Print Preview looks nearly the same as with default Page Margins. (attach 3, qimage13.gif).

Even more puzzled

/old Fox

I suppose my problem should be possible to solve with Qimage. Qimage "knows" the Epson physical margins. All it has do is to include them when calculating the final output.

Last attachment here:

Hi Oldfox,
These are the same as the driver produces; Qimage is just reporting what they are, not a separate or additional value - until you edit in Qimage.
Just to say, Page Margins are not the same as Image Borders!
I don't think you can use negative values when using a normal page set-up, it only works on borderless. That's why the print preview looks the same, It tried it once.
Yes, it does that already.
Yes, you'll have to live with that I'm afraid.
Your Qimage 11 attachment, did that have negative margins? I can't see why some border should be missing on the rhs. This should all be very straightforward
I wonder if Brian could try to simulate this set-up on his 3800 and see what the preview looks like; I can't see why your preview should not look like mine on an R800 - reply #3, second attachment, above.
Terry

Hi Ernst,
Yes, but others had said otherwise elsewhere, perhaps I should have posted there instead.
It's not so bad when you are aware of it. I recently checked the FPE and the arrow keys give incremental movements of 0.15 mm. Once you know that, good precision can be obtained.
Terry

Terry,

If you work with metrics daily it is a xxx*. To the degree that I once considered to use Qimage in imperial mode. I'm on a computersystem with a GUI, I'm not going to count the number of arrow key actions. If I select an image in the print page editor I like to see its position in the blink of an  eye at 0.25 mm precise or alike. Like you can when you use inches. I'm living in a metric world. Not too lazy to translate measurements in inches when in an email discussion as I otherwise get a reply what the size is in inches. Which shows that while we can compute everything there is a desire to keep things simple in daily practice. My daily practice is metric.


met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

Ernst
I agree with everything you have said, I can see the differences in borders you mentioned and can trim even borders without measuring, but few people without printing (or engineering  :D ) experience can. It's always nice to be exact, and I always try to be, but sometimes it's not necessary depending on the end user. When I said trimming to 1mm could be a problem,  I was thinking of some of the folks I used to work with  >:( (I've recently retired).

My original point was that the 3800 paper feed is not precise or exact, although the printing is. When I worked with a 7600 I could print or overprint and be totally confident of the feed, this is not the case with the 3800 (in my experience).

Terry, I'll try your suggestion later.

Brian

RM is Fixed=3.034mm, Additional=-8mm, Total -4.966mm
All others Additonal=0, Total 2.963mm (bottom 3.034mm)

This gave me hope because the right hand side of the print looks ok to me. (after changing all border colors to white).


You cannot change the margins in the Epson driver. According to the manual you can do it in Mac, but not in Win, which I have.


I think not. Otherwise there would not be problem at all. Attachment shows how the margins and borders should come out in this case.

I have come to the conclusion that my "problem" cannot be solved within Qimage.

/old Fox

Hi Oldfox,

It's pity this diagram was not shown earlier on, it makes it quite clear what your requirements are.

You are right. Your requirement shows in effect unequal borders, 2 at 7mm and 2 at 10mm and Qimage does not do unequal borders.
I assumed, because borderless does not work, that with having to have page margins, you would either accept a smaller image or smaller borders and trim the page margins from the print.
Oh well, after a few diversions we got there but probably not to your satisfaction.

I just remembered, all is not lost because there is a way to do it in Qimage. It's a bit complicated and involves having a plain white image behind each picture, which is larger and offset from the picture. The white background image creates the border. If this is a regular requirement, a template can be made, so just the picture images need be added.
If you want more explanation, come back to me.
Terry

Brian,
No need to bother now. See Oldfox's last post and the diagram, the requirement is in fact unequal borders which can't be done with Borders. It can be fiddled though, see my post.
Terry

I thought that the first picture in the first post would show what I wanted. Sorry that it was not clear enough.

Thx for your help. This exercise has been useful to me, still.

Hi Oldfox,

I'm hesitating to add anything at this late stage, but it may work better if you 'fudge' the paper size in the epson printer - e.g. make a custom size, say 6mm bigger than A4, but still load a4 paper. Also, you can possibly put a spacer to shift the paper 3mm away from the existing lhs guide - a few bits of card and sticky tape would be good enough to try. Of course, you may have to do somersaults in the calculations you make.

Just an idea you may want to try, no guarantee it'll work for you, or even worth trying (just don't blame me if you get tape stuck to the printhead/whatever  ???).

Best wishes,

Ray

Cunning idea. It could work for the sides of the paper, what about the leading edge (short side)? It would be difficult to fool the printer edge sensing for that  ::)
Terry

Brian,
I just checked with Mike, and after 100 posts, you need to provide an icon image of yourself that  represents your visage after you retired from your old job which we all took to be the boatman, Charon, taking lost souls across the river Styx.
Now that you admit to having retired, AKA joining the Old Folks, it is a forum rule that you provide a more current image.
Everyone is waiting for your personal representation.
After all, I posted one of me, Terry did a movie of himself, and last but not least, Jeff, the Grumpy one! ;D ;D
Your turn!

Fred

Hi Terry,


If it's a white border, (hopefully narrower than the desired border around each photo) then you will most likely have to turn off any printed crop marks, and reposition the templates, but probably easier to just make the required extra guillotine cut. An alternative would be to mount the A4 sheet on a larger carrier, similar to folk who manage to print on canvas using printers not designed for that, but that is even more fiddly. If you print on a roll, however, it can probably be ignored (but the op is not using roll paper).

But, if it is white borders you turn off the borders/crop marks in qimage then the problem goes away. I think you could generate a custom template and simply position the images as required and use a stop on the guillotine to give the even borders. ( I may have missed something here, the original question is not in my memory ???)

Best wishes,

Ray

Ray,
Ummm, yes you have, too complicated to explain, read Oldfox's stuff again.   :o I think this is finished with now.  ;)

Terry

I give explanation :'(

Photo taken after a lawn picknic - my usual state after a bottle of wine :)

Grumpy is down to Google.  When setting up a Blog  http://grumpy-jeff.blogspot.com/  I could not get Google to accept a sensible name and after about a dozen attempts I was getting quite 'Grumpy' so I entered grumpy-jeff and for a password I gave penis, they accepted the name but rejected the password - not long enough must be 8 char.  That really upset me so I gave cobblers spelled with a B and finally got my blog address.

Jeff  

That's what I also thought yesterday. This morning I started to play with margins and other variables involved. And I managed to get a satisfactory result. It may not be accurate in terms of millimeters (because Epson does not tell what is the difference between Min. Mid and Max Enlargement with Borderless) but it is good enough for my purposes.

Auto cropping:   On
Borderless:        On, Expansion: Min
Borders:            1=5mm, 2=5mm

Margins:

Left     Top   Right  Bottom

Fixed        0,000   0,000   0,070   0,070

Additional  0,000   1,000   1,000   2,000

Total        0,000   1,000   1,070   2,070

See attachment.

/(young again) Fox

Hi (young)Fox  ;)
I'm pleased to hear you got it sorted. I hope you've saved that set-up for future use.
Thanks for letting us know.
Terry

Fred

I'll have a look in my attic, there's a picture of me in which I seem to be getting older every year . . .

Brian

That picture is Dorian Gray; the one thet keeps getting older..... tut tut, as Terry would say!   :D :o