and now a technical discussion on print size

Having discussed the artistic merit of selecting a print size, and also with Sean's comment, the consideration of what happens when it all falls apart, could anyone shed any light on what is necessary to take a print beyond where you would norally go - so to speak.

I've been playing with enlarging sections of Gx100 images to an equivalent of 48 by 32 inches or thereabouts and after putting the file through qimage am getting maze artifacts, particularly in areas with fine texture.

My thoughts are that the following may help_

- reduce the level of sharpening beore sending to Qimage - perhaps I have been too aggressive in capture sharpening?

- clean up noise more aggressively before resampling

- do a small resample with bicubic smoother to increase the size of the detail and give Qimage something easier to work on

- use silkypix instead of LR to give a smoother raw conversion

Obviously there is a resolution issue here in that the fiels are only 9Mp to start with, but any thoughts on the specific problem and the more general how to work on files for big prints would be really appreciated.

Thanks

Mike

Mike

any images, that would show us the artefacts?

Michael, that was quick!

Here is a scan of a section of 48 by 32 print - the section is about 2.25 inches square and is scanned at 200dpi with no sharpening applied.



You can see the maze like artifacts in an area that is actually gravel.

Thanks

Mike

Mike Shimwell said:

I've been playing with enlarging sections of Gx100 images to an equivalent of 48 by 32 inches or thereabouts and after putting the file through qimage am getting maze artifacts, particularly in areas with fine texture.

Click to expand...

Hi Mike,

Which of Qimage's resampling algorithms did you use, and are you sure the artifacts were not caused by earlier sharpening steps?

Bart

looks weird....

I don't know Qimage, but it looks like it puts that maze-structure over the image to cover noise, pixelartefacts, etc

- Could you print a section without Qimage? Straight out of PS or friends?

I dont' see sharpening artefacts; nor pixel one's - therefore my guess.
A month ago, I uprez from my 17 MP-cam to the same size, with a Lamda-printer.

Even at very close inspection, none of the 7 prints had that structure...

I must admit I wondered about earlier steps in the process. I'll do another print tomorrow evening and post the results.

It's time for bed now - early start

Thanks

Mike

Hi Mike;
I have had the same issues with some of my early 30x40 prints. After some experimentation I found that the issues were in how I handled my original Raw files and their conversion. I started out using CS3 and the genuine fractals plugin. I found the image ok but too much of the maze like structures. I then went back to Qimage which I had used a lot when I had a 2200. I found that converting with CaptureNX and then printing with Qimage give me my best results. I also found that I had to dial back my base sharpening by at least 50% or even more, my eye quickly found that “sweet spot” of sharpening where uprezzing in Qimage didn’t case any noticeable effect to my eye. The photo editor of the local daily disagrees and tells me there is some “crunchieness” which is what he calls over sharpening. I have found individual images require different sharpening methods and styles. My current workflow is RAW to CaptureNX producing 16 bit tiffs and then printing in Qimage to a Epson 9880. I am of course still learning and find myself spending a lot of time reading online.

Mark Prins
Inanda Images
Whitehorse Yukon
Canada

Hi

Thanks for your replies. The inital problem was the raw conversion and sharpening - now I've just gotto work out how to fix it.

I've got to log off now as I'm too busy, but will add more when I get the chance to do some more work on the image.

Agains, Thanks for now

Mike