IQ4 output does not fit Epson print sizes that well!

[narikin]

Let me start out by saying I make big prints, (which is why I use IQ backs) With that in mind, it was disappointing to find that the IQ4 does not fit well into Epson's larger paper rolls/printer sizes, i.e. the 44" & 60" rolls.

Using the new generation Epson Printers, which, like Canon, have native 150/300dpi heads (rather than the older 180/360), and printing at round divisions of native dpi, to avoid artifacts, the IQ3-100 fitted perfectly onto both 44" and 60" paper: 150dpi gave beautiful x-large 58x77" prints, a great fit on 60" rolls, and 200dpi gave 43.5x58", perfectly suited to either 44" or 60" rolls. Amazing! (a 0.25 to 1" border is very useful for handling, mounting, etc)

With the IQ4 however nothing really works without an excess of waste:
200dpi is 51x70.8" so you're wasting a lot of a 60" wide roll
250dpi is 41x56" (least waste on 44" roll)
300dpi is 34x46" (46" does not go on 44" roll)
(150dpi would give a lovely 68.7x94" if only Epson made a 72" printer! Please please, Epson?)

Of course this is not Phase/Sony's fault, just the way it works out, but... disappointing. I was naively thinking it would all fit nicely at a higher dpi. Nope. Put another way, this is a decent reason to stay with 100mp if you print at this sort of scale.

(yes I get that buying an IQ4 back and then counting the pennies on paper waste is anachronistic, but at ~$440/39ft for 60" of the good stuff, you see your $ literally going into the trash every week, and that hurts!  )

[faberryman]

Of course this is not Phase/Sony's fault, just the way it works out, but... disappointing. I was naively thinking it would all fit nicely at a higher dpi. Nope. Put another way, this is a decent reason to stay with 100mp if you print at this sort of scale.

So you would forego the extra resolution of the IQ4 150 to minimize paper waste? "Anachronistic" isn't the right word for it.

[MichaelEzra]

No disrespect, but this sounds rediculois..

What is wrong with using a decent interpolation to whatever pixel size you require?

[narikin]

You are missing the simple point I'm trying to share: For large prints, the 100Mp back happens to perfectly fit to the paper sizes that are now standard across all Large Format printers. The 150mp back does not. That's it.

If you print large (and presumably that's one of the principal reasons people buy 100 or 150mp backs) then the 100 works better without up or down rezing files.  Accepted wisdom is that its better to send the printer some exact number corresponding to its native resolution. 100mp gives you exactly that, as I outline above. 150mp does not.  Send the printer 150 or 300ppi exactly, and you get better results than sending it some random number like, oh, 183.5ppi.

Does the extra resolution of the IQ4 make up for the less 'printer friendly' files? that's a test that remains to be made. 

(I don't uprez/interpolate, because I always find that makes everything softer, which negates all my effort in getting a remarkably sharp image)

[aaronchan]

I feel like you just need to learn how to properly up/down res and sharpen your image.

Aaron


Sent from my iPhone using Tapatalk

[MichaelEzra]

Exactly.

[narikin]

I feel like you just need to learn how to properly up/down res and sharpen your image.

Hmmm.  I've tried many of the up-res programs out there, and found the results marginally inferior to what I'm seeing in the untouched file. My mantra is to keep as close to the source image as possible, native pixels always being better than invented ones.  But YMMV, of course.

Sharpening: I've had plenty of people like college lecturers, other printers, write to me after exhibitions, to enquire what my print sharpening routine is. So my feeling is I'm not doing anything wrong there. But there's always room to improve!

Am not trying to generate disagreement here, just wanted to point out the 150Mp back does not natively fit well with the standard paper sizes, when printing big. The 100mp one does. That's all! Cheers.

[hubell]

I feel like you just need to learn how to properly up/down res and sharpen your image.

Aaron


Sent from my iPhone using Tapatalk

+1. I have never heard of someone focused on whether their camera’s native file size fits a particular print size without the need for upsampling.

[Joe Towner]

Sounds like someone needs a bigger printer... 80" wide?

[chrismuc]

Every pixel of a tiff image from a digital camera sensor has a certain depth (DR) of for example 8 bit or 16 bit times 3 colors RGB and the pixels are distributed in a square grid. An ink jet printer print head prints a number of cmyk+ (4 or more) colors in 1 bit in a pseudo-stochastic distribution of very high resolution, much higher than 150 or 300 dpi. Therefore the 'resolution' of a sensor of tiff image is not related to the 'resolution' of a printer. Never 1 pixel of an image would be printed on 1 dot by a printer.

Besides: Nearly never, the image of a sensor stays without correction/interpolation for print: Mostly the first interpolation nowadays is already in the camera when the firmware interpolates the image for lens correction. Second, in post mostly (especially in architecture or interior design photography) one would slightly adjust a tilted view and apply keystone correction = another interpolation. Then one would possibly crop the image to taste and to the prefered width/height ratio, so the original resolution would not remain.

Therefore the whole topic to relate a sensor resolution directly to a pinter resolution is meaningless.

[eronald]

He may be right about the quality issues, or he might not (which most assume) which would lead to the question why. Double blind testing would be interesting.

Edmund

Every pixel of a tiff image from a digital camera sensor has a certain depth (DR) of for example 8 bit or 16 bit times 3 colors RGB and the pixels are distributed in a square grid. An ink jet printer print head prints a number of cmyk+ (4 or more) colors in 1 bit in a pseudo-stochastic distribution of very high resolution, much higher than 150 or 300 dpi. Therefore the 'resolution' of a sensor of tiff image is not related to the 'resolution' of a printer. Never 1 pixel of an image would be printed on 1 dot by a printer.

Besides: Nearly never, the image of a sensor stays without correction/interpolation for print: Mostly the first interpolation nowadays is already in the camera when the firmware interpolates the image for lens correction. Second, in post mostly (especially in architecture or interior design photography) one would slightly adjust a tilted view and apply keystone correction = another interpolation. Then one would possibly crop the image to taste and to the prefered width/height ratio, so the original resolution would not remain.

Therefore the whole topic to relate a sensor resolution directly to a pinter resolution is meaningless.

[narikin]

Every pixel of a tiff image from a digital camera sensor has a certain depth (DR) of for example 8 bit or 16 bit times 3 colors RGB and the pixels are distributed in a square grid. An ink jet printer print head prints a number of cmyk+ (4 or more) colors in 1 bit in a pseudo-stochastic distribution of very high resolution, much higher than 150 or 300 dpi. Therefore the 'resolution' of a sensor of tiff image is not related to the 'resolution' of a printer. Never 1 pixel of an image would be printed on 1 dot by a printer.

Besides: Nearly never, the image of a sensor stays without correction/interpolation for print: Mostly the first interpolation nowadays is already in the camera when the firmware interpolates the image for lens correction. Second, in post mostly (especially in architecture or interior design photography) one would slightly adjust a tilted view and apply keystone correction = another interpolation. Then one would possibly crop the image to taste and to the prefered width/height ratio, so the original resolution would not remain.

Therefore the whole topic to relate a sensor resolution directly to a pinter resolution is meaningless.

There are pages and pages of established tests showing that printers have input resolutions that work better with their internal computations (input pixels to ink drops) For Epson it used to be 360 dpi or round multiples/divisions thereof. The latest generation, and others like Canon, use 300dpi. I'm not saying anything radical or new here. It's a pretty accepted wisdom to get the ultimate output, that this or some easily divisible/multiplicable factor of it is ideal.

I work in a very pure way. No interpolation/upres of files, no keystone correction, no cropping, no correcting converging verticals. (I appreciate it's absolutely required for other kinds of photography) . Hence my delight at the 100mp back giving near perfect large print sizes at the printers preferred 'native' resolutions. The new 150 doesn't work out so neatly, that's all I was pointing out.

There's lots of other fields, like HiFi, where people want to keep as close to the original source material as possible. That's me, here. Of course there's de-mosaicing, dithering, etc, but if I can remove one or two steps of futzing the pixels, on my way to making a print, I prefer to do that. No idea why this causes consternation to some here, but it's the way I like to work. And I feel it does just that: works. 

[Chris Barrett]

Well, you won't lose any image quality by DOWNrezzing, so you could simply print at 150 and downsize from 71" to 58".

[fdisilvestro]

Downsampling is not free of issues either. It has do be done properly, otherwise high frequency detail in the original image could cause artifacts in the downsampled version.

He may be right about the quality issues, or he might not (which most assume) which would lead to the question why. Double blind testing would be interesting.

Edmund

Indeed!

[faberryman]

Let me start out by saying I make big prints, (which is why I use IQ backs) With that in mind, it was disappointing to find that the IQ4 does not fit well into Epson's larger paper rolls/printer sizes, i.e. the 44" & 60" rolls.
With the IQ4 however nothing really works without an excess of waste:
200dpi is 51x70.8" so you're wasting a lot of a 60" wide roll
250dpi is 41x56" (least waste on 44" roll)
300dpi is 34x46" (46" does not go on 44" roll)
(150dpi would give a lovely 68.7x94" if only Epson made a 72" printer! Please please, Epson?)

Not sure why you settled on 50dpi increments. At 180dpi, you get 59.2" (10652/280) which fits on 60" paper with less than 1/2" border. and even at 250dpi, my calculations are 42.6" (10652/250) which fits on 44" paper with less than a 3/4" border. I'm not seeing the waste.

[Wayne Fox]

Using the new generation Epson Printers, which, like Canon, have native 150/300dpi heads (rather than the older 180/360)

The only Epson wide format printers that use this resolution are the p10000 and the p20000.  The p800 and p5000-9000 are still all 360 based devices.

I have the p800, p5000, p9000, and I have a p20000 at my store. I’ve printed on Canon’s in years past.  My print are based on the specific sizes I have chosen to sell, I never worry about how the file resolution relates, and really have never had any issues. It would take some pretty close examination (such as with a loupe)  to see any difference between a 90” print printed on the p9000 vs the p20000.

[boinger]

I think you are fundamentally misunderstanding how printers work.

1 pixel from your camera will have a definite color value somewhere on the rgb spectrum.

A printer dot is one of multiple dots of the various pigments that exist in the printers library of colors. It uses these dots to make a color pixel or image.

It takes multiple dots from a printer to equal 1 pixel from your camera as you would want to see it in print. 

The two are not at all equivalent and should thus not bother being compared.