Digital BW, The Print

I sent a 720 ppi file to my 1280 then resampled it to 360 and printed
it again. No difference to my naked eye, possible miniscule
differences with a very good loupe but it's a hard call. 

It was a very sharp detailed 6x7 neg scanned at 3200 ppi in a Minolta
MultiPro. I sized it to 8x10 and it's native resolution was 896 ppi. I
then downsampled it to 720 ppi in PS and printed it on my 1280 using
Eboni BO on EEM at 2880 dpi. I marked the print on the back. Then I
downsampled the file to 360 ppi in PS and reprinted it the same. The
only difference was one sent to the printer as 720 ppi, the other at
360 ppi. No visual difference from the Epson driver upsampling the 360
ppi back to 720 for printing. 

My experience had already told them that there was nothing to be
gained from scanning at a rez tht yeilded more than 360 ppi at the
print size but I had to try the same image both ways. FWIW, I see
almost no diffeence printing 2880 & 1440 dpi. 

Now I'm more a "way the prints look" analytical type than a scientific
one. My memory of the "scientific method" was 40+ years ago in school.
Someone tell me if my approach is valid? Sometime I will download
Ernst's suggested test targets and try. But I already know I'm more
interested in printing quality images than quality test targets. 

Bob Michaels

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Bob Michaels"
<bob@b...> wrote:
> 
> I sent a 720 ppi file to my 1280 then resampled it to 360 and printed
> it again. No difference to my naked eye...

Only the newer model Epsons have native res higher than 360, so there
wouldn't be.
Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tyler Boley"
<tyler@t...> wrote:
> 
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Bob Michaels"
> <bob@b...> wrote:
> > 
> > I sent a 720 ppi file to my 1280 then resampled it to 360 and printed
> > it again. No difference to my naked eye...
> 
> Only the newer model Epsons have native res higher than 360, so there
> wouldn't be.
> Tyler

That's not true of desktop Epson printers, which I believe the 1280 is
regarded as being. As I understand it (from an Epson Technical Support
Bulletin) Epson has used 720 ppi as the native resolution for all
desktop printers. Their large format (professional) printers--until
recent models--were native at 360 ppi.

Anyway, the fact that one might not be able to discern differences in
iamge quality (by comparing two prints) is not particularly relevant
to the question of what the printer's native resolution happens to be.

Phil

Phil: I wasn't trying to establish what the native resolution was,
only to see if the prints looked any better. I previously thought it
was 360 until someone came up with some authorative support for 720.
It still appears to be a non issue as the prints look the same if sent
to the printer around 360 ppi or anything much larger. That was my
original supposition (i.e. most very high rez scans are unnecessary).

Bob Michaels 

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Phil Rose"
<pjrose@f...> wrote:
> 
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tyler Boley"
> <tyler@t...> wrote:
> > 
> > --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Bob Michaels"
> > <bob@b...> wrote:
> > > 
> > > I sent a 720 ppi file to my 1280 then resampled it to 360 and
printed

> > > it again. No difference to my naked eye...
> > 
> > Only the newer model Epsons have native res higher than 360, so there
> > wouldn't be.
> > Tyler
> 
> That's not true of desktop Epson printers, which I believe the 1280 is
> regarded as being. As I understand it (from an Epson Technical Support
> Bulletin) Epson has used 720 ppi as the native resolution for all
> desktop printers. Their large format (professional) printers--until
> recent models--were native at 360 ppi.
> 
> Anyway, the fact that one might not be able to discern differences in
> iamge quality (by comparing two prints) is not particularly relevant
> to the question of what the printer's native resolution happens to be.
> 
> Phil

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Phil Rose"
<pjrose@f...> wrote:
...
> That's not true of desktop Epson printers, which I believe the 1280 is
> regarded as being...

ANyone with a 1280 and Qimage could verify one way or the other very
quickly. I don't have it.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Bob Michaels"
<bob@b...> wrote:
> 
> Phil: I wasn't trying to establish what the native resolution was,
> only to see if the prints looked any better. I previously thought it
> was 360 until someone came up with some authorative support for 720.
> It still appears to be a non issue as the prints look the same if sent
> to the printer around 360 ppi or anything much larger. That was my
> original supposition (i.e. most very high rez scans are unnecessary).

This issue is confounded by the by the particulars of downsampling.
How the Epson driver downsamples vrs how you downsampled in PS may be
a factor. Sometimes it's better to start with 2 scans, one twice the
res of the other, resizing (NOT resampling) the smaller to 360 and the
larger to 720. They will be the same physical dimension on paper, and
how the printer deals with it will be the only issue.
If in fact the native res is 720, you may (or may not) see some
benifit with your original test by downsizing with Photoshop's newer
bicubic sharper mode. However if may be a mute point if you are
attempting to avoid large scans to begin with.
Ernst has reported superior resampling with Qimage for Epson printing,
over the driver itself.
Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Bob Michaels"
<bob@b...> wrote:
> 
> I sent a 720 ppi file to my 1280 then resampled it to 360 and
printed
> it again. No difference to my naked eye, possible miniscule
> differences with a very good loupe but it's a hard call. 
> 
> It was a very sharp detailed 6x7 neg scanned at 3200 ppi in a
Minolta
> MultiPro.

Try resampling the 360ppi image back to 720ppi and then compare them
in Photoshop.  You may find that there's little if any difference, so
you wouldn't expect to see a difference in the print either.

Film scanners scan at a much higher resolution than the resolution you
can actually get out of film.   Many scanners can generate a 20
megapixel scan from 35mm film, but in real life you can get more
resolution out of a 6-megapixel digital SLR than from 35mm film in
most cases.  I've done experiments where I downsampled film scans to
1/4 their size and back up again and found very little difference. 
I've also upsampled 6-megapixel digital SLR files to 20 megapixel and
they had more detail than the 20-megapixel film scans.

It would make more sense to try the experiement with a true 720ppi
image from a digital SLR or with an artificially-generated
test image that clearly had 720ppi worth of detail.

--
Steve

Tyler Boley wrote:

>  Sometimes it's better to start with 2 scans, one twice the
> res of the other, resizing (NOT resampling) the smaller to 360 and the
> larger to 720. They will be the same physical dimension on paper, and
> how the printer deals with it will be the only issue.

I guess the best way would be a vector design from Illustrator 
with Siemens stars etc and rasterise that to 360 PPI  + 720 PPI 
in Photoshop. An artificial photo like landscape made in a vector 
program would be a nice addition to that design. Starting from 
that you know that the rasterisation is a one step process both 
for the 360 and the 720 PPI version.

I like the idea that my 9000's and the 10000 don't have more than 
360 PPI native resolution, doesn't raise the question whether I 
should feed them more than that.

Ernst

Tyler Boley wrote:

>
>Only the newer model Epsons have native res higher than 360, so there
>wouldn't be.
>Tyler
>
>  
>
Tyler, in all politeness and with all due respect, that's completely 
incorrect information.

Read the document PS #2002.01.001  (specifically Note 2 as the querant 
here was speaking of a 1280) carefully at...
http://files.support.epson.com/pdf/pro10a/pro10aps.pdf

The EPSON desktop models from the 1270 forward all have at least a 
native resolution of 720 dpi...

Unfortunately, those used to using EPSON's larger format EPSON's often 
give inapt advice to those using desktop models, b/c they are basing 
their information on what they know of the larger model line.

Keith

 
Keith Krebs

"Just some guy," caretaker of the Multiverse's largest EPSON printer 
User Community (highly recommended by Vogon Poets and MegaDodo 
Publications), at:
http://groups.yahoo.com/group/EPSON_Printers/
and  the Multiverse's largest Canon printer User  Community at:
http://groups.yahoo.com/group/Canon-printers
"For the rest of you out there, the secret is to bang the rocks together 
guys"

Tyler Boley wrote:

>--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Phil Rose"
><pjrose@f...> wrote:
>..
>  
>
>>That's not true of desktop Epson printers, which I believe the 1280 is
>>regarded as being...
>>    
>>
>
>ANyone with a 1280 and Qimage could verify one way or the other very
>quickly. I don't have it.
>
>  
>
Yup, I have QImage too.. It's 720..
Keith

 
Keith Krebs

"Just some guy," caretaker of the Multiverse's largest EPSON printer 
User Community (highly recommended by Vogon Poets and MegaDodo 
Publications), at:
http://groups.yahoo.com/group/EPSON_Printers/
and  the Multiverse's largest Canon printer User  Community at:
http://groups.yahoo.com/group/Canon-printers
"For the rest of you out there, the secret is to bang the rocks together 
guys"

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Bob Michaels"
<bob@b...> wrote:
> 
> Phil: I wasn't trying to establish what the native resolution was,
> only to see if the prints looked any better.

Yes, but my comment was prompted by Tyler's statement to the effect
that a printer native resolution of only 360 ppi would explain your
result.

>I previously thought it
> was 360 until someone came up with some authorative support for 720.

I'd say that info from Epson's own technical bulletin ought to be
about as authoritative about that as we're likely to find. ;-)


> It still appears to be a non issue as the prints look the same if sent
> to the printer around 360 ppi or anything much larger. That was my
> original supposition (i.e. most very high rez scans are >unnecessary).

I expect that you're right. The liklihood of producing a "noticable"
improvement (by printing data input at 720 ppi) would be strongly
dependent on the type of image, and also possibly dependent on whether
or not the alternative was to use some "odd" resolution value that was
a non-factor of 720 (e.g., 275, 318, etc., etc.)  With just wishful
thinking to base it on, I've developed reluctance to use resolutions
that were merely "around 360 ppi" (rather than exactly 360).

Finally (for anyone coming in late to this thread), if anyone
resamples to 720 in Photoshop using the nearest-neighbor interpolation
rather than something like bicubic (or better) then there'd be no
reason at all to expect improvement over letting the driver
interpolate. That, of course, assumes the Epson driver uses nearest
neighbor interpolation, which is a supposition that could use some
"authoritative support". As we know, Mike Chaney (Qimage) is one
source of support for that supposition.

Phil

Phil,

I seem to remember Kennedy McEwan saying that the basic interpolation method 
of Epson drivers seemed to be Nearest Neighbour, but that newer drivers 
often had a DCC box that could be checked (Digital Camera Correction) to 
switch to some other interpolation (Bilinear probably) that would reduce 
jaggies on low MP pictures.
I think he said it was perfectly easy to check whether the driver is using 
Nearest Neighbour (NN) - just make an image of black and white squares and 
print it. If the squares have nice sharp edges in the print, it is using NN. 
If the driver is using Bilinear or Bicubic, the edges will have been 
softened.

Bob Frost.

----- Original Message ----- 
From: "Phil Rose" <pjrose@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>


Finally (for anyone coming in late to this thread), if anyone
resamples to 720 in Photoshop using the nearest-neighbor interpolation
rather than something like bicubic (or better) then there'd be no
reason at all to expect improvement over letting the driver
interpolate. That, of course, assumes the Epson driver uses nearest
neighbor interpolation, which is a supposition that could use some
"authoritative support". As we know, Mike Chaney (Qimage) is one
source of support for that supposition.

Dear some guy, I was not basing my info on what I know of large format
Epsons, and have used desktops for years. I passed on info based on
faulty memory, that's all.
By the way, I have never believed much of what Epson says regarding
this and other issues, and place much more  faith in what expert
users, and non-aligned tech wizards, have had to say over the years.
If you ever have an Epson rep want to pass on tech info, make him buy
you some drinks, nod and smile a lot, and leave the info at the bar.
Sorry for the misinformation, I thought I read 720 began for desktops
with the 2200.
Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Editor P.O.V.
Image Service" <editor@p...> wrote:
> Tyler Boley wrote:
> 
> >
> >Only the newer model Epsons have native res higher than 360, so there
> >wouldn't be.
> >Tyler
> >
> >  
> >
> Tyler, in all politeness and with all due respect, that's completely 
> incorrect information.
> 
> Read the document PS #2002.01.001  (specifically Note 2 as the querant 
> here was speaking of a 1280) carefully at...
> http://files.support.epson.com/pdf/pro10a/pro10aps.pdf
> 
> The EPSON desktop models from the 1270 forward all have at least a 
> native resolution of 720 dpi...
> 
> Unfortunately, those used to using EPSON's larger format EPSON's often 
> give inapt advice to those using desktop models, b/c they are basing 
> their information on what they know of the larger model line.
> 
> Keith
> 
>  
> Keith Krebs
> 
> "Just some guy," caretaker of the Multiverse's largest EPSON printer 
> User Community (highly recommended by Vogon Poets and MegaDodo 
> Publications), at:
> http://groups.yahoo.com/group/EPSON_Printers/
> and  the Multiverse's largest Canon printer User  Community at:
> http://groups.yahoo.com/group/Canon-printers
> "For the rest of you out there, the secret is to bang the rocks
together 
> guys"

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Editor P.O.V.
Image Service" <editor@p...> wrote:
...
> >ANyone with a 1280 and Qimage could verify one way or the other very
> >quickly. I don't have it.
> >
> >  
> >
> Yup, I have QImage too.. It's 720..

There you have it, l my apologies.
Tyler

This speaks to something I've wondered about:  I've read in various 
places that a digital camera needs to record 12-14 Mp in order to 
exceed the information available from a 35mm negative.  Yet my own 8 
Mp camera routinely produces images which allow excellent prints at 
12 x 16, sometimes even to 18 x 24.  I could never say that about any 
of my 35mm lenses (all of which are at least 15 years old).  Some of 
this I'm sure is due to the improvement in lens quality recently, but 
is it really true film is beaten at 6 Mp?  Thanks

eric perkins 



--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "kanefsky" 
<steve@k...> wrote:
> 
> 
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Bob Michaels"
> <bob@b...> wrote:
> > 
> > I sent a 720 ppi file to my 1280 then resampled it to 360 and
> printed
> > it again. No difference to my naked eye, possible miniscule
> > differences with a very good loupe but it's a hard call. 
> > 
> > It was a very sharp detailed 6x7 neg scanned at 3200 ppi in a
> Minolta
> > MultiPro.
> 
> Try resampling the 360ppi image back to 720ppi and then compare them
> in Photoshop.  You may find that there's little if any difference, 
so
> you wouldn't expect to see a difference in the print either.
> 
> Film scanners scan at a much higher resolution than the resolution 
you
> can actually get out of film.   Many scanners can generate a 20
> megapixel scan from 35mm film, but in real life you can get more
> resolution out of a 6-megapixel digital SLR than from 35mm film in
> most cases.  I've done experiments where I downsampled film scans to
> 1/4 their size and back up again and found very little difference. 
> I've also upsampled 6-megapixel digital SLR files to 20 megapixel 
and

> they had more detail than the 20-megapixel film scans.
> 
> It would make more sense to try the experiement with a true 720ppi
> image from a digital SLR or with an artificially-generated
> test image that clearly had 720ppi worth of detail.
> 
> --
> Steve

Tyler Boley wrote:

>Dear some guy, I was not basing my info on what I know of large format
>Epsons, and have used desktops for years. I passed on info based on
>faulty memory, that's all.
>  
>
No problem.  I do that myself.. It's an occupational hazard of being human.

>By the way, I have never believed much of what Epson says regarding
>this and other issues, 
>
ditto

>and place much more  faith in what expert
>users, and non-aligned tech wizards, have had to say over the years.
>If you ever have an Epson rep want to pass on tech info, make him buy
>you some drinks, nod and smile a lot, and leave the info at the bar.
>  
>
LOL

HOW TRUE!

Thanks Ty!
Keith

 
Keith Krebs

"Just some guy," caretaker of the Multiverse's largest EPSON printer 
User Community (highly recommended by Vogon Poets and MegaDodo 
Publications), at:
http://groups.yahoo.com/group/EPSON_Printers/
and  the Multiverse's largest Canon printer User  Community at:
http://groups.yahoo.com/group/Canon-printers
"For the rest of you out there, the secret is to bang the rocks together 
guys"