couple of questions about scanning photos

Don’t know about the rest of the folks here, but I always scan photos that I intend to store long-term at the maximum DPI the scanner allows.

I’ve been able to make almost perfect copies for my sister after her house fire last year from CD-R’s that I had burned from her album scans.

—
Tiny
=================================================
The biggest troublemaker you’ll probably ever have to deal with watches you shave his face in the mirror every morning.
=================================================
"Nidhal" wrote in message

I’m scanning a bunch of family pics for storage on my computer. My main question is about DPI. I’ve been scanning the photos at 1200 DPI, because that is the maximum that my scanner will allow. It’s a lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI because I want to retain the best quality possible. Anyway, I’m just looking for advice as to whether that is overkill or not. The photos I’m scanning are a pretty even split of color and black & white, should that matter. Also, some of the photos are 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so. Just thought I’d ask for some advice. Thanks in advance for any help you can pass along.

it depends what you will use them for.
unless you are doing drastic enlargements, 600 dpi is plenty.

if you plan to reprint them at original size, 300 dpi would be good, 600 is very high quality, and the difference between 600 and 1200 would not be noticeable on most inkjet photo printers.

however, if you ever plan to enlarge them, and you have the storage space, you might as well go for the max your scanner can handle.

Michael Evangelista
Southern Utah Web Design
www.suwebs.com
—
"Nidhal" wrote in message

I’m scanning a bunch of family pics for storage on my computer. My main question is about DPI. I’ve been scanning the photos at 1200 DPI, because that is the maximum that my scanner will allow. It’s a lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI because I want to retain the best quality possible. Anyway, I’m just looking for advice as to whether that is overkill or not. The photos I’m scanning are a pretty even split of color and black & white, should that matter. Also, some of the photos are 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so. Just thought I’d ask for some advice. Thanks in advance for any help you can pass along.

On Tue, 28 Dec 2004 20:45:50 -0800, "Nidhal"
wrote:

I’m scanning a bunch of family pics for storage on my computer. My main question is about DPI. I’ve been scanning the photos at 1200 DPI, because that is the maximum that my scanner will allow. It’s a lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI because I want to retain the best quality possible. Anyway, I’m just looking for advice as to whether that is overkill or not. The photos I’m scanning are a pretty even split of color and black & white, should that matter. Also, some of the photos are 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so. Just thought I’d ask for some advice. Thanks in advance for any help you can pass along.

You must be getting hugh file sizes. What are you scanning them for? I would assume you want to either archive them or burn them to a DVD for viewing. I think the 1200 dpi is overkill. If you find any definitive awnser please post it.

Tiny wrote:

Don’t know about the rest of the folks here, but I always scan photos that I intend to store long-term at the maximum DPI the scanner allows.
I’ve been able to make almost perfect copies for my sister after her house fire last year from CD-R’s that I had burned from her album scans.

hehe. Have you any idea how LARGE a 2400dpi scnned TIFF file of a regular photo is? You dont get many of them on a CDR 🙂

—

x theSpaceGirl (miranda)

# lead designer @ http://www.dhnewmedia.com #
# remove NO SPAM to email, or use form on website #

"Nidhal" wrote in message I’m scanning a bunch of family pics for storage on my computer. My main question is about DPI. I’ve been scanning the photos at 1200 DPI, because that is the maximum that my scanner will allow. It’s a lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI because I want to retain the best quality possible. Anyway, I’m just looking for advice as to whether that is overkill or not. The photos I’m scanning are a pretty even split of color and black & white, should that matter. Also, some of the photos are 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so. Just thought I’d ask for some advice. Thanks in advance for any help you can pass along.

What file format are you using to save these photos? Since you are scanning at 1,200 DPI, are you saving them as TIFF?

—

"Nidhal" wrote in message

[…] The photos I’m scanning are a pretty even split of color and black & white, should that matter. Also, some of the photos are 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so. Just thought I’d ask for some advice. Thanks in advance for any help you can pass along.

You are doing the right thing. If the majority are 3×5" or smaller, then higher resolution is a good thing because you will have images that will print adequately at 4x the original size – if that’s what you wish to do. If you want to make web images of them eventually, then you can resize automatically later.

"SpaceGirl" wrote in message

hehe. Have you any idea how LARGE a 2400dpi scnned TIFF file of a regular photo is? You dont get many of them on a CDR 🙂

Sure. His pictures are mostly 3"x5", or about 3600×6000 pixels, maybe 22mb each, and depending upon how well they compress with RLE, he can store at least 30 pictures to a CDROM or 200 or so to a DVD.

Of course he could save in JPEG at a much smaller ‘resolution’ and have crap. Is that what you recommend?

Tiny wrote:

Don’t know about the rest of the folks here, but I always scan photos that I intend to store long-term at the maximum DPI the scanner allows.

If you scan photos (meaning prints), anything higher than 600 dpi is just overkill. You get more pixels and more megabytes, but not more detail.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

jjs wrote:

"SpaceGirl" wrote in message

hehe. Have you any idea how LARGE a 2400dpi scnned TIFF file of a regular photo is? You dont get many of them on a CDR 🙂

Sure. His pictures are mostly 3"x5", or about 3600×6000 pixels, maybe 22mb each, and depending upon how well they compress with RLE, he can store at least 30 pictures to a CDROM or 200 or so to a DVD.

Of course he could save in JPEG at a much smaller ‘resolution’ and have crap. Is that what you recommend?

Not at all. Just scan at a lower resolution, unless space is unlimited.

—

x theSpaceGirl (miranda)

# lead designer @ http://www.dhnewmedia.com #
# remove NO SPAM to email, or use form on website #

"SpaceGirl" wrote in message

Not at all. Just scan at a lower resolution, unless space is unlimited.

His pictures are mostly 3"x5". If he wants to make good enlargements he has no choice but to scan with at least as many pixels as the (nominal) 300ppi of the target image size.

"Nidhal" wrote in message
: I’m scanning a bunch of family pics for storage on my computer. My main : question is about DPI. I’ve been scanning the photos at 1200 DPI, because : that is the maximum that my scanner will allow. It’s a lot slower than if I
: scan at 600 DPI, but I’ve stuck with 1200 DPI because I want to retain the : best quality possible. Anyway, I’m just looking for advice as to whether : that is overkill or not. The photos I’m scanning are a pretty even split of
: color and black & white, should that matter. Also, some of the photos are : 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow,
: if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so.
: Just thought I’d ask for some advice. Thanks in advance for any help you : can pass along.

First check to see what the actual optical resolution of your printer is. If its
600 x 600 and you’re getting the extra resolution through interpolation then don’t scan at the higher rate. Wait until you need the pics and use Photoshop
or something else to enlarge.

It used to be the rule to scan at 1 1/2 to 2 times the target resolution. I still go
by that. Smaller pictures need more scanning power because the larger ones already are at a higher resolution in the real world.

"Johan W. Elzenga" wrote in message
: Tiny wrote:
:
: > Don’t know about the rest of the folks here, but I always scan photos that I
: > intend to store long-term at the maximum DPI the scanner allows. :
: If you scan photos (meaning prints), anything higher than 600 dpi is : just overkill. You get more pixels and more megabytes, but not more : detail.
:
:
: —
: Johan W. Elzenga johan<<at>>johanfoto.nl : Editor / Photographer http://www.johanfoto.nl/

Depends on the scanner, no?

jjs wrote:

"SpaceGirl" wrote in message

Not at all. Just scan at a lower resolution, unless space is unlimited.

His pictures are mostly 3"x5". If he wants to make good enlargements he has no choice but to scan with at least as many pixels as the (nominal) 300ppi of the target image size.

Yeah. 600dpi would be good I think.

—

x theSpaceGirl (miranda)

# lead designer @ http://www.dhnewmedia.com #
# remove NO SPAM to email, or use form on website #

"formerly known as ‘cat arranger’" wrote
in message

"Johan W. Elzenga" wrote in message
: Tiny wrote:
:
: > Don’t know about the rest of the folks here, but I always scan photos that I
: > intend to store long-term at the maximum DPI the scanner allows. :
: If you scan photos (meaning prints), anything higher than 600 dpi is : just overkill. You get more pixels and more megabytes, but not more : detail.
:
:
: —
: Johan W. Elzenga johan<<at>>johanfoto.nl : Editor / Photographer http://www.johanfoto.nl/

Depends on the scanner, no?

It may depend more on the original source of the photo. Was it a Kodak 110 instamatic or a 35mm SLR?
There is definitely a point of diminishing returns. Maybe try scanning the same small portion of one photo at different resolutions and see where the differences becomes insignificant.

Peadge 🙂

Nidhal wrote:

I’m scanning a bunch of family pics for storage on my computer. My main question is about DPI. I’ve been scanning the photos at 1200 DPI, because that is the maximum that my scanner will allow. It’s a lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI because I want to retain the best quality possible. Anyway, I’m just looking for advice as to whether that is overkill or not. The photos I’m scanning are a pretty even split of color and black & white, should that matter. Also, some of the photos are 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so. Just thought I’d ask for some advice. Thanks in advance for any help you can pass along.

1200 ppi is probably not overkill for the 3×5’s. They are contact prints that may contain a lot of detail. Enlarging hese images can be fun because no one has seen them at that size before.

The run of the mill 4×6 snapshots are probably enlargements from the 60’s and later. There is probably no benefit to scanning them at more than 300 ppi. You can verify this for yourself by picking out one of the sharpest 4×6’s and scanning it at several resolutions. If the image looks blurry at 300 ppi, there is no point going to a higher resolution.

Save time and space by scanning your black and white images in grayscale, rather than going to a lower resolution. Storage is cheap, but if space is an issue, you will get more quality if you use jpeg compression rather than going to a lower ppi.

If you haven’t already, check out Wayne Fulton’s www.scantips.com for good info on scanning issues.
—

Mike Russell
www.curvemeister.com
www.geigy.2y.net

Mike Russell wrote:

Nidhal wrote:

I’m scanning a bunch of family pics for storage on my computer. My main question is about DPI. I’ve been scanning the photos at 1200 DPI, because that is the maximum that my scanner will allow. It’s a lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI because I want to retain the best quality possible. Anyway, I’m just looking for advice as to whether that is overkill or not. The photos I’m scanning are a pretty even split of color and black & white, should that matter. Also, some of the photos are 4" x 6" or larger but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s an added benefit to scanning at 1200 DPI, I’ll continue to do so. Just thought I’d ask for some advice. Thanks in advance for any help you can pass along.

1200 ppi is probably not overkill for the 3×5’s. They are contact prints that may contain a lot of detail. Enlarging hese images can be fun because no one has seen them at that size before.

The run of the mill 4×6 snapshots are probably enlargements from the 60’s and later. There is probably no benefit to scanning them at more than 300 ppi. You can verify this for yourself by picking out one of the sharpest 4×6’s and scanning it at several resolutions. If the image looks blurry at 300 ppi, there is no point going to a higher resolution.
Save time and space by scanning your black and white images in grayscale, rather than going to a lower resolution. Storage is cheap, but if space is an issue, you will get more quality if you use jpeg compression rather than going to a lower ppi.

If you haven’t already, check out Wayne Fulton’s www.scantips.com for good info on scanning issues.

Wayne Fulton usually advises people to scan Prints at 300 dpi because commercial prints do not contain any more information than that. The paper on which prints are made does not have the fine grain that negatives and slides do. If you scan ordinary prints at any higher resolution you will just be getting a bigger file but no additional image information.
If you want to enlarge the image (larger than the original) just resample it in photoshop or your photo editor.
If you DEFINITELY do not want to make prints from the scanned image but only want to view it on a computer or TV screen, you could even get by with scanning your 3x5s at 200 dpi. That would give you a 600 x 1000 pixel image, just about right if your computer monitor is set to display at 600 x 800 pixels.
My recommendation is: Scan at 300 dpi and save the image in highest quality jpeg. Then burn the images to a CD. From there you can play the CD on your TV via a DVD player, create a slide show on your computer, or print enlargements on your inkjet.
Bob Williams

"Peadge" wrote in message
:
: "formerly known as ‘cat arranger’"
wrote
: in message : >
: > "Johan W. Elzenga" wrote in message
: > : > : Tiny wrote:
: > :
: > : > Don’t know about the rest of the folks here, but I always scan photos
: > that I
: > : > intend to store long-term at the maximum DPI the scanner allows. : > :
: > : If you scan photos (meaning prints), anything higher than 600 dpi is : > : just overkill. You get more pixels and more megabytes, but not more : > : detail.
: > :
: > :
: > : —
: > : Johan W. Elzenga johan<<at>>johanfoto.nl : > : Editor / Photographer http://www.johanfoto.nl/ : >
: > Depends on the scanner, no?
: >
: >
:
: It may depend more on the original source of the photo. Was it a Kodak 110 : instamatic or a 35mm SLR?
: There is definitely a point of diminishing returns. Maybe try scanning the : same small portion of one photo at different resolutions and see where the : differences becomes insignificant.
:
: Peadge 🙂

That’s probably the best suggestion; to experiment. : -) I do think though that the optical vs. interpolation is a factor than can be figured easily and makes a definite
difference.

"Mike Russell" wrote in message
: Nidhal wrote:
: > I’m scanning a bunch of family pics for storage on my computer. My : > main question is about DPI. I’ve been scanning the photos at 1200 : > DPI, because that is the maximum that my scanner will allow. It’s a : > lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI : > because I want to retain the best quality possible. Anyway, I’m just : > looking for advice as to whether that is overkill or not. The photos : > I’m scanning are a pretty even split of color and black & white, : > should that matter. Also, some of the photos are 4" x 6" or larger : > but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s : > an added benefit to scanning at 1200 DPI, I’ll continue to do so. : > Just thought I’d ask for some advice. Thanks in advance for any help : > you can pass along.
:
: 1200 ppi is probably not overkill for the 3×5’s. They are contact prints : that may contain a lot of detail. Enlarging hese images can be fun because
: no one has seen them at that size before.
:
: The run of the mill 4×6 snapshots are probably enlargements from the 60’s : and later. There is probably no benefit to scanning them at more than 300 : ppi. You can verify this for yourself by picking out one of the sharpest : 4×6’s and scanning it at several resolutions. If the image looks blurry at
: 300 ppi, there is no point going to a higher resolution. :
: Save time and space by scanning your black and white images in grayscale, : rather than going to a lower resolution. Storage is cheap, but if space is
: an issue, you will get more quality if you use jpeg compression rather than
: going to a lower ppi.
:
: If you haven’t already, check out Wayne Fulton’s www.scantips.com for good
: info on scanning issues.
: —
:
: Mike Russell
: www.curvemeister.com
: www.geigy.2y.net

Interesting. Yes, doing different crops on the same picture, even in a slide show or maybe particularly in a slide show type presentation can add a lot of interest, so don’t assume that all of your pics will be presented full frame, or whatever that is called. And using the "Ken Burns Effect" is a lot of fun. I wish I had the time and energy to experiment with a bunch of different presentations.

"Bob Williams" wrote in message
:
:
: Mike Russell wrote:
: > Nidhal wrote:
: >
: >>I’m scanning a bunch of family pics for storage on my computer. My : >>main question is about DPI. I’ve been scanning the photos at 1200 : >>DPI, because that is the maximum that my scanner will allow. It’s a : >>lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI : >>because I want to retain the best quality possible. Anyway, I’m just : >>looking for advice as to whether that is overkill or not. The photos : >>I’m scanning are a pretty even split of color and black & white, : >>should that matter. Also, some of the photos are 4" x 6" or larger : >>but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s : >>an added benefit to scanning at 1200 DPI, I’ll continue to do so. : >>Just thought I’d ask for some advice. Thanks in advance for any help : >>you can pass along.
: >
: >
: > 1200 ppi is probably not overkill for the 3×5’s. They are contact prints
: > that may contain a lot of detail. Enlarging hese images can be fun because
: > no one has seen them at that size before.
: >
: > The run of the mill 4×6 snapshots are probably enlargements from the 60’s
: > and later. There is probably no benefit to scanning them at more than 300
: > ppi. You can verify this for yourself by picking out one of the sharpest
: > 4×6’s and scanning it at several resolutions. If the image looks blurry at
: > 300 ppi, there is no point going to a higher resolution. : >
: > Save time and space by scanning your black and white images in grayscale,
: > rather than going to a lower resolution. Storage is cheap, but if space is
: > an issue, you will get more quality if you use jpeg compression rather than
: > going to a lower ppi.
: >
: > If you haven’t already, check out Wayne Fulton’s www.scantips.com for good
: > info on scanning issues.
:
: Wayne Fulton usually advises people to scan Prints at 300 dpi because : commercial prints do not contain any more information than that. The : paper on which prints are made does not have the fine grain that : negatives and slides do. If you scan ordinary prints at any higher : resolution you will just be getting a bigger file but no additional : image information.
: If you want to enlarge the image (larger than the original) just : resample it in photoshop or your photo editor.
: If you DEFINITELY do not want to make prints from the scanned image but : only want to view it on a computer or TV screen, you could even get by : with scanning your 3x5s at 200 dpi. That would give you a 600 x 1000 : pixel image, just about right if your computer monitor is set to display : at 600 x 800 pixels.
: My recommendation is: Scan at 300 dpi and save the image in highest : quality jpeg. Then burn the images to a CD. From there you can play the : CD on your TV via a DVD player, create a slide show on your computer, : or print enlargements on your inkjet.
: Bob Williams

I disagree. You can always downsize the file but you can’t upsize it without some loss compared to the original scan at higher resolution. This is especially true on small photos.

Really, there should be a kind of matrix with the factors listed, and the idea of Pleadge to experiment is probably the definitive statement.

formerly known as ‘cat arranger’
wrote:

: If you scan photos (meaning prints), anything higher than 600 dpi is : just overkill. You get more pixels and more megabytes, but not more : detail.

Depends on the scanner, no?

No, that is exactly my point. It does NOT depend on the scanner. The resolution of a photographic print is not higher than 600 dpi, normally it’s about 300 to 400 dpi. So, no matter what scanner (or what negative was used to produce the print!), it’s useless to scan at very high resolution. What isn’t there is the first place, cannot be captured by increasing the scan resolution.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

"Johan W. Elzenga" wrote in message
: formerly known as ‘cat arranger’
: wrote:
:
: > : If you scan photos (meaning prints), anything higher than 600 dpi is : > : just overkill. You get more pixels and more megabytes, but not more : > : detail.
: >
: > Depends on the scanner, no?
:
: No, that is exactly my point. It does NOT depend on the scanner. The : resolution of a photographic print is not higher than 600 dpi, normally : it’s about 300 to 400 dpi. So, no matter what scanner (or what negative : was used to produce the print!), it’s useless to scan at very high : resolution. What isn’t there is the first place, cannot be captured by : increasing the scan resolution.
:
:
: —
: Johan W. Elzenga johan<<at>>johanfoto.nl : Editor / Photographer http://www.johanfoto.nl/

A photograph is continuous tone. There are no dpi. You must be talking about a print from computer or a half tone and not an actual photo. And as far as transparencies go, they can be scanned at several thousand samples/inch, may be higher. I has been a while since I did this kind of work but I’m pretty sure.

formerly known as ‘cat arranger’ wrote:

"Johan W. Elzenga" wrote in message

formerly known as ‘cat arranger’
wrote:

If you scan photos (meaning prints), anything higher than 600 dpi is just overkill. You get more pixels and more megabytes, but not more detail.

Depends on the scanner, no?

No, that is exactly my point. It does NOT depend on the scanner. The resolution of a photographic print is not higher than 600 dpi, normally it’s about 300 to 400 dpi. So, no matter what scanner (or what negative was used to produce the print!), it’s useless to scan at very high resolution. What isn’t there is the first place, cannot be captured by increasing the scan resolution.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

A photograph is continuous tone. There are no dpi. You must be talking about a print from computer or a half tone and not an actual photo. And as far as transparencies go, they can be scanned at several thousand samples/inch, may be higher. I has been a while since I did this kind of work but I’m pretty sure.

This is correct. There is nothing magic about 300 ppi. In particular, I have scanned glossy contact prints from the 40’s and earlier that had incredible amounts of detail, as well as some recent tintypes, which are negatives in their own right.
—

Mike Russell
www.curvemeister.com
www.geigy.2y.net

"Mike Russell" wrote in message

This is correct. There is nothing magic about 300 ppi. In particular, I have scanned glossy contact prints from the 40’s and earlier that had incredible amounts of detail, as well as some recent tintypes, which are negatives in their own right.

Mike, while we are near the subject, I’ve had no luck whatsoever scanning glass negatives on my Epson 3200. Any tips?

formerly known as ‘cat arranger’ wrote:

"Bob Williams" wrote in message
:
:
: Mike Russell wrote:
: > Nidhal wrote:
: >
: >>I’m scanning a bunch of family pics for storage on my computer. My : >>main question is about DPI. I’ve been scanning the photos at 1200 : >>DPI, because that is the maximum that my scanner will allow. It’s a : >>lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI : >>because I want to retain the best quality possible. Anyway, I’m just : >>looking for advice as to whether that is overkill or not. The photos : >>I’m scanning are a pretty even split of color and black & white, : >>should that matter. Also, some of the photos are 4" x 6" or larger : >>but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s : >>an added benefit to scanning at 1200 DPI, I’ll continue to do so. : >>Just thought I’d ask for some advice. Thanks in advance for any help : >>you can pass along.
: >
: >
: > 1200 ppi is probably not overkill for the 3×5’s. They are contact prints
: > that may contain a lot of detail. Enlarging hese images can be fun because
: > no one has seen them at that size before.
: >
: > The run of the mill 4×6 snapshots are probably enlargements from the 60’s
: > and later. There is probably no benefit to scanning them at more than 300
: > ppi. You can verify this for yourself by picking out one of the sharpest
: > 4×6’s and scanning it at several resolutions. If the image looks blurry at
: > 300 ppi, there is no point going to a higher resolution. : >
: > Save time and space by scanning your black and white images in grayscale,
: > rather than going to a lower resolution. Storage is cheap, but if space is
: > an issue, you will get more quality if you use jpeg compression rather than
: > going to a lower ppi.
: >
: > If you haven’t already, check out Wayne Fulton’s www.scantips.com for good
: > info on scanning issues.
:
: Wayne Fulton usually advises people to scan Prints at 300 dpi because : commercial prints do not contain any more information than that. The : paper on which prints are made does not have the fine grain that : negatives and slides do. If you scan ordinary prints at any higher : resolution you will just be getting a bigger file but no additional : image information.
: If you want to enlarge the image (larger than the original) just : resample it in photoshop or your photo editor.
: If you DEFINITELY do not want to make prints from the scanned image but : only want to view it on a computer or TV screen, you could even get by : with scanning your 3x5s at 200 dpi. That would give you a 600 x 1000 : pixel image, just about right if your computer monitor is set to display : at 600 x 800 pixels.
: My recommendation is: Scan at 300 dpi and save the image in highest : quality jpeg. Then burn the images to a CD. From there you can play the : CD on your TV via a DVD player, create a slide show on your computer, : or print enlargements on your inkjet.
: Bob Williams

I disagree. You can always downsize the file but you can’t upsize it without some loss compared to the original scan at higher resolution. This is especially true on small photos.

True, that upsampling does not add new information to the image file.
[Otherwise everybody would shoot pictures at 640 x 480 pixels (0.3 MP)
and resample to 2560 x 1920 (4.9MP) to print gorgeous 8 x 10s]. However, the bicubic resampling algorithm in Photoshop and other good photo editors is very competent as you might expect.
A typical observer would see no difference at all between two 8×10 prints made from the same 4×6 original:
1) One scanned at 300 dpi (really ppi) and resampled to 600 ppi in PS
2) One scanned at 600 dpi and printed with no resampling.
And unless you intend to PRINT these images, any scanning above 300 dpi is a total non-issue.
Bob Williams

Really, there should be a kind of matrix with the factors listed, and the idea of Pleadge to experiment is probably the definitive statement.

jjs wrote:

"Mike Russell" wrote in message

This is correct. There is nothing magic about 300 ppi. In particular, I have scanned glossy contact prints from the 40’s and earlier that had incredible amounts of detail, as well as some recent tintypes, which are negatives in their own right.

Mike, while we are near the subject, I’ve had no luck whatsoever scanning glass negatives on my Epson 3200. Any tips?

I do recall a discussion by Ansel Adams in his biography. In the mid 50’s Adams got the rather daunting job of printing Arnold Genthe’s images of the 1906 SF earthquake. The emulsions of that time were extermely contrasty, designed for printing out paper. This process has a tremendously flat toe because as the exposed paper darkens, light is absorbed, effectively reducing the sensitivity of the emusion for the darker areas of the image. I believe Adams used No 0 paper combined with a different developer Here is one of the prints from the Eastman collection:

http://www.geh.org/taschen/htmlsrc7/m197300430001_ful.html

Contrast that with this image, from roughly the same time period, typical of images printed with no allowance for this contrast difference. The quartertone and halftone are ok, but the brighter parts of the image – the center of interest – are blocked.
http://www.geh.org/taschen/htmlsrc7/m198130510748_ful.html#t opofimage

Back to your point, though, about scanning such an image (if that is indeed the problem you are facing!). I would start by scanning two or three different exposures, then use a combination of curve adjustment layers, and layer blending modes to combine each plate. Once you have this set up, scanning each additional image should be relatively straightforward. —

Mike Russell
www.curvemeister.com
www.geigy.2y.net

"Mike Russell" wrote in message

http://www.geh.org/taschen/htmlsrc7/m197300430001_ful.html

That man was a great printer. NB I used to be a very good printer, or at least good enough to be contracted by the best B&W custom printer in the country. But that was a long time ago.

My problem is not with contrast – the photographer was highly advanced and made great negatives. No kidding. A real find. The problem is placing the plates in the scanner. If I put them flat on the glass, I get nothing but noise. It is very odd. This weekend I may make a special holder. It will be a challenge. I will ask again if I fail again.

Until then, I’ll continue to print them conventionally.

I’ve had success scanning 60-year-old 3-color (carbro) separations by simply placing them emulsion side down on the glass in my Epson 4870.

How about telling us a little bit more about what you’re doing. For instance: Are you trying to scan *through* the glass plate (IMHO: Bad Idea), or are you trying to scan emulsion side down?

"jjs" <john&#064;xstafford.net> wrote in message

"Mike Russell" wrote in message

http://www.geh.org/taschen/htmlsrc7/m197300430001_ful.html

That man was a great printer. NB I used to be a very good printer, or at least good enough to be contracted by the best B&W custom printer in the country. But that was a long time ago.

My problem is not with contrast – the photographer was highly advanced

and

made great negatives. No kidding. A real find. The problem is placing the plates in the scanner. If I put them flat on the glass, I get nothing but noise. It is very odd. This weekend I may make a special holder. It will

be

a challenge. I will ask again if I fail again.

Until then, I’ll continue to print them conventionally.

"Mike Russell" wrote in message
: formerly known as ‘cat arranger’ wrote:
: > "Johan W. Elzenga" wrote in message
: > : >> formerly known as ‘cat arranger’
: >> wrote:
: >>
: >>>> If you scan photos (meaning prints), anything higher than 600 dpi : >>>> is just overkill. You get more pixels and more megabytes, but not : >>>> more detail.
: >>>
: >>> Depends on the scanner, no?
: >>
: >> No, that is exactly my point. It does NOT depend on the scanner. The : >> resolution of a photographic print is not higher than 600 dpi, : >> normally it’s about 300 to 400 dpi. So, no matter what scanner (or : >> what negative was used to produce the print!), it’s useless to scan : >> at very high resolution. What isn’t there is the first place, cannot : >> be captured by increasing the scan resolution. : >>
: >>
: >> —
: >> Johan W. Elzenga johan<<at>>johanfoto.nl : >> Editor / Photographer http://www.johanfoto.nl/ : >
: > A photograph is continuous tone. There are no dpi. You must : > be talking about a print from computer or a half tone and not : > an actual photo. And as far as transparencies go, they can be : > scanned at several thousand samples/inch, may be higher. I has : > been a while since I did this kind of work but I’m pretty sure. :
: This is correct. There is nothing magic about 300 ppi. In particular, I : have scanned glossy contact prints from the 40’s and earlier that had : incredible amounts of detail, as well as some recent tintypes, which are : negatives in their own right.
: —
:
: Mike Russell
: www.curvemeister.com
: www.geigy.2y.net
:
:

The one thing about 300 dpi scans is that it used to be the highest optical resolution on scanners and anything higher was done by software which was not as good as Photoshop, so it didn’t make sense to scan any higher than the actual resolution of the scanner. Some then came out with 600×300 which kind of made me wonder about what difference that made. I guess you could scan at 600 x 600 and then reduce the vertical, it was vertical wasn’t it?, by half and then interpolate in Photoshop… anyway that’s a moot point now.

Again, though, the best idea I’ve heard so far is to experiment. And not just with resolution. The oldest apple scanners used to be able to let you select areas that were scanned at highest resolution than the rest of the source… weird, I guess they interpolated the areas that weren’t designated as important… anyway the black and white scan of photos had a diffusion pattern that made some beautiful bitmaps of grayscale that I’ve never been able to duplicate… just experimenting. Good luck to all : -)

"RSD99" wrote in message
: I’ve had success scanning 60-year-old 3-color (carbro) separations by : simply placing them emulsion side down on the glass in my Epson 4870. :
: How about telling us a little bit more about what you’re doing. For : instance: Are you trying to scan *through* the glass plate (IMHO: Bad : Idea), or are you trying to scan emulsion side down?
:
:
:
:
:
:
: "jjs" <john&#064;xstafford.net> wrote in message : : >
: > "Mike Russell" wrote in message
: > : >
: > > http://www.geh.org/taschen/htmlsrc7/m197300430001_ful.html : >
: > That man was a great printer. NB I used to be a very good printer, or at : > least good enough to be contracted by the best B&W custom printer in the : > country. But that was a long time ago.
: >
: > My problem is not with contrast – the photographer was highly advanced : and
: > made great negatives. No kidding. A real find. The problem is placing the
: > plates in the scanner. If I put them flat on the glass, I get nothing but
: > noise. It is very odd. This weekend I may make a special holder. It will
: be
: > a challenge. I will ask again if I fail again.
: >
: > Until then, I’ll continue to print them conventionally.

I have no experience with the kind of high quality stuff you are working with but would a high resolution camera mounted to take small areas at a time work. I’m doing that with a lot of photos and it really speeds things up. I imagine that you could do a bunch of pics from each and stitch them together… BTW, what happened to Bill Gates project of scanning fine art? Did he make it into a money making proposition or was it part of his charitable work? : -)

"Bob Williams" wrote in message
:
:
: formerly known as ‘cat arranger’ wrote:
: > "Bob Williams" wrote in message
: > : > :
: > :
: > : Mike Russell wrote:
: > : > Nidhal wrote:
: > : >
: > : >>I’m scanning a bunch of family pics for storage on my computer. My : > : >>main question is about DPI. I’ve been scanning the photos at 1200 : > : >>DPI, because that is the maximum that my scanner will allow. It’s a : > : >>lot slower than if I scan at 600 DPI, but I’ve stuck with 1200 DPI : > : >>because I want to retain the best quality possible. Anyway, I’m just
: > : >>looking for advice as to whether that is overkill or not. The photos
: > : >>I’m scanning are a pretty even split of color and black & white, : > : >>should that matter. Also, some of the photos are 4" x 6" or larger : > : >>but the overwhelming majority are 3×5 or smaller. Anyhow, if there’s
: > : >>an added benefit to scanning at 1200 DPI, I’ll continue to do so. : > : >>Just thought I’d ask for some advice. Thanks in advance for any help
: > : >>you can pass along.
: > : >
: > : >
: > : > 1200 ppi is probably not overkill for the 3×5’s. They are contact : > prints
: > : > that may contain a lot of detail. Enlarging hese images can be fun : > because
: > : > no one has seen them at that size before.
: > : >
: > : > The run of the mill 4×6 snapshots are probably enlargements from the : > 60’s
: > : > and later. There is probably no benefit to scanning them at more than
: > 300
: > : > ppi. You can verify this for yourself by picking out one of the : > sharpest
: > : > 4×6’s and scanning it at several resolutions. If the image looks blurry
: > at
: > : > 300 ppi, there is no point going to a higher resolution. : > : >
: > : > Save time and space by scanning your black and white images in : > grayscale,
: > : > rather than going to a lower resolution. Storage is cheap, but if space
: > is
: > : > an issue, you will get more quality if you use jpeg compression rather
: > than
: > : > going to a lower ppi.
: > : >
: > : > If you haven’t already, check out Wayne Fulton’s www.scantips.com for
: > good
: > : > info on scanning issues.
: > :
: > : Wayne Fulton usually advises people to scan Prints at 300 dpi because : > : commercial prints do not contain any more information than that. The : > : paper on which prints are made does not have the fine grain that : > : negatives and slides do. If you scan ordinary prints at any higher : > : resolution you will just be getting a bigger file but no additional : > : image information.
: > : If you want to enlarge the image (larger than the original) just : > : resample it in photoshop or your photo editor.
: > : If you DEFINITELY do not want to make prints from the scanned image but
: > : only want to view it on a computer or TV screen, you could even get by : > : with scanning your 3x5s at 200 dpi. That would give you a 600 x 1000 : > : pixel image, just about right if your computer monitor is set to display
: > : at 600 x 800 pixels.
: > : My recommendation is: Scan at 300 dpi and save the image in highest : > : quality jpeg. Then burn the images to a CD. From there you can play the
: > : CD on your TV via a DVD player, create a slide show on your computer, : > : or print enlargements on your inkjet.
: > : Bob Williams
: >
: >
: > I disagree. You can always downsize the file but you can’t upsize it : > without some loss compared to the original scan at higher resolution. : > This is especially true on small photos.
:
: True, that upsampling does not add new information to the image file. : [Otherwise everybody would shoot pictures at 640 x 480 pixels (0.3 MP) : and resample to 2560 x 1920 (4.9MP) to print gorgeous 8 x 10s]. : However, the bicubic resampling algorithm in Photoshop and other good : photo editors is very competent as you might expect.
: A typical observer would see no difference at all between two 8×10 : prints made from the same 4×6 original:
: 1) One scanned at 300 dpi (really ppi) and resampled to 600 ppi in PS : 2) One scanned at 600 dpi and printed with no resampling. : And unless you intend to PRINT these images, any scanning above 300 dpi : is a total non-issue.
: Bob Williams

I agree with your terminilogy but someone in another thread convinced me to use spi (samples per inch) for scans. I like the idea of having different
designations for monitor, print, and scans.

It may be absolutely true that no one could tell the difference between a scan
at 300 dpi and one at 600.. you know it used to be that you scanned at 1 1/2 the line screen which was at 200 in the best of magazines, so 300 was plenty.
I’m not so sure about the linescreen but it’s in the ball park. But again, and I
don’t want to be dogmatic (but this is one of the few things I know anything about : – ) there is the considreation of upsizing, especially smaller pictures.
: – )

: >
: > Really, there should be a kind of matrix with the factors listed, and : > the idea of Pleadge to experiment is probably the definitive statement. : >
: >
:

"formerly known as ‘cat arranger’" asked:
"…
BTW, what happened
to Bill Gates project of scanning fine art? Did he make it into a money making proposition or was it part of his charitable work? …."

Corbis

"RSD99" wrote:

BTW, what happened
to Bill Gates project of scanning fine art? Did he make it into a money making proposition or was it part of his charitable work?

You didn’t hear? The moron jumped right into Production Mode and tried to auto-sheet feed paintings into the super-scanner.

Jeez, what trival reply to a logical question!

On Thu, 30 Dec 2004 18:53:20 -0600, "jjs" wrote:

"RSD99" wrote:

BTW, what happened
to Bill Gates project of scanning fine art? Did he make it into a money making proposition or was it part of his charitable work?

You didn’t hear? The moron jumped right into Production Mode and tried to auto-sheet feed paintings into the super-scanner.

formerly known as ‘cat arranger’
wrote:

A photograph is continuous tone. There are no dpi. You must be talking about a print from computer or a half tone and not an actual photo. And as far as transparencies go, they can be scanned at several thousand samples/inch, may be higher. I has been a while since I did this kind of work but I’m pretty sure.

A photograph is indeed continuous tone, but the fact remains that the detail can be described in line pairs per millimeter. Scanning at more than twice that resolution seems useless. The larger the print, the lower the resolution, because the optical resolution of the enlarger lens wil not increase when the projection distance increases.

It also depends on the source. Todays photos from 1 hour labs are almost all printed digitally, even if they are printed from negatives. Contact prints from large negatives are different, as are transparencies. But they do have a resolution as well (be it higher).

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

Johan W. Elzenga wrote:

formerly known as ‘cat arranger’
wrote:

A photograph is continuous tone. There are no dpi. You must be talking about a print from computer or a half tone and not an actual photo. And as far as transparencies go, they can be scanned at several thousand samples/inch, may be higher. I has been a while since I did this kind of work but I’m pretty sure.

A photograph is indeed continuous tone, but the fact remains that the detail can be described in line pairs per millimeter. Scanning at more than twice that resolution seems useless. The larger the print, the lower the resolution, because the optical resolution of the enlarger lens wil not increase when the projection distance increases.
It also depends on the source. Todays photos from 1 hour labs are almost all printed digitally, even if they are printed from negatives. Contact prints from large negatives are different, as are transparencies. But they do have a resolution as well (be it higher).

Just adding some URLs to confirm what I’m syaing:

http://www.epi-centre.com/reports/9607cs.html
"To get real detail we need a scanner with good optical resolution. Bear in mind, however, that a photographic print only holds a maximum of 200 line pairs of resolution per inch, so a 300ppi scanner would be adequate, 600ppi more than adequate."

http://www.pcplus.co.uk/tutorials/default.asp?articleid=1969 4&page=2&pag etypeid=4&subsectionid=375
"No photographic print offers ‘infinite’ resolution. In practice, even good photo lab prints offer only about 300dpi-worth of resolution. In other words, you won’t get any more detail out of them by scanning them at 600dpi – you’ll just get bigger files."

http://64.233.183.104/search?q=cache:Ld4VtlZMdToJ:classes.ul eth.ca/20030 1/fa2020a/PDF/Jan13-16.pdf+%22optical+resolution%22+%22photo graphic+prin t%22&hl=en
" If a photographic print will not yield any more information beyond 300 spi, then why would anyone want a scanner that scans at 2400 spi?"

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

"Johan W. Elzenga" wrote in message
SNIP

A photograph is indeed continuous tone, but the fact remains that the detail can be described in line pairs per millimeter. Scanning at more than twice that resolution seems useless.

Correct, the so-called Nyquist limit dictates that one needs to sample at twice the spatial frequency of the finest detail that needs to be reliably resolved. It therefore boils down to the question; "how much resolution is present in the photograph"?

The larger the print, the lower the resolution, because the optical resolution of the enlarger lens wil not increase when the projection distance increases.

It also depends on the source. Todays photos from 1 hour labs are almost all printed digitally, even if they are printed from negatives.

Correct, and printing resolution is limited to 300-400 ppi, so scanning those requires 600-800ppi tops, assuming the paper used can resolve such detail. 600 ppi seems adequate for the vast majority of continuous tone images available.

Contact prints from large negatives are different, as are transparencies. But they do have a resolution as well
(be it higher).

Yes, they can exceed resolutions of 80 lp/mm (camera lens + film limiting resolution), which would require more than 4000 ppi to capture, but there is no paper that could provide such resolution. Therefore the question in the special case of contact prints, is more about the resolution limits of the photographic paper.

Some more background for the technically inclined can be found at: <http://www.normankoren.com/Tutorials/MTF3.html>, although the tests on that page were performed on an inkjet print, which has higher resolution than a minilab print, and only measures horizontal/vertical resolution (diagonal resolution can be higher!). For those considering to print the targets on that page, do note that not all inkjet printers produce the same resolution.

It is possible to derive the resolution of a print, by printing this target: <http://www.xs4all.nl/~bvdwolf/main/downloads/JTF120cy.jpg> Use a minilab/lightjet/lambda or an inkjet printer at its highest native resolution to print, and then evaluate the print. The target’s center will blur into an unresolved disk like shape (or elliptical one if the resolution is asymmetrical), from which the resolution can be calculated. The formula to use is: 120 / pi / diameter = cycles/mm, where pi=3.1415… and the diameter of the unresolved center is measured in millimetres (you may need a loupe and a caliper or other accurate measuring device). Cycles/mm is roughly speaking identical to lp/mm for this purpose.
For a contact print or conventional enlarger test, one can use a high quality film-writer or print on an inkjet printer at 600/720 ppi and photograph that from a distance of, say, 50x the focal length.

Bart

"Johan W. Elzenga" wrote in message

http://www.epi-centre.com/reports/9607cs.html
"To get real detail we need a scanner with good optical resolution. Bear in mind, however, that a photographic print only holds a maximum of 200 line pairs of resolution per inch, so a 300ppi scanner would be adequate, 600ppi more than adequate."

Of course I disagree.

The transition from lp/mm to ppi is slippery and requires clarification. Note that while the human eye has been ‘measured’ to count only 6 to 8 lp/mm it can discriminate sharpness (accutance) of much higher resolution, for example a person with experienced expectations and professional experience can sense the superiority of 1200 lp/mm over 300 lp/mm (but not count the lines.)

But the important point omitted in your quotation above is the _degree of enlargement_. A 35mm film scanned at the (nominal) 600ppi will only enlarge adequately to about 2" x 3", which is rather small, so more is better – if you can get more. Drum scans can get more sharp pixels, but amateurs don’t often use drum scanners. And they don’t care.

Finally, it is highly unlikely that anyone here has a camera, lens and technique that can achieve 200 lp/mm on the film plane, and conventional printing knocks down the combined lens resolution of the taking/printing lenses to about 70% of the mean. Further, the average photographer doesn’t use a tripod and it has been shown that they achieve as little as 40% of their actual sharpness potential on a regular basis due to camera movement and focusing errors. The average digital mavens fare even worse due to low shutter speeds, poor lenses, filtration aberations, autofocus errors, and especially lower expectations set by screen resolution limits – in other words, most have never even seen a sharp image and think they are doing just fine.

"Bart van der Wolf" wrote in message

"Johan W. Elzenga" wrote in message
SNIP

A photograph is indeed continuous tone, but the fact remains that the detail can be described in line pairs per millimeter. Scanning at more than twice that resolution seems useless.

Correct, the so-called Nyquist limit dictates that one needs to sample at twice the spatial frequency of the finest detail that needs to be reliably resolved. It therefore boils down to the question; "how much resolution is present in the photograph"?

Yes, there is the Nyquist limit but there are other factors that influence perceived sharpness. For example, adjacency effects combined with the nature of the subject and lighting. Round, gently contoured figures won’t give the impression of sharpness that the same hardware might show with contrasty, fine subject objects.

Correct, and printing resolution is limited to 300-400 ppi, so scanning those requires 600-800ppi tops, assuming the paper used can resolve such detail. 600 ppi seems adequate for the vast majority of continuous tone images available.

One must include the degree of enlargement.

Therefore the question in the special case of contact prints, is more about the resolution limits of the photographic paper.

We don’t have a lot of people shooting 8×10 film.

It is possible to derive the resolution of a print, by printing this target: <http://www.xs4all.nl/~bvdwolf/main/downloads/JTF120cy.jpg>

I don’t see any optical test charts hanging on gallerywalls. (Seriously, that’s an probably a better test than the military high-rez targets for many real-life images.)

jjs wrote:

"Johan W. Elzenga" wrote in message

http://www.epi-centre.com/reports/9607cs.html
"To get real detail we need a scanner with good optical resolution. Bear in mind, however, that a photographic print only holds a maximum of 200 line pairs of resolution per inch, so a 300ppi scanner would be adequate, 600ppi more than adequate."

Of course I disagree.

The transition from lp/mm to ppi is slippery and requires clarification. Note that while the human eye has been ‘measured’ to count only 6 to 8 lp/mm it can discriminate sharpness (accutance) of much higher resolution, for example a person with experienced expectations and professional experience can sense the superiority of 1200 lp/mm over 300 lp/mm (but not count the lines.)

But the important point omitted in your quotation above is the _degree of enlargement_. A 35mm film scanned at the (nominal) 600ppi will only enlarge adequately to about 2" x 3", which is rather small, so more is better – if you can get more. Drum scans can get more sharp pixels, but amateurs don’t often use drum scanners. And they don’t care.

Finally, it is highly unlikely that anyone here has a camera, lens and technique that can achieve 200 lp/mm on the film plane, and conventional printing knocks down the combined lens resolution of the taking/printing lenses to about 70% of the mean. Further, the average photographer doesn’t use a tripod and it has been shown that they achieve as little as 40% of their actual sharpness potential on a regular basis due to camera movement and focusing errors. The average digital mavens fare even worse due to low shutter speeds, poor lenses, filtration aberations, autofocus errors, and especially lower expectations set by screen resolution limits – in other words, most have never even seen a sharp image and think they are doing just fine.

So, if I understand you correctly, the first say you DISAGREE, and then write a long essay which only confirms what I said?

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

"Johan W. Elzenga" wrote in message

jjs wrote:
[…]
So, if I understand you correctly, the first say you DISAGREE, and then write a long essay which only confirms what I said?

I disagree with the sense of the thread evinced by what you said here, "so a 300ppi scanner would be adequate, 600ppi more than adequate."

jjs wrote:

Of course I disagree.

The transition from lp/mm to ppi is slippery and requires clarification. Note that while the human eye has been ‘measured’ to count only 6 to 8 lp/mm it can discriminate sharpness (accutance) of much higher resolution, for example a person with experienced expectations and professional experience can sense the superiority of 1200 lp/mm over 300 lp/mm (but not count the lines.)

It would be interesting to see any proof of this statement. AFAIK, there is no photographic paper in the world that can record 1200 lp/mm or come even close to that, so I cannot see how you can claim what you just wrote. Remember, we are talking about PRINTS, not film.

But the important point omitted in your quotation above is the _degree of enlargement_. A 35mm film scanned at the (nominal) 600ppi will only enlarge adequately to about 2" x 3", which is rather small, so more is better – if you can get more. Drum scans can get more sharp pixels, but amateurs don’t often use drum scanners. And they don’t care.

We are not talking about the scan resolution of a negative, we are talking about the scan resolution of a PRINT. Because a print is already an enlargement of your negative, a 10x enlargement (A4 size print) automatically means that you have only 10% of the resolution of the original negative, even if the enlarging process itself would not cause any further loss of resolution (which it does). Let’s say that a negative should be scanned at 4000 ppi to get all the detail, then an A4 size print of this negative should be scanned at 400 ppi to get the same.

Finally, it is highly unlikely that anyone here has a camera, lens and technique that can achieve 200 lp/mm on the film plane, and conventional printing knocks down the combined lens resolution of the taking/printing lenses to about 70% of the mean. Further, the average photographer doesn’t use a tripod and it has been shown that they achieve as little as 40% of their actual sharpness potential on a regular basis due to camera movement and focusing errors. The average digital mavens fare even worse due to low shutter speeds, poor lenses, filtration aberations, autofocus errors, and especially lower expectations set by screen resolution limits – in other words, most have never even seen a sharp image and think they are doing just fine.

Which only confirms my statement that scanning PRINTS of these lousy cameras at 600 ppi is more than enough to get all the detail.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

jjs wrote:

"Johan W. Elzenga" wrote in message

jjs wrote:
[…]
So, if I understand you correctly, the first say you DISAGREE, and then write a long essay which only confirms what I said?

I disagree with the sense of the thread evinced by what you said here, "so a 300ppi scanner would be adequate, 600ppi more than adequate."

I cancelled above message, but apparently not quick enough. Don’t confuse my own statements with quotations, though. I never said that, I just quoted it from a web page about scannning prints. The reason for quoting it was not this particular phrase, but the optical resolution of photographic paper which was mentioned as well.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

"Johan W. Elzenga" wrote in message

jjs wrote:

Of course I disagree.

The transition from lp/mm to ppi is slippery and requires clarification. Note that while the human eye has been ‘measured’ to count only 6 to 8 lp/mm
it can discriminate sharpness (accutance) of much higher resolution, for example a person with experienced expectations and professional experience
can sense the superiority of 1200 lp/mm over 300 lp/mm (but not count the lines.)

It would be interesting to see any proof of this statement.

I screwed up there, Johan. I should have been speaking of nominal ppi as you find in printers.

But the important point omitted in your quotation above is the _degree of
enlargement_. A 35mm film scanned at the (nominal) 600ppi will only enlarge
adequately to about 2" x 3", which is rather small, so more is better – if
you can get more. Drum scans can get more sharp pixels, but amateurs don’t
often use drum scanners. And they don’t care.

We are not talking about the scan resolution of a negative, we are talking about the scan resolution of a PRINT. Because a print is already an enlargement of your negative, a 10x enlargement (A4 size print) automatically means that you have only 10% of the resolution of the original negative, even if the enlarging process itself would not cause any further loss of resolution (which it does). Let’s say that a negative should be scanned at 4000 ppi to get all the detail, then an A4 size print of this negative should be scanned at 400 ppi to get the same.

Okay. You are saying, for example, a scan of an 8×10 print. I was speaking of scanning negatives. Sorry for the misunderstanding.

Which only confirms my statement that scanning PRINTS of these lousy cameras at 600 ppi is more than enough to get all the detail.

Yes indeed.

"Johan W. Elzenga" wrote in message

jjs wrote:

I disagree with the sense of the thread evinced by what you said here, "so a
300ppi scanner would be adequate, 600ppi more than adequate."

I cancelled above message, but apparently not quick enough. Don’t confuse my own statements with quotations, though. I never said that, I just quoted it from a web page about scannning prints. The reason for quoting it was not this particular phrase, but the optical resolution of photographic paper which was mentioned as well.

Understood. I am also sorry for the confusion.

Very Best,
John

On Thu, 30 Dec 2004 15:30:19 -0800, "formerly known as ‘cat arranger’" wrote:

BTW, what happened
to Bill Gates project of scanning fine art? Did he make it into a money making proposition or was it part of his charitable work? : -)

Corbis 🙂

—

Hecate – The Real One

veni, vidi, reliqui

"Bart van der Wolf" wrote in message
:
: "Johan W. Elzenga" wrote in message
: : SNIP
: > A photograph is indeed continuous tone, but the fact remains : > that the detail can be described in line pairs per millimeter. : > Scanning at more than twice that resolution seems useless.

I’ve never heard this before and doubt that it comes into play in actual printing. I would appreciate being educated more about it.

: Correct, the so-called Nyquist limit dictates that one needs to sample : at twice the spatial frequency of the finest detail that needs to be : reliably resolved. It therefore boils down to the question; "how much : resolution is present in the photograph"?

: > The larger the print, the lower the resolution, because the : > optical resolution of the enlarger lens wil not increase when : > the projection distance increases.
: >
: > It also depends on the source. Todays photos from 1 hour : > labs are almost all printed digitally, even if they are printed : > from negatives.
:
: Correct, and printing resolution is limited to 300-400 ppi, so : scanning those requires 600-800ppi tops, assuming the paper used can : resolve such detail. 600 ppi seems adequate for the vast majority of : continuous tone images available.

Printing, at least commercial printing is not done in ppi, but dpi and is really measured in line screen which is called halftone. It used to be that newspapers were about the 70-80 line screen but now are probably more than 100. According to http://dx.sheridan.com/advisor/line_screen.html a 300-600 dpt laser printer can put out a line screen lower quality than the old newspaper standard… for gray scale. That doesn’t seem right to me for 600 but is correct to my experice with 300 dpi.

Color is printed in 4 colors and the angle of the printing of each color is different and is beyond me in that respect…

: > Contact prints from large negatives are different, as are : > transparencies. But they do have a resolution as well : > (be it higher).
:
: Yes, they can exceed resolutions of 80 lp/mm (camera lens + film : limiting resolution), which would require more than 4000 ppi to : capture, but there is no paper that could provide such resolution. : Therefore the question in the special case of contact prints, is more : about the resolution limits of the photographic paper.

: -) what are lp/mm ?

: Some more background for the technically inclined can be found at: : <http://www.normankoren.com/Tutorials/MTF3.html>, although the tests : on that page were performed on an inkjet print, which has higher : resolution than a minilab print, and only measures horizontal/vertical : resolution (diagonal resolution can be higher!). For those considering : to print the targets on that page, do note that not all inkjet : printers produce the same resolution.

Inkjet printer use a different system than printers. I guess part of the discussion here is a matter of terms and whether we are talking about printing presses and the printers who run them as opposed to the printers that most of us have now. I don’t know how the newer ‘printers’ work… Do they also use patterns printed at distinct angles? Probably no because of the way the printheads appear. Do they print different size dots? Does the ink mix? This is all aside from the OPs question about scanning but obviously the scan resolution might be affected by the target output device and we should be sure we are talking about the sam kind of ‘printer’.

: It is possible to derive the resolution of a print, by printing this : target: <http://www.xs4all.nl/~bvdwolf/main/downloads/JTF120cy.jpg> : Use a minilab/lightjet/lambda or an inkjet printer at its highest : native resolution to print, and then evaluate the print. The target’s : center will blur into an unresolved disk like shape (or elliptical one : if the resolution is asymmetrical), from which the resolution can be : calculated. The formula to use is: 120 / pi / diameter = cycles/mm, : where pi=3.1415… and the diameter of the unresolved center is : measured in millimetres (you may need a loupe and a caliper or other : accurate measuring device). Cycles/mm is roughly speaking identical to : lp/mm for this purpose.
: For a contact print or conventional enlarger test, one can use a high : quality film-writer or print on an inkjet printer at 600/720 ppi and : photograph that from a distance of, say, 50x the focal length. :
: Bart
:

formerly known as ‘cat arranger’
wrote:

"Bart van der Wolf" wrote in message
:
: "Johan W. Elzenga" wrote in message
: : SNIP
: > A photograph is indeed continuous tone, but the fact remains : > that the detail can be described in line pairs per millimeter. : > Scanning at more than twice that resolution seems useless.
I’ve never heard this before and doubt that it comes into play in actual printing. I would appreciate being educated more about it.

Who’s talking about printing? We are talking about SCANNING a (photographic, not inkjet) print.

: Correct, and printing resolution is limited to 300-400 ppi, so : scanning those requires 600-800ppi tops, assuming the paper used can : resolve such detail. 600 ppi seems adequate for the vast majority of : continuous tone images available.

Printing, at least commercial printing is not done in ppi, but dpi and is really measured in line screen which is called halftone.

Who’s talking about printing. We are talking about SCANNING a (photographic, not inkjet) print?

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

"Johan W. Elzenga" wrote in message
: formerly known as ‘cat arranger’
: wrote:
:
: > "Bart van der Wolf" wrote in message
: > : > :
: > : "Johan W. Elzenga" wrote in message
: > : : > : SNIP
: > : > A photograph is indeed continuous tone, but the fact remains : > : > that the detail can be described in line pairs per millimeter. : > : > Scanning at more than twice that resolution seems useless. : >
: > I’ve never heard this before and doubt that it comes into play in : > actual printing. I would appreciate being educated more about it. :
: Who’s talking about printing? We are talking about SCANNING a : (photographic, not inkjet) print.

I do not know how you can define line pairs in a grayscale image without introducing some kind of arbitray threshold level, but like I said, I’m not that knowledgable about photography. However to say that no one is talking about printing or pringers seems incorrect. Right below what I’m writing now is someone’s comment about printer resolution and its effect on the choice of original scan resolution…

Can you explain how you use line pairs and their measurement in your editing and photographic work?

: > : Correct, and printing resolution is limited to 300-400 ppi, so : > : scanning those requires 600-800ppi tops, assuming the paper used can : > : resolve such detail. 600 ppi seems adequate for the vast majority of : > : continuous tone images available.
: >
: > Printing, at least commercial printing is not done in ppi, but dpi and : > is really measured in line screen which is called halftone. :
: Who’s talking about printing. We are talking about SCANNING a : (photographic, not inkjet) print?

You have to know the target of the scan, ie what you are going to present it on to know what kind of resolution you need. As was mentioned before, if you are going for the web you will need a lot less resolution than if you are going to blow up a small photo for a world class magazine printed on a 4 color press.

: Johan W. Elzenga johan<<at>>johanfoto.nl : Editor / Photographer http://www.johanfoto.nl/

"formerly known as ‘cat arranger’"
wrote in message SNIP

Printing, at least commercial printing is not done in ppi, but dpi and
is really measured in line screen which is called halftone.

That is the difference between halftone printing and continuous tone printing. Continuous tone, which is what we’re talking about when discussing photos, can have any color (from a range within the dye/pigment gamut) at each pixel position. Halftone needs to simulate colors inbetween the given ink colors. Spotcolor can be used to get an exact color, but you can’t have 16 million spot colors in a process print image, where you (theoretically) can in a continuous tone image. SNIP

: -) what are lp/mm ?

Line pairs per millimetre (2 lines / mm).

SNIP

Inkjet printer use a different system than printers.

Yes, they use an advanced kind of stochastic dithering, and the better ones use 6 types of ink that gets mixed/blended AND stochastically displaced/dithered at a *very* high (like is used in typesetters) resolution. Minilabs can use photochemical paper, upon which (per pixel position, thus pixels per inch) either R/G/B laser output or the projection of a CRT image is mixed in the right proportion for "any" possible color in the dye’s gamuth. Other systems (LED, LVT, etc.) to project an image on photopaper are being used as well, but less common.

Bart

formerly known as ‘cat arranger’
wrote:

I do not know how you can define line pairs in a grayscale image without introducing some kind of arbitray threshold level, but like I said, I’m not that knowledgable about photography. However to say that no one is talking about printing or pringers seems incorrect. Right below what I’m writing now is someone’s comment about printer resolution and its effect on the choice of original scan resolution…

A photographic print is a real photograph, made on photographic paper using an old fashioned enlarger. There is no ‘printer resolution’ to deal with if you want to scan it, just the optical resolution of the image.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

"Johan W. Elzenga" wrote in message

[…]
A photographic print is a real photograph, made on photographic paper using an old fashioned enlarger. There is no ‘printer resolution’ to deal with if you want to scan it, just the optical resolution of the image.

Further, a conventional ‘wet darkroom’ print has had the original scene’s brightness range compressed, making it far easier to accomodate than scanning a negative or slide. I did a number of scans of 11×14" B&W prints done by a well published photographer and was almost crazy-easy compared to the usual jobs.

"Bart van der Wolf" wrote in message
:
: "formerly known as ‘cat arranger’"
: wrote in message : SNIP
: > Printing, at least commercial printing is not done in ppi, but dpi : > and
: > is really measured in line screen which is called halftone. :
: That is the difference between halftone printing and continuous tone : printing. Continuous tone, which is what we’re talking about when : discussing photos, can have any color (from a range within the : dye/pigment gamut) at each pixel position. Halftone needs to simulate : colors inbetween the given ink colors. Spotcolor can be used to get an : exact color, but you can’t have 16 million spot colors in a process : print image, where you (theoretically) can in a continuous tone image. : SNIP
: > : -) what are lp/mm ?
:
: Line pairs per millimetre (2 lines / mm).
:
: SNIP
: > Inkjet printer use a different system than printers. :
: Yes, they use an advanced kind of stochastic dithering, and the better : ones use 6 types of ink that gets mixed/blended AND stochastically : displaced/dithered at a *very* high (like is used in typesetters) : resolution. Minilabs can use photochemical paper, upon which (per : pixel position, thus pixels per inch) either R/G/B laser output or the : projection of a CRT image is mixed in the right proportion for "any" : possible color in the dye’s gamuth. Other systems (LED, LVT, etc.) to : project an image on photopaper are being used as well, but less : common.
:
: Bart

Interesting, sounds like some randomness/ fractal has been introduced into the printing process.

"Johan W. Elzenga" wrote in message
: formerly known as ‘cat arranger’
: wrote:
:
: > I do not know how you can define line pairs in a grayscale image : > without introducing some kind of arbitray threshold level, but like : > I said, I’m not that knowledgable about photography. However to : > say that no one is talking about printing or pringers seems incorrect. : > Right below what I’m writing now is someone’s comment about : > printer resolution and its effect on the choice of original scan : > resolution…
:
: A photographic print is a real photograph, made on photographic paper : using an old fashioned enlarger. There is no ‘printer resolution’ to : deal with if you want to scan it, just the optical resolution of the : image.
:
:
: —
: Johan W. Elzenga johan<<at>>johanfoto.nl : Editor / Photographer http://www.johanfoto.nl/

Did you snip that part that mentioned printer resolution?

formerly known as ‘cat arranger’
wrote:

: > I do not know how you can define line pairs in a grayscale image : > without introducing some kind of arbitray threshold level, but like : > I said, I’m not that knowledgable about photography. However to : > say that no one is talking about printing or pringers seems incorrect. : > Right below what I’m writing now is someone’s comment about : > printer resolution and its effect on the choice of original scan : > resolution…
:
: A photographic print is a real photograph, made on photographic paper : using an old fashioned enlarger. There is no ‘printer resolution’ to : deal with if you want to scan it, just the optical resolution of the : image.

Did you snip that part that mentioned printer resolution?

Yes, because I’mm trying to explain that a real photograph does not have a ‘printer resolution’. Apparently, in vain though (<sigh>…

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

"Johan W. Elzenga" wrote in message
: formerly known as ‘cat arranger’
: wrote:
:
: > : > I do not know how you can define line pairs in a grayscale image : > : > without introducing some kind of arbitray threshold level, but like : > : > I said, I’m not that knowledgable about photography. However to : > : > say that no one is talking about printing or pringers seems incorrect.
: > : > Right below what I’m writing now is someone’s comment about : > : > printer resolution and its effect on the choice of original scan : > : > resolution…
: > :
: > : A photographic print is a real photograph, made on photographic paper : > : using an old fashioned enlarger. There is no ‘printer resolution’ to : > : deal with if you want to scan it, just the optical resolution of the : > : image.
:
: > Did you snip that part that mentioned printer resolution? :
: Yes, because I’mm trying to explain that a real photograph does not have : a ‘printer resolution’. Apparently, in vain though (<sigh>… :
:
: —
: Johan W. Elzenga johan<<at>>johanfoto.nl : Editor / Photographer http://www.johanfoto.nl/

When photographers talk about graininess and how film that allows images taken in less light, is that analogous to resolution?

formerly known as ‘cat arranger’
wrote:

When photographers talk about graininess and how film that allows images taken in less light, is that analogous to resolution?

Indirectly yes, but not entirely. A grainier, more sensitive film usually has a lower resolution, but there is no one-to-one relation between grain and optical resolution AFAIK. That’s because there is no simple value for ‘graininess’.

—
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/