DPlingScan

[spanner]

I have been scanning some old phot's and saving them as Tiff's for editing in Lightroom or Photolab. On loading them they are only in 8bit. I have tried to find a way of saving them in 16bit as there is more information to edit with and have not found any thing. Is there a way of saving in16bit

[News]

Post by spanner
I have been scanning some old phot's and saving them as Tiff's for
editing in Lightroom or Photolab. On loading them they are only in
8bit. I have tried to find a way of saving them in 16bit as there
is more information to edit with and have not found any thing. Is
there a way of saving in16bit

In DPlngScan menu over image. Select Colour-Change format.

In the dbox that opens: for new format choose 16 or 24 bpp and for
palette leave it at RISC OS.

[M Harding]

Post by spanner
I have been scanning some old phot's and saving them as Tiff's for
editing in Lightroom or Photolab. On loading them they are only in
8bit. I have tried to find a way of saving them in 16bit as there
is more information to edit with and have not found any thing. Is
there a way of saving in16bit

If these are prints, I've found that scanning generally is less
effective than re-photographing the print. There seems to be more
realism in camera-sourced versions, more variations of light and
shade. (Thanks to whoever first passed on to me this tip.)

My technique has been to prop up the photo (in my case, on top of an
amplifier with CD player above it!) and have the camera [a Panasonic
TZ70] on a tiny tripod so that it's in line with the photo for ease
of copying many photos. Then having set the focus to close-up, the
shutter's set to delayed operation, so that I don't shake the camera
when it triggers.

Keeping the photo upright rather than horizontal seems best for me,
because otherwise tripods become difficult. I can sit at my desk while
I re-photo them!

I've used Photoshop Elements 15 very successfully with this technique.
A bit of cropping's usually needed. I don't often need to correct the
perspective. The result if you start off in RAW can be amazing.

Documents scanned - yes. Photos - I reckon a camera's better.

Michael Harding
Rev. Preb. M.D. Harding ***@mdharding.org.uk

[Harriet Bazley]

On 16 Jan 2022 as I do recall,

Post by spanner
I have been scanning some old phot's and saving them as Tiff's for
editing in Lightroom or Photolab. On loading them they are only in
8bit. I have tried to find a way of saving them in 16bit as there is
more information to edit with and have not found any thing. Is there a
way of saving in16bit

If you're scanning them from DPlngScan, there should be an option
(probably depending on the Twain drivers for your scanner) to select the
colour depth at that point: either black and white, 8bit, or in my case
24bpp (the drivers don't offer 16bit).

'Promoting' them to 16bit after they have been scanned won't create any
extra information as such.

[spanner]

Post by Harriet Bazley
On 16 Jan 2022 as I do recall,

Post by spanner
I have been scanning some old phot's and saving them as Tiff's for
editing in Lightroom or Photolab. On loading them they are only in
8bit. I have tried to find a way of saving them in 16bit as there is
more information to edit with and have not found any thing. Is there a
way of saving in16bit

If you're scanning them from DPlngScan, there should be an option
(probably depending on the Twain drivers for your scanner) to select the
colour depth at that point: either black and white, 8bit, or in my case
24bpp (the drivers don't offer 16bit).
'Promoting' them to 16bit after they have been scanned won't create any
extra information as such.
--
Harriet Bazley == Loyaulte me lie ==
Nostalgia isn't what it used to be.

I have tried colour change format 24 to 16 same outcome.
I have read Tiff are either 8bit or 16bit per channel with the 16bit having more data to play with.
This gives a better printed photo.
I do photo them most of the time but these are old and wont lie flat and sticking them down is not an option.
I use DPlingScan as it is better than the Epson's software. The scanner is a perfection 4870 photo with version 1.33.
I photograph most in raw and can change them to 16bit tiff in the software to use in some software that will not load raw files.

Philip May

[News]

Post by spanner
I have read Tiff are either 8bit or 16bit per channel with the
16bit having more data to play with.

This wasn't clear from your first post. There is a big difference
between an 8-bit image and an 8-bit per channel image. The latter is
what is generally referred to in RISC OS as a 24-bit image, i.e.
8-bits for each of the R,G, and B channels.

As far as I am aware, under RISC OS, (a) there is no way of getting a
scan using DPScan of more than 24-bit (8-bit per channel) and (b)
assuming you managed to obtain such an image of 48-bits (16-bit per
channel) there is no app available that could deal with such an image.

On other OSs images of more than 8-bits per channel can be handled.

[spanner]

Post by News

Post by spanner
I have read Tiff are either 8bit or 16bit per channel with the
16bit having more data to play with.

This wasn't clear from your first post. There is a big difference
between an 8-bit image and an 8-bit per channel image. The latter is
what is generally referred to in RISC OS as a 24-bit image, i.e.
8-bits for each of the R,G, and B channels.
As far as I am aware, under RISC OS, (a) there is no way of getting a
scan using DPScan of more than 24-bit (8-bit per channel) and (b)
assuming you managed to obtain such an image of 48-bits (16-bit per
channel) there is no app available that could deal with such an image.
On other OSs images of more than 8-bits per channel can be handled.
--
Chris Johnson

Thank you for clearing up that. Thinking 8bit 16bit 24bit was the same was overall.
For 16bit I will need 48bit (16bit per channel).
I will have to go to the windows side and use something like Vuescan. One less reason to use RiscOs.

Philip May

[David Higton]

Post by spanner

Post by News

I have read Tiff are either 8bit or 16bit per channel with the 16bit
having more data to play with.

This wasn't clear from your first post. There is a big difference
between an 8-bit image and an 8-bit per channel image. The latter is
what is generally referred to in RISC OS as a 24-bit image, i.e. 8-bits
for each of the R,G, and B channels.
As far as I am aware, under RISC OS, (a) there is no way of getting a
scan using DPScan of more than 24-bit (8-bit per channel) and (b)
assuming you managed to obtain such an image of 48-bits (16-bit per
channel) there is no app available that could deal with such an image.
On other OSs images of more than 8-bits per channel can be handled.
-- Chris Johnson

Thank you for clearing up that. Thinking 8bit 16bit 24bit was the same was
overall. For 16bit I will need 48bit (16bit per channel). I will have to
go to the windows side and use something like Vuescan. One less reason to
use RiscOs.

Good luck trying to tell the difference between 8 bits per colour and
any higher number.

David

[druck]

Post by David Higton

Post by spanner
Thank you for clearing up that. Thinking 8bit 16bit 24bit was the same was
overall. For 16bit I will need 48bit (16bit per channel). I will have to
go to the windows side and use something like Vuescan. One less reason to
use RiscOs.

Good luck trying to tell the difference between 8 bits per colour and
any higher number.

There is a very noticeable difference between 8 and 10 or 12 bits per
component, you can see it in grey scales and some parts of the colour
space are far less banded - BUT you not only have to have an image with
the additional information, but your graphics application, OS, graphics
card and monitor all have to handle greater than 8 bits per component,
and that tends to only be on a pro or semi-pro setup.

---druck

[Jim Lesurf]

Post by druck

Post by David Higton

Post by spanner
Thank you for clearing up that. Thinking 8bit 16bit 24bit was the
same was overall. For 16bit I will need 48bit (16bit per channel). I
will have to go to the windows side and use something like Vuescan.
One less reason to use RiscOs.

Good luck trying to tell the difference between 8 bits per colour and
any higher number.

There is a very noticeable difference between 8 and 10 or 12 bits per
component, you can see it in grey scales and some parts of the colour
space are far less banded - BUT you not only have to have an image with
the additional information, but your graphics application, OS, graphics
card and monitor all have to handle greater than 8 bits per component,
and that tends to only be on a pro or semi-pro setup.

Also, with 'real' images if you use a high resolution can tend to 'dither'
the values from sample to adjacent sample, and your eye can then average
over that.

Jim

[Jim Lesurf]

Post by David Higton
Good luck trying to tell the difference between 8 bits per colour and
any higher number.

My thought also. Provided I've chosen to set the gamma, etc, sensibly, I
get quite satisfactory results using a scanner and DPingScan. The only snag
being that sometimes getting these parameters right is more fiddly than
with xsane + gimp or similar.

Jim

[Sprow]

Post by spanner
I have read Tiff are either 8bit or 16bit per channel with the
16bit having more data to play with.

This wasn't clear from your first post [...]
As far as I am aware, under RISC OS, [...] there is no app available that could deal with such an image.

ChangeFSI can read 16b per plane Alpiar images, and perhaps more usefully 48bpp PNGs. Internally the RGB planes are manipulated with 28 fractional bits so no issues holding 16b planes. None of the output formats support 48bpp though so there's always a lossy final step,
Sprow.