IrfanView (or other): can I do this?

  • Thread starter Thread starter Terry Pinnell
  • Start date Start date
T

Terry Pinnell

I've recently started making DVD movies for showing on the family TV,
using a program called MemoriesOnTV. Typically they are holiday photos
with a sound track. The TV screen proportions, and the DVD settings
are 4:3. In practice I'm finding it difficult to avoid cutting off the
edges. It obviously helps if I get each image format to follow the 4:3
ratio too. Apart from trial and error (adding/removing borders and
reaching for the calculator at each stage), can anyone recommend a
quick way of achieving this please, from JPEG images of disparate size
and formats?
 
I've recently started making DVD movies for showing on the family TV,
using a program called MemoriesOnTV. Typically they are holiday photos
with a sound track. The TV screen proportions, and the DVD settings
are 4:3. In practice I'm finding it difficult to avoid cutting off the
edges. It obviously helps if I get each image format to follow the 4:3
ratio too. Apart from trial and error (adding/removing borders and
reaching for the calculator at each stage), can anyone recommend a
quick way of achieving this please, from JPEG images of disparate size
and formats?
Click to expand...

Yes in Irfanview using Create custom Selection from the edit
Menu: there you can set the 4:3 ratio on the right hand side and set
you own width or height. you can move the selection with the right
mouse button and alter its size, maintaining the ratio using the Ctrl
key. Forget the calculator
Personally I'd use another program, but that's not freeware,
but I know you have it. :-)
 
David Gilbert said:
Yes in Irfanview using Create custom Selection from the edit
Menu: there you can set the 4:3 ratio on the right hand side and set
you own width or height. you can move the selection with the right
mouse button and alter its size, maintaining the ratio using the Ctrl
key. Forget the calculator
Personally I'd use another program, but that's not freeware,
but I know you have it. :-)
Click to expand...

Excellent - thanks, David.

The other program? Do tell!
 
FWIW, I read on a graphics forum that when resizing images, it's better to
do it in several smaller increments rather that one big one, particularly
when making them bigger.

Since the proggie must interpolate, using smaller increments reduces the
error involved. No idea if it's true or not, but sounds good in theory.

M
 
Michael Laplante said:
FWIW, I read on a graphics forum that when resizing images, it's better to
do it in several smaller increments rather that one big one, particularly
when making them bigger.

Since the proggie must interpolate, using smaller increments reduces the
error involved. No idea if it's true or not, but sounds good in theory.

M
Click to expand...

I'd be interested in seeing some careful comparisons. I see the
rationale, but on the other hand it goes against the maxim I'd
understood was generally accepted, that there's degradation with
*every* edit.

In this case though, it's irrelevant. What I need to do is not resize,
but add a border - i.e. change canvas size. That way, the main image
gets displayed on the TV screen.
 
FWIW, I read on a graphics forum that when resizing images, it's better to
do it in several smaller increments rather that one big one, particularly
when making them bigger.

Since the proggie must interpolate, using smaller increments reduces the
error involved. No idea if it's true or not, but sounds good in theory.
Click to expand...
Sounds good but, in my opinion, the best way is to convert the image
to some lossless format, that is NOT jpeg, do the resizing and any
other tweaks necessary and then convert back to jpeg.
 
David said:
Sounds good but, in my opinion, the best way is to convert the image
to some lossless format, that is NOT jpeg, do the resizing and any
other tweaks necessary and then convert back to jpeg.
Click to expand...

Which format would be your recommendation? Is there any loss in info during
the conversion to/from this other format?

M
 
Sounds good but, in my opinion, the best way is to convert the image
to some lossless format, that is NOT jpeg, do the resizing and any
other tweaks necessary and then convert back to jpeg.
Click to expand...

I just shot off a reply without thinking a little more carefully about your
advice. Regardless of which format you are using, there will be some
inaccuracy that results from the resizing process. This has nothing to do
with format -- it's just that any interpolation will introduce errors. I
think the theory is that by doing it in smaller increments you decrease the
overall error rate rather than doing it in one big jump.

Aside from interpolation errors, there are also errors resulting from
compression. So, to that extent, I think your advice is still good. Probably
converting photos to bmp format, then resizing in a series of smaller
increments is probably the best thing to do. (That may introduce other
issues to overcome such as loss of exif data.)

Any further thoughts on this?

M
 
David said:
Sounds good but, in my opinion, the best way is to convert the image
to some lossless format, that is NOT jpeg, do the resizing and any
other tweaks necessary and then convert back to jpeg.
Click to expand...

Just got back from alt.html where, coincidentally, there is a thread re this
issue as well. According to one member there, it's his understanding that
any jpeg resizing involves first uncompressing the jpeg to bitmap format,
resizing, then re-converting to jpeg format. If that's true then the step
you describe would be unnecessary.

Anyone in this ng into digital imaging who could perhaps shed some light on
this issue?

M
 
Michael said:
Just got back from alt.html where, coincidentally, there is a thread re this
issue as well. According to one member there, it's his understanding that
any jpeg resizing involves first uncompressing the jpeg to bitmap format,
resizing, then re-converting to jpeg format. If that's true then the step
you describe would be unnecessary.

Anyone in this ng into digital imaging who could perhaps shed some light on
this issue?
Click to expand...

I can point you to a very helpful web site:

A few scanning tips by Wayne Fulton
http://www.scantips.com/

JPG Image File Format:

http://www.scantips.com/basics9j.html

Susan
--
Posted to alt.comp.freeware
Search alt.comp.freeware (or read it online):
http://google.ca/advanced_group_search?q=+group:alt.comp.freeware
Pricelessware & ACF: http://www.pricelesswarehome.org
Pricelessware: http://www.pricelessware.org (not maintained)
 
Michael said:
Just got back from alt.html where, coincidentally, there is a thread re this
issue as well. According to one member there, it's his understanding that
any jpeg resizing involves first uncompressing the jpeg to bitmap format,
resizing, then re-converting to jpeg format. If that's true then the step
you describe would be unnecessary.

Anyone in this ng into digital imaging who could perhaps shed some light on
this issue?

M
Click to expand...

The true test is to take an image and use both methods and then see if
you can tell the difference.
Stan
 
Which format would be your recommendation? Is there any loss in info during
the conversion to/from this other format?
Click to expand...
If the source is JPEG, loss is already present in the image. This
cannot be recovered.

Which format to be used depends on the number of colours in the source
and the program used for conversion. I would probably suggest PNG or
TIFF depending on available space. TIFF and BMP files are generally
huge. GIF has colour limitations. The native format of the conversion
program is generally lossless. eg. PSP for Paint Shop Pro, etc.

If you've obtained the image from a camera it should be resized before
the JPEG conversion. This will cut down on the initial definition
loss.
 
I just shot off a reply without thinking a little more carefully about your
advice. Regardless of which format you are using, there will be some
inaccuracy that results from the resizing process. This has nothing to do
with format -- it's just that any interpolation will introduce errors. I
think the theory is that by doing it in smaller increments you decrease the
overall error rate rather than doing it in one big jump.

Aside from interpolation errors, there are also errors resulting from
compression.
Click to expand...

GIF and PNG are lossless in compression. PNG has more colours
available. I think some others are also lossless but I can't think
which without having to look it up and I'm being lazy.
So, to that extent, I think your advice is still good. Probably
converting photos to bmp format, then resizing in a series of smaller
increments is probably the best thing to do. (That may introduce other
issues to overcome such as loss of exif data.)

Any further thoughts on this?
Click to expand...
Expanding will produce more errors than reduction because you have to
interpolate. Actually the program takes a look at what is around the
area and makes a reasonable assumption. It sometimes gets it wrong.
Doing it in smallish increments could allow other refinements, such as
"Sharpen" or "Blur", to be applied to the intervening image as well as
allowing manual editing,
 
Terry said:
I've recently started making DVD movies for showing on the family TV,
using a program called MemoriesOnTV. Typically they are holiday photos
with a sound track. The TV screen proportions, and the DVD settings
are 4:3. In practice I'm finding it difficult to avoid cutting off the
edges. It obviously helps if I get each image format to follow the 4:3
ratio too. Apart from trial and error (adding/removing borders and
reaching for the calculator at each stage), can anyone recommend a
quick way of achieving this please, from JPEG images of disparate size
and formats?
Click to expand...

Take a look at JPegCrops <http://ekot.dk/programmer/JPEGCrops/>

I use it all the time to do what you are doing. There is a "4:3
Monitor" setting already defined. Make it your default and you can
open and process a whole ton of pictures quickly.

Keep in mind that most digital cameras take pictures that are MUCH
bigger than you need for a DVD (MemoriesOnTV will actually reduce the
size as it works). You can crop a lot out without losing any quality.

Eric
 
Eric said:
Take a look at JPegCrops <http://ekot.dk/programmer/JPEGCrops/>

I use it all the time to do what you are doing. There is a "4:3
Monitor" setting already defined. Make it your default and you can
open and process a whole ton of pictures quickly.

Keep in mind that most digital cameras take pictures that are MUCH
bigger than you need for a DVD (MemoriesOnTV will actually reduce the
size as it works). You can crop a lot out without losing any quality.
Click to expand...
Thanks, Eric. You probably missed my follow-up. I later realised,
after trying some resizing in IV and other programs like PaintShop Pro
and Thumbs Plus, that I was mistaken in my goal. What I actually need
to do is not resize but change canvas size, i.e. add a border (or
borders). So then the lost area on the TV screen won't contain crucial
edges of the wanted image. Here's an example
http://www.terrypin.dial.pipex.com/Images/MOT2.gif
You can see that even with the 'TV safe area' set to 85% (losing a
whopping 15% of the content at each side), that heading would still
not be seen. And in fact on my actual TV it seems worse that 15% loss.
So I would in this case add a top border - particularly easy with
IrfanView.

It's a PITA that MemoriesOnTV itself has no clever facilities to do it
though.
 
Back
Top