Reducing picture size with same quality.

On Mon, 26 Mar 2012 08:16:35 +0100, "J. P. Gilliver (John)"
wrote:

In message , Peter Jason
writes:
[
I think I have fixed it by using the Irfanview "Resample" option. The
photo below is about 160KB, resampled down from the original of 1MB.
Note the effective maintaining of the caption at the lower left that
compression always trashes to a furry blur.

http://imageshack.us/photo/my-images...leirfanvi.jpg/

When I click on the magnifying glass, or just the image, on that, I get
something that looks as if it is trying to get me to register; without
that I can see that the caption is there, but not read it.

I know, Imageshack is not exactly superior; do you know of a better
one?


The original photo was a JPEG 5.01MB, then imported into PShop CS5,

Original being what the camera gave you, I take it. What is it in
pixels?

I check it out and repost.

cropped 6x4 (for wide-screen monitors), exposure adjusted and

Or even 3:2 (-:

Yes, or even 12x8; but the Dear People here go by the photo-kiosk
displays. :-))

sharpened. The resulting JPEG was 1MB and this Irfanview reduced to

(IrfanView can crop and sharpen by the way.) What was the 1MB image in
pixels?

I check it out and repost.

161KB. I emailed this 161KB jpeg size to myself as an "attachment"
and it returned the same size with no loss of quality. However if I

So it should. What size was the 161KB image in pixels?

I check it out and repost.

emailed this as a picture insertion the result was somehow returned as
only about 50KB and quality was much less and the caption unreadable.

What size was that in pixels?

I check it out and repost.

Why a loss of size as an "insertion" but not with "attachment"?

Peter

That's a function of your email software. I don't know what you're
using, but I have seen email software - or even just the right-click and
send-to email recipient function in Windows Explorer (under XP, anyway)
- offer to shrink images; if there's a tickbox to always do this and
don't ask me again, you might have ticked that at some point. (If so, I
don't know how to undo it.) It may be a fixed option in your email
software.

I use Outlook10 though I have not yet plumbed all its plutonic depths.

Call me pixel-obsessed if you wish, but it's important to distinguish
between pixel reduction and compression.

I check it out and repost.

On 3/26/2012, J. P. Gilliver (John) posted:
cropped 6x4 (for wide-screen monitors), exposure adjusted and

Or even 3:2 (-:

Oh come on, what's wrong with 9:6? :-)

But I would have expected (and recommended) 32:18 or some such.

--
Gene E. Bloch (Stumbling Bloch)

On Sun, 25 Mar 2012 22:33:16 -0400, Paul wrote:

Peter Jason wrote:
On Sun, 25 Mar 2012 09:44:46 +0100, "J. P. Gilliver (John)"
wrote:

In message , Peter Jason
writes:
On Sat, 24 Mar 2012 15:09:36 -0700, Gene E. Bloch
wrote:
[]
He says he'll need to do some research. OTOH, elsewhere in this thread
Bob L recommends using IrfanView's resampling instead of resizing,
which I agree with. If the OP uses other software, maybe he can find a
similar choice in that software.
I have the Irfanview and the size reduction from jpgs was less than
stellar. What is resampling?
What do you mean by "the size reduction from jpgs"?

I can think of several things you might mean, and it'd be useful to know
which you mean:

o the size reduction by saving in jpg (at the default 80%) compared to a
raw format such as bitmap, uncompressed tiff, or raw

o the size reduction by saving in jpg at a lower quality

o the size reduction by saving at a lower pixel size (e. g. "Half")

It'd be useful to know if the "smaller" size pictures your camera
produces are the same size in pixels but just a higher compression, or
actually a smaller number of pixels. Similarly was the reduction in
filesize you were obtaining in Photoshop or whatever it is, a reduction
in number of pixels or just higher compression. (You said you were
happier with the "small" pictures you got from the camera than the ones
you'd reduced in the software.)

As for resampling/resizing, I _think_ - as IrfanView uses the terms - it
means this: say you're reducing to one third the size (in each axis, i.
e. one ninth the number of pixels). A simple way would be to take every
third pixel. This can lose significant aspects - say you had three
pixels that were white black white, representing a single black line;
the line could disappear altogether. I think that might be what IV means
by resizing, since it says that's faster. The other way involves taking
some account of the content of the discarded pixels.

Thanks

I think I have fixed it by using the Irfanview "Resample" option. The
photo below is about 160KB, resampled down from the original of 1MB.
Note the effective maintaining of the caption at the lower left that
compression always trashes to a furry blur.

http://imageshack.us/photo/my-images...leirfanvi.jpg/

The original photo was a JPEG 5.01MB, then imported into PShop CS5,
cropped 6x4 (for wide-screen monitors), exposure adjusted and
sharpened. The resulting JPEG was 1MB and this Irfanview reduced to
161KB. I emailed this 161KB jpeg size to myself as an "attachment"
and it returned the same size with no loss of quality. However if I
emailed this as a picture insertion the result was somehow returned as
only about 50KB and quality was much less and the caption unreadable.
Why a loss of size as an "insertion" but not with "attachment"?

Peter

On an "insertion", did you use Windows Live Mail and the picture
is actually stored on a Skydrive ? If so, perhaps Microsoft shrank
the picture a bit when stored on their server. The thing about
"attachment", is no other things should play with it, when it's sent.
The "attachment" option should make you more in control of the situation.
If you weren't using Windows Live Mail, then insertion and attachment
could end up working the same way.

I'm using Outlook10

You could try placing the caption on it, later in the process,
to avoid a bit of fuzziness but not all of it. If you use JPEG for
the final compression, that has to fuzz the text. You could also
store the caption in the metadata of the JPEG. I see an EXIF section
in the picture, with my hex editor, where the camera "E-5" is
identified. Perhaps there'd be some way to put the
caption in there.

The caption has always been a problem. When I scanned in about 2000
family photos, some going back to 1885, a large proportion were
anonymous and there was no-one left alive to identify the people, time
or location. I resolved that this would never happen with mine. At
first I had a discreet border at the bottom of the photo on which I
placed the text, but the photo kiosks have a knack of trimming this
off partly or completely. Hence I plaster the text on the body of
the photo, and not too near the edges either. There are
precedents....
http://hoocher.com/Henry_VIII/Elizabeth_of_York.jpg

Naturally I want to do the caption only once, so I need to have the
caption readable at reasonable reductions.

As a test, you could edit the 5.01MB JPEG, and when saving out your
email copy (not your archival copy), select a quality setting of 10%,
for a compression of 50:1. That would shrink the picture to around
the same size as the 90KB I just got from your Imageshack.us link.
You'd get to keep all the pixels, but with degradation in shading.
You should be able to do that in Photoshop (although, in this dialog,
I don't understand how a Q=10 is the same thing as "Maximum", unless
they mean maximum compression).

http://photoshopcall.com/wp-content/...utorial-03.jpg

Paul

In message , Peter Jason
writes:
[]
The caption has always been a problem. When I scanned in about 2000
family photos, some going back to 1885, a large proportion were

(Ah, genealogical work? Hobby of mine too!)

anonymous and there was no-one left alive to identify the people, time
or location. I resolved that this would never happen with mine. At
first I had a discreet border at the bottom of the photo on which I
placed the text, but the photo kiosks have a knack of trimming this
off partly or completely. Hence I plaster the text on the body of
the photo, and not too near the edges either. There are
precedents....
http://hoocher.com/Henry_VIII/Elizabeth_of_York.jpg

(-:

Naturally I want to do the caption only once, so I need to have the
caption readable at reasonable reductions.

You can embed it - as text - into a JPEG file; in IrfanView, have the
image on screen, then press I then C. This has the advantage that you
can add a sizeable paragraph of text (I don't know if there is actually
any limit). I _think_ the writing back of the file (to save the text
comment) _doesn't_ involve recompressing the image and hence adding
another layer of corruption, though I'm not 100% sure of that. (You
could easily find out with an image comparison routine like
DupDetector.)
[]
--
J. P. Gilliver. UMRA: 1960/1985 MB++G.5AL-IS-P--Ch++(p)Ar@T0H+Sh0!:`)DNAf

What's awful about weird views is not the views. It's the intolerance. If
someone wants to worship the Duke of Edinburgh or a pineapple, fine. But don't
kill me if I don't agree. - Tim Rice, Radio Times 15-21 October 2011.

On 26/03/2012 6:12 PM, J. P. Gilliver (John) wrote:
You can embed it - as text - into a JPEG file; in IrfanView, have the
image on screen, then press I then C. This has the advantage that you
can add a sizeable paragraph of text (I don't know if there is actually
any limit). I _think_ the writing back of the file (to save the text
comment) _doesn't_ involve recompressing the image and hence adding
another layer of corruption, though I'm not 100% sure of that. (You
could easily find out with an image comparison routine like DupDetector.)

What you see on the screen is a decompressed image, a bitmap in fact (of
course - think about it!) So Save in JPEG format always recompresses the
image. In Irfan view and XnView, you can set the default compression,
which IMO should be set to high quality.

HTH,
Wolf K.

In message , Wolf K
writes:
On 26/03/2012 6:12 PM, J. P. Gilliver (John) wrote:
You can embed it - as text - into a JPEG file; in IrfanView, have the
image on screen, then press I then C. This has the advantage that you
can add a sizeable paragraph of text (I don't know if there is actually
any limit). I _think_ the writing back of the file (to save the text
comment) _doesn't_ involve recompressing the image and hence adding
another layer of corruption, though I'm not 100% sure of that. (You
could easily find out with an image comparison routine like DupDetector.)

What you see on the screen is a decompressed image, a bitmap in fact
(of course - think about it!) So Save in JPEG format always

Yes, I know (-:.

recompresses the image. In Irfan view and XnView, you can set the
default compression, which IMO should be set to high quality.

HTH,
Wolf K.

I _think_ adding the comment field via I, C in IrfanView - assuming you
were starting with a JPEG image in the first place - doesn't recode. I
think it just moves the - compressed - data further down the file to
make room for the comment text.
--
J. P. Gilliver. UMRA: 1960/1985 MB++G.5AL-IS-P--Ch++(p)Ar@T0H+Sh0!:`)DNAf

Hit any user to continue.

On 26/03/2012 7:49 PM, J. P. Gilliver (John) wrote:
In message , Wolf K
writes:
On 26/03/2012 6:12 PM, J. P. Gilliver (John) wrote:
You can embed it - as text - into a JPEG file; in IrfanView, have the
image on screen, then press I then C. This has the advantage that you
can add a sizeable paragraph of text (I don't know if there is actually
any limit). I _think_ the writing back of the file (to save the text
comment) _doesn't_ involve recompressing the image and hence adding
another layer of corruption, though I'm not 100% sure of that. (You
could easily find out with an image comparison routine like
DupDetector.)

What you see on the screen is a decompressed image, a bitmap in fact
(of course - think about it!) So Save in JPEG format always

Yes, I know (-:.

recompresses the image. In Irfan view and XnView, you can set the
default compression, which IMO should be set to high quality.

HTH,
Wolf K.

I _think_ adding the comment field via I, C in IrfanView - assuming you
were starting with a JPEG image in the first place - doesn't recode. I
think it just moves the - compressed - data further down the file to
make room for the comment text.

Visually, what happens is that you place the text (with or without
transparent background) anywhere you like. This implies that it's
overlaid on the image data at that location, and the whole thing is
recompressed when the image is saved. Any other way of doing it would
limit correct decoding of the imge data to software that understood how
Irfanview or XnView added text.

BTW, font size is measured in pixels, and is absolute, so that if you
use the same font with large (say 4000x3000) and resized (say
2400x1800), the text will take up a larger portion of the smaller image.
This too implies that the text data is overlaid.

HTH
Wolf K.

In message , Wolf K
writes:
On 26/03/2012 7:49 PM, J. P. Gilliver (John) wrote:
[]
I _think_ adding the comment field via I, C in IrfanView - assuming you
were starting with a JPEG image in the first place - doesn't recode. I
think it just moves the - compressed - data further down the file to
make room for the comment text.

Visually, what happens is that you place the text (with or without
transparent background) anywhere you like. This implies that it's
overlaid on the image data at that location, and the whole thing is
recompressed when the image is saved. Any other way of doing it would
limit correct decoding of the imge data to software that understood how
Irfanview or XnView added text.

BTW, font size is measured in pixels, and is absolute, so that if you
use the same font with large (say 4000x3000) and resized (say
2400x1800), the text will take up a larger portion of the smaller
image. This too implies that the text data is overlaid.

HTH
Wolf K.

Adding text to the comment field does not make it visible in the image.
It places text in a place within the file that can be retrieved by any
viewer capable of looking at that field. It's like the EXIF data that
records camera details etcetera. Thus it need not - and I believe does
not, at least when done in IrfanView - affect the image in any way.
--
J. P. Gilliver. UMRA: 1960/1985 MB++G.5AL-IS-P--Ch++(p)Ar@T0H+Sh0!:`)DNAf

"When I was young I used to scintillate
now I only sin 'til ten past three" (Ogden Nash) [via Andy Breen]

On 27/03/2012 3:06 AM, J. P. Gilliver (John) wrote:
Adding text to the comment field does not make it visible in the image.
It places text in a place within the file that can be retrieved by any
viewer capable of looking at that field. It's like the EXIF data that
records camera details etcetera. Thus it need not - and I believe does
not, at least when done in IrfanView - affect the image in any way.

Sorry, I overlooked that "comment field". I don't use it, put basic data
(copyright notice and often subject/date/location) in/on image itself.
But thanks for the reminder, will have to experiment with comment field.
AIUI, it's metadata, like EXIF.

Wolf K.