Tom's Hardware > Forum > Digital Camera > Digital Camera General > which compression level ?
Word :    Username :           
 

Archived from groups: rec.photo.digital (More info?)

 

Hi,

I am new to digital photography. I've got my Casio exilim P700 just a
few days, ago. The camera has about 7 M pixels and offers 3 levels of
JPEG compress (fine, standard and economy). With 'fine' JPEG compression
4.4 Mb (instead of 20 Mb) are required to store an image - that's a
ration of about 4.6. The 'standard' compression level compresses to
2.2 Mb ( a factor 9.2 ).
Now I have taken pictures of a pond in my garden with reflections and
lots of details - once with 'fine' compression and then with 'standard'
compression.
I have compared these two visually side-by-side on a high quality CRT
with a magnification factor up to 5. And I didn't notice any difference.

So, which sort of motifs require a 'fine' (i.e. low) compression?

Many thanks for a hint,

Helmut Jarausch

Sponsored Links
Register or log in to remove.

Archived from groups: rec.photo.digital (More info?)

 

Helmut Jarausch wrote:

> Hi,
>
> I am new to digital photography. I've got my Casio exilim P700 just a
> few days, ago. The camera has about 7 M pixels and offers 3 levels of
> JPEG compress (fine, standard and economy). With 'fine' JPEG compression
> 4.4 Mb (instead of 20 Mb) are required to store an image - that's a
> ration of about 4.6. The 'standard' compression level compresses to
> 2.2 Mb ( a factor 9.2 ).
> Now I have taken pictures of a pond in my garden with reflections and
> lots of details - once with 'fine' compression and then with 'standard'
> compression.
> I have compared these two visually side-by-side on a high quality CRT
> with a magnification factor up to 5. And I didn't notice any difference.
>
> So, which sort of motifs require a 'fine' (i.e. low) compression?

The amount of compression that is acceptable depends on what *you*
like personally. As you've noticed there isn't a big difference.

The only reason one would use the higher compression levels was
if they were trying to conserve memory card space. That's about
it.

With the cost of memory cards today, there really shouldn't be a
need to use anything other than the the FINE mode. I'd just leave
the setting there and forget about it.

Reply to Anonymous

Archived from groups: rec.photo.digital (More info?)

 

Hi,

It is not really possible to judge the sharpness and detail present in a
digital image, merely from viewing it on a CRT or an LCD screen. Both these
viewing tools are coarse, and the brightness of the image can make you think
that it is sharper than it really is. For evaluating sharpness and detail,
make prints of a nice size, e.g. 8 x 10" = 20x25cm., on good quality photo
quality paper.

On the other hand, if CRT images or small prints are all that you need, more
compression may be suitable.

However, as already posted by another, storage media are cheap. You never know
when you will need an image enlarged or cropped. It is probably best to shoot
all images at lowest compression and finest detail.

Good luck.

Morton



Jim Townsend wrote:

> Helmut Jarausch wrote:
>
> > Hi,
> >
> > I am new to digital photography. I've got my Casio exilim P700 just a
> > few days, ago. The camera has about 7 M pixels and offers 3 levels of
> > JPEG compress (fine, standard and economy). With 'fine' JPEG compression
> > 4.4 Mb (instead of 20 Mb) are required to store an image - that's a
> > ration of about 4.6. The 'standard' compression level compresses to
> > 2.2 Mb ( a factor 9.2 ).
> > Now I have taken pictures of a pond in my garden with reflections and
> > lots of details - once with 'fine' compression and then with 'standard'
> > compression.
> > I have compared these two visually side-by-side on a high quality CRT
> > with a magnification factor up to 5. And I didn't notice any difference.
> >
> > So, which sort of motifs require a 'fine' (i.e. low) compression?
>
> The amount of compression that is acceptable depends on what *you*
> like personally. As you've noticed there isn't a big difference.
>
> The only reason one would use the higher compression levels was
> if they were trying to conserve memory card space. That's about
> it.
>
> With the cost of memory cards today, there really shouldn't be a
> need to use anything other than the the FINE mode. I'd just leave
> the setting there and forget about it.

Reply to mort

Archived from groups: rec.photo.digital (More info?)

 

A good question. My recommendation is to try to do some triage when you
are worried about filling your memory card or overloading your hard
drive. On extended travel when I am not in a good backup situation (eg.
traveling light abroad) I tend to shoot at either the SHQ or HQ
settings on my 5.0 and 8.0 megapix cameras, depending on whether I have
any intention of doing an 8x10 or higher print. Since I rarely do that
I generally shoot at HQ and, with a little post-processing, can get a
very good 8x10 anyhow. On the other hand, when I am close to home, have
my laptop with me for backup, or doing some semi-commercial work I
always shoot at either raw or SHQ.

Incidentally, while it is generally true that you cannot see
differences in compression very easily on a CRT, you CAN if you do high
zooms on them. In my case, Olympus 5060 and E-800, it is virtually
impossible, even when zooming in closely, to see the difference between
SHQ and HQ, and that has been confirmed in some independent tests.

You'll have to do a little more testing -- by printing or zooming -- to
reach some comfort levels.

Reply to Anonymous

Archived from groups: rec.photo.digital (More info?)

 

Helmut,

I did quite a few tests on my Canon S30 and found the
following:

At anything from 0-10 times compression, I was able to
see color differences in individual pixels between the
images if I zoomed to the point where I could see each
pixel. But I could NOT see any difference when viewed
at normal size (one image pixel = 1 screen pixel).

If you asked me which of the highly zoomed images
compressed at 5:1 and 10:1 was more accurate, I wouldn't
be able to say. And at normal size, I couldn't see
the difference at all.

At 20 times compression I could see differences at
normal size in some, but not all, photographs. However
the differences were generally not objectionable.

Above 20 times compression I more often saw objectionable
compressions - i.e., compression artefacts around sharp
edges and sharp color transitions. It is those kinds of
photos (high contrast, sharp edges) that seem to me to
suffer most from over compression.

Your results may vary from this if 1) your perceptions
are sharper than mine or 2) your camera is more or less
sophisticated in its JPEG compression algorithm.

My conclusion was to set the compression level at Canon's
mid-level - which gave me an average of about 10:1.

By the way, the Canon experience is not universal. I also
own a Pentax Optio 750Z. The Canon produces very different
file sizes for different images. The Pentax produces
much more uniform file sizes. They're clearly using very
different algorithms. However the 10:1 compression seemed
pretty good to me on the Pentax too.

Alan

Reply to Anonymous

Archived from groups: rec.photo.digital (More info?)

 

Alan Meyer wrote:
>
> Above 20 times compression I more often saw objectionable
> compressions - i.e., compression artefacts around sharp
> edges and sharp color transitions. It is those kinds of
> photos (high contrast, sharp edges) that seem to me to
> suffer most from over compression.


One of the biggest problem areas for me with jpeg compression is in
gradually fading colored skys, the colors get compressed and you see
curved bands like a rainbow: posterizing. This is true for my D70 & the
old Oly.


>
> Your results may vary from this if 1) your perceptions
> are sharper than mine or 2) your camera is more or less
> sophisticated in its JPEG compression algorithm.

Reply to Paul

Archived from groups: rec.photo.digital (More info?)

 

Alan Meyer wrote:
> Helmut,
>
> I did quite a few tests on my Canon S30 and found the
> following:
>
> At anything from 0-10 times compression, I was able to
> see color differences in individual pixels between the
> images if I zoomed to the point where I could see each
> pixel. But I could NOT see any difference when viewed
> at normal size (one image pixel = 1 screen pixel).
>
> If you asked me which of the highly zoomed images
> compressed at 5:1 and 10:1 was more accurate, I wouldn't
> be able to say. And at normal size, I couldn't see
> the difference at all.
>
> At 20 times compression I could see differences at
> normal size in some, but not all, photographs. However
> the differences were generally not objectionable.
>
> Above 20 times compression I more often saw objectionable
> compressions - i.e., compression artefacts around sharp
> edges and sharp color transitions. It is those kinds of
> photos (high contrast, sharp edges) that seem to me to
> suffer most from over compression.
>
> Your results may vary from this if 1) your perceptions
> are sharper than mine or 2) your camera is more or less
> sophisticated in its JPEG compression algorithm.
>
> My conclusion was to set the compression level at Canon's
> mid-level - which gave me an average of about 10:1.
>
> By the way, the Canon experience is not universal. I also
> own a Pentax Optio 750Z. The Canon produces very different
> file sizes for different images. The Pentax produces
> much more uniform file sizes. They're clearly using very
> different algorithms. However the 10:1 compression seemed
> pretty good to me on the Pentax too.
>
> Alan

Hi...

If it's at all helpful.... here's a means by which we
can actually see in real time the actual "switch" from
uncompressed and any level of jpeg compression...

Get ahold of (for first tests, anyway) a tiff or bitmap
or any uncompressed format picture.

Open it in paint shop pro...

No need to modify a thing, just go to save as. Choose
jpeg from the drop down box. Click options. From there
click Run Optimizer.

You'll see the pic in two separate panes. The left is
uncompressed original, the right is the compressed jpeg.

Now you can drag the set compression level to bar, and see
exactly what the effect is. Change the zoom level to see
the effect at other sizes.

And best... if you'll put the mouse cursor (hand) on either
of the panes and drag the picture just a tad... even a pixel
worth will do... and hold the left button down, the right
pane will stay uncompressed. The instant you release the
right button the picture will change to the compressed version,
so you get to compare one for one in the same box.

Try it with different compression levels.

Then try it with other already compressed jpeg's and so on.

Hope it's helpful, and that I'm not re-inventing the wheel,
or preaching to the choir. :)

Take care.

Ken

Reply to Anonymous

Archived from groups: rec.photo.digital (More info?)

 

Alan Meyer wrote:
> Helmut,
>
> I did quite a few tests on my Canon S30 and found the
> following:
>
> At anything from 0-10 times compression, I was able to
> see color differences in individual pixels between the
> images if I zoomed to the point where I could see each
> pixel. But I could NOT see any difference when viewed
> at normal size (one image pixel = 1 screen pixel).
>
> If you asked me which of the highly zoomed images
> compressed at 5:1 and 10:1 was more accurate, I wouldn't
> be able to say. And at normal size, I couldn't see
> the difference at all.
>
> At 20 times compression I could see differences at
> normal size in some, but not all, photographs. However
> the differences were generally not objectionable.
>
> Above 20 times compression I more often saw objectionable
> compressions - i.e., compression artefacts around sharp
> edges and sharp color transitions. It is those kinds of
> photos (high contrast, sharp edges) that seem to me to
> suffer most from over compression.
>
> Your results may vary from this if 1) your perceptions
> are sharper than mine or 2) your camera is more or less
> sophisticated in its JPEG compression algorithm.
>
> My conclusion was to set the compression level at Canon's
> mid-level - which gave me an average of about 10:1.
>
> By the way, the Canon experience is not universal. I also
> own a Pentax Optio 750Z. The Canon produces very different
> file sizes for different images. The Pentax produces
> much more uniform file sizes. They're clearly using very
> different algorithms. However the 10:1 compression seemed
> pretty good to me on the Pentax too.
>
> Alan
>
Different manufacturers, and different models seem to have different
compression algorithms. Sometimes I feel that my camera is
overcompressing. Given a choice, I would tend to favor larger files,
even an 'uncompressed' option. Most of the time, compression isn't an
objectionable factor, but Kodak's compression of things like trees, and
grass isn't attractive. Worse, sharpening it in PSE seems to make the
problem much worse. I would much rather have NO compression as an
option. SD cards are cheap.


--
Ron Hunter rphunter@charter.net

Reply to Anonymous

Archived from groups: rec.photo.digital (More info?)

 

"Helmut Jarausch" <jarausch@skynet.be> wrote in message
news:42665bf1$0$327$ba620e4c@news.skynet.be...
> Hi,
>
> I am new to digital photography. I've got my Casio exilim P700 just a few
> days, ago. The camera has about 7 M pixels and offers 3 levels of
> JPEG compress (fine, standard and economy). With 'fine' JPEG compression
> 4.4 Mb (instead of 20 Mb) are required to store an image - that's a ration
> of about 4.6. The 'standard' compression level compresses to
> 2.2 Mb ( a factor 9.2 ).
> Now I have taken pictures of a pond in my garden with reflections and lots
> of details - once with 'fine' compression and then with 'standard'
> compression.
> I have compared these two visually side-by-side on a high quality CRT
> with a magnification factor up to 5. And I didn't notice any difference.
>
> So, which sort of motifs require a 'fine' (i.e. low) compression?
>
> Many thanks for a hint,
>
> Helmut Jarausch

The way JPEG compression works, as I understand it, is that instead of
recording the exact RGB values for each pixel, it will record something like
"x pixels in this row are this RGB value, or very close to it." The higher
the degree of compression, the more lax is the "very close to it." So where
you are most likely to notice the difference is in some area of the same (or
nearly the same) color, such as an expanse of clear sky or the side of a
house painted the same color, etc.

Personally I use "standard" (intermediate) compression for most general
photography because, like you, I can't see any difference between that and
"fine." Many people advise always using "fine" because they believe it must
be better, and of course it is true that memory cards (at least the most
popular types, SD and CompactFlash) are now very low in cost per megabyte.
On the other hand, saving lots of "fine" shots will take up more hard drive
space, take a bit longer to transfer, etc. If there's some really great shot
you expect to make enlargements of, then what the heck, might as well use
"fine." But I have found "standard" satisfactory for up to letter-size
prints anyway.

If you expect to do a good deal of post-processing, then "fine" would
probably be better, but in that case an uncompressed TIFF file would be
best, or a RAW file if your camera has that capability. Either TIFF or RAW
are lossless formats; they preserve everything, whereas JPEG is always
"lossy." If your camera does TIFF files they may be called something else,
e.g. Minolta used to call them "superfine."

N.

Reply to Anonymous
Tom's Hardware > Forum > Digital Camera > Digital Camera General > which compression level ?
Go to:

There are 961 identified and unidentified users. To see the list of identified users, Click here.

Please mind

You are about to answer a thread that has been inactive for more than 6 months.
If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.

Add a reply Cancel
Sponsored links
  • Ask the community now
  • Publish
Ad
They won a badge
Join us in greeting them