«

My camera has various JPEG compression modes. The manual says that
higher compression means lower image quality, but I haven't seen any
difference in the image quality of the two compression levels that I
have been using: FINE (1:4) and NORMAL (1:. However, I did see a
signigicant difference in file size. What am I missing here?

thanks

James Ramaley wrote:

> My camera has various JPEG compression modes. The manual says that
> higher compression means lower image quality, but I haven't seen any
> difference in the image quality of the two compression levels that I
> have been using: FINE (1:4) and NORMAL (1:. However, I did see a
> signigicant difference in file size. What am I missing here?

You're not missing anything

JPEG is an extremely effective method of compression. It's done
by discarding some of the color information.

If there are say, 20 pixels adjacent to one another that are very
close to the same color, it *will* make them the exact same color.
Once that is done, it can represent these 10 pixels as a few bytes
rather than 30 bytes. (This is a *very* rough description of how it
works

If you look at the size your images, you'll see the ones with large areas
of the same color (ie a large smooth wall) compress better than images with
lots of detail. (A gravel road).

The reason this works is that in a photographic image, these changes are
too subtle to be seen by the eye... Of course this is only true to a certain
level.. If you compress too far, the images will be noticeably worse.

There *will* be a difference in images with different levels of
compression, but you'll have to blow them up to 400% or more and look
hard to really see the difference.

Look at the areas of the image where there are large transitions
between light and dark. You should see some dark distortion there.
This is called 'artifacting' Another thing to look for is banding
in large areas of similar color. (like the sky).

James Ramaley wrote:
> My camera has various JPEG compression modes. The manual says that
> higher compression means lower image quality, but I haven't seen any
> difference in the image quality of the two compression levels that I
> have been using: FINE (1:4) and NORMAL (1:. However, I did see a
> signigicant difference in file size. What am I missing here?
>
> thanks

Best option to see the difference is to shoot some text and see it - but
usually you'll have to look it magnified, to see the difference.
There's distortion around text. But i agree that there's barely noticeable
difference between fine and normal, there's a bit more between normal and
coarse(or low, whatever your camera says).
Interesting point is that cameras use exactly opposite of usual words -
usually it's said High as high compression (and so low quality), while all
cameras have high as high quality...

More noticeable on straight edges and fine detail - signs, textures etc.
Zoom in enough and you'll see the difference, but at 'normal' scales it's
probably so minor a difference as to be negligible. Maybe becomes more
noticeable if you have to lift the shadows too...

Cheers, Jason (remove ... to reply)
Video & Gaming: http://gadgetaus.com

James Ramaley wrote:
> My camera has various JPEG compression modes. The manual says that
> higher compression means lower image quality, but I haven't seen any
> difference in the image quality of the two compression levels that I
> have been using: FINE (1:4) and NORMAL (1:. However, I did see a
> signigicant difference in file size. What am I missing here?
>
> thanks

There are differences, but you might not be aware of them in any single
photo. Try compressing a picture of your lawn at both compressions
levels, and then compare them blown up onscreen to 200%. That should
show you how they are different, and what effect excess JPEG compression
has on the image quality.

On 24 Oct 2004 22:30:16 -0700, (James Ramaley)
wrote:

>My camera has various JPEG compression modes. The manual says that
>higher compression means lower image quality, but I haven't seen any
>difference in the image quality of the two compression levels that I
>have been using: FINE (1:4) and NORMAL (1:. However, I did see a
>signigicant difference in file size. What am I missing here?

It all depends on what size your viewing at. Zoom in on the image,
and you'll soon see.

(James Ramaley) writes:
>My camera has various JPEG compression modes. The manual says that
>higher compression means lower image quality, but I haven't seen any
>difference in the image quality of the two compression levels that I
>have been using: FINE (1:4) and NORMAL (1:. However, I did see a
>signigicant difference in file size. What am I missing here?

Try shooting a high-contrast subject, like bare tree branches against an
overcast sky. Then look at the images at 100%. You'll probably see
some artifacts along the boundary between tree branch and sky in all
of the images, even the highest-quality one. In most photos, the effect
is less visible because the contrast is less, but now that you know what
to look for you'll see it in other places.

I did this once with my G2, and decided that memory cards and CDRs (for
archiving) are cheap, so I always shoot in either RAW (no lossy
compression at all), or JPEG at maximum size and least compression
(Superfine). I've never used the higher-compression JPEG modes except
for these tests.

Dave

On Mon, 25 Oct 2004 00:50:52 -0500, Jim Townsend <>
wrote:

>James Ramaley wrote:
>
>> My camera has various JPEG compression modes. The manual says that
>> higher compression means lower image quality, but I haven't seen any
>> difference in the image quality of the two compression levels that I
>> have been using: FINE (1:4) and NORMAL (1:. However, I did see a
>> signigicant difference in file size. What am I missing here?
>
>You're not missing anything
>
>JPEG is an extremely effective method of compression. It's done
>by discarding some of the color information.
>
>If there are say, 20 pixels adjacent to one another that are very
>close to the same color, it *will* make them the exact same color.
>Once that is done, it can represent these 10 pixels as a few bytes
>rather than 30 bytes. (This is a *very* rough description of how it
>works
>
>If you look at the size your images, you'll see the ones with large areas
>of the same color (ie a large smooth wall) compress better than images with
>lots of detail. (A gravel road).
>
>The reason this works is that in a photographic image, these changes are
>too subtle to be seen by the eye... Of course this is only true to a certain
>level.. If you compress too far, the images will be noticeably worse.
>
>There *will* be a difference in images with different levels of
>compression, but you'll have to blow them up to 400% or more and look
>hard to really see the difference.
>
>Look at the areas of the image where there are large transitions
>between light and dark. You should see some dark distortion there.
>This is called 'artifacting' Another thing to look for is banding
>in large areas of similar color. (like the sky).
>

Jim Townsend,
great job @ describing JPEG photo compression in an
easy to understand way. You did such a good job that there is little
else to add but I will offer a related side bar.

If 1 believes or knows that a given picture(s) might or will
need to be heavily cropped or a very large print made of them, then
selecting the lowest JPEG compression (best picture quality) is the
way to go. However if you should find yourself running low on memory
card storage space @ a given photo shoot & are using the lowest
compression, switching to a higher compression to fit more photos in
the remaining space on the memory card becomes a viable option.

Often people ask if switching to a smaller size (resolution)
would be better. Since JPEG works so well in most cases, it's better
to remain @ the higher resolution & increase the JPEG compression to
reduce the file size. This is largely true because many of the
artifacts can later be edited out but lost resolution can not be
replaced. However we are fortunate that advances in flash memory is
continuing to drive memory priced down, speed & capacity up. So to
many it may seem like a 512MB or a 1GB memory card is very large, but
in a year or so, they are likely to be considered the smallest size
worth buying or using in many situations.

Last note, consider also the price/size ratio of any memory
card. With higher resolution cameras becoming increasingly common,
larger cards are also becoming more common & less costly. Thus it may
be lass expensive to purchase a 1GB card rather than 2 512MB cards.
Yes somebody will point out the logic of "not having all you eggs in 1
basket" & there is some validity to this even with very reliable
memory cards but the same applies to a hard drive in a computer. Most
consumer computers have only a single hard drive in them, "all there
eggs in 1 basket". Yes you can & should back-up this DATA
periodically but unless your running 2 drives on a RAID system, if
your drive suddenly suffers an unrecoverable failure between back-ups,
you have lost all of that information added since your last back-up.

There is no 100% safe way to store your DATA of any kind but
spreading it out & backing up often can greatly minimize the chance of
loss. This is also why "I" don't buy generic memory from the cheapest
vendor. Yes most of that is highly reliable as well but I prefer to
only buy from a manufacture that offers a "lifetime warrantee". Lexar
12X & faster CF memory cards all have such a "lifetime warrantee" as
do other companies.

Best of luck & again as a general rule, my preference & advice
is to almost always shoot @ the highest resolution & minimum JPEG
compression whenever possible. However as noted above, I will
increase compression if I am running low on memory storage space or if
I an almost certain that the picture will never be printed or need to
be printed any larger than 5x7 cropped more than a little".

Respectfully, DHB

(James Ramaley) writes:

> My camera has various JPEG compression modes. The manual says that
> higher compression means lower image quality, but I haven't seen any
> difference in the image quality of the two compression levels that I
> have been using: FINE (1:4) and NORMAL (1:. However, I did see a
> signigicant difference in file size. What am I missing here?

There will be large differences in quality between those modes. You
probably won't see them until you view the image at 1:1 magnification
(actual pixels) on screen. You'll spot it most easily in large areas
of fairly even color, like a blue sky.
--
David Dyer-Bennet, <mailto:dd->, <http://www.dd-b.net/dd-b/>
RKBA: <http://noguns-nomoney.com/> <http://www.dd-b.net/carry/>
Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/>
Dragaera/Steven Brust: <http://dragaera.info/>

(Dave Martindale) wrote in news:clj5tj$9ho$1
@mughi.cs.ubc.ca:

> Try shooting a high-contrast subject, like bare tree branches against an
> overcast sky. Then look at the images at 100%. You'll probably see
> some artifacts along the boundary between tree branch and sky in all
> of the images, even the highest-quality one. In most photos, the effect
> is less visible because the contrast is less, but now that you know what
> to look for you'll see it in other places.
>

I think tree branches are about the best place to spot artifacts, but
telephone wires work too.

In the tree branches, you will frequently see dark blue mush in the higher
compressions where there are black details in the low.

Bob

--
Delete the inverse SPAM to reply