Sign in with
Sign up | Sign in
Your question

canon 20D - soft images

Last response: in Digital Camera
Share
Anonymous
April 7, 2005 2:34:53 AM

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

Been using the canon 20D and find that my previous canon powershotG1 gave me
the impression of sharper detail. Is this a setup issue or a characteristic
of SLR's versus point & shoot?

More about : canon 20d soft images

Anonymous
April 7, 2005 9:35:53 AM

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

User error....
:-p



"deloid" <lambert@cableone.net> wrote in message
news:1159e3hfhnavo84@corp.supernews.com...
> Been using the canon 20D and find that my previous canon powershotG1 gave
me
> the impression of sharper detail. Is this a setup issue or a
characteristic
> of SLR's versus point & shoot?
>
>
>
Anonymous
April 7, 2005 10:07:46 AM

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

"deloid" <lambert@cableone.net> wrote in message
news:1159e3hfhnavo84@corp.supernews.com...
> Been using the canon 20D and find that my previous canon powershotG1 gave
> me the impression of sharper detail. Is this a setup issue or a
> characteristic of SLR's versus point & shoot?
>
>

Even at the highest setting, Canon's DSLRs have less sharpening than their
digicams. A little post-processing will bring out the best in the image.

Mark
Related resources
Can't find your answer ? Ask !
Anonymous
April 7, 2005 10:17:02 AM

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

deloid <lambert@cableone.net> wrote:
>
> Been using the canon 20D and find that my previous canon powershotG1 gave me
> the impression of sharper detail. Is this a setup issue or a characteristic
> of SLR's versus point & shoot?

This is a characteristic of the *lenses* involved, not the bodies.
Which lens do you have attached to the 20D? At what aperture and
focal distance are you shooting?

--
Zed Pobre <zed@resonant.org> a.k.a. Zed Pobre <zed@debian.org>
PGP key and fingerprint available on finger; encrypted mail welcomed.
Anonymous
April 7, 2005 10:21:39 AM

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

A P&S camera will have a lot of sharpening done in camera, with the 20D and
other D-SLR's you want complete control over the end product so will do your
own sharpening in whatever post tools you use.


"deloid" <lambert@cableone.net> wrote in message
news:1159e3hfhnavo84@corp.supernews.com...
> Been using the canon 20D and find that my previous canon powershotG1 gave
> me the impression of sharper detail. Is this a setup issue or a
> characteristic of SLR's versus point & shoot?
>
>
>
April 7, 2005 11:30:03 AM

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

The 20D has defaults that are quite soft compared to the average P&S as the
below poster pointed out. If you read your manual you will see quite a good
section on set-up and the numerous ways of setting up your camera to meet
your needs. Yes, surprise surprise its all in the manual. I have found
that if you prefer to have P&S type output with out then having to post
process (assuming you know what you are doing with your photographic
technique etc) the 20D gives great results when you do the relevant
comparison after printing to home printer without having to spend much time
with such processing tools such as Photoshop, Breeze Browser or another of
the numerous packages available. However, if you wish to enlarge your shots
past say A4 then you can notice the differences creeping in and in my humble
opinion its best to shoot soft and process out of camera.

Hope you are enjoying your 20D as much as I am.

regards

Don from Down Under


"Pete D" <no@email.com> wrote in message
news:Tj45e.2723$5F3.1275@news-server.bigpond.net.au...
>A P&S camera will have a lot of sharpening done in camera, with the 20D and
>other D-SLR's you want complete control over the end product so will do
>your own sharpening in whatever post tools you use.
>
>
> "deloid" <lambert@cableone.net> wrote in message
> news:1159e3hfhnavo84@corp.supernews.com...
>> Been using the canon 20D and find that my previous canon powershotG1 gave
>> me the impression of sharper detail. Is this a setup issue or a
>> characteristic of SLR's versus point & shoot?
>>
>>
>>
>
>
Anonymous
April 7, 2005 11:51:02 AM

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

>Been using the canon 20D and find that my previous canon powershotG1
gave
>me the impression of sharper detail. Is this a setup issue or a
characteristic
>of SLR's versus point & shoot?

Canon recommends a first pass USM (sharpening) run at 300% (amount),
0.3 radius, 0 threshold to counteract the effects of the AA filter with
the dSLRs. These are suggested starting points and should be adequate
for many images but not all. Try that and see if it adds snap.
Anonymous
April 7, 2005 1:42:34 PM

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

Zed Pobre wrote:

> deloid <lambert@cableone.net> wrote:
>>
>> Been using the canon 20D and find that my previous canon powershotG1 gave me
>> the impression of sharper detail. Is this a setup issue or a characteristic
>> of SLR's versus point & shoot?
>
> This is a characteristic of the *lenses* involved, not the bodies.
> Which lens do you have attached to the 20D? At what aperture and
> focal distance are you shooting?

That's a good point. The built in G1 lens probably has far better
optical quality than say the 20D kit lens wide open. (It's not
as hard to build quality into a tiny lens).

Considering that the lens actually makes the picture, it's an important
consideration. The camera only captures what the lens saw.

As others have mentioned.. Good post processing will sharpen up
the images. DSLR's don't sharpen as much as P&S cameras. This
is a common complaint.

I find shooting RAW and doing the sharpening on RAW files first
provides the best results.
Anonymous
April 7, 2005 3:16:22 PM

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

Chances are a little work with sharpening in post exposure (on your
computer) should make the result shine. As the others have noted the 20D
appears to be softer by design. For most uses I like it. I feel the P&S
tend to over do it.

--
Joseph Meehan

Dia's Muire duit
"deloid" <lambert@cableone.net> wrote in message
news:1159e3hfhnavo84@corp.supernews.com...
> Been using the canon 20D and find that my previous canon powershotG1 gave
> me the impression of sharper detail. Is this a setup issue or a
> characteristic of SLR's versus point & shoot?
>
>
>
Anonymous
April 7, 2005 3:16:23 PM

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

Thanks for all the great answers- I understand now.

Dean
Anonymous
April 7, 2005 3:16:24 PM

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

deloid wrote:
> Thanks for all the great answers- I understand now.
>
> Dean

Unfortunately that doesn't eliminate the possibility of other factors
determining a "soft" appearance. Your "soft" may be another's "Quite
good", as noted by a couple of posters; however, it could also mean
"Flat unacceptable", so for your own peace of mind you'll need to
discover or invent a means of comparing your camera/lens output with
known "sharp" photos.

I'd really appreciate an expert-generated recipe built to accomplish
such a project.


--
Frank ess
Anonymous
April 7, 2005 3:40:28 PM

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

Bill Hilton wrote:

>>Been using the canon 20D and find that my previous canon powershotG1
>
> gave
>
>>me the impression of sharper detail. Is this a setup issue or a
>
> characteristic
>
>>of SLR's versus point & shoot?
>
>
> Canon recommends a first pass USM (sharpening) run at 300% (amount),
> 0.3 radius, 0 threshold to counteract the effects of the AA filter with
> the dSLRs. These are suggested starting points and should be adequate
> for many images but not all. Try that and see if it adds snap.


Is that on a RAW image?

Seems very high (300%). USM effectiveness is very dependant on the
detail in the image.



--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- slr-systems FAQ project: http://tinyurl.com/6m9aw
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: Remove FreeLunch.
Anonymous
April 7, 2005 7:54:16 PM

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

"Joseph Meehan" <sligojoe_Spamno@hotmail.com> writes:
> Chances are a little work with sharpening in post exposure (on your
>computer) should make the result shine. As the others have noted the 20D
>appears to be softer by design. For most uses I like it. I feel the P&S
>tend to over do it.

All of the Canon P&S cameras that I've used have the ability to turn
down the amount of in-camera sharpening. On some (G2) there are
individual settings for sharpness, colour saturation, etc. On others
with less control, there is a "special effect" mode called Low
Sharpening. Either way, if you set the camera for less in-camera
sharpening and then carefully apply Unsharp Mask later, the results are
better. All my Canons are routinely configured this way.

So Canon *knows* that less in-camera sharpening provides better results,
and provides for that on even relatively low-end cameras as well as it
being the default on the DSLRs. But low sharpening isn't the default on
the P&S cameras because some users want to be able to use pictures right
out of the camera with no processing, and a default of low sharpening
would make Canon's images look soft compared to Nikon, etc.

Dave
Anonymous
April 7, 2005 8:05:40 PM

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

In article <1159e3hfhnavo84@corp.supernews.com>, deloid
<lambert@cableone.net> writes
>Been using the canon 20D and find that my previous canon powershotG1 gave me
>the impression of sharper detail. Is this a setup issue or a characteristic
>of SLR's versus point & shoot?
>
>
>
To add to the answers you have received (which I agree with) I suggest
you look at the very similar thread entitled "DSLR harder to handhold
than P&S?" which started last week.

David
--
David Littlewood
Anonymous
April 8, 2005 5:06:56 AM

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

"Alan Browne" <alan.browne@FreeLunchVideotron.ca> wrote in message
news:Ivc5e.458$sj1.6246@wagner.videotron.net...
SNIP
> Seems very high (300%).

Not on such a small radius.

> USM effectiveness is very dependant on the detail in the image.

You attempt to reverse the effect of the AA-filter and lens
aberrations, not having much effect on e.g. out-of-focus edges at
final output size, yet.

Bart
Anonymous
April 8, 2005 5:06:57 AM

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

"Bart van der Wolf" <bvdwolf@no.spam> wrote in message
news:4255bd12$0$154$e4fe514c@news.xs4all.nl...
>
> "Alan Browne" <alan.browne@FreeLunchVideotron.ca> wrote in message
> news:Ivc5e.458$sj1.6246@wagner.videotron.net...
> SNIP
>> Seems very high (300%).
>
> Not on such a small radius.
>
>> USM effectiveness is very dependant on the detail in the image.
>
> You attempt to reverse the effect of the AA-filter and lens aberrations,
> not having much effect on e.g. out-of-focus edges at final output size,
> yet.
>
> Bart
I've found that 150%, 0.3 radius, 0 threshold is a little more pleasing to
my eye, but that's just my opinion.

--
Skip Middleton
http://www.shadowcatcherimagery.com
Anonymous
April 8, 2005 5:06:58 AM

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

In message <nxj5e.58$lz2.28@fed1read07>,
"Skip M" <shadowcatcher@cox.net> wrote:

>I've found that 150%, 0.3 radius, 0 threshold is a little more pleasing to
>my eye, but that's just my opinion.

Which camera? I find that the 20D needs less than the 10D, as it has a
less aggressive AA filter.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
April 8, 2005 8:59:59 AM

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

<JPS@no.komm> wrote in message
news:qmhb51doo364gdh3h2i89donqte0pktoop@4ax.com...
> In message <nxj5e.58$lz2.28@fed1read07>,
> "Skip M" <shadowcatcher@cox.net> wrote:
>
>>I've found that 150%, 0.3 radius, 0 threshold is a little more pleasing
>>to
>>my eye, but that's just my opinion.
>
> Which camera? I find that the 20D needs less than the 10D, as it has a
> less aggressive AA filter.
> --
>
> <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
> John P Sheehy <JPS@no.komm>
> ><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><

20D. That 300% figure is from the 1D suggestions.
http://www.photoworkshop.com/canon/EOS_Digital.pdf

--
Skip Middleton
http://www.shadowcatcherimagery.com
Anonymous
April 8, 2005 12:47:39 PM

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

>>Canon recommends a first pass USM (sharpening) run at 300% (amount),
>>0.3 radius, 0 threshold to counteract the effects of the AA filter

>Alan Browne writes
>
>Seems very high (300%)

Not with that small a radius. This was the suggested starting point
for the Professional bodies, with 1.3x and 1.0x crop factors so it
might be a bit high for the 1.6x models, but it's Canon's "official"
word.

>USM effectiveness is very dependant on the
>detail in the image.

Wow, and some of us thought you knew nothing about digital. Guess we
were wrong ... I did point out "These are suggested starting points and
should be adequate for many images but not all." but I guess you didn't
get to the second sentence?

Bill
Anonymous
April 8, 2005 6:32:42 PM

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

"Skip M" <shadowcatcher@cox.net> wrote in message
news:3nu5e.18153$lz2.15626@fed1read07...
SNIP
> That 300% figure is from the 1D suggestions.
> http://www.photoworkshop.com/canon/EOS_Digital.pdf

Yes, but I sometimes even use 500%, 0.3 radius, but then I don't just
use a simple USM. I apply the USM on a copied layer (allows to mask
out areas), and Luminosity blend it with the original (and perhaps
adjust the opacity) thus avoiding halo and providing maximum control.

Bart
Anonymous
April 9, 2005 1:54:22 AM

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

"Bart van der Wolf" <bvdwolf@no.spam> wrote in message
news:42567b5b$0$139$e4fe514c@news.xs4all.nl...
>
> "Skip M" <shadowcatcher@cox.net> wrote in message
> news:3nu5e.18153$lz2.15626@fed1read07...
> SNIP
>> That 300% figure is from the 1D suggestions.
>> http://www.photoworkshop.com/canon/EOS_Digital.pdf
>
> Yes, but I sometimes even use 500%, 0.3 radius, but then I don't just use
> a simple USM. I apply the USM on a copied layer (allows to mask out
> areas), and Luminosity blend it with the original (and perhaps adjust the
> opacity) thus avoiding halo and providing maximum control.
>
> Bart
The idea is to use whatever works best, but the 300%/0.3/0 is just a
starting place...

--
Skip Middleton
http://www.shadowcatcherimagery.com
Anonymous
April 12, 2005 4:00:18 PM

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

Bart van der Wolf wrote:

>
> "Skip M" <shadowcatcher@cox.net> wrote in message
> news:3nu5e.18153$lz2.15626@fed1read07...
> SNIP
>
>> That 300% figure is from the 1D suggestions.
>> http://www.photoworkshop.com/canon/EOS_Digital.pdf
>
>
> Yes, but I sometimes even use 500%, 0.3 radius, but then I don't just
> use a simple USM. I apply the USM on a copied layer (allows to mask out
> areas), and Luminosity blend it with the original (and perhaps adjust
> the opacity) thus avoiding halo and providing maximum control.

I avoid halo by simply keeping the sharpening down to just below halo
formation at clean edges. I don't know that this is *the* way to do it,
but it does seem effective for me. It is not often that I use
sharpenning of over 100%, and rare that it exceeds 200%, regardless of
radius.

Cheers,
Alan

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- slr-systems FAQ project: http://tinyurl.com/6m9aw
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: Remove FreeLunch.
Anonymous
April 12, 2005 4:03:14 PM

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

Bill Hilton wrote:

>>>Canon recommends a first pass USM (sharpening) run at 300% (amount),
>>>0.3 radius, 0 threshold to counteract the effects of the AA filter
>
>
>>Alan Browne writes
>>
>>Seems very high (300%)
>
>
> Not with that small a radius. This was the suggested starting point
> for the Professional bodies, with 1.3x and 1.0x crop factors so it
> might be a bit high for the 1.6x models, but it's Canon's "official"
> word.
>
>
>>USM effectiveness is very dependant on the
>>detail in the image.
>
>
> Wow, and some of us thought you knew nothing about digital. Guess we
> were wrong ... I did point out "These are suggested starting points and
> should be adequate for many images but not all." but I guess you didn't
> get to the second sentence?

Temper, temper. I was just surprised at the high sharpenning factor,
not the details of what you said. As others have pointed out, the AA
filter has a lot to with what the manuf selects for sharpening, and
possibly that high factor is very appropriate to the Canon in question.
FWIW I have no idea about the characteristics of the AA filter on the
Max 7D that I own, and I have yet to see a clear case of aliasing.

Cheers,
Alan.

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- slr-systems FAQ project: http://tinyurl.com/6m9aw
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: Remove FreeLunch.
Anonymous
April 13, 2005 7:19:04 AM

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

"Alan Browne" <alan.browne@freelunchVideotron.ca> wrote:

> FWIW I have no idea about the characteristics of the AA filter on the
> Max 7D that I own, and I have yet to see a clear case of aliasing.

Bart has a radial test chart that you can print out. It works great for
checking for aliasing.

If Bart could chip in with the link to the chart...

Here's the chart in action, but scanners don't show much aliasing.

http://www.xs4all.nl/~bvdwolf/main/foto/scan/se5400/se5...

David J. Littleboy
Tokyo, Japan
Anonymous
April 13, 2005 7:19:05 AM

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

"David J. Littleboy" <davidjl@gol.com> wrote in message
news:D 3h3ib$j79$1@nnrp.gol.com...
>
> "Alan Browne" <alan.browne@freelunchVideotron.ca> wrote:
>
>> FWIW I have no idea about the characteristics of the AA
>> filter on the Max 7D that I own, and I have yet to see a clear
>> case of aliasing.

Aliasing can "enhance" the visibility of grain/noise, and in general
reproduces fine spatial detail as larger detail, which thus becomes
more visible (larger than life). It depends on the image content if
and how it shows.

> Bart has a radial test chart that you can print out. It works
> great for checking for aliasing.
>
> If Bart could chip in with the link to the chart...

Sure, I have created a target file that is well suited for the testing
of Digcam sensor resolution (and aliasing tendency) with simple means,
you can make your own capture target from it at home with a decent
inkjet printer.
For HP/Canon inkjet printers (3.8MB):
http://www.xs4all.nl/~bvdwolf/main/downloads/Jtf60cy-10...
For Epson inkjet printers (5.3MB):
http://www.xs4all.nl/~bvdwolf/main/downloads/Jtf60cy-10...

Print it at the indicated ppi (highest/best quality, user settings)
without printer enhancements on glossy Photopaper which should produce
a 100x100mm target, and shoot it with your (digi)cam from a
(non-critical) distance like between 25-50x the focal length.

> Here's the chart in action, but scanners don't show much aliasing.
>
> http://www.xs4all.nl/~bvdwolf/main/foto/scan/se5400/se5...

Indeed, although that version of the target is even more (if not too)
aliasing critical than the newer one. The newer one, linked above,
which is better for comparing analog versus digital. The resolution
accuracy is the same (if not better). The JTF60cy* versions are better
suited for discrete sampling systems (like digicams).

Aliasing shows as hyperbolic deviations from the predictable radials
(because the aliases are always "larger than life", as smaller than 1
pixel detail cannot be reliably visualized other than as (mid-)gray
from this target). Any central blur "detail" that shows within the
38.2 pixels central diameter, is aliased (larger than life) detail.

Bart
Anonymous
April 13, 2005 2:17:20 PM

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

"Bart van der Wolf" <bvdwolf@no.spam> wrote:
> "David J. Littleboy" <davidjl@gol.com> wrote:
>
> > Bart has a radial test chart that you can print out. It works
> > great for checking for aliasing.
> >
> > If Bart could chip in with the link to the chart...
>
> Sure, I have created a target file that is well suited for the testing
> of Digcam sensor resolution (and aliasing tendency) with simple means,
> you can make your own capture target from it at home with a decent
> inkjet printer.
> For HP/Canon inkjet printers (3.8MB):
> http://www.xs4all.nl/~bvdwolf/main/downloads/Jtf60cy-10...
> For Epson inkjet printers (5.3MB):
> http://www.xs4all.nl/~bvdwolf/main/downloads/Jtf60cy-10...
>
> Print it at the indicated ppi (highest/best quality, user settings)
> without printer enhancements on glossy Photopaper which should produce
> a 100x100mm target, and shoot it with your (digi)cam from a
> (non-critical) distance like between 25-50x the focal length.

Thanks again for the chart.

> > Here's the chart in action, but scanners don't show much aliasing.
> >
> > http://www.xs4all.nl/~bvdwolf/main/foto/scan/se5400/se5...
>
> Indeed, although that version of the target is even more (if not too)
> aliasing critical than the newer one. The newer one, linked above,
> which is better for comparing analog versus digital. The resolution
> accuracy is the same (if not better). The JTF60cy* versions are better
> suited for discrete sampling systems (like digicams).
>
> Aliasing shows as hyperbolic deviations from the predictable radials
> (because the aliases are always "larger than life", as smaller than 1
> pixel detail cannot be reliably visualized other than as (mid-)gray
> from this target). Any central blur "detail" that shows within the
> 38.2 pixels central diameter, is aliased (larger than life) detail.

Your chart is nice for testing digital cameras for aliasing because the
aliasing jumps out of the image and hits you like a sledgehammer. I'd love
to see D100 vs. D70 images of that chart<g>. FWIW, any reasonably sharp lens
on the 300D shows gross aliasing with that chart. Sigh. Inversely, the lack
of aliasing shows gross softness on the part of the lens.

The cool thing about that chart is that it is _self-scaling_ in that at
whatever subject distance you image it (as long as it's close enough that
the outer edge area is resolved and far enough that the inner area is _not_
resolved), the resultant center part of the image will be identical. Much
easier than futzing with bar charts and calculating distances.

David J. Littleboy
Tokyo, Japan
Anonymous
April 13, 2005 2:17:21 PM

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

In message <d3hrri$pjs$1@nnrp.gol.com>,
"David J. Littleboy" <davidjl@gol.com> wrote:

>The cool thing about that chart is that it is _self-scaling_ in that at
>whatever subject distance you image it (as long as it's close enough that
>the outer edge area is resolved and far enough that the inner area is _not_
>resolved), the resultant center part of the image will be identical. Much
>easier than futzing with bar charts and calculating distances.

You have to watch the very center, though; it is messy in the center,
and if your magnification is high enough, you'll be looking at the
bitmap artifacts instead of your camera and lens' artifacts.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
April 13, 2005 2:40:34 PM

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

<JPS@no.komm> wrote:
> "David J. Littleboy" <davidjl@gol.com> wrote:
>
> >The cool thing about that chart is that it is _self-scaling_ in that at
> >whatever subject distance you image it (as long as it's close enough that
> >the outer edge area is resolved and far enough that the inner area is
_not_
> >resolved), the resultant center part of the image will be identical. Much
> >easier than futzing with bar charts and calculating distances.
>
> You have to watch the very center, though; it is messy in the center,
> and if your magnification is high enough, you'll be looking at the
> bitmap artifacts instead of your camera and lens' artifacts.

Exactly. That's why I said "and far enough that the inner area is _not_
resolved".

David J. Littleboy
Tokyo, Japan
Anonymous
April 13, 2005 2:40:35 PM

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

In message <d3ht6t$ptu$1@nnrp.gol.com>,
"David J. Littleboy" <davidjl@gol.com> wrote:

><JPS@no.komm> wrote:
>> "David J. Littleboy" <davidjl@gol.com> wrote:
>>
>> >The cool thing about that chart is that it is _self-scaling_ in that at
>> >whatever subject distance you image it (as long as it's close enough that
>> >the outer edge area is resolved and far enough that the inner area is
>_not_
>> >resolved), the resultant center part of the image will be identical. Much
>> >easier than futzing with bar charts and calculating distances.
>>
>> You have to watch the very center, though; it is messy in the center,
>> and if your magnification is high enough, you'll be looking at the
>> bitmap artifacts instead of your camera and lens' artifacts.
>
>Exactly. That's why I said "and far enough that the inner area is _not_
>resolved".

Sorry, I read your post quickly with taxes on my mind, and I never saw
anyone else make that distinction before, so it seemed that it needed to
be said.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
April 14, 2005 7:17:05 AM

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

<JPS@no.komm> wrote in message
news:o 91p51pp13jl82hv5ab36p7mjghr1he0s6@4ax.com...
> In message <d3ht6t$ptu$1@nnrp.gol.com>,
> "David J. Littleboy" <davidjl@gol.com> wrote:
>
>><JPS@no.komm> wrote:
SNIP
>>> You have to watch the very center, though; it is messy in
>>> the center, and if your magnification is high enough, you'll
>>> be looking at the bitmap artifacts instead of your camera
>>> and lens' artifacts.
>>
>>Exactly. That's why I said "and far enough that the inner
>> area is _not_ resolved".
>
> Sorry, I read your post quickly with taxes on my mind, and
> I never saw anyone else make that distinction before, so it
> seemed that it needed to be said.

My recommendation was/is to shoot the target at a distance of at least
25-50x the focal length. Longer is better, but with telelenses one may
run out of maneuvering space (especially indoors). This assumes that
the target is printed with adequate resolution. My prints show a
smaller than 2mm unresolved printed target center diameter.

If the printed target's unresolved center diameter is 2mm or less,
there are 9.55 cycles/mm (~485 ppi) of resolution, or more. The
targets are based on 600 or 720 ppi input, so there is a safety margin
for lower quality paper or slight nozzle misalignment. At 1:25
magnification that would mean an on-sensor projected resolution of 239
cycles/mm, which to me already seems a challenge for even the best
camera lenses. Shooting from a distance larger than 25x focal length
will only increase the safety marging for lower quality prints.

Therefore, *any* detail within a limiting resolution (Nyquist)
diameter of 38.2 pixels in the result will be aliasing or pseudo
detail. Due to the layout of the radial gradients, it will be easy to
see the aliasing because of the apparent hyperbolic change in
direction, phase, and increasing size of the "spokes".

So any "messy" center detail, *is* aliasing.

Bart

P.S. good luck with the tax forms :-(
Anonymous
April 14, 2005 12:08:21 PM

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

One other point on this ... the Canon USM recommendations assume the
output file wasn't sharpened during RAW conversion, yet many of the RAW
converters apply a default sharpening during conversion and you should
turn that off if you're going to apply a default USM to your images
later. For example the new RSE program has a default sharpening
setting of zero but that still applies some sharpening (which seems
counter-intuitive but there you go) ... you have to set it to -50 to
turn it off completely, or set up the preferences to skip sharpening
altogether. Phase One Capture One is similar in that a default
sharpening is applied unless you turn it off.

So if you're applying USM per the Canon instructions and the images
look over-sharpened with their settings then it might be because you're
actually applying sharpening twice, once with the converter.

Bill
Anonymous
April 20, 2005 2:33:19 AM

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

On Wed, 13 Apr 2005 10:17:20 +0900, "David J. Littleboy"
<davidjl@gol.com> wrote:

>
>"Bart van der Wolf" <bvdwolf@no.spam> wrote:
>> "David J. Littleboy" <davidjl@gol.com> wrote:
>>
>> > Bart has a radial test chart that you can print out. It works
>> > great for checking for aliasing.
>> >
>> > If Bart could chip in with the link to the chart...
>>
>> Sure, I have created a target file that is well suited for the testing
>> of Digcam sensor resolution (and aliasing tendency) with simple means,
>> you can make your own capture target from it at home with a decent
>> inkjet printer.
>> For HP/Canon inkjet printers (3.8MB):
>> http://www.xs4all.nl/~bvdwolf/main/downloads/Jtf60cy-10...
>> For Epson inkjet printers (5.3MB):
>> http://www.xs4all.nl/~bvdwolf/main/downloads/Jtf60cy-10...
>>
>> Print it at the indicated ppi (highest/best quality, user settings)
>> without printer enhancements on glossy Photopaper which should produce
>> a 100x100mm target, and shoot it with your (digi)cam from a
>> (non-critical) distance like between 25-50x the focal length.
>
>Thanks again for the chart.
>
>> > Here's the chart in action, but scanners don't show much aliasing.
>> >
>> > http://www.xs4all.nl/~bvdwolf/main/foto/scan/se5400/se5...
>>
>> Indeed, although that version of the target is even more (if not too)
>> aliasing critical than the newer one. The newer one, linked above,
>> which is better for comparing analog versus digital. The resolution
>> accuracy is the same (if not better). The JTF60cy* versions are better
>> suited for discrete sampling systems (like digicams).
>>
>> Aliasing shows as hyperbolic deviations from the predictable radials
>> (because the aliases are always "larger than life", as smaller than 1
>> pixel detail cannot be reliably visualized other than as (mid-)gray
>> from this target). Any central blur "detail" that shows within the
>> 38.2 pixels central diameter, is aliased (larger than life) detail.
>
>Your chart is nice for testing digital cameras for aliasing because the
>aliasing jumps out of the image and hits you like a sledgehammer. I'd love
>to see D100 vs. D70 images of that chart<g>. FWIW, any reasonably sharp lens
>on the 300D shows gross aliasing with that chart. Sigh. Inversely, the lack
>of aliasing shows gross softness on the part of the lens.
>
>The cool thing about that chart is that it is _self-scaling_ in that at
>whatever subject distance you image it (as long as it's close enough that
>the outer edge area is resolved and far enough that the inner area is _not_
>resolved), the resultant center part of the image will be identical. Much
>easier than futzing with bar charts and calculating distances.
>
>David J. Littleboy
>Tokyo, Japan
>
>
Although this subject has quit getting posts, I would like to reopen
it, for my own education. Frankly, I've heard about aliasing and
anti-aliasing for years, without ever really understanding what it
meant. SO - after downloading Bart's chart, and encouraging a friend
to do likewise, and we both made prints and photographed them with our
various digital cameras - I finally did a Google search to find out
just what those terms meant. OK, so now I know what anti-aliasing
does to fonts, etc. But, in spite of trying to translate all the
comments in the posts above, I have NO idea how to interpret any
photos we took of Bart's chart.

I took several as he decreed (actually 4 at 16 inch distance, with
different focal lengths) to match the 25X to 50X criteria he
specified. These first were with my Nikon CP-995. Then later I took
several with other cameras, a Samsung 3 MP fixed focus camera and an
early Canon 350 fixed focus camera. Naturally these had differences
in and around the center section, except for the 4 with the 995, which
were essentially identical. In all there was some blur of the spokes
out from the center. If the center was 38.2 pixels as Bart stated, I
checked it out as closer to 30 pixels on his original chart, but don't
make an issue here. In all the CP 995 shots, there was blurring out
to roughly 150 pixels diameter, and all were similar.

Now, my friend who used a Nikon 8800 (which has the RAW capability)
did something similar with it, using the Epson chart. He provided me
with several files, the one where he converted the RAW into a .TIF
file was almost identical to the original file from Bart. Even at
maximum magnification my viewer allowed. BUT when he let the camera
output as a .JPG file, and gave me that image, it was quite similar to
those from my CP-995. This even though the 8800 is an 8 MP camera,
and the 995 is only a 3 MP.

So - can either David or Bart explain to us how we should be
interpreting our results? We frankly don't know what we've
accomplished nor what it means. That paragraph that begins with
"Aliasing shows as hyperbolic deviation .........." seems to be the
key but we can't translate the meaning of the entire statement.

Would you please help us? Thanks.

Olin McDaniel
Anonymous
April 20, 2005 1:32:02 PM

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

"Olin K. McDaniel" <omcdaniel.abcd@mindspring.com> wrote:
> >
> Although this subject has quit getting posts, I would like to reopen
> it, for my own education. Frankly, I've heard about aliasing and
> anti-aliasing for years, without ever really understanding what it
> meant. SO - after downloading Bart's chart, and encouraging a friend
> to do likewise, and we both made prints and photographed them with our
> various digital cameras - I finally did a Google search to find out
> just what those terms meant. OK, so now I know what anti-aliasing
> does to fonts, etc. But, in spite of trying to translate all the
> comments in the posts above, I have NO idea how to interpret any
> photos we took of Bart's chart.

I sounds as though you haven't found the right definition of aliasing
yet<g>.

The keywords you need to use for google are "discrete sampling", "Nyquist",
and of course "aliasing". The basic idea is best explained not in terms of
images, but in terms of signals such as audio signals. In a digital image, a
line drawn through the image intersects varying levels of brightness. When
we talk about frequency in a digital image, we're talking about the
frequency of the variations in intensity along lines drawn in the image.

Here's the problem. If you sample an analog signal at points a fixed
distance (or time) apart (discrete sampling), the resultant digital signal
is only capable of representing frequencies up to the _Nyquist frequency_.
If the original analog signal includes frequencies above the Nyquist
frequncy, then those frequencies may appear in the digital signal as
frequencies within the range of frequencies that can be represented. Thus
the high frequencies appear as lower frequency "aliases". Which is why it's
called aliasing. To correctly sample an analog signal, you must low-pass
filter the analog signal so that the sampling system does not see the higher
frequencies. That's why a digital camera must include an antialiasing filter
in front of the sensor.

The short answer to your question is that if you look at the image, the
frequency of the image along the circumference of a circle with a 38.2 pixel
diameter (centered at the center of the pattern) will be the Nyquist
frequency for the sensor, and if the image is _not_ a smooth gray along that
circle (and inside that circle), some amount of aliasing will be occurring.
Similarly, the frequency of the image along the circumference of a circle
with a 76.4 pixel diameter circle will be one half the Nyquist frequency.
The camera should render the pattern nicely at that point. If it does not,
the lens resolution is probably the problem.

Does the above help???

David J. Littleboy
Tokyo, Japan
Anonymous
April 20, 2005 8:51:54 PM

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

On Wed, 20 Apr 2005 09:32:02 +0900, "David J. Littleboy"
<davidjl@gol.com> wrote:

>
>"Olin K. McDaniel" <omcdaniel.abcd@mindspring.com> wrote:
>> >
>> Although this subject has quit getting posts, I would like to reopen
>> it, for my own education. Frankly, I've heard about aliasing and
>> anti-aliasing for years, without ever really understanding what it
>> meant. SO - after downloading Bart's chart, and encouraging a friend
>> to do likewise, and we both made prints and photographed them with our
>> various digital cameras - I finally did a Google search to find out
>> just what those terms meant. OK, so now I know what anti-aliasing
>> does to fonts, etc. But, in spite of trying to translate all the
>> comments in the posts above, I have NO idea how to interpret any
>> photos we took of Bart's chart.
>
>I sounds as though you haven't found the right definition of aliasing
>yet<g>.
>
>The keywords you need to use for google are "discrete sampling", "Nyquist",
>and of course "aliasing". The basic idea is best explained not in terms of
>images, but in terms of signals such as audio signals. In a digital image, a
>line drawn through the image intersects varying levels of brightness. When
>we talk about frequency in a digital image, we're talking about the
>frequency of the variations in intensity along lines drawn in the image.
>
>Here's the problem. If you sample an analog signal at points a fixed
>distance (or time) apart (discrete sampling), the resultant digital signal
>is only capable of representing frequencies up to the _Nyquist frequency_.
>If the original analog signal includes frequencies above the Nyquist
>frequncy, then those frequencies may appear in the digital signal as
>frequencies within the range of frequencies that can be represented. Thus
>the high frequencies appear as lower frequency "aliases". Which is why it's
>called aliasing. To correctly sample an analog signal, you must low-pass
>filter the analog signal so that the sampling system does not see the higher
>frequencies. That's why a digital camera must include an antialiasing filter
>in front of the sensor.
>
>The short answer to your question is that if you look at the image, the
>frequency of the image along the circumference of a circle with a 38.2 pixel
>diameter (centered at the center of the pattern) will be the Nyquist
>frequency for the sensor, and if the image is _not_ a smooth gray along that
>circle (and inside that circle), some amount of aliasing will be occurring.
>Similarly, the frequency of the image along the circumference of a circle
>with a 76.4 pixel diameter circle will be one half the Nyquist frequency.
>The camera should render the pattern nicely at that point. If it does not,
>the lens resolution is probably the problem.
>
>Does the above help???
>
>David J. Littleboy
>Tokyo, Japan
>
>
Yes, it helps on the question of exactly how to utilize this chart
from an aliasing standpoint. (I think! But will have to do more
thinking yet, hopefully to BETTER grasp the concept.)

But can you address the other part of my post - why does my friend's
Nikon 8800 produce a .TIF image, from the RAW, that is almost
identical to the original image from Bart's .GIF file - yet any form
of .JPG from this same camera has what you define as a "lens
resolution" problem? It was the same lens in both instances. His
jpg image is no better than any of the 4 that I got with my 3 MP
Nikon CP 995, as far as that 150 pixel diameter blurred area is
concerned.

SO - what are we learning from our various images? Is Anti Aliasing
desired, or not?

Thanks,

Olin
Anonymous
April 21, 2005 1:01:42 PM

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

"Olin K. McDaniel" <omcdaniel.abcd@mindspring.com> wrote:
> >
> >Does the above help???
> >
> Yes, it helps on the question of exactly how to utilize this chart
> from an aliasing standpoint. (I think! But will have to do more
> thinking yet, hopefully to BETTER grasp the concept.)
>
> But can you address the other part of my post - why does my friend's
> Nikon 8800 produce a .TIF image, from the RAW, that is almost
> identical to the original image from Bart's .GIF file - yet any form
> of .JPG from this same camera has what you define as a "lens
> resolution" problem?

I've not met or heard of a camera who's jpegs were that bad. Are you sure
you didn't use a lower resolution for the jpegs???

> It was the same lens in both instances. His
> jpg image is no better than any of the 4 that I got with my 3 MP
> Nikon CP 995, as far as that 150 pixel diameter blurred area is
> concerned.

That 150 pixel diameter blur is seriously bad, and would correspond to
completely unacceptable images, so if the jpegs are working as photographs,
then you've done something wrong with your test.

> SO - what are we learning from our various images? Is Anti Aliasing
> desired, or not?

If the images show Moiré patterns, then the antialiasing filter is on the
inadequate side.

David J. Littleboy
Tokyo, Japan
Anonymous
April 25, 2005 1:07:23 AM

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

On Thu, 21 Apr 2005 09:01:42 +0900, "David J. Littleboy"
<davidjl@gol.com> wrote:

>
>"Olin K. McDaniel" <omcdaniel.abcd@mindspring.com> wrote:
>> >
>> >Does the above help???
>> >
>> Yes, it helps on the question of exactly how to utilize this chart
>> from an aliasing standpoint. (I think! But will have to do more
>> thinking yet, hopefully to BETTER grasp the concept.)
>>
>> But can you address the other part of my post - why does my friend's
>> Nikon 8800 produce a .TIF image, from the RAW, that is almost
>> identical to the original image from Bart's .GIF file - yet any form
>> of .JPG from this same camera has what you define as a "lens
>> resolution" problem?
>
>I've not met or heard of a camera who's jpegs were that bad. Are you sure
>you didn't use a lower resolution for the jpegs???
>
>> It was the same lens in both instances. His
>> jpg image is no better than any of the 4 that I got with my 3 MP
>> Nikon CP 995, as far as that 150 pixel diameter blurred area is
>> concerned.
>
>That 150 pixel diameter blur is seriously bad, and would correspond to
>completely unacceptable images, so if the jpegs are working as photographs,
>then you've done something wrong with your test.
>
>> SO - what are we learning from our various images? Is Anti Aliasing
>> desired, or not?
>
>If the images show Moiré patterns, then the antialiasing filter is on the
>inadequate side.
>
>David J. Littleboy
>Tokyo, Japan
>


David,

Over the last several days since your last post, I've done some
additional "studying" on the various subjects you suggested, but more
important, I discovered that some of the statements I made were not as
correct as I'd first thought.

So, let me clarify where I found I was wrong. First of all, my
friend's camera was not the Nikon 8800, rather it was (and is) the
Nikon D70. The lens he has is a zoom lens, from what I recall, about
24 mm to perhaps 70 mm.

Next mistake I made was a misunderstanding of what he said about the
..TIF image. (In fact, he was also mixed up) The TIF image that
looked so great was simply the original .GIF he had downloaded and run
through PhotoShop to convert from .GIF to .TIF.

After discovering all these mistakes, we agreed to set up another
session - working together, so as to see what each was doing, and
repeat some of the shots. I took pics of both my target printed on my
Canon i950 and of his target printed on his Epson 1280. And he took
similar shots of both our targets with his D70, in the RAW mode - plus
one more of his as a .JPG from the start. He saved the RAW files as
..TIF. I brought these 5 files home on a CD and examined them in
considerable detail. I conclude several things - first, my target is
noticeably sharper than his. Secondly, there is very little
difference between my .JPG from my CP-995 and his .TIF from his D70.
And these 2 are the only ones worthy of studying further. His .JPG
from the D70 is clearly inferior to either of the above. And all
taken of his target are inferior, regardless of camera. But neither
he nor I have adequately sorted out what is to be learned from these
photos of this target!!!! What physical measurement or any other
evaluation can be utilized to make meaningful judgment of all this?

At your suggestion several days ago, I began searching on the 3 key
phrases you suggested to get a better grasp of the science involved
(i.e. Aliasing, Discrete Sampling, and Nyquist). My background
includes electronics beginning in pre-WWII days, US Navy radar school,
an engineering degree in Chemical Engineering, and a career with
DuPont, plus a Radio Amateur since 1948 with particular emphasis on
constructing equipment and utilizing oscilloscopes and many other
pieces of hitech test equipment. So, I'm familiar with lots of Analog
theories, but admittedly have not kept up with the field since digital
became dominant. Thus, all this Nyquist, M.T.F., aliasing and non
aliasing discussion is Greek to me! One of the most useful web sites
I found in my search for knowledge was Norman Koren's site headed
"Understanding Image Sharpness", especially the sections that stressed
Resolution and M.T.F curves. But, unfortunately, I'm too far over the
hill to fully grasp how to APPLY it.

When I got involved in this "original" subject, shown above, I thought
that by downloading Bart's chart, and using it according to his
instructions - I would somehow magically learn how to evaluate my
several digital cameras. I'm still unsure what if anything useful has
come of all this effort - I still do not know what is good and what is
mediocre, using this test chart. When my CP-995 images seem so
similar to the D70 images, I really feel like we are missing something
important, but have no idea what. Or perhaps it's really not
important at all, if we get good photographs! But being of an
experimental inclination, I had to give it a try.

If this is boring and wearisome to you, I'll stop with this post, but
will say thanks for your tolerance so far.

Olin McDaniel, W4PFZ
(my somewhat outdated web site is:
http://home.mindspring.com/~omcdaniel/Index.htm)
Anonymous
April 25, 2005 1:13:11 PM

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

"Olin K. McDaniel" <omcdaniel.abcd@mindspring.com> wrote:
>
> At your suggestion several days ago, I began searching on the 3 key
> phrases you suggested to get a better grasp of the science involved
> (i.e. Aliasing, Discrete Sampling, and Nyquist). My background
> includes electronics beginning in pre-WWII days, US Navy radar school,
> an engineering degree in Chemical Engineering, and a career with
> DuPont, plus a Radio Amateur since 1948 with particular emphasis on
> constructing equipment and utilizing oscilloscopes and many other
> pieces of hitech test equipment.

Neat! I'm a bit of an analog/digital AC/DC type: as an undergrad, I avoided
the digital circuits lab all the other computer geeks were taking and took
the analog lab, and getting a radio license has been on my list of things to
do since high school (but never got around to). Nowadays I translate
technical documentation on ICs (Japanese to English), and since one of my
customers does a lot of motor driver ICs, the analog stuff comes in useful.

> So, I'm familiar with lots of Analog
> theories, but admittedly have not kept up with the field since digital
> became dominant. Thus, all this Nyquist, M.T.F., aliasing and non
> aliasing discussion is Greek to me! One of the most useful web sites
> I found in my search for knowledge was Norman Koren's site headed
> "Understanding Image Sharpness", especially the sections that stressed
> Resolution and M.T.F curves. But, unfortunately, I'm too far over the
> hill to fully grasp how to APPLY it.

Keep at it. One of the things I remember reading in a psychology class is
that human intellectual capacity peaks at 23 or so, but that (unless one
suffers major disease or explicit physical damage to the brain) the decrease
in capacity from 23 to 90 is much less than the increase from 17 to 23.

Age is simply not an excuse.

> If this is boring and wearisome to you, I'll stop with this post, but
> will say thanks for your tolerance so far.

The basic ideas of digital signal processing and imaging really do not go
beyond anything in the analog realm. Fourier figured it all out, and that's
long before even your time<g>. The only thing that's different in digital is
that there's an upper limit to the frequencies that a digital system can
represent (the Nyquist frequency), and that presenting a digital sensor with
an analog frequency above that limit causes the digital system to generate
artifacts (which is why you need an antialiasing filter (which is just a
low-pass filter)).

Bart's chart is simply an easy way of looking for those artifacts. If you
see anything inside a circle 38.2 pixels in diameter, it's an artifact. The
pattern outside that circle should look like a lower contrast version of the
original chart; patterns that differ from the chart pattern are artifacts.

David J. Littleboy
Tokyo, Japan
Anonymous
April 29, 2005 2:59:12 PM

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

"Olin K. McDaniel" <omcdaniel.abcd@mindspring.com> wrote in message
news:426bcfd1.7584583@news.east.earthlink.net...
....
> noticeably sharper than his. Secondly, there is very little
> difference between my .JPG from my CP-995 and his .TIF from his D70.
> And these 2 are the only ones worthy of studying further. His .JPG
> from the D70 is clearly inferior to either of the above. And all
....

Olin:

Two things you might want to consider:

1. Conditions. Under ideal conditions (good lighting), point and shoot
cameras do a marvelous job, as do DSLR cameras. But as conditions
get worse, the far more capable sensors found in DSLRs will begin
to assert themselves. Particularly in low light situations, you should
see that the DSLR image will have far less noise than the point and
shoot camera image.

2. Settings. By default, point and shoot cameras do considerable
processing (color balance, saturation, sharpness, contrast, brightness)
in camera. DSLRs by default often don't apply those adjustments,
prefering to leave them for post-processing. Applying similar adjustments
to the DSLR image will probably "improve" it in the same way that
the point and shoot camera "improved" its image before you saw it.


--
Dan (Woj...) [dmaster](no space)[at](no space)[lucent](no space)[dot](no
space)[com]
===============================
"The sky turned to black / Would he ever come back?
They would climb a high dune / They would pray to the moon
But he'd never return / So the sisters would burn
As their eyes searched the land / With their cups full of sand"
Anonymous
May 1, 2005 6:21:58 AM

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

On Fri, 29 Apr 2005 10:59:12 -0500, "Dan Wojciechowski"
<too.much@spam.com> wrote:

>"Olin K. McDaniel" <omcdaniel.abcd@mindspring.com> wrote in message
>news:426bcfd1.7584583@news.east.earthlink.net...
>...
>> noticeably sharper than his. Secondly, there is very little
>> difference between my .JPG from my CP-995 and his .TIF from his D70.
>> And these 2 are the only ones worthy of studying further. His .JPG
>> from the D70 is clearly inferior to either of the above. And all
>...
>
>Olin:
>
>Two things you might want to consider:
>
>1. Conditions. Under ideal conditions (good lighting), point and shoot
> cameras do a marvelous job, as do DSLR cameras. But as conditions
> get worse, the far more capable sensors found in DSLRs will begin
> to assert themselves. Particularly in low light situations, you should
> see that the DSLR image will have far less noise than the point and
> shoot camera image.
>
>2. Settings. By default, point and shoot cameras do considerable
> processing (color balance, saturation, sharpness, contrast, brightness)
> in camera. DSLRs by default often don't apply those adjustments,
> prefering to leave them for post-processing. Applying similar adjustments
> to the DSLR image will probably "improve" it in the same way that
> the point and shoot camera "improved" its image before you saw it.
>
>
>--
>Dan (Woj...) [dmaster](no space)[at](no space)[lucent](no space)[dot](no
>space)[com]


Dan,

Thanks for your comments. I accept and agree with almost everything
you mentioned, except perhaps one. And I emphasize the "perhaps",
since I haven't the slightest bit of experience with the D70. My
natural presumption was that when he set it up to do a .JPG, that it
would then automatically do some of the "in camera" tweaking that
point and shoot cameras do. If the D70 does not, then that probably
explains the comparatively poor .JPG he obtained. Naturally, his
RAW shots converted to .TIF were not supposed to do any of this, as
you point out. So, comparing my .JPGs to his .TIFs is really not
reasonable, is it?

Olin
!