Sign in with
Sign up | Sign in
Your question

Canon 20D Low Noise?

Last response: in Digital Camera
Share
July 2, 2005 1:22:07 PM

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

Is it the sensor that has the lower noise or has it been proven that
software is getting the noise lower?
Is this why the pictures are soft.

More about : canon 20d low noise

July 2, 2005 7:17:54 PM

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

SteveJ wrote:

> Is it the sensor that has the lower noise or has it been proven that
> software is getting the noise lower?
>

Seems to be a mix of both sensor and "firmware". Canon users will claim that
there is no NR being done in the "digic" chip, I don't buy it. If you look
at resolution tests done at various ISO levels, something is going on.

--

Stacey
Anonymous
July 2, 2005 8:30:13 PM

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

SteveJ napisa³(a):
> Is it the sensor that has the lower noise or has it been proven that
> software is getting the noise lower?
> Is this why the pictures are soft.

sensor technology...
heh i don't think that there is a heavier noise reduction than for
example in 300d, d70.

It goes like this that on iso 100 you also don't have noise and also the
photos look silky - so it's rather the matter of noise that
photographies look plastic/soft or not...

--
..........Marek Mollin "rogus".........
...http://rogus.atspace.com/da/ad.jpg..
.............coming soon...............
...............Pozdrawiam..............
Related resources
Anonymous
July 3, 2005 4:08:12 AM

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

SteveJ wrote:
> Is it the sensor that has the lower noise or has it been proven that
> software is getting the noise lower?
> Is this why the pictures are soft.
The jpeg files from all cameras are very much processed. The raw files
are not, at least to the best of my knowledge, until you convert them
to a image file.

I will say that I have not seen the soft pictures at higher ISO that
you are talking about.

Scott
Anonymous
July 3, 2005 8:16:20 AM

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

Ryadia wrote:
> Stacey wrote:
> > SteveJ wrote:
> >
> >
> >> Is it the sensor that has the lower noise or has it been proven that
> >>software is getting the noise lower?
> >>
> >
> >
> > Seems to be a mix of both sensor and "firmware". Canon users will claim that
> > there is no NR being done in the "digic" chip, I don't buy it. If you look
> > at resolution tests done at various ISO levels, something is going on.
> >
>
>
> Either Canon are post processing in the camera to reduce noise or the
> higher the ISO, the lower the resolving power of their 20D sensor.
>
> Either way, Canon are deceiving the public. If they are applying a
> de-noising algorithm, the image is losing detail in the process and it
> can't be switched off. If they are not, the Sensor cannot resolve the
> same detail at 400 or 800 ISO as it does at 100 ISO. Choose your poison.
>
> Please yourself which one. Interesting enough, I can enlarge high noise,
> Nikon D100 images shot at ISO 400, larger than I can get a 400 ISO, 20D
> image and produce fine detail. The Nikon images have a hell of a lot of
> noise but they also have a lot of detail.
>
> Use neat image on the file and you get rid of the noise but keep most of
> the detail. I can also coax up a 400 ISO 10D image and get a more highly
> detailed result than I can a 20D image.
>
> http://www.technoaussie.com/examples.htm Our business is enlarging
> digital images. We print between 5 and 20 super enlargements a week (48"
> and 42" wide) from all kinds of digital and scanned images.
>
> Please yourself if you choose to believe or not believe we have a clue
> about what we make a living from. Assuming the former, you might be
> surprised which images from which cameras and what size (Megapixel)
> files make the clearest, sharpest poster prints.
>
> Douglas

Here is the problem that many of us are having, you say that the D100
has a lot more detail then the 20D, you put this forth as a fact but
don't tell us what if any testing you have don't to come to this
conclusion, you offer no images to look at. Is it posible the images
you got from the 20D used a poor lens or there was a focus problem or
motion blur?

We look at any review out there where they have shot the same scene
with the 20D and the D100 and the 20D clearly is showing more detail.

I look at the test photos at higher ISOs and still the 20D has more
detail.

Can anyone here point us to any review that shows the D100 showing more
detail then the 20D?

Scott
Anonymous
July 3, 2005 1:02:43 PM

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

J...@no.komm wrote:
> In message <1120374492.498421.218750@g47g2000cwa.googlegroups.com>,
> "Scott W" <biphoto@hotmail.com> wrote:
>
> >I will say that I have not seen the soft pictures at higher ISO that
> >you are talking about.
>
> We can look at RAW data, but the difference in the levels between the 3
> channels causes a checker pattern, greater in intensity than the noise.
> It would be rather difficult to scale the channels without affecting the
> noise. One way to look at noise might be to use a light source colored
> to cause equal response in all channels for a given sensor, assuming
> perfect linearity of transmission at all intensities. Short of that,
> you could take the green channel and rotate it to regular bitmap form,
> but that does not contain horizontal or vertical neighboring pixels from
> the original.
>
> You could convert to color in IRIS, but IRIS interpolates each channel
> independently, rather than demosaicing, I believe, so it's basically 3
> soft color channels layered.
> --
I was not thinking of looking at the raw data directly, but if you did
the easiest thing to do is just pick on pixel out of every 2 x 2 group,
so you would make a photo that was 1/2 in size in both vertical and
horizontal.

Scott
July 3, 2005 1:36:04 PM

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

"Scott W" <biphoto@hotmail.com> wrote in message
news:1120374492.498421.218750@g47g2000cwa.googlegroups.com...
> SteveJ wrote:
>> Is it the sensor that has the lower noise or has it been proven that
>> software is getting the noise lower?
>> Is this why the pictures are soft.
> The jpeg files from all cameras are very much processed. The raw files
> are not, at least to the best of my knowledge, until you convert them
> to a image file.
>
> I will say that I have not seen the soft pictures at higher ISO that
> you are talking about.
>
> Scott
>

I'm not even sure he means at higher ISO's, but in general.

I've looked at sample images in steve's digicams and compared 20d vs 70s (as
an example), the 20D noise is better
and I believe there is definitely more detail in the 20D examples at various
ISOs, as expected.
Try looking at the temperature gauge in the still life pics.

Who cares if they use software, hardware or black magic, to improve the
pics, IMO it's the output that matters.
Anonymous
July 3, 2005 7:57:25 PM

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

In message <1120374492.498421.218750@g47g2000cwa.googlegroups.com>,
"Scott W" <biphoto@hotmail.com> wrote:

>I will say that I have not seen the soft pictures at higher ISO that
>you are talking about.

We can look at RAW data, but the difference in the levels between the 3
channels causes a checker pattern, greater in intensity than the noise.
It would be rather difficult to scale the channels without affecting the
noise. One way to look at noise might be to use a light source colored
to cause equal response in all channels for a given sensor, assuming
perfect linearity of transmission at all intensities. Short of that,
you could take the green channel and rotate it to regular bitmap form,
but that does not contain horizontal or vertical neighboring pixels from
the original.

You could convert to color in IRIS, but IRIS interpolates each channel
independently, rather than demosaicing, I believe, so it's basically 3
soft color channels layered.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 3, 2005 8:49:27 PM

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

In message <1120406563.672418.255290@g14g2000cwa.googlegroups.com>,
"Scott W" <biphoto@hotmail.com> wrote:

>I was not thinking of looking at the raw data directly, but if you did
>the easiest thing to do is just pick on pixel out of every 2 x 2 group,
>so you would make a photo that was 1/2 in size in both vertical and
>horizontal.

Easily done, and I've already done it. Just load an uncompressed .dng
as .raw in PS, and use filter factory to separate the bayer components
into separate quadrants. Or, render the RAW to .bmp with IRIS, rather
than make the intermediate .dng.

The contrast of noise at the nyquist is missing, though, and the
difference between the noise at half the nyquist and at the nyquist is
exactly the kind of thing that might help determine if NR were occuring.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 3, 2005 9:44:46 PM

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

Randall Ainsworth wrote:
> In article <3irai6Fn13poU3@individual.net>, Stacey <fotocord@yahoo.com>
> wrote:
>
> > You'll never convince a canon user, they'll just find a "Pro canon" review
> > site that backs up their purchase choice.
>
> Most of us don't give two squats about the technical mumbo-jumbo. We
> look at the end results. I have 16x20s and 20x24s from my 10D that tell
> me the system is working just fine. And it will beat a Sigma any day of
> the week.

I was going to say beating the Sigma was not setting the bar very high,
but that would be a cheap shot so I won't say that. In truth the Sigma
is a bit behind on just about all aspects of image quality when
compared to the 10D, but not all that much behind.

I am not sure where we are supposed to find a non pro cannon site,
Stacey has yet to pointus to that site.

Scott
Anonymous
July 3, 2005 10:53:46 PM

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

Stacey wrote:
> SteveJ wrote:
>
>
>> Is it the sensor that has the lower noise or has it been proven that
>>software is getting the noise lower?
>>
>
>
> Seems to be a mix of both sensor and "firmware". Canon users will claim that
> there is no NR being done in the "digic" chip, I don't buy it. If you look
> at resolution tests done at various ISO levels, something is going on.
>


Either Canon are post processing in the camera to reduce noise or the
higher the ISO, the lower the resolving power of their 20D sensor.

Either way, Canon are deceiving the public. If they are applying a
de-noising algorithm, the image is losing detail in the process and it
can't be switched off. If they are not, the Sensor cannot resolve the
same detail at 400 or 800 ISO as it does at 100 ISO. Choose your poison.

Please yourself which one. Interesting enough, I can enlarge high noise,
Nikon D100 images shot at ISO 400, larger than I can get a 400 ISO, 20D
image and produce fine detail. The Nikon images have a hell of a lot of
noise but they also have a lot of detail.

Use neat image on the file and you get rid of the noise but keep most of
the detail. I can also coax up a 400 ISO 10D image and get a more highly
detailed result than I can a 20D image.

http://www.technoaussie.com/examples.htm Our business is enlarging
digital images. We print between 5 and 20 super enlargements a week (48"
and 42" wide) from all kinds of digital and scanned images.

Please yourself if you choose to believe or not believe we have a clue
about what we make a living from. Assuming the former, you might be
surprised which images from which cameras and what size (Megapixel)
files make the clearest, sharpest poster prints.

Douglas
July 3, 2005 10:53:47 PM

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

> Either Canon are post processing in the camera to reduce noise or the
> higher the ISO, the lower the resolving power of their 20D sensor.
>
> Either way, Canon are deceiving the public. If they are applying a
> de-noising algorithm, the image is losing detail in the process and it
> can't be switched off. If they are not, the Sensor cannot resolve the same
> detail at 400 or 800 ISO as it does at 100 ISO. Choose your poison.
>
> Please yourself which one. Interesting enough, I can enlarge high noise,
> Nikon D100 images shot at ISO 400, larger than I can get a 400 ISO, 20D
> image and produce fine detail. The Nikon images have a hell of a lot of
> noise but they also have a lot of detail.

Taking example images from steve's digicams for 20D and 100D at 400 iso
doesn't bear out your comments.
Do you have some examples ?

see www.knighttrain.freeserve.co.uk/20d100.htm
July 3, 2005 11:27:02 PM

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

Ryadia wrote:

>
>
> Either Canon are post processing in the camera to reduce noise or the
> higher the ISO, the lower the resolving power of their 20D sensor.
>
> Either way, Canon are deceiving the public. If they are applying a
> de-noising algorithm, the image is losing detail in the process and it
> can't be switched off. If they are not, the Sensor cannot resolve the
> same detail at 400 or 800 ISO as it does at 100 ISO. Choose your poison.
>


You'll never convince a canon user, they'll just find a "Pro canon" review
site that backs up their purchase choice.
--

Stacey
Anonymous
July 3, 2005 11:27:03 PM

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

In article <3irai6Fn13poU3@individual.net>, Stacey <fotocord@yahoo.com>
wrote:

> You'll never convince a canon user, they'll just find a "Pro canon" review
> site that backs up their purchase choice.

Most of us don't give two squats about the technical mumbo-jumbo. We
look at the end results. I have 16x20s and 20x24s from my 10D that tell
me the system is working just fine. And it will beat a Sigma any day of
the week.
Anonymous
July 4, 2005 6:45:55 AM

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

In message <3irai6Fn13poU3@individual.net>,
Stacey <fotocord@yahoo.com> wrote:

>You'll never convince a canon user, they'll just find a "Pro canon" review
>site that backs up their purchase choice.

No; I look at the RAW data from a blackframe, and it looks like noise to
me. I look at high ISO images form other coameras, and don't see any
more detail.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 5, 2005 8:09:26 PM

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

Stacey wrote:
> JPS@no.komm wrote:
>
> > I look at high ISO images form other coameras, and don't see any
> > more detail.
>
> So the detail at 100ISO is equal to the 1600 ISO ones after being processed
> in the camera? That is the question here, people claiming there is -NO- NR
> being done in the camera.
>
> --
>
> Stacey

I don't see any loss in detail between ISO 100 and ISO 1600. Of course
there are more limits on the sharpening that I can do but the photos
out of the camera look pretty much the same.

Can you point us to the photos that show this loss?

Scott
July 5, 2005 10:20:36 PM

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

JPS@no.komm wrote:

> I look at high ISO images form other coameras, and don't see any
> more detail.

So the detail at 100ISO is equal to the 1600 ISO ones after being processed
in the camera? That is the question here, people claiming there is -NO- NR
being done in the camera.

--

Stacey
Anonymous
July 6, 2005 1:01:07 AM

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

"dylan" <no@nowhere.com> wrote in message
news:D a8822$jv3$1@newsg4.svr.pol.co.uk...

> I'm not even sure he means at higher ISO's, but in general.
>
> I've looked at sample images in steve's digicams and compared 20d vs 70s
(as
> an example), the 20D noise is better
> and I believe there is definitely more detail in the 20D examples at
various
> ISOs, as expected.
> Try looking at the temperature gauge in the still life pics.
>
> Who cares if they use software, hardware or black magic, to improve the
> pics, IMO it's the output that matters.
>

ROFL - it must be black magic but I'll take it! They should charge even
more and advertise the magical powers of this camera on the box!
Anonymous
July 6, 2005 5:02:12 AM

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

In message <3j0fdkFnt12mU4@individual.net>,
Stacey <fotocord@yahoo.com> wrote:

>JPS@no.komm wrote:

>> I look at high ISO images form other coameras, and don't see any
>> more detail.

>So the detail at 100ISO is equal to the 1600 ISO ones after being processed
>in the camera?

Noise masks detail, period, regardless of whether any AI noise reduction
or median filtering, or whatever, is done. Noise raises the uncertainty
of pixel values. The lower the SNR, the more likely that that speck was
not on the subject, but is an artifact of amplifying a low signal.

>That is the question here, people claiming there is -NO- NR
>being done in the camera.

What was being said here was that Canon DSLRs that sport low noise at
high ISOs do so by low-quality noise reduction. The thing to compare is
detail in other cameras at high ISOs, *not* detail at high and low ISOs
on the same camera.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
July 6, 2005 8:19:19 PM

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

JPS@no.komm wrote:

>*not* detail at high and low ISOs
> on the same camera.

Why not? If the details are being smeared, then they are just doing NR in
camera not some "Magical technology" like most canon users want to believe.
And yes I see more -detail- in nikon files at high ISO than I do in the
canon samples I've seen.

--

Stacey
July 6, 2005 8:34:27 PM

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

Scott W wrote:


>
> I don't see any loss in detail between ISO 100 and ISO 1600.


The above is one of the most absurd claims I've seen anyone make here!

http://www.dpreview.com/reviews/canoneos350d/page19.asp

Look at the ISO 100 shots vs the ISO 1600 shots, look at her crown and if
you can't see the loss of detail, yes you are blind!

--

Stacey
Anonymous
July 6, 2005 9:08:46 PM

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

Stacey dribbles on his bib:

> JPS@no.komm wrote:
>
> >*not* detail at high and low ISOs
> > on the same camera.
>
> Why not? If the details are being smeared, then they are just doing NR in
> camera not some "Magical technology" like most canon users want to believe.

Read what JPS wrote: noise degrades detail. Take any low-noise image,
add noise, and observe the "loss of detail".

> And yes I see more -detail- in nikon files at high ISO than I do in the
> canon samples I've seen.

You've seen _noise_ in Nikon files, nitwit.
Anonymous
July 7, 2005 2:05:27 AM

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

"Stacey" <fotocord@yahoo.com> wrote in message
news:3j2tikFmpbn7U1@individual.net...
> Scott W wrote:
>
>
> >
> > I don't see any loss in detail between ISO 100 and ISO 1600.
>
>
> The above is one of the most absurd claims I've seen anyone make here!
>
> http://www.dpreview.com/reviews/canoneos350d/page19.asp
>
> Look at the ISO 100 shots vs the ISO 1600 shots, look at her crown and if
> you can't see the loss of detail, yes you are blind!
>
> --
>
> Stacey

"Image sharpness is also virtually identical with a slight softening at ISO
800 and 1600 but NOTHING LIKE THE LEVELS WE HAVE SEEN FROM OTHER CAMERAS. "
quote from your link....
Anonymous
July 7, 2005 3:42:39 AM

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

"Dirty Harry" <NOJUNK@FU.ca> wrote:

>"Image sharpness is also virtually identical with a slight softening at ISO
>800 and 1600 but NOTHING LIKE THE LEVELS WE HAVE SEEN FROM OTHER CAMERAS. "
>quote from your link....


If you were sponsored by a camera manufacturer, you wouldn't want to
upset them by posting derogatory remarks about their equipment on your
web site, would you?

;-)
July 7, 2005 3:42:40 AM

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

Tony Polson wrote:

> "Dirty Harry" <NOJUNK@FU.ca> wrote:
>
>>"Image sharpness is also virtually identical with a slight softening at
>>ISO 800 and 1600 but NOTHING LIKE THE LEVELS WE HAVE SEEN FROM OTHER
>>CAMERAS. " quote from your link....
>
>
> If you were sponsored by a camera manufacturer, you wouldn't want to
> upset them by posting derogatory remarks about their equipment on your
> web site, would you?
>


Exactly, if you actually look at the images rathen then reading their -PAID
FOR- text advertising, you see a different story..

I like the "Image sharpness is virtually identical except for slight
softening" like how can it be the same if it's softer?

--

Stacey
July 7, 2005 5:59:04 AM

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

eawckyegcy@yahoo.com wrote:

> Stacey dribbles on his bib:
>
>> JPS@no.komm wrote:
>>
>> >*not* detail at high and low ISOs
>> > on the same camera.
>>
>> Why not? If the details are being smeared, then they are just doing NR in
>> camera not some "Magical technology" like most canon users want to
>> believe.
>
> Read what JPS wrote: noise degrades detail. Take any low-noise image,
> add noise, and observe the "loss of detail".

Or it's smeared to hide the noise. Or do you agree with Scott that there is
ZERO quality difference between ISO 100 and ISO 1600?

>
>> And yes I see more -detail- in nikon files at high ISO than I do in the
>> canon samples I've seen.
>
> You've seen _noise_ in Nikon files, nitwit.

And I seen smearing of details in Canon files.

Back to personal insults I see?

--

Stacey
Anonymous
July 7, 2005 1:21:28 PM

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

Stacey sputters:

>> Read what JPS wrote: noise degrades detail. Take any low-noise image,
>> add noise, and observe the "loss of detail".
>
> Or it's smeared to hide the noise.

One more time: read what JPS wrote. Noise degrades detail. It doesn't
matter what you do with the image once the noise is in. If you
anisotropically remove some of the noise, detail is lost. If you just
leave the image as is, detail is lost. If you send the image to be
blessed by the Pope, detail is lost.

> Or do you agree with Scott that there is
> ZERO quality difference between ISO 100 and ISO 1600?

I have no idea who this "Scott" is, nor what his claims actually are
(your proven difficulties with reading comprehension, and your patchy
treatment of the truth cause me to disbelieve almost all of your
assertions). Even if I was acutely aware of his position, whatever it
may be, whether or not I agree with it has absolutely no impact on the
fact that detail is lost.

> >> And yes I see more -detail- in nikon files at high ISO than I do in the
> >> canon samples I've seen.
> >
> > You've seen _noise_ in Nikon files, nitwit.
>
> And I seen smearing of details in Canon files.

You are an ineducable nitwit.

> Back to personal insults I see?

The truth is never an insult.
Anonymous
July 7, 2005 3:08:47 PM

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

Stacey wrote:
> eawckyegcy@yahoo.com wrote:
>
> > Stacey dribbles on his bib:
> >
> >> JPS@no.komm wrote:
> >>
> >> >*not* detail at high and low ISOs
> >> > on the same camera.
> >>
> >> Why not? If the details are being smeared, then they are just doing NR in
> >> camera not some "Magical technology" like most canon users want to
> >> believe.
> >
> > Read what JPS wrote: noise degrades detail. Take any low-noise image,
> > add noise, and observe the "loss of detail".
>
> Or it's smeared to hide the noise. Or do you agree with Scott that there is
> ZERO quality difference between ISO 100 and ISO 1600?
>
> >
> >> And yes I see more -detail- in nikon files at high ISO than I do in the
> >> canon samples I've seen.
> >
> > You've seen _noise_ in Nikon files, nitwit.
>
> And I seen smearing of details in Canon files.
>
> Back to personal insults I see?
>
> --
>
> Stacey

Well let's start with what I said
"I don't see any loss in detail between ISO 100 and ISO 1600. Of course

there are more limits on the sharpening that I can do but the photos
out of the camera look pretty much the same."

I did not say "ZERO quality difference between ISO 100 and ISO 1600"
there is more noise on the ISO 1600 and I did point out this will limit
shapening.

So do the photos look pretty much the same, I will let others decide.

I did two quick snap shots, one at iso 100 and one at iso 1600

ISO 100
http://www.sewcon.com/photos/IMG_8276.JPG

ISO 1600
http://www.sewcon.com/photos/IMG_8281.JPG

Print these out at 8 x 10 and I would bet you would have a hard time
telling them apart.

Scott
Anonymous
July 7, 2005 4:36:21 PM

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

Stacey blithers some more:

> > One more time: read what JPS wrote. Noise degrades detail.
>
> Duh? Smeared details and grainy/noisy details aren't the same thing when
> you're looking ate final print.

I'll repeat myself: you are an ineducable nitwit.

But hey, if you "like" looking at noise, you are certainly free to add
some back to the image. Why, I'll even write a program to do it! I'll
call it the "DetailEmphasizer(TM)" and make it a PhotoSlop plug-in, and
sell it to brainless nitwits like you for $100 a pop. Maybe a special
$500 "Professional" version -- for those so-called 'pros' who need that
'film' look...
Anonymous
July 7, 2005 5:22:16 PM

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

Stacey <fotocord@yahoo.com> wrote:
>
>Exactly, if you actually look at the images rathen then reading their -PAID
>FOR- text advertising, you see a different story..
>
>I like the "Image sharpness is virtually identical except for slight
>softening" like how can it be the same if it's softer?


They are weasel words.

They are the kind of words that say nothing, but if the writer is ever
accused of not stating something significant, he/she can point to them
and say that he/she did state it.

For example, using the words above, no-one could accuse the writer of
not mentioning the softness. However, it is mentioned - very
carefully - in such a way that no-one would think it significant.

More important, it is mentioned in such a way that no sponsor could
possibly be offended.

No wonder that the most damning criticism to be seen is a rating of
"Above Average".
July 7, 2005 6:04:56 PM

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

eawckyegcy@yahoo.com wrote:

> Stacey sputters:
>
>>> Read what JPS wrote: noise degrades detail. Take any low-noise image,
>>> add noise, and observe the "loss of detail".
>>
>> Or it's smeared to hide the noise.
>
> One more time: read what JPS wrote. Noise degrades detail.

Duh? Smeared details and grainy/noisy details aren't the same thing when
you're looking ate final print.

> It doesn't
> matter what you do with the image once the noise is in.

Seems to some people it does. Because it's been removed in the camera, they
assume it never existed. I'd rather be able to do this myself, selectively
if needed, some people I suppose can't fathom this is easily done?


> You are an ineducable nitwit.


Hope this gave you a much needed boost to your ego for the day!
--

Stacey
Anonymous
July 7, 2005 8:17:03 PM

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

Stacey <fotocord@yahoo.com> wrote:
> Scott W wrote:
>>
>> I don't see any loss in detail between ISO 100 and ISO 1600.
>
> The above is one of the most absurd claims I've seen anyone make here!
>
> http://www.dpreview.com/reviews/canoneos350d/page19.asp
>
> Look at the ISO 100 shots vs the ISO 1600 shots, look at her crown and if
> you can't see the loss of detail, yes you are blind!

Bad comparison. You're comparing a shot at 1/100s to a shot at
1/1600s -- of course you're going to get a slight loss of detail with
four stops less light to work with. Try shooting that again at
1/1600s at ISO100 and then compare how much detail you get; I bet the
detail at ISO1600 will be much greater.

--
Zed Pobre <zed@resonant.org> a.k.a. Zed Pobre <zed@debian.org>
PGP key and fingerprint available on finger; encrypted mail welcomed.
July 7, 2005 8:17:04 PM

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

Zed Pobre wrote:

> Stacey <fotocord@yahoo.com> wrote:
>> Scott W wrote:
>>>
>>> I don't see any loss in detail between ISO 100 and ISO 1600.
>>
>> The above is one of the most absurd claims I've seen anyone make here!
>>
>> http://www.dpreview.com/reviews/canoneos350d/page19.asp
>>
>> Look at the ISO 100 shots vs the ISO 1600 shots, look at her crown and if
>> you can't see the loss of detail, yes you are blind!
>
> Bad comparison. You're comparing a shot at 1/100s to a shot at
> 1/1600s -- of course you're going to get a slight loss of detail with
> four stops less light to work with.

Well that's what the higher ISO is for, shooting with less light. Do you
shoot at ISO 1600 and then over expose it 4 stops rather than use ISO 100?


> Try shooting that again at
> 1/1600s at ISO100 and then compare how much detail you get;

You lost me there, what is the point in underexposing the ISO100 shot? Or do
you mean open up the lens to get the correct exposure at 1/1600s which
isn't going to do anything either.

--

Stacey
Anonymous
July 7, 2005 10:42:33 PM

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

RichA wrote:
> On 7 Jul 2005 11:08:47 -0700, "Scott W" <biphoto@hotmail.com> wrote:
>
> >Stacey wrote:
> >> eawckyegcy@yahoo.com wrote:
> >>
> >> > Stacey dribbles on his bib:
> >> >
> >> >> JPS@no.komm wrote:
> >> >>
> >> >> >*not* detail at high and low ISOs
> >> >> > on the same camera.
> >> >>
> >> >> Why not? If the details are being smeared, then they are just doing NR in
> >> >> camera not some "Magical technology" like most canon users want to
> >> >> believe.
> >> >
> >> > Read what JPS wrote: noise degrades detail. Take any low-noise image,
> >> > add noise, and observe the "loss of detail".
> >>
> >> Or it's smeared to hide the noise. Or do you agree with Scott that there is
> >> ZERO quality difference between ISO 100 and ISO 1600?
> >>
> >> >
> >> >> And yes I see more -detail- in nikon files at high ISO than I do in the
> >> >> canon samples I've seen.
> >> >
> >> > You've seen _noise_ in Nikon files, nitwit.
> >>
> >> And I seen smearing of details in Canon files.
> >>
> >> Back to personal insults I see?
> >>
> >> --
> >>
> >> Stacey
> >
> >Well let's start with what I said
> >"I don't see any loss in detail between ISO 100 and ISO 1600. Of course
> >
> >there are more limits on the sharpening that I can do but the photos
> >out of the camera look pretty much the same."
> >
> >I did not say "ZERO quality difference between ISO 100 and ISO 1600"
> >there is more noise on the ISO 1600 and I did point out this will limit
> >shapening.
> >
> >So do the photos look pretty much the same, I will let others decide.
> >
> >I did two quick snap shots, one at iso 100 and one at iso 1600
> >
> >ISO 100
> >http://www.sewcon.com/photos/IMG_8276.JPG
> >
> >ISO 1600
> >http://www.sewcon.com/photos/IMG_8281.JPG
> >
> >Print these out at 8 x 10 and I would bet you would have a hard time
> >telling them apart.
> >
> >Scott
>
> You know what the horrifying is? You have MORE detail at 1600 than at
> 100!! Take a look at the blue real estate sign. Look at the "For
> Sale" part. That is Weird. I see the tell-tale chromatic noise that
> tells me the 1600ISO image is at a high speed, but what's with the
> "inverted" detail level at the two different ISOs? Did you
> noise-process the 100ISO image?
> -Rich

These are straight out of the camera. These are not meant to be any
kind of real test, but rather to show what I see.

Unlike what some would claim I have not been stating that there is no
difference at ISO 1600, I have been saying that I have not seen it,
there is not much more then I can do.

The ISO 1600 photo to me does not look at all bad, what can I say, I
pointed the camera pushed the button and this is what I got. The EXIF
data is all intact for anyone who want to see the particulars.

Scott
Anonymous
July 7, 2005 11:17:02 PM

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

Stacey <fotocord@yahoo.com> wrote:
>
>
> Zed Pobre wrote:
>
>> Stacey <fotocord@yahoo.com> wrote:
>>> Scott W wrote:
>>>>
>>>> I don't see any loss in detail between ISO 100 and ISO 1600.
>>>
>>> The above is one of the most absurd claims I've seen anyone make here!
>>>
>>> http://www.dpreview.com/reviews/canoneos350d/page19.asp
>>>
>>> Look at the ISO 100 shots vs the ISO 1600 shots, look at her crown and if
>>> you can't see the loss of detail, yes you are blind!
>>
>> Bad comparison. You're comparing a shot at 1/100s to a shot at
>> 1/1600s -- of course you're going to get a slight loss of detail with
>> four stops less light to work with.
>
> Well that's what the higher ISO is for, shooting with less light. Do you
> shoot at ISO 1600 and then over expose it 4 stops rather than use ISO 100?

The context here is attempting to test whether or not the camera
firmware is degrading the image after light hits the sensor when ISO
levels are increased. To do that, you have to control for the amount
of light hitting the sensor. That means running all comparisons at
exactly the same aperture and speed -- otherwise, all you're doing is
measuring how much more detail you get with more light, with obvious
results on *any* camera.


>> Try shooting that again at
>> 1/1600s at ISO100 and then compare how much detail you get;
>
> You lost me there, what is the point in underexposing the ISO100 shot? Or do
> you mean open up the lens to get the correct exposure at 1/1600s which
> isn't going to do anything either.

The point is so that you're measuring the same amount of real detail
as it hits the sensor, and seeing if you're losing detail by using
analog gain and a bad filtering algorithm in the firmware, as opposed
to using a better filtering algorithm boosting the EV in software on a
computer later.

Or, alternately, you can do direct camera-to-camera resolution
comparisons with identical lenses.

Given the same amount of light, Scott is correct, though: there is
almost always more detail, not less, at ISO1600 evenly exposed than
there is at ISO100 4 stops under.

--
Zed Pobre <zed@resonant.org> a.k.a. Zed Pobre <zed@debian.org>
PGP key and fingerprint available on finger; encrypted mail welcomed.
Anonymous
July 8, 2005 12:18:49 AM

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

On 7 Jul 2005 11:08:47 -0700, "Scott W" <biphoto@hotmail.com> wrote:

>Stacey wrote:
>> eawckyegcy@yahoo.com wrote:
>>
>> > Stacey dribbles on his bib:
>> >
>> >> JPS@no.komm wrote:
>> >>
>> >> >*not* detail at high and low ISOs
>> >> > on the same camera.
>> >>
>> >> Why not? If the details are being smeared, then they are just doing NR in
>> >> camera not some "Magical technology" like most canon users want to
>> >> believe.
>> >
>> > Read what JPS wrote: noise degrades detail. Take any low-noise image,
>> > add noise, and observe the "loss of detail".
>>
>> Or it's smeared to hide the noise. Or do you agree with Scott that there is
>> ZERO quality difference between ISO 100 and ISO 1600?
>>
>> >
>> >> And yes I see more -detail- in nikon files at high ISO than I do in the
>> >> canon samples I've seen.
>> >
>> > You've seen _noise_ in Nikon files, nitwit.
>>
>> And I seen smearing of details in Canon files.
>>
>> Back to personal insults I see?
>>
>> --
>>
>> Stacey
>
>Well let's start with what I said
>"I don't see any loss in detail between ISO 100 and ISO 1600. Of course
>
>there are more limits on the sharpening that I can do but the photos
>out of the camera look pretty much the same."
>
>I did not say "ZERO quality difference between ISO 100 and ISO 1600"
>there is more noise on the ISO 1600 and I did point out this will limit
>shapening.
>
>So do the photos look pretty much the same, I will let others decide.
>
>I did two quick snap shots, one at iso 100 and one at iso 1600
>
>ISO 100
>http://www.sewcon.com/photos/IMG_8276.JPG
>
>ISO 1600
>http://www.sewcon.com/photos/IMG_8281.JPG
>
>Print these out at 8 x 10 and I would bet you would have a hard time
>telling them apart.
>
>Scott

You know what the horrifying is? You have MORE detail at 1600 than at
100!! Take a look at the blue real estate sign. Look at the "For
Sale" part. That is Weird. I see the tell-tale chromatic noise that
tells me the 1600ISO image is at a high speed, but what's with the
"inverted" detail level at the two different ISOs? Did you
noise-process the 100ISO image?
-Rich
Anonymous
July 8, 2005 2:40:08 AM

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

In message <3j58ksFodgmjU1@individual.net>,
Stacey <fotocord@yahoo.com> wrote:

>Well that's what the higher ISO is for, shooting with less light. Do you
>shoot at ISO 1600 and then over expose it 4 stops rather than use ISO 100?

For a low-key scene, yes, *I* do, and they come out much more detailed
at ISO 1600.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 8, 2005 2:40:09 AM

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

On Thu, 07 Jul 2005 22:40:08 GMT, JPS@no.komm wrote:

>In message <3j58ksFodgmjU1@individual.net>,
>Stacey <fotocord@yahoo.com> wrote:
>
>>Well that's what the higher ISO is for, shooting with less light. Do you
>>shoot at ISO 1600 and then over expose it 4 stops rather than use ISO 100?
>
>For a low-key scene, yes, *I* do, and they come out much more detailed
>at ISO 1600.

You must be dealing with scenes that have no more than 4 contrast
levels then, otherwise, you'd radically burn out some parts of the
image. If you mean night shooting, there better not be any lights
in the scene.
Anonymous
July 8, 2005 2:48:27 AM

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

In message <3j2sm7Fnu67pU1@individual.net>,
Stacey <fotocord@yahoo.com> wrote:

>JPS@no.komm wrote:

>>*not* detail at high and low ISOs
>> on the same camera.

>Why not? If the details are being smeared,

How the hell can you tell that they're "smeared", when there is "fuzz"
all over them from noise?

You can't seriously believe that an ISO 1600 image with default exposure
compensation has the same level of detail as an ISO 100 image with the
same EC on the same camera? That is ridiculous.

>then they are just doing NR in
>camera not some "Magical technology" like most canon users want to believe.
>And yes I see more -detail- in nikon files at high ISO than I do in the
>canon samples I've seen.

Show me some examples.

Anyway, this is all talk. Put a Canon, an Olympus, a Nikon, or whatever
you want on a t-mount lens on a tripod, and take the same shot, with the
target distance adjusted for crop differences, using the same shutter
speed and aperture, and ISO setting, and see what detail they have in
the RAW data.

--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 8, 2005 2:51:48 AM

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

In message <3j3ul8Fo0hhqU1@individual.net>,
Stacey <fotocord@yahoo.com> wrote:

>Or it's smeared to hide the noise. Or do you agree with Scott that there is
>ZERO quality difference between ISO 100 and ISO 1600?

You really need to see a shrink. You always treat people who take
similar positions on an issue as a unit; that is not healthy. Is this
an "us (or me) and them" thing to you? It is good policy to address
individuals based on their individual statements.

I don't know what Scott meant by that; it could mean that he already has
an intuition for how much detail is reduced by noise, and taking that
into account, the ISO 1600 images don't look generically noise-reduced.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 8, 2005 4:36:24 AM

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

"Ryadia" <just@the.group> wrote in message
news:42c7a79d$1@dnews.tpgi.com.au...
SNIP
> Either Canon are post processing in the camera to reduce noise or
> the higher the ISO, the lower the resolving power of their 20D
> sensor.

You may be correct when, and only when, judging in-camera JPEGs.
I spent some time testing your hypotheses, and I don't agree for
non-JPEG conversions.

I used a 20D and shot at Raw+Large JPEGs (Parameter 2 settings), and
processed the Raws with DPP and RawShooter Essentials 2005 (v1.13
build 15), so the results can be verified independently with free
tools. The DPP conversions were linear Raw (16-b/ch, no sharpening),
and the RSE conversions were as-shot, but with Sharpening and Detail
extraction processing bias both set to -50 (my personal preference).

Testing was done with "Imatest", in particular the SFR option because
I wanted to test both the effect on resolution and noise spectrum.

To summarize, for the Raw conversions the 10-90% edge profile *and*
modulation at 50% *and* at Nyquist the resolution is *unaffected* by
the ISO settings (100 - 3200), in other words overall resolution is
NOT compromised!!! This may of course differ between camera
brands/models.
In the JPEGs however, the 20D's 1600 and H (3200) settings show a 32%
to 50% reduction versus lower ISO settings at Nyquist. So for ISO
100 - 400/800 there is NO reduction of resolution, regardless of the
choice of the Bayer CFA reconstruction method, and only in-camera
JPEGs lose resolution at the 800/1600 - 3200 settings.

As I've stated before, there are differences in overall resolution and
suppression of various artifacts between Raw converters. However,
resolution is only compromised in (20D) in-camera JPEGs at the 2/3
highest ISO settings, Raws are unaffected.

The effects of noise reduction software of course depends on the
actual program and settings used. Given the fact that resolution is
generally unaffected (except for JPEGs at the highest settings) the
efficiency of different Noise Reduction programs seems to be the
limiting factor when NR is applied. My
experiences with NeatImage, which is highly tunable, are very
favorable with regards to keeping the resolution intact.

Bart
July 8, 2005 6:18:02 AM

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

Zed Pobre wrote:

>>>
>>> Bad comparison. You're comparing a shot at 1/100s to a shot at
>>> 1/1600s -- of course you're going to get a slight loss of detail with
>>> four stops less light to work with.
>>
>> Well that's what the higher ISO is for, shooting with less light. Do you
>> shoot at ISO 1600 and then over expose it 4 stops rather than use ISO
>> 100?
>
> The context here is attempting to test whether or not the camera
> firmware is degrading the image after light hits the sensor when ISO
> levels are increased.

Funny you say that. Someone here just tested this with RAW files at ISO 1600
from a 20D and found the detail was there, in the in camera jpegs it was
gone, just like the DPreview images.

> To do that, you have to control for the amount
> of light hitting the sensor. That means running all comparisons at
> exactly the same aperture and speed -- otherwise, all you're doing is
> measuring how much more detail you get with more light, with obvious
> results on *any* camera.

What is the point in overexposing the image 4 stops at ISO 1600 and looking
for details? The camera is applying NR for the in camera jpegs at ISO 1600
and over exposing the image isn't going to make it stop doing that. I don't
follow this "more light = more details" or people would just overexpose
everything, which they don't.


>>
>> You lost me there, what is the point in underexposing the ISO100 shot? Or
>> do you mean open up the lens to get the correct exposure at 1/1600s which
>> isn't going to do anything either.
>
> The point is so that you're measuring the same amount of real detail
> as it hits the sensor, and seeing if you're losing detail by using
> analog gain and a bad filtering algorithm in the firmware,

"real detail"? Like somehow the detail changes because there isn't as much
light on it?

>
> Given the same amount of light, Scott is correct, though: there is
> almost always more detail, not less, at ISO1600 evenly exposed than
> there is at ISO100 4 stops under.
>

????? Who said it was a good idea to underexpose 4 stops? BTW SCott is the
one claiming he's getting detail 7 stops under! I'm sure though as a canon
cheerleader you're not going to concede your camera can't do magic tricks
for you...

--

Stacey
Anonymous
July 8, 2005 7:38:08 AM

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

In message <42cdae6a$0$93486$e4fe514c@news.xs4all.nl>,
"Bart van der Wolf" <bvdwolf@no.spam> wrote:

>You may be correct when, and only when, judging in-camera JPEGs.
>I spent some time testing your hypotheses, and I don't agree for
>non-JPEG conversions.

Here is a 20D ISO 1600 shot, an 800*500 crop from a full 8MP image,
converted in ACR with NR settings at 0 in ACR, and everything default
except exposure, which was set to +0.25 stops, making the EI actually
1900 for this image, as displayed:

http://www.pbase.com/jps_photo/image/45889233/original

The only sharpening was the default "25" of ACR.

Those colors are actually on the bill.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
July 8, 2005 7:38:09 AM

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

JPS@no.komm wrote:

> In message <42cdae6a$0$93486$e4fe514c@news.xs4all.nl>,
> "Bart van der Wolf" <bvdwolf@no.spam> wrote:
>
>>You may be correct when, and only when, judging in-camera JPEGs.
>>I spent some time testing your hypotheses, and I don't agree for
>>non-JPEG conversions.
>
> Here is a 20D ISO 1600 shot, an 800*500 crop from a full 8MP image,
> converted in ACR with NR settings at 0 in ACR, and everything default
> except exposure, which was set to +0.25 stops, making the EI actually
> 1900 for this image, as displayed:
>
> http://www.pbase.com/jps_photo/image/45889233/original
>


So you and ALLEN share this "Shooting money" as an example of the image
quality a camera can produce? LOL!

--

Stacey
Anonymous
July 8, 2005 8:30:54 AM

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

"Stacey" <fotocord@yahoo.com> wrote in message
news:3j6k9mFoeqjuU2@individual.net...
> JPS@no.komm wrote:
>

>>
>> Here is a 20D ISO 1600 shot, an 800*500 crop from a full 8MP image,
>> converted in ACR with NR settings at 0 in ACR, and everything default
>> except exposure, which was set to +0.25 stops, making the EI actually
>> 1900 for this image, as displayed:
>>
>> http://www.pbase.com/jps_photo/image/45889233/original
>>
>
>
> So you and ALLEN share this "Shooting money" as an example of the image
> quality a camera can produce? LOL!
>
> --
>
> Stacey

I get it now, Stacey. You attack a camera's quality, when someone posts an
image that seems to refute your assertion, you deride the subject, not the
quality of the image. Very apropos to the discussion.

--
Skip Middleton
http://www.shadowcatcherimagery.com
July 8, 2005 4:52:42 PM

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

Skip M wrote:


>
> I get it now, Stacey. You attack a camera's quality, when someone posts
> an image that seems to refute your assertion, you deride the subject, not
> the
> quality of the image. Very apropos to the discussion.
>

No because it's the best possible scenario to shoot as far as hiding noise.
Paper with colored dots in it already, shot with a flash instead of ambient
light and in all likelihood overexposed due to how close they are to the
object no matter what they have the camera set on.

Where is the shadows that would actually show noise? Of course the whole
image is compressed into a few stops on the right so you see it "At it's
best". This isn't a real world example unless this is what you plan to
shoot and then why would ISO 1600 even matter? If you want to show low
noise performance, post an image in available light like tungsten lighting
with some shadows in the image.

if he was trying to show the detail is the same at ISO 100 and 1600, where
is the shot to compare it to? Yes this is a useless example unless you
think money is a good subject matter for your photography?

--

Stacey
Anonymous
July 9, 2005 12:35:27 AM

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

In article <Mttze.7521$HV1.6077@fed1read07>,
"Skip M" <shadowcatcher@cox.net> wrote:

>
> I get it now, Stacey. You attack a camera's quality, when someone posts an
> image that seems to refute your assertion, you deride the subject, not the
> quality of the image. Very apropos to the discussion.

Its a flat subject without DOF.

--
LF Website @ http://members.verizon.net/~gregoryblank

"To announce that there must be no criticism of the President,
or that we are to stand by the President, right or wrong,
is not only unpatriotic and servile, but is morally treasonable
to the American public."--Theodore Roosevelt, May 7, 1918
Anonymous
July 9, 2005 12:36:28 AM

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

In message <3j6k9mFoeqjuU2@individual.net>,
Stacey <fotocord@yahoo.com> wrote:

>So you and ALLEN share this "Shooting money" as an example of the image
>quality a camera can produce? LOL!

Who the hell is "ALLEN"?

You really need to get new glasses that let you see individuals.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 9, 2005 2:26:30 AM

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

In message <3j6k4rFoeqjuU1@individual.net>,
Stacey <fotocord@yahoo.com> wrote:

>What is the point in overexposing the image 4 stops at ISO 1600 and looking
>for details? The camera is applying NR for the in camera jpegs at ISO 1600
>and over exposing the image isn't going to make it stop doing that.

How do you know that? How do you know that the JPEG routines aren't
simply smoothing the image more in the lower sensor levels? If that is
the case, a high-key, "exposed to the right" 1600 JPEG is not going to
be very smoothed.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
July 9, 2005 2:35:41 AM

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

In message <greg-D66DC8.16450708072005@news.verizon.net>,
Gregory Blank <greg@greg_____photo.com> wrote:

>In article <Mttze.7521$HV1.6077@fed1read07>,
> "Skip M" <shadowcatcher@cox.net> wrote:

>> I get it now, Stacey. You attack a camera's quality, when someone posts an
>> image that seems to refute your assertion, you deride the subject, not the
>> quality of the image. Very apropos to the discussion.
>>Its a flat subject without DOF.

Have you been following along? The charge has been by Stacey, and
Ryadia, and the other odd posters, that the Canon DSLRs, especially the
20D, use general noise reduction that loses lots of detail at high ISOs
in the RAW files. There is no detail in the OOF areas to lose.

Stacey keeps changing his/her argument, to avoid being wrong.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
!