«

www.digitalcamerainfo.com gave the FZ5 a very poor writeup when it
comes to NOISE. It seems that noise (or the lack of it) it at the heart
of getting great results in a variety of circumstances. While I was
getting ready to consider the FZ5 because of its lower weight and bulk,
compared to the FZ20, I feel now I should wait for another crop of
announcements.

Now dpreview.com does not hit the camera on this issue as hard. Why
would there be such a difference in objective reporting?

Anyway, what do we consider from here?

measekite wrote:
> www.digitalcamerainfo.com gave the FZ5 a very poor writeup when it
> comes to NOISE. It seems that noise (or the lack of it) it at the
> heart of getting great results in a variety of circumstances. While
> I was getting ready to consider the FZ5 because of its lower weight
> and bulk, compared to the FZ20, I feel now I should wait for another
> crop of announcements.

In small sensors, there is a limit to how low the noise can be. It is
unlikely you will see great improvements in the short term. You may find
yourself waiting for ever!

> Now dpreview.com does not hit the camera on this issue as hard. Why
> would there be such a difference in objective reporting?

Different measurement techniques, none of which actually relate directly
to how humans perceive the images. [Justification for comment: in audio
the SNR measurement is weighted to the ear's differing sensitivity at
different frequencies. I see no camera noise measurements which does this
for the eye's spatial frequency response.]

> Anyway, what do we consider from here?

How much does noise matter? In times past, some grain enhanced the
character of low-light shots - it added atmosphere, but this view is
currently unfashionable. How do you [propose to] view your images?
Screen, Web site, printed? What size?

Cheers,
David

I've seen a colour noise test in a UK computer magazine, PCW I think, where
they measured the noise on a few hundred different colours and aggregated
the results statistically to give an overall result for each camera. It was
a good way to compare cameras. The problem was that the winning camera was
a Kodak 6490 which has excessive in-camera noise reduction which obliterates
fine detail, so you have to be very careful when comparing.



"David J Taylor" <david->
wrote in message news:kS66e.8733$ ...
> measekite wrote:
>> www.digitalcamerainfo.com gave the FZ5 a very poor writeup when it
>> comes to NOISE. It seems that noise (or the lack of it) it at the
>> heart of getting great results in a variety of circumstances. While
>> I was getting ready to consider the FZ5 because of its lower weight
>> and bulk, compared to the FZ20, I feel now I should wait for another
>> crop of announcements.
>
> In small sensors, there is a limit to how low the noise can be. It is
> unlikely you will see great improvements in the short term. You may find
> yourself waiting for ever!
>
>> Now dpreview.com does not hit the camera on this issue as hard. Why
>> would there be such a difference in objective reporting?
>
> Different measurement techniques, none of which actually relate directly
> to how humans perceive the images. [Justification for comment: in audio
> the SNR measurement is weighted to the ear's differing sensitivity at
> different frequencies. I see no camera noise measurements which does this
> for the eye's spatial frequency response.]
>
>> Anyway, what do we consider from here?
>
> How much does noise matter? In times past, some grain enhanced the
> character of low-light shots - it added atmosphere, but this view is
> currently unfashionable. How do you [propose to] view your images?
> Screen, Web site, printed? What size?
>
> Cheers,
> David
>

SteveB wrote:
> I've seen a colour noise test in a UK computer magazine, PCW I think,
> where they measured the noise on a few hundred different colours and
> aggregated the results statistically to give an overall result for
> each camera. It was a good way to compare cameras. The problem was
> that the winning camera was a Kodak 6490 which has excessive
> in-camera noise reduction which obliterates fine detail, so you have
> to be very careful when comparing.

So, in fact, it was not "a good way to compare cameras"!

David

David J Taylor wrote:

>measekite wrote:
>
>
>>www.digitalcamerainfo.com gave the FZ5 a very poor writeup when it
>>comes to NOISE. It seems that noise (or the lack of it) it at the
>>heart of getting great results in a variety of circumstances. While
>>I was getting ready to consider the FZ5 because of its lower weight
>>and bulk, compared to the FZ20, I feel now I should wait for another
>>crop of announcements.
>>
>>
>
>In small sensors, there is a limit to how low the noise can be. It is
>unlikely you will see great improvements in the short term. You may find
>yourself waiting for ever!
>
>

Do you own one. Are you saying that results at ISO 200 and ISO 400
produce very acceptable results and I should disregard the
digitalcamerainfo report. If that is the case are they lying or just
plain wrong. And if that is true can they not be trusted as a sourced on
anything?

>
>
>>Now dpreview.com does not hit the camera on this issue as hard. Why
>>would there be such a difference in objective reporting?
>>
>>
>
>Different measurement techniques, none of which actually relate directly
>to how humans perceive the images. [Justification for comment: in audio
>the SNR measurement is weighted to the ear's differing sensitivity at
>different frequencies. I see no camera noise measurements which does this
>for the eye's spatial frequency response.]
>
>
>
>>Anyway, what do we consider from here?
>>
>>
>
>How much does noise matter? In times past, some grain enhanced the
>character of low-light shots - it added atmosphere, but this view is
>currently unfashionable. How do you [propose to] view your images?
>Screen, Web site, printed? What size?
>
>

I am hoping for a camera capable of producing striking low noise
enlargements up to 8.5x11 that are cropped 20 to 30%.

>Cheers,
>David
>
>
>
>

SteveB wrote:

>I've seen a colour noise test in a UK computer magazine, PCW I think, where
>they measured the noise on a few hundred different colours and aggregated
>the results statistically to give an overall result for each camera. It was
>a good way to compare cameras. The problem was that the winning camera was
>a Kodak 6490 which has excessive in-camera noise reduction which obliterates
>fine detail, so you have to be very careful when comparing.
>
>

Noise reduction that obliterates fine detail is not the same as having a
decent sensor size ratio to the MP and not having noise at all. I think
the issue with the PS cameras is not having a sufficient sensor size and
cramming more MP on the sensor then they should.

>
>
>"David J Taylor" <david->
>wrote in message news:kS66e.8733$ ...
>
>
>>measekite wrote:
>>
>>
>>>www.digitalcamerainfo.com gave the FZ5 a very poor writeup when it
>>>comes to NOISE. It seems that noise (or the lack of it) it at the
>>>heart of getting great results in a variety of circumstances. While
>>>I was getting ready to consider the FZ5 because of its lower weight
>>>and bulk, compared to the FZ20, I feel now I should wait for another
>>>crop of announcements.
>>>
>>>
>>In small sensors, there is a limit to how low the noise can be. It is
>>unlikely you will see great improvements in the short term. You may find
>>yourself waiting for ever!
>>
>>
>>
>>>Now dpreview.com does not hit the camera on this issue as hard. Why
>>>would there be such a difference in objective reporting?
>>>
>>>
>>Different measurement techniques, none of which actually relate directly
>>to how humans perceive the images. [Justification for comment: in audio
>>the SNR measurement is weighted to the ear's differing sensitivity at
>>different frequencies. I see no camera noise measurements which does this
>>for the eye's spatial frequency response.]
>>
>>
>>
>>>Anyway, what do we consider from here?
>>>
>>>
>>How much does noise matter? In times past, some grain enhanced the
>>character of low-light shots - it added atmosphere, but this view is
>>currently unfashionable. How do you [propose to] view your images?
>>Screen, Web site, printed? What size?
>>
>>Cheers,
>>David
>>
>>
>>
>
>
>
>

measekite wrote:
[]
> Do you own one. Are you saying that results at ISO 200 and ISO 400
> produce very acceptable results and I should disregard the
> digitalcamerainfo report. If that is the case are they lying or just
> plain wrong. And if that is true can they not be trusted as a sourced
> on anything?

I have an FZ5 (recently acquired) and my wife has an FZ20 (some months
old). I don't find the noise on the FZ20 pictures unacceptable, and I
expect my FZ5 to be the same. If noise is critical, I would avoid the
higher ISOs.

[]
> I am hoping for a camera capable of producing striking low noise
> enlargements up to 8.5x11 that are cropped 20 to 30%.

Here are the sums - crop by 25% (say) so your image uncropped would be
14.6 x 11.3 inches. A rule of thumb is 300 pixels per inch, so you would
need 15 MP sensor resolution. As I recall, only a full-frame Canon DSLR
offers so many pixels. In practice, I have seen goo A4-sized prints
(about the same size) from a 3MP Nikon 990, so I would certainly expect
acceptable images from a 5MP FZ5. Doubtless, images from a 15MP camera
would look sharper, particularly if you take a magnifying glass to the
print....

See if you can borrow an FZ5 for a short while and try it. I don't think
you will be disappointed.

Cheers,
David

measekite wrote:
> David J Taylor wrote:
>
> >measekite wrote:
> >
> >
> >>www.digitalcamerainfo.com gave the FZ5 a very poor writeup when it
> >>comes to NOISE. It seems that noise (or the lack of it) it at the
> >>heart of getting great results in a variety of circumstances.
While
> >>I was getting ready to consider the FZ5 because of its lower weight
> >>and bulk, compared to the FZ20, I feel now I should wait for
another
> >>crop of announcements.
> >>

SNIP

If you read the DRP review closely, they're differentiating between
chrominance and luminance noise - the K-M Z5 has very high chroma noise
(color noise) whereas the Pana FZ-5 has very high Luminance noise(
light and dark noise). The K-M Z5 images are seriously over-processed
in-camera, a lot of info is being lost; they still have quite nasty
chroma noise at ISO 400, too. Panasonic tends to under-process their
pics, especially when it comes to noise, so that you can do it
yourself. The noise from the Pana FZ-5 is much "finer-grained", too -
it tends to be less objectionable that way.

> >
> >How much does noise matter? In times past, some grain enhanced the
> >character of low-light shots - it added atmosphere, but this view is

> >currently unfashionable. How do you [propose to] view your images?
> >Screen, Web site, printed? What size?
> >
>
> I am hoping for a camera capable of producing striking low noise
> enlargements up to 8.5x11 that are cropped 20 to 30%.
>

You need to look to a dSLR then - there aren't any consumer digicams
these days that you can use above ISO 100 without significant noise.
For your cropping and enlarging requirements you'll probably need at
least 6 Mpixels. The Canon 350 XT would be good, although I'm not
trying to start a brand war; CMOS sensors in general are supposed to be
good for low noise, though. I've seen very striking images from several
low-end dSLR's recently; the Pentax *istDS, Oly E300, Nikon D70, Canon
dRebel/dRebel XT, whatever - with a good lens they'll all do what you
want. Lighting and a bright lens are vital, so that you're shooting
with as low an ISO as possible.

Really, the FZ-5 is aimed at a less demanding audience - it's a nice
lightweight 5 Mpixel IS long zoom consumer-level camera; good for
snapshots, vacation pics and amateur nature buffs/birders/etc.
long-zoom pics. It's not meant to be used for art shots; if you're
lucky and know how to take pictures properly, you'll occasionally get
one worth putting on the wall, though.

ECM

"David J Taylor" <david->
wrote in message news:BOa6e.8941$ ...
> SteveB wrote:
>> I've seen a colour noise test in a UK computer magazine, PCW I think,
>> where they measured the noise on a few hundred different colours and
>> aggregated the results statistically to give an overall result for
>> each camera. It was a good way to compare cameras. The problem was
>> that the winning camera was a Kodak 6490 which has excessive
>> in-camera noise reduction which obliterates fine detail, so you have
>> to be very careful when comparing.
>
> So, in fact, it was not "a good way to compare cameras"!
>
> David
>
That's correct. What it means is that any other method of measuring which
is bound to be less thorough than that PCW test has to be treated with
doubt.

Here is what I did do. First I set aside the digitalcamera.com review
that said the noise level on the FZ5 was not acceptable. I went to the
dpreview.com site and as well I temporarily set aside their opinions.

I went to the samples section of dpreview.com for the FZ5 review and
downloaded all of the images. Unfortunately most were at ISO 80 but
there were a couple at ISO400. The images were large and the resolution
was set for 72ppi. I then went to image size in Photoshop and clicked on
the auto button for 133 lines of resolution and check the best quality
radio button. The resolution jumped to about 266. I am telling you
what I did even thought I only very very vaguely understand it.

I then used the cropping tool for either a 4x6 or a 6x4 and the
resolution jumped to over 600 ppi. I then set the view to the Canon PR1
profile in Photoshop. I printed it with my Canon IP4000 using OEM ink
on Costco/Kirkland Paper using a setting in the Canon Driver of Photo
Paper Pro, and added all of the effects like vivid and noise reduction -
strong etc. The print looked fabulous. I have never gotten a wet
process photo print from anything other than a professional lab that
looked this good. I still need to find out what would happen going up
to full letter size.

Using one of the photos I printed it out using the same setting using
Canon Photo Paper Pro and the Costco/Kirkland. On the Canon there was a
hair less stipple texture. Both were fantastic. A 4x6 Costco Print is
35 cents cheaper than the Canon Photo Paper Pro. 39 cents vs 4 cents
plus ink Cost. As for the actual Colors, I could not tell, even after
starring for a few minutes, that there was any difference at all.

I just hope that what I downloaded from dpreview was really work from a
Panasonic Lumix FZ5. I know that they were done by a Pro and edited by
a Pro to achieve all of the perfection they could get.

David J Taylor wrote:

>measekite wrote:
>[]
>
>
>>Do you own one. Are you saying that results at ISO 200 and ISO 400
>>produce very acceptable results and I should disregard the
>>digitalcamerainfo report. If that is the case are they lying or just
>>plain wrong. And if that is true can they not be trusted as a sourced
>>on anything?
>>
>>
>
>I have an FZ5 (recently acquired) and my wife has an FZ20 (some months
>old). I don't find the noise on the FZ20 pictures unacceptable, and I
>expect my FZ5 to be the same. If noise is critical, I would avoid the
>higher ISOs.
>
>[]
>
>
>>I am hoping for a camera capable of producing striking low noise
>>enlargements up to 8.5x11 that are cropped 20 to 30%.
>>
>>
>
>Here are the sums - crop by 25% (say) so your image uncropped would be
>14.6 x 11.3 inches. A rule of thumb is 300 pixels per inch, so you would
>need 15 MP sensor resolution. As I recall, only a full-frame Canon DSLR
>offers so many pixels. In practice, I have seen goo A4-sized prints
>(about the same size) from a 3MP Nikon 990, so I would certainly expect
>acceptable images from a 5MP FZ5. Doubtless, images from a 15MP camera
>would look sharper, particularly if you take a magnifying glass to the
>print....
>
>See if you can borrow an FZ5 for a short while and try it. I don't think
>you will be disappointed.
>
>Cheers,
>David
>
>
>
>