Rockwell on DSLR vs. P&S

> We've got to hand it to you for being ironic.
>
> If Ken Rockwell had used a better downsampling algorithm,
> rather than Photoshop bicubic (I'm assuming), maybe just maybe
> the DSLR image would look better than the P&S digicam image.
>


I would like to know, what are the better downsamplers?

--
Giant_Alex
not my site: http://www.e-sword.net/
 >> Stay informed about: Rockwell on DSLR vs. P&S 

In rec.photo.digital AAvK <Idondodat DeleteThis @wahoo.com> wrote:
>>
>> If Ken Rockwell had used a better downsampling algorithm,
>> rather than Photoshop bicubic (I'm assuming), maybe just maybe
>> the DSLR image would look better than the P&S digicam image.
>>
> I would like to know, what are the better downsamplers?

At whole-fraction scaling factors (25%, 33%, 50%, 67%, 75%)
most downsampling algorithms give identical results.

Otherwise Lanczos is far better than Bicubic for most images.
Lanczos leaves smooth areas smooth and keeps edges sharp.

It's possible Bicubic is good if you're changing aspect ratio
(different scale for x / y). I have not investigated.
 >> Stay informed about: Rockwell on DSLR vs. P&S 

"Roger N. Clark (change username to rnclark)" <username RemoveThis @qwest.net> wrote:
>Walter Hancock wrote:
>> On Thu, 4 Oct 2007 19:20:51 -0400, "David Ruether" <rpn1 RemoveThis @no-junk.cornell.edu>
>> wrote:
>>
>>>
>>> "David Ruether" <rpn1 RemoveThis @no-junk.cornell.edu> wrote in message news:fe3qgr$lgq$1@ruby.cit.cornell.edu...
>>>> "Bob Williams" <mytbobnospam RemoveThis @cox.net> wrote in message news:kddNi.154923$Mu5.25689@newsfe15.phx...
>>>>> What I'd like to see.....Maybe Roger can do this
>>>>> for us.....Is to shoot two identical subjects under
>>>>> BRIGHT conditions where both cameras capture an
>>>>> image at about f=4.0 - 5.6, ISO = 80-100, speed
>>>>> 1/250-1/400 sec.
>>>>> These conditions are fairly typical of outdoor
>>>>> lighting on a partly cloudy day and utilize a
>>>>> camera parameters where BOTH cameras are operating
>>>>> near their optimum conditions.
>>>>> Bob
>>>> I should have this up in an hour or so, but I will put up comparison
>>>> photos at www.doughicksphotography.com/comparison.htm.
>>>> of a D1x with an 18-70mm at 18mm and f4 and a Sony 707 at
>>>> 10mm and f4.5 (about the same angle of view). The files will be
>>>> very large...
>>>> --
>>>> David Ruether
>>> The photos are now up (dull, and it would have been nice to have
>>> had the sharpening at normal on the Sony, and/or to have used the
>>> Fuji S700, but anyway...). I will remove this page soon...
>> Ouch! (saying that on behalf of all dSLR owners out
>> there)
>> Yet again the P&S camera clearly wins, over a Nikon
>> dSLR no less.
>> Not even a contest.
>>
>What I see are halos around all edges in the Sony image, which to the
>untrained eye gives the impression of sharpness. To others it looks like
>horrible artifacts. The advantage of the Nikon image is not shown.
>It should be processed through a raw converter, where it would show
>sharper edges and even less noise.

In addition to that, the D1x image is overexposed.
E.g., the lack of detail in the yellow lines compared to
the Sony image is because either the R or G channel is
blown out in almost every pixel. (I'd also expect the
RAW file from the D1x was a Compressed NEF, which does
indeed lose some detail in the highlights.) And that is
not to mention that the "detail" seen in the Sony image
appears to be more artifact than scene detail.

The above is made worse because apparently there was a
bit of cloudy haze in front of the sun for the Sony
image that was not there for the D1x image. It shows up
in the shadows and contrast on the wooden rail in the
lower left corner of the image. (I'm assuming that is a
difference in lighting as opposed to the possibility
that the D1x images simply have that much more contrast
than the Sony.)

I don't understand the idea that there is more detail in
the Sony image. That "detail" appears to mostly be the
artifacts, as you point out. It isn't just at the edges
of the image either. The parked cars aren't in the
center, but they have horrible fringing on them. The
far sidewalk, which is relatively close to center also
has fringing.

It's hard to make comparisons on those images though,
because as the EXIF data shows, neither is directly from
the camera. I'm not sure what the program that produced
them is, but RAW files, or at least JPEG's with the
original EXIF data would be more interesting.

Also, while it makes the content more interesting,
from an evaluation point of view it would have been
better not use each zoom lense so close to the limit of
its range. Closer to the center would be less a measure
of lenses and more a measure of the camera.

--
Floyd L. Davidson <http://www.apaflo.com/floyd_davidson>
Ukpeagvik (Barrow, Alaska) floyd RemoveThis @apaflo.com
 >> Stay informed about: Rockwell on DSLR vs. P&S 

"Bob Williams" <mytbobnospam DeleteThis @cox.net> wrote in message news:kddNi.154923$Mu5.25689@newsfe15.phx...

> What I'd like to see.....Maybe Roger can do this for us.....Is to shoot two identical subjects under BRIGHT conditions where both
> cameras capture an image at about f=4.0 - 5.6, ISO = 80-100, speed 1/250-1/400 sec.
> These conditions are fairly typical of outdoor lighting on a partly cloudy day and utilize a camera parameters where BOTH cameras
> are operating near their optimum conditions.
> Bob

I should have this up in an hour or so, but I will put up comparison
photos at www.doughicksphotography.com/comparison.htm.
of a D1x with an 18-70mm at 18mm and f4 and a Sony 707 at
10mm and f4.5 (about the same angle of view). The files will be
very large...
--
David Ruether
d_ruether DeleteThis @hotmail.com
http://www.donferrario.com/ruether
 >> Stay informed about: Rockwell on DSLR vs. P&S 

In rec.photo.digital David Ruether <rpn1 DeleteThis @no-junk.cornell.edu> wrote:
>
> I should have this up in an hour or so, but I will put up comparison
> photos at www.doughicksphotography.com/comparison.htm.
> of a D1x with an 18-70mm at 18mm and f4 and a Sony 707 at
> 10mm and f4.5 (about the same angle of view). The files will be
> very large...

Certainly not more than $10 to $50 worth of difference between these.
Isn't the Sony 707 a really old camera?

I like the Sony 707 picture better because of the kid with a red shirt
riding a scooter. That's the main difference for me. The Nikon D1x
with 18-70 has less purple fringing in the corners, but fuzzy infinity
and less saturated greens. So it's 6 of one, half dozen of the other.
 >> Stay informed about: Rockwell on DSLR vs. P&S 

"David Ruether" <rpn1.RemoveThis@no-junk.cornell.edu> wrote in message
news:fe3qgr$lgq$1@ruby.cit.cornell.edu...
>
> "Bob Williams" <mytbobnospam.RemoveThis@cox.net> wrote in message
> news:kddNi.154923$Mu5.25689@newsfe15.phx...
>
>> What I'd like to see.....Maybe Roger can do this for us.....Is to shoot
>> two identical subjects under BRIGHT conditions where both cameras capture
>> an image at about f=4.0 - 5.6, ISO = 80-100, speed 1/250-1/400 sec.
>> These conditions are fairly typical of outdoor lighting on a partly
>> cloudy day and utilize a camera parameters where BOTH cameras are
>> operating near their optimum conditions.
>> Bob
>
> I should have this up in an hour or so, but I will put up comparison
> photos at www.doughicksphotography.com/comparison.htm.
> of a D1x with an 18-70mm at 18mm and f4 and a Sony 707 at
> 10mm and f4.5 (about the same angle of view). The files will be
> very large...
> --

Yep, the Sony P&S won this round of comparison. Less glass, more sass!

Thanks!

NM
 >> Stay informed about: Rockwell on DSLR vs. P&S 

"David Ruether" <rpn1.RemoveThis@no-junk.cornell.edu> wrote in message news:fe3qgr$lgq$1@ruby.cit.cornell.edu...
> "Bob Williams" <mytbobnospam.RemoveThis@cox.net> wrote in message news:kddNi.154923$Mu5.25689@newsfe15.phx...

>> What I'd like to see.....Maybe Roger can do this for us.....Is to shoot two identical subjects under BRIGHT conditions where both
>> cameras capture an image at about f=4.0 - 5.6, ISO = 80-100, speed 1/250-1/400 sec.
>> These conditions are fairly typical of outdoor lighting on a partly cloudy day and utilize a camera parameters where BOTH cameras
>> are operating near their optimum conditions.
>> Bob

> I should have this up in an hour or so, but I will put up comparison
> photos at www.doughicksphotography.com/comparison.htm.
> of a D1x with an 18-70mm at 18mm and f4 and a Sony 707 at
> 10mm and f4.5 (about the same angle of view). The files will be
> very large...
> --
> David Ruether

The photos are now up (dull, and it would have been nice to have
had the sharpening at normal on the Sony, and/or to have used the
Fuji S700, but anyway...). I will remove this page soon...
--
David Ruether
d_ruether.RemoveThis@hotmail.com
http://www.donferrario.com/ruether
 >> Stay informed about: Rockwell on DSLR vs. P&S 

On Oct 4, 6:50 pm, "David Ruether" <r....RemoveThis@no-junk.cornell.edu> wrote:
>
> I should have this up in an hour or so, but I will put up comparison
> photos atwww.doughicksphotography.com/comparison.htm.
> of a D1x with an 18-70mm at 18mm and f4 and a Sony 707 at
> 10mm and f4.5 (about the same angle of view).

Oh boy, dueling Mavicas!
 >> Stay informed about: Rockwell on DSLR vs. P&S 

On Oct 4, 8:05 pm, Walter Hancock <spams....DeleteThis@unknown.com> wrote:
>
> Yet again the P&S camera clearly wins, over a Nikon dSLR no less.
>
> Not even a contest

I couldn't look much past the grotesque skyline on the Sony shot.
That P&S sure does have lotsa colors .... even when there is no color
there.
 >> Stay informed about: Rockwell on DSLR vs. P&S 

GetAClue wrote:
> On Thu, 04 Oct 2007 15:54:04 GMT, "David J Taylor"
> <david-taylor.DeleteThis@blueyonder.not-this-bit.nor-this-bit.co.uk> wrote:
>
>> The benefits I find with the DSLR include faster operation, and much
>> better low-light performance. I can shoot at ISO 1600 and have almost
>> noise-free pictures, whereas with the compact camera I needed to stick
>> with ISO 100 to keep the noise down.
>
> You wouldn't have this problem if you had a smaller sensor for which it is much
> easier to make larger aperture lenses. How is shooting at ISO1600 better than
> shooting at lower noise-free ISOs if your lens provides more light? You can pay
> $500 for a Canon 70-300mm f/4.5-f/5.6, just the lens alone, and only get 112mm
> to 480mm focal length range, which forces you to use ISO1600 under many lighting
> situations. Or pay $300 for a whole camera plus lens and use ISO 200 at f/2.8
> with 36mm to 432mm focal lengths for the same lighting situations.

You forgot some critical facts. The DSLR is 12 to 16x more sensitive
than the small pixel P&S so an f/5.6 lens on a DSLR is like an
f/1.4 to f/1.8 lens on a P&S regarding light gathered.
Then buy an f/2.8 lens for your DSLR, and its like having
an f/0.7 on the small pixel P&S.

> Some P&S
> cameras providing even larger apertures and wider zoom ranges than that.

Faster than f/2? Faster than f/1?


> Sure, you can get wider apertures for that DSLR, then you are faced with having
> to change them to compose each shot for the maximum resolution needed for that
> subject. If you don't change that 50mm f1.2 lens for the 500mm f/4.0, then that
> bird further away is going to have to be cropped to less than the resolution of
> a cell-phone camera. Wait, you won't even get that. It already flew away while
> your camera forced you to change lenses.

You mean like this:
http://www.clarkvision.com/galleries/gallery.bird/web/lilac-breasted.r...er.c01.
http://www.clarkvision.com/galleries/gallery.bird/web/eagle.c09.11.200...Z3F4717
http://www.clarkvision.com/galleries/gallery.bird/web/c01.14.2003.img_...3.egret
http://www.clarkvision.com/galleries/gallery.bird/web/black-shouldered...te.c01.

Let's see your P&S get images like that. Oh, and by the way, those
make impressively sharp 16x24-inch prints (the last one I've only sold
in galleries up to 16x18 inches).


> Do the math. You actually get more by paying for less using a P&S camera.

No, you get less with a P&S, not more. But you do get great
technology, relatively speaking, and can take very nice pictures
of certain things under good conditions. I actually carry a
P&S more hours per month than a DSLR. But I know the difference
between P&S and DSLR and use the best tool for the job.

> Your
> high ISOs are needed because of the larger sensor.

No. High ISO are enabled ny the large sensor. You can still
use low ISOs and get images with 4x lower noise (or better) than
small pixel cameras.

> It all averages out where
> there is absolutely no gain by using a larger sensor for lower noise. I take
> that back, there is a gain. The DSLR and lens makers' bank accounts. That's the
> only gain that I've ever found in buying a DSLR.

Absolutely incorrect.
>
> As we've recently seen by a comparison of moon photos between Roger Clark's
> $7000 MkII and someone else with a $400 Lumix FZ18, his DSLR and $1,200 L-glass
> lens couldn't even provide as much resolution as a P&S camera with a 28mm-504mm
> f/2.8 zoom lens, the P&S being hand-held at that. (Thanks for posting your
> photos Roger, what a wonderful source for comparisons to show how much better
> the P&S cameras are these days!)

Again incorrect.

Lets see any P&S camera with CAMERA LENSES do this good:
http://www.clarkvision.com/galleries/gallery.astrophoto-1/web/moon-JZ3...58-60-c


> I appears that DSLR owners take pride in trying to outdo each other in displays
> of ignorance and stupidity. On the plus side they are more than ready to provide
> examples for the rest of us to see (i.e. Roger Clark for example). Those that
> can see through their smoke-screen that they keep displaying, to justify why
> they spent so much have no other clear choice but to run and buy the best P&S
> cameras.

Your lies are getting old.
>
> All this arguing about it boils down to just one thing: People who didn't do
> their homework because they blindly followed some self-appointed authority and
> self-proclaimed "pro" are now trying to justify why they spent so much money to
> get a decent picture.
>
Absolutely false.

Roger
 >> Stay informed about: Rockwell on DSLR vs. P&S 

>> I would like to know, what are the better downsamplers?
>
> At whole-fraction scaling factors (25%, 33%, 50%, 67%, 75%)
> most downsampling algorithms give identical results.
>
> Otherwise Lanczos is far better than Bicubic for most images.
> Lanczos leaves smooth areas smooth and keeps edges sharp.
>
> It's possible Bicubic is good if you're changing aspect ratio
> (different scale for x / y). I have not investigated.
>

OK thanks for the reply! I guess the closest I've got to Lanczos is Irfanview
and xnview. But I've read that Irfanview's version is not the whole version
that is installed in a software program, PhotoLine32 has it full versioned,
written in Germany. I am not gonna buy it.

--
Giant_Alex
not my site: http://www.e-sword.net/
 >> Stay informed about: Rockwell on DSLR vs. P&S 

Annika1980 wrote:
> My buddy, Ken Rockwell, did a recent comparison of high-ISO
> performance between DSLRs that can be found here:
> http://kenrockwell.com/tech/iso-comparisons/2007-10/index.htm
>
> Note what he says about the Point & Shoots near the end of the
> article:
>
> "I was too lazy to include a compact camera, which as I showed last
> year, is abysmal compared to any DSLR. A typical compact camera, like
> the Canon SD700 I use all the time, is ten times worse than any DSLR.
> My SD700 at its lowest ISO 80 looks about the same as any of these
> DSLRs at ISO 800! "
>
> LOL! Even Ken Rockwell says, "Sucks to be you, Point and Shitters !!!"

I put Ken Rockwell right up there with ASAAR in terms of providing
accurate information, but even Rockwell gets it right occasionally.

Don't you love how he writes "Personally I get my goodies at Ritz,
Amazon and Adorama," without disclosing that the hyperlink to each of
those uses his affiliate code. There are many of use on
rec.photo.digital that have web sites with links to affiliate accounts,
but at least we clearly disclose that fact on the web sites.

Then he "apologizes" for the third party ads, that he allows to be
displayed on his site from the dealtime.com/shopping.com affiliate program.
 >> Stay informed about: Rockwell on DSLR vs. P&S 

GetAClue wrote:
[]
> All this arguing about it boils down to just one thing: People who
> didn't do their homework because they blindly followed some
> self-appointed authority and self-proclaimed "pro" are now trying to
> justify why they spent so much money to get a decent picture.

No, having 16 times the sensitivity is about getting photos, the cost was
about the same.

David
 >> Stay informed about: Rockwell on DSLR vs. P&S