Sigma publishes an Image Sample Gallery

[DavidMillier]

I'm viewing this image through firefox at perfectly normal resolution. I can see the rope effects on the image below perfectly clearly at about a 10 inch viewing distance from the monitor.

Some examples

I can see a 'rope' type artifact on one of the down elements of the
lamp and strange diagonal stripes in the blinds that might possibly
be moire.

I didn't "see" them because I am looking at 100 percent. And yes,
at 300 percent I can see them. I can also see jaggies at 300
percent on all my digital photo's...

On the SD9/10 I could see jaggies at 100 percent. No need for
zooming...

Dave

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[Ralf Ronander]

Not even with a really badly smearing interpolation method like that are you able to get rid of the jaggies

The originals are far from bad, and certainly not "deserving" a treatment like that.

I´m just pointing out the obvious; there are jaggies and sometimes they are very disturbing, I wouldn´t have noticed them otherwise.

 

[Lin Evans]

Are you using an LCD monitor? If so, the vast majority of LCD monitors, though producing very sharp images, are incapable of displaying the full gamut of colors or greyscale. Often I can see jaggies on my wife's 20 inch (and very expensive) LCD monitor which I can't see on any of my calibrated CRT's or on my best friend's 9 megapixel IBM monitor.

I suspect the reason you are seeing any jaggies at 100% viewing size has to do with the monitor's inability to display the full gamut of greyscale pixels which fill in the gaps by darker pixels on the edges. This is normal. On the other hand if you are seeing them on a CRT could you tell me what screen resolution and which video card you are using?

Best regards,

Lin

I doubt you would see it in any reasonable size print though but it
is perfectly clear on screen.

If you are talking about normal stairstep aliasing which can really
only be seen in these samples by using excessive zoom .... snip

What? Excessive zoom? These are nothing but 100%, certainly not
"excessive zoom"!

And that's the point.

Dave

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[DavidMillier]

Hi Lin

I recently switched from a 17 inch CRT to using a Samsung 215TW 21inch LCD at 1680* 1050. Generally since I've had it I notice things I couldn't see on the blurry CRT - this is the first time I've heard anyone explain away digital artifacts by blaming the screen but I'm willing to learn!

Video card is one of those nVidia gaming things, can never remember the number they change so fast. Does so many billion vertex pixel things per second through whoknowshowmanypipelines in parallel....

I suspect the reason you are seeing any jaggies at 100% viewing
size has to do with the monitor's inability to display the full
gamut of greyscale pixels which fill in the gaps by darker pixels
on the edges. This is normal. On the other hand if you are seeing
them on a CRT could you tell me what screen resolution and which
video card you are using?

Best regards,

Lin

I doubt you would see it in any reasonable size print though but it
is perfectly clear on screen.

If you are talking about normal stairstep aliasing which can really
only be seen in these samples by using excessive zoom .... snip

What? Excessive zoom? These are nothing but 100%, certainly not
"excessive zoom"!

And that's the point.

Dave

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[DavidMillier]

You know how this sounds...the sweet irony!

If you had a monitor with single pixel resolution you would be able to see the jaggies. The fact you prefer the nutella CRT effect...

I'm sure Laurence would enjoy the joke ;-)

I suspect the reason you are seeing any jaggies at 100% viewing
size has to do with the monitor's inability to display the full
gamut of greyscale pixels which fill in the gaps by darker pixels
on the edges. This is normal. On the other hand if you are seeing
them on a CRT could you tell me what screen resolution and which
video card you are using?

Best regards,

Lin

I doubt you would see it in any reasonable size print though but it
is perfectly clear on screen.

If you are talking about normal stairstep aliasing which can really
only be seen in these samples by using excessive zoom .... snip

What? Excessive zoom? These are nothing but 100%, certainly not
"excessive zoom"!

And that's the point.

Dave

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[Erik Ohlson]

.....nothing here makes me want to toss out my PAnasonic FZ4 & TZ1

BTW That reservoir shot is taken very near where one of my dogs got bitten by a rattlesnake.

Dog is doing fine!

 

[Lin Evans]

I suspected that might be the case. It's not really explaining away any artifacts, but simply that you probably can't see the entire gamut of greyscale on the Samsung. The edge roll-off when examined at single pixel levels shows a series of parallel pixels starting with the very dark edge pixels and diminishing with each layer until a match with the backdrop is seen. Bayer cameras show more parallel pixel layers making up this edge due to the AA filtering so edges are softened in appearance. When USM is applied to edges it increases relative contrast between the adjacent gradations which results in drawing our eyes to these edges thus giving the effect of enhanced "sharpness".

The X3 capture has fewer of these adjacent graduations and when a monitor doesn't discern between small differences in tonality the effect is to make the edge appearances more harsh. In other words most LCD monitors exacerbate jaggies while most CRT monitors ameliorate them via revealing more shades of grey as well as more shades of color.

There are distinct advantages to LCD monitors for many applications but also disadvantages. Determining the true amount of "jaggies" in an image is not a strength of LCD monitors.

The video card and display resolution can somewhat affect our perceptions of edge quality but it sounds as if the one you have is likely quite capable and not a substantive variable.

Best regards,

Lin

I recently switched from a 17 inch CRT to using a Samsung 215TW
21inch LCD at 1680* 1050. Generally since I've had it I notice
things I couldn't see on the blurry CRT - this is the first time
I've heard anyone explain away digital artifacts by blaming the
screen but I'm willing to learn!

Video card is one of those nVidia gaming things, can never remember
the number they change so fast. Does so many billion vertex pixel
things per second through whoknowshowmanypipelines in parallel....

I suspect the reason you are seeing any jaggies at 100% viewing
size has to do with the monitor's inability to display the full
gamut of greyscale pixels which fill in the gaps by darker pixels
on the edges. This is normal. On the other hand if you are seeing
them on a CRT could you tell me what screen resolution and which
video card you are using?

Best regards,

Lin

I doubt you would see it in any reasonable size print though but it
is perfectly clear on screen.

If you are talking about normal stairstep aliasing which can really
only be seen in these samples by using excessive zoom .... snip

What? Excessive zoom? These are nothing but 100%, certainly not
"excessive zoom"!

And that's the point.

Dave

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[Peter G]

I suspect the reason you are seeing any jaggies at 100% viewing
size has to do with the monitor's inability to display the full
gamut of greyscale pixels which fill in the gaps by darker pixels
on the edges. This is normal. On the other hand if you are seeing
them on a CRT could you tell me what screen resolution and which
video card you are using?

I have dual monitors, one LCD, one CRT. I move the picture back and forth. Nothing changes. This are just pointless rationalizations to deny a well known effect of not having an AA filter.

Any grid imaging device without AA filtering will produce jagged edges and other aliasing artifacts. It was painfully obvious on the SD9-10 even in 8x10" prints.

The laws of physics don't change with the new sensor. I will agree that with more pixels this is less of an issue, but lets not pretend this is not a actual symptom of the lack of AA and try to blame it on the monitor.

 

[DavidMillier]

Hmmm

Not convinced. What I see looks exactly like what I saw in SD10 images on my old CRT. Identical.

However, the extra resolution does mean it is a little smaller. I doubt very much whether it would be visible on any size print I would print but I don't think you can justifiably wish it away...

The X3 capture has fewer of these adjacent graduations and when a
monitor doesn't discern between small differences in tonality the
effect is to make the edge appearances more harsh. In other words
most LCD monitors exacerbate jaggies while most CRT monitors
ameliorate them via revealing more shades of grey as well as more
shades of color.

There are distinct advantages to LCD monitors for many applications
but also disadvantages. Determining the true amount of "jaggies" in
an image is not a strength of LCD monitors.

The video card and display resolution can somewhat affect our
perceptions of edge quality but it sounds as if the one you have is
likely quite capable and not a substantive variable.

Best regards,

Lin

I recently switched from a 17 inch CRT to using a Samsung 215TW
21inch LCD at 1680* 1050. Generally since I've had it I notice
things I couldn't see on the blurry CRT - this is the first time
I've heard anyone explain away digital artifacts by blaming the
screen but I'm willing to learn!

Video card is one of those nVidia gaming things, can never remember
the number they change so fast. Does so many billion vertex pixel
things per second through whoknowshowmanypipelines in parallel....

I suspect the reason you are seeing any jaggies at 100% viewing
size has to do with the monitor's inability to display the full
gamut of greyscale pixels which fill in the gaps by darker pixels
on the edges. This is normal. On the other hand if you are seeing
them on a CRT could you tell me what screen resolution and which
video card you are using?

Best regards,

Lin

I doubt you would see it in any reasonable size print though but it
is perfectly clear on screen.

If you are talking about normal stairstep aliasing which can really
only be seen in these samples by using excessive zoom .... snip

What? Excessive zoom? These are nothing but 100%, certainly not
"excessive zoom"!

And that's the point.

Dave

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[Lin Evans]

Well, that's nice that in your experience there is no difference. In my experience and that of countless others there is a difference between what is seen on "most" LCD monitors and what is seen on CRT monitors in terms of "jaggies".

We've been down this road before (you and I) and nothing will be gained by arguing. I was probably the very first person to point out "jaggies" on my original SD9 - one which Laurence owns now. No one has said that "jaggies" don't exist and if you read my earlier explanation carefully, something you "should" do before criticizing, you will see that I explained why they are greater in visibility on X3 captures than on CFA captures.

The issue was seeing them at 100% and the "majority" of people do not see them with the SD14 images at 100%. Several of us who post here have used the SD14 and taken many thousands of images between us and have had ample opportunity to do what you have not and that is look long and hard at many, many images from the camera.

Jaggies are simply not of any primary concern. If you feel otherwise then state your case and as far as I'm concerned that's the end of it. I'm not going to be drawn into a silly argument with you about something which may be a personal bias you have.

I suspect the reason you are seeing any jaggies at 100% viewing
size has to do with the monitor's inability to display the full
gamut of greyscale pixels which fill in the gaps by darker pixels
on the edges. This is normal. On the other hand if you are seeing
them on a CRT could you tell me what screen resolution and which
video card you are using?

I have dual monitors, one LCD, one CRT. I move the picture back and
forth. Nothing changes. This are just pointless rationalizations to
deny a well known effect of not having an AA filter.

Any grid imaging device without AA filtering will produce jagged
edges and other aliasing artifacts. It was painfully obvious on the
SD9-10 even in 8x10" prints.

The laws of physics don't change with the new sensor. I will agree
that with more pixels this is less of an issue, but lets not
pretend this is not a actual symptom of the lack of AA and try to
blame it on the monitor.

I believe I know the "laws of physics" quite as well as you........

Lin

 

[Lin Evans]

Is it "exactly" or is it "smaller" - you can't have it both ways Of course it's not "exactly" the same - it's exactly as it's expected to be and the amount of visiblitiy at 100% always depends on the number of pixels representing a linear dimension. To the vast majority it's a non issue.

You can either believe me or not or do research on LCD monitors where you can learn the relevant differences - it's really not material. The important thing is that you can also see it on your Kodak 14 or on any other digital camera. What is important is that it's a non issue except for those who want something else to pick apart.

Best regards,

Lin

Not convinced. What I see looks exactly like what I saw in SD10
images on my old CRT. Identical.

However, the extra resolution does mean it is a little smaller. I
doubt very much whether it would be visible on any size print I
would print but I don't think you can justifiably wish it away...

The X3 capture has fewer of these adjacent graduations and when a
monitor doesn't discern between small differences in tonality the
effect is to make the edge appearances more harsh. In other words
most LCD monitors exacerbate jaggies while most CRT monitors
ameliorate them via revealing more shades of grey as well as more
shades of color.

There are distinct advantages to LCD monitors for many applications
but also disadvantages. Determining the true amount of "jaggies" in
an image is not a strength of LCD monitors.

The video card and display resolution can somewhat affect our
perceptions of edge quality but it sounds as if the one you have is
likely quite capable and not a substantive variable.

Best regards,

Lin

I recently switched from a 17 inch CRT to using a Samsung 215TW
21inch LCD at 1680* 1050. Generally since I've had it I notice
things I couldn't see on the blurry CRT - this is the first time
I've heard anyone explain away digital artifacts by blaming the
screen but I'm willing to learn!

Video card is one of those nVidia gaming things, can never remember
the number they change so fast. Does so many billion vertex pixel
things per second through whoknowshowmanypipelines in parallel....

I suspect the reason you are seeing any jaggies at 100% viewing
size has to do with the monitor's inability to display the full
gamut of greyscale pixels which fill in the gaps by darker pixels
on the edges. This is normal. On the other hand if you are seeing
them on a CRT could you tell me what screen resolution and which
video card you are using?

Best regards,

Lin

I doubt you would see it in any reasonable size print though but it
is perfectly clear on screen.

If you are talking about normal stairstep aliasing which can really
only be seen in these samples by using excessive zoom .... snip

What? Excessive zoom? These are nothing but 100%, certainly not
"excessive zoom"!

And that's the point.

Dave

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[aruzinsky]

1. Your talk is cheap. Let's see you do better.

2. Although my method has a higher PSNR than popular methods, it is not my best method, just the best for not introducing jaggies.

3. If anything, my method did introduce jaggies. I used it only because it introduces less jaggies than other methods while retaining the original pixels

Not even with a really badly smearing interpolation method like
that are you able to get rid of the jaggies
The originals are far from bad, and certainly not "deserving" a
treatment like that.
I´m just pointing out the obvious; there are jaggies and sometimes
they are very disturbing, I wouldn´t have noticed them otherwise.

--
Author of SAR Image Processor and anomic sociopath
http://www.general-cathexis.com

 

[Deecy]

--Lighten up!

Deecy//

"There are specific brand forums available."

I think the way this thread has drifted off into discussions on
jaggies and the hurling of insults rather proves my point. NO NEWS
HERE.

Mustafa

Thanks for reinforcing my last point.

Mustafa

--If you think what I suggested was an insult you've lived a very sheltered life my friend!

T.
Deecy//



http://www.flickr.com/photos/provocative/

http://tomdeecy.blogspot.com/

 

[Chato]

--Lighten up!

Deecy//

"There are specific brand forums available."

I think the way this thread has drifted off into discussions on
jaggies and the hurling of insults rather proves my point. NO NEWS
HERE.

Mustafa

Thanks for reinforcing my last point.

Mustafa

--If you think what I suggested was an insult you've lived a very
sheltered life my friend!

Yes indeed....

This gallery is News. And of course whether of not the Sigma's produce jaggies, whether pro or con, is part of the news story. While I don't plan to buy a Sigma, it is threads like these which might change my mind. Indeed, I don't even look in the Sigma forum, because I don't own one.

Certainly if I bought a Sigma I would post there - But my posting now, would be a kind of trollish thing to do...

Dave

 

[Mustafa]

--If you think what I suggested was an insult you've lived a very
sheltered life my friend!

Maybe impertinent would be a better word.

Mustafa

 

[Lin Evans]

Here are some links in which you may find some more information on the difference between CRT and LCD display characteristics.

LCD monitors come in many "flavors". Some very badly exacerbate jaggies even to the point of producing these artifacts. Others with sub-pixel algorithms can actually smooth edges, but only in the horizontal aspect of display. If one uses the monitor in the vertical direction then sub-pixel ant-aliasing is inoperable on LCD monitors.

The difference in what we often see and discuss and even argue about on these forums as far as pixel peeping is concerned is that we do not all "see" the same things on our monitors. Some of this is due to browser differences, some is due to monitor differences and some is undoubtedly due to individual perception differences. The only thing which is truly important to the photographer is how the print looks - at least to those photographers who actually print their captures. There are probably a large number who never go further than displaying their captures on the web.

So what may be glaringly "apparent" to one may be totally "invisible" to another and it's entirely possible that the truth lies between these extremes. These links may help to explain the reasons for descrepancies.

Lin

http://www.grc.com/ctwhat.htm

http://www.dansdata.com/pv720.htm

http://www.hometheatermag.com/lcds/106maxent/

http://www.gcn.com/print/16_27/32204-1.html

 

[DavidMillier]

Here are some links in which you may find some more information on
the difference between CRT and LCD display characteristics.

LCD monitors come in many "flavors". Some very badly exacerbate
jaggies even to the point of producing these artifacts. Others with
sub-pixel algorithms can actually smooth edges, but only in the
horizontal aspect of display. If one uses the monitor in the
vertical direction then sub-pixel ant-aliasing is inoperable on LCD
monitors.

The difference in what we often see and discuss and even argue
about on these forums as far as pixel peeping is concerned is that
we do not all "see" the same things on our monitors. Some of this
is due to browser differences, some is due to monitor differences
and some is undoubtedly due to individual perception differences.
The only thing which is truly important to the photographer is how
the print looks - at least to those photographers who actually
print their captures. There are probably a large number who never
go further than displaying their captures on the web.

So what may be glaringly "apparent" to one may be totally
"invisible" to another and it's entirely possible that the truth
lies between these extremes. These links may help to explain the
reasons for descrepancies.

Lin

http://www.grc.com/ctwhat.htm

http://www.dansdata.com/pv720.htm

http://www.hometheatermag.com/lcds/106maxent/

http://www.gcn.com/print/16_27/32204-1.html

--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm

 

[Ralf Ronander]

MY talk is cheap? You say more for the same price so by definition your talk is cheaper

Do better at what? Interpolation? Do you want to introduce jaggies or not?

I don´t know what you mean by "do better", interpolation is beside the point here. The point being, that jaggies are visible at 100%.

 

[JacobC]

http://www.sigma-sd14.com/sample-photo/index.html

That is the direct link to Sigma's site. It's usually looked upon as trying to generate hits for a site your affiliated with if you don't link to the original content, and instead go through another site/blog etc.

I'm just trying to nicely point it out not flame ya!

 

[Peter G]

More pixels, same technology, same behavior, except this time it appears that the jaggies previously caused by not having an AA filter are now caused by LCD monitors. I guess LCD's weren't popular enough to use as an excuse for jaggies when the SD-9 came out.

For the record I have a dual monitor setup with an LCD and CRT and I see the same jaggies on each. I resized the images and the same jaggies are still present.

Thus it is not the LCD monitor that is to blame.

That aside, with more pixels, the jaggies should be less obvious in prints. I am potentially interested in the Sigma DP-1 with the same sensor.