SD14 JPEG interpolation

[LMC54]

Why oh why do you want interpolated images from Bayer to start with no matter what they do :/ amazing how we hold on to old Bayer technology LOL

--
http://www.fredmiranda.com/hosting/showgallery.php?ppuser=235&cat=500
http://www.pbase.com/lmc54/sd10

 

[Ante Vukorepa]

Because THIS is what a Canon 5D crop looks like:



Are you telling me you seriously expect the interpolated 14MP from SD14 will look like that at 100%?!

Don't take me wrong, Foveon is great and i'm sure SD14 will live up to it's expectations, but let's just stop with the dreams here...

IF... the SD14 can produce a 14 MP JPEG that is comparable in
resolved detail to, say, a Canon 5D, then why wouldn't people buy
an SD14 for half the price?

--
--------------------------------------------
Ante Vukorepa

 

[Ante Vukorepa]

I don't!

I just want people to stop mistaking color resolution for spatial resolution and stop daydreaming about 14MP interpolated files that rival 13MP native spatial resolution files

That, and it really would be nice (if impractical) to have a really real 14MP (i.e. 3x14MP) Foveon sensor, but that's just me dreaming again

--
--------------------------------------------
Ante Vukorepa

 

[markspan]

I don't!
I just want people to stop mistaking color resolution for spatial
resolution and stop daydreaming about 14MP interpolated files that
rival 13MP native spatial resolution files

what is measured at those spatial locations? its not colour is it? and let colour be what the picture makes.... (yes obi wan)

That, and it really would be nice (if impractical) to have a really
real 14MP (i.e. 3x14MP) Foveon sensor, but that's just me dreaming
again

--
--------------------------------------------
Ante Vukorepa

 

[markspan]

Because THIS is what a Canon 5D crop looks like:



Are you telling me you seriously expect the interpolated 14MP from
SD14 will look like that at 100%?!

Most of us are eager to find out....

Don't take me wrong, Foveon is great and i'm sure SD14 will live up
to it's expectations, but let's just stop with the dreams here...

IF... the SD14 can produce a 14 MP JPEG that is comparable in
resolved detail to, say, a Canon 5D, then why wouldn't people buy
an SD14 for half the price?

--
--------------------------------------------
Ante Vukorepa

 

[LMC54]

I don't!
I just want people to stop mistaking color resolution for spatial
resolution and stop daydreaming about 14MP interpolated files that
rival 13MP native spatial resolution files

That, and it really would be nice (if impractical) to have a really
real 14MP (i.e. 3x14MP) Foveon sensor, but that's just me dreaming
again

--
--------------------------------------------
Ante Vukorepa

Good, then get a Foveon and get real colors instead of interpolated slop.

--
http://www.fredmiranda.com/hosting/showgallery.php?ppuser=235&cat=500
http://www.pbase.com/lmc54/sd10

 

[Peter G]

Pardon my french, but - boll* ks!
It is an overrepresentation, but in color resolution.

Besides, things aren't as grim with Bayer as they are usually
portrayed. New Bayer interpolation algorithms (like adaptive
homogeneity-directed demosaicing algorithm) have gone leaps and
bounds over the plain vanilla Bayer demosaic.

The truth is Sigma is true 4.6 MP sensor, it captures all colors and will not be affected by subject matter. It can always capture full 4.6MP individual samples if the lens can deliver.

I am no zeolot for either side. I use Bayer and I have tried many converters, RSE, Silkypix, ACR, RIT and DCRAW are currently installed. I have found no resolution advantage with the G6 from DCRAW. If you could provide a RAW that demonstrates this, I would like to be enlightened.

It is a fact that because of interpolation and the necessary AA filter Bayer will lose resolution. I don't want to get into an agument about the factors with zealots from either camp. The factor will vary depending on what subject matter bayer encounters and the strength of the AA filter. I find as a rule of thumb dividing bayer/2 works very well most of the time.

So a 7MPx3 would equal a bayer 14MPx1 (more or less).

You can't think your camera would be equal in resolution to a 14MP x3 sensor?

 

[Roland Karlsson]

Good, then get a Foveon and get real colors instead of interpolated
slop.

Now Larry - did it look good or not?

It was a 100% crop - i.e. every pixel in the image corresponded to a pixel in the Bayer sensor. And to me it looked good - very good. So good that I would not have been surprised if someone here said it proved the quality of X3 technology. And that is a high praise.

You can get 10 miillion of those pixels for $900.

With a SD14 you can get 4.7 million of not upscaled color pixels in your picture for twice that sum. Thats 4 times the cost per pixel.

So - to show that the SD14 is worth the money you have to show that every non upscaled pixel is 4 times better than a CFA Bayer pixel. You just got the bench mark to compare to posted here. 4 times better.

--
Roland
http://klotjohan.mine.nu/~roland/

 

[Jay Turberville]

9-11 is OK. Its just 20% difference and there will not be heated
discussion regarding if 9 or 11 is OK. Enough people is mature
enough for avoiding that, I hope

Or you could just say "about 10" or "around 10" etc.

--
Jay Turberville
http://www.jayandwanda.com

 

[Roland Karlsson]

So a 7MPx3 would equal a bayer 14MPx1 (more or less).

Yepp - thats the common belief. The same as 5x3 equalling 10x1. Unfortunately 5x3 will cost $1600 and 10x1 cost $800 now. Quite an advantage I would say.

You can't think your camera would be equal in resolution to a 14MP
x3 sensor?

Hmmm ... some cameras have rather weak AA filters to emphasis sharpness before supressing artefacts.

But in general - no - a Bayer with AA filter cannot get the same resolution as an X3 pixel per pixel.

--
Roland
http://klotjohan.mine.nu/~roland/

 

[Jay Turberville]

Now Larry - did it look good or not?

Yes. It looks very good. Very non-sloppish.

It was a 100% crop - i.e. every pixel in the image corresponded to
a pixel in the Bayer sensor. And to me it looked good - very good.
So good that I would not have been surprised if someone here said
it proved the quality of X3 technology. And that is a high praise.

Why did it look good? It has very little in the way of saturated color.

It is also without much pedigree.

You can get 10 miillion of those pixels for $900.

Where? They are smaller on the new 10Mp camera. So you can get 10Mp similar pixels for that price, but not quite the same.

With a SD14 you can get 4.7 million of not upscaled color pixels in
your picture for twice that sum. Thats 4 times the cost per pixel.

So - to show that the SD14 is worth the money you have to show that
every non upscaled pixel is 4 times better than a CFA Bayer pixel.
You just got the bench mark to compare to posted here. 4 times
better.

Silly math. I'm assuming you aren't really serious.

--
Jay Turberville
http://www.jayandwanda.com

 

[LMC54]

Good, then get a Foveon and get real colors instead of interpolated
slop.

Now Larry - did it look good or not?

It was a 100% crop - i.e. every pixel in the image corresponded to
a pixel in the Bayer sensor. And to me it looked good - very good.
So good that I would not have been surprised if someone here said
it proved the quality of X3 technology. And that is a high praise.

You can get 10 miillion of those pixels for $900.

With a SD14 you can get 4.7 million of not upscaled color pixels in
your picture for twice that sum. Thats 4 times the cost per pixel.

So - to show that the SD14 is worth the money you have to show that
every non upscaled pixel is 4 times better than a CFA Bayer pixel.
You just got the bench mark to compare to posted here. 4 times
better.

--
Roland
http://klotjohan.mine.nu/~roland/

No it did not look good to me... looks just like what comes from my Bayer sensors after being sharpened to death. Problem with sharpening is that it does micro contrating which leads to this metallic video look. Sorry looks like bayer sharpened image.
regards,
Larry

--
http://www.fredmiranda.com/hosting/showgallery.php?ppuser=235&cat=500
http://www.pbase.com/lmc54/sd10

 

[Roland Karlsson]

Where? They are smaller on the new 10Mp camera. So you can get
10Mp similar pixels for that price, but not quite the same.

Fair enough.

Silly math. I'm assuming you aren't really serious.

Yepp - silly math. I just got somewhat tired of hearing that he should really use this or that technology if he wants better pictures. In this case in the light that he just have showed a very good picture using the competing technology.

--
Roland
http://klotjohan.mine.nu/~roland/

 

[Roland Karlsson]

No it did not look good to me... looks just like what comes from my
Bayer sensors after being sharpened to death. Problem with
sharpening is that it does micro contrating which leads to this
metallic video look. Sorry looks like bayer sharpened image.

OK - it is a heavily compressed JPEG.

Can we have a lesser compressed picture or even a tiff Ante?

Then Larry - point me to the micro contrating

It is here were the possible openings for SD14 starts. By analyzing real Bayer imaging faults not found in direct color imaging. Not by silli pixel counting.

--
Roland
http://klotjohan.mine.nu/~roland/

 

[LMC54]

No it did not look good to me... looks just like what comes from my
Bayer sensors after being sharpened to death. Problem with
sharpening is that it does micro contrating which leads to this
metallic video look. Sorry looks like bayer sharpened image.

OK - it is a heavily compressed JPEG.

Can we have a lesser compressed picture or even a tiff Ante?

Then Larry - point me to the micro contrating

It is here were the possible openings for SD14 starts. By analyzing
real Bayer imaging faults not found in direct color imaging. Not by
silli pixel counting.

--
Roland
http://klotjohan.mine.nu/~roland/

What do think sharpening is? Sharpening changes the constrat of detail edges... send me the actual RAW and I will show you what the freaking thing looks like at 100% without sharpening. It'll be soft but the colors and contrasts would look a weeee better.
regards,
larry

--
http://www.fredmiranda.com/hosting/showgallery.php?ppuser=235&cat=500
http://www.pbase.com/lmc54/sd10

 

[Roland Karlsson]

What do think sharpening is? Sharpening changes the constrat of
detail edges... send me the actual RAW and I will show you what the
freaking thing looks like at 100% without sharpening. It'll be
soft but the colors and contrasts would look a weeee better.

This picture is so heavily JPEG-ed that you cannot see any problems with any sharpening IMHO. Look at the thing att 400% and it will look like a satelite photo over mid america full of mile roads.

--
Roland
http://klotjohan.mine.nu/~roland/

 

[Kendall Helmstetter Gelner]

Pardon my french, but - boll* ks!
It is an overrepresentation, but in color resolution.

Besides, things aren't as grim with Bayer as they are usually
portrayed. New Bayer interpolation algorithms (like adaptive
homogeneity-directed demosaicing algorithm) have gone leaps and
bounds over the plain vanilla Bayer demosaic.

Here's an example:
http://o.orcinus.googlepages.com/Zeiss_2.4_35_f8_crop.jpg

I invite people to examine all crops carefully at 400% before they submit them for revue, as it's a lot easier to see the issues in an image that way.

1) Are those vertical blinds on the windows, or are they really plain white sheets/blinds? The greyness of them compared to the white frames iplies blinds and there are hints of occasional vertical lines.

2) What is up with that maze on the far left or your crop, in the window on the edge of the roofline?

3) Where the top of the shadowed tree intersects the window in the bottom left corner, there are some interesting color artifacts (also in home highlights here and there off lights and a little bit in a faw balcony railings)

4) Is that bottom satellite dish really partly blue with a pink rim? Judging by the other ones in the picture I think not.

5) There's a faceless man in the window where the green roof ends on the right side of the picture.

6) The sky has some kind of odd cross-hatch pattern

7) Why does the wire leading to the satellite dish on the edge of the balcony (bottom middle) taper off? It looks like it's going to the dish, but never makes it...

8) Really curious if those are two sticks with wires between them in the white window with the shades raised pretty much in the center of the crop. The wires kind of fade in and out, the shadows are more clear....

Yes it has done a decent job giving you back some sharpness, but we are also seeing some of the classic bayer artifatcs and early detail extinction. You are reminding me all over again why I waited for the SD-14.

--
---> Kendall
http://InsideAperture.com
http://www.pbase.com/kgelner
http://www.pbase.com/sigmasd9/user_home

 

[LMC54]

What do think sharpening is? Sharpening changes the constrat of
detail edges... send me the actual RAW and I will show you what the
freaking thing looks like at 100% without sharpening. It'll be
soft but the colors and contrasts would look a weeee better.

This picture is so heavily JPEG-ed that you cannot see any problems
with any sharpening IMHO. Look at the thing att 400% and it will
look like a satelite photo over mid america full of mile roads.

--
Roland
http://klotjohan.mine.nu/~roland/

I don't care if it is RAW! I know what a Bayer produces. IMHO I don't like it.
regards,
Larry

--
http://www.fredmiranda.com/hosting/showgallery.php?ppuser=235&cat=500
http://www.pbase.com/lmc54/sd10

 

[Kendall Helmstetter Gelner]

What do think sharpening is? Sharpening changes the constrat of
detail edges... send me the actual RAW and I will show you what the
freaking thing looks like at 100% without sharpening. It'll be
soft but the colors and contrasts would look a weeee better.

This picture is so heavily JPEG-ed that you cannot see any problems
with any sharpening IMHO. Look at the thing att 400% and it will
look like a satelite photo over mid america full of mile roads.

I have been and while some artifacts might be from JPG (The sky?), other things like the maze in the far left window and color errors cannot be caused by JPG - JPG will just flatten out some gradients, not add new colors where none were before.

See my other message for a fuller list of issues, I don't think most of them are JPG related.

--
---> Kendall
http://InsideAperture.com
http://www.pbase.com/kgelner
http://www.pbase.com/sigmasd9/user_home

 

[Kendall Helmstetter Gelner]

Also, I don't know why I passed over it before but take a look at all of the antennas on the roof! The tops of them all just fade in and out of existance, my own personal pet peeve for the most annoying Bayer behaviour right after color mosiacing. It's not even that unnoticable at 100%.

That's exactly where an image could use a little "false" detail to let you know exactly how far the antenna extends.

The image gives the illusion of per-pixel sharpness without all of the benefits.

--
---> Kendall
http://InsideAperture.com
http://www.pbase.com/kgelner
http://www.pbase.com/sigmasd9/user_home