D100 Macro Sample

[Tom899]

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I then saved to tif, opened in Photo Shop, reduced resolution using Auto, picking 72dpi. This cut the size in half, then saved as jpg with a setting of 10.
Tom B.

 

[finkployd]

Holy Shmoly,

Nice lighting too....

more more more more more more more

 

[Moshe Vainer]

Hi Tom, this is 1.3mp image. Sorry to sound harsh, but it would be sharp even if you shoot with a bottom of a bottle... Every pixel here is about equal to 4 pixels of the original.

Rgrds,
Moshe

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

 

[XYZXYZXYZ]

Very nice photo Tom.

How much do the packages cost that you used to enhance the photo?

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

 

[Tom899]

Hi Moshe, I'm sorry, I don't understand your statement. Should I have saved it some other way? Please give me your way of saving and I will redo it and repost.
Tom B.

Rgrds,
Moshe

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

 

[Tom899]

Thanks finkployd,
Tom B.

Holy Shmoly,

Nice lighting too....

more more more more more more more

 

[Tom899]

Thanks TYROX, Nikon Capture is what I used, but only the trial version. I do plan on buying it though after the 30 day trial period ends. The other program is PhotoShop 7.0, around $150. US for the upgrade.
Tom B.

How much do the packages cost that you used to enhance the photo?

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

 

[Moshe Vainer]

Tom, please let me say again i am sorry if i sounded harsh...

It was my understanding that you send this image as an example of how sharp an image you can get from the combination of your lens/camera, especially in light of recent "d100 blurriness" threads.

What i wanted to stress, is that you use a 6mp camera to produce a 6mp image. You then downsample it to 1.3mp image. The process of downsampling would automatically make the image look much sharper, but in fact it would hold even less detail than the original.

As far as i can tell, Nikon choose to put a very strong antialias filter in front of their sensor. Their goal must have been to completely remove color moire, which they (almost) succeeded to do. Of course, in a bayer sensor, it is a mathematical certainty, that in order to anti-alias color moire completely, you HAVE to set your filter to cover about 4 pixels. That is why on the one hand, d100 images have very little color moire, on the other, they produce images that have roughly the detail of 1.5mp non-bayer camera (3 shot db, or foveon etc...)

If i take any 6mp image, even one that is somewhat out of focus, and reduce it to 1.3mp, i would get a cleaner, sharper on-screen image. However, trying to print it to any larger than 4x6 size would immediately reveal that it does not have enough detail.

Rgrds,
Moshe

P.S.

The shot itself is great, and my comments are simply regarding the technicalities of the issue.

Rgrds,
Moshe

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

 

[alfred_schiele]

Hi!

What i wanted to stress, is that you use a 6mp camera to produce a
6mp image. You then downsample it to 1.3mp image. The process of
downsampling would automatically make the image look much sharper,
but in fact it would hold even less detail than the original.

Well, no 6MP bayer pattern sensor is able to produce a blazting 6MP output file... and if sharpness looks really a bit better it is for the price of additional colour moiree risk...

As far as i can tell, Nikon choose to put a very strong antialias
filter in front of their sensor. Their goal must have been to
completely remove color moire, which they (almost) succeeded to do.

I agree. What is bad in this? Did you ever try to remove colour moiree??

Of course, in a bayer sensor, it is a mathematical certainty, that
in order to anti-alias color moire completely, you HAVE to set your
filter to cover about 4 pixels. That is why on the one hand, d100
images have very little color moire, on the other, they produce
images that have roughly the detail of 1.5mp non-bayer camera (3
shot db, or foveon etc...)

Doesn't this apply to ALL digital cameras based on bayer pattern?

I thank you for repeating this issue.

But I believe your practical reduction to 1.5 is too harsh either. I think a good compromise is to say a 6MP camera outputs a good (so not perfect) 3MP image... (-:

If i take any 6mp image, even one that is somewhat out of focus,
and reduce it to 1.3mp, i would get a cleaner, sharper on-screen
image. However, trying to print it to any larger than 4x6 size
would immediately reveal that it does not have enough detail.

Again agreed, but this applies to all images, not only D100 ones, doesn't it?

Regards, A. Schiele

 

[Moshe Vainer]

Hi Alfred,

What i wanted to stress, is that you use a 6mp camera to produce a
6mp image. You then downsample it to 1.3mp image. The process of
downsampling would automatically make the image look much sharper,
but in fact it would hold even less detail than the original.

Well, no 6MP bayer pattern sensor is able to produce a blazting 6MP
output file... and if sharpness looks really a bit better it is for
the price of additional colour moiree risk...

Definately, did i said anything to contradict it?

As far as i can tell, Nikon choose to put a very strong antialias
filter in front of their sensor. Their goal must have been to
completely remove color moire, which they (almost) succeeded to do.

I agree. What is bad in this? Did you ever try to remove colour
moiree??

I routinely do. I did not said it is bad or good. Some manufacturers choose moire over softness, others softness over moire. It is up to the user to decide what is more important to him. I just said that IMHO, nikon made a clear choice here, and went all the way to the softness vs the moire. It is neither good nor bad, just something to be aware of.

Of course, in a bayer sensor, it is a mathematical certainty, that
in order to anti-alias color moire completely, you HAVE to set your
filter to cover about 4 pixels. That is why on the one hand, d100
images have very little color moire, on the other, they produce
images that have roughly the detail of 1.5mp non-bayer camera (3
shot db, or foveon etc...)

Doesn't this apply to ALL digital cameras based on bayer pattern?

Yes to a different degree, as i said, some choose moire to produce sharper results.

I thank you for repeating this issue.

But I believe your practical reduction to 1.5 is too harsh either.
I think a good compromise is to say a 6MP camera outputs a good (so
not perfect) 3MP image... (-:

That is what i thought at first, but lately, i've been following a thread in a Pro forum that shows some DB, even 1 shot, and i have to downgrade our cameras even further :-(... Even bayer DBs (or at least some of them) are much (MUCH) sharper and cleaner, with no hint on moire when compared to 35mm digital. And i am not talking about more pixels, i am talking about per/pixel information.

If i take any 6mp image, even one that is somewhat out of focus,
and reduce it to 1.3mp, i would get a cleaner, sharper on-screen
image. However, trying to print it to any larger than 4x6 size
would immediately reveal that it does not have enough detail.

Again agreed, but this applies to all images, not only D100 ones,
doesn't it?

Definitely, i am not bashing the D100, i am considering it for myself vs the S2, and am in a kind of "internal debate" of softness vs moire.

Regards, A. Schiele

Rgrds,
Moshe

 

[alfred_schiele]

Hi

I enjoy you opinon. It is very close to mine

I agree. What is bad in this? Did you ever try to remove colour
moiree??

I routinely do.

How do you manage this without detail loss?

I tried to do this on S1 examples with strange hair moiree and didn't get any satisfying results at all...

I did not said it is bad or good. Some
manufacturers choose moire over softness, others softness over
moire. It is up to the user to decide what is more important to
him. I just said that IMHO, nikon made a clear choice here, and
went all the way to the softness vs the moire. It is neither good
nor bad, just something to be aware of.

I agree, Nikon decided for aggressive LP.

Yes to a different degree, as i said, some choose moire to produce
sharper results.

From my point of view, a digitizing system without appropriate LP according to the nyquist theoreme is simply not well done, regardless if the "artifacts" suggest sharpness...

But I believe your practical reduction to 1.5 is too harsh either.
I think a good compromise is to say a 6MP camera outputs a good (so
not perfect) 3MP image... (-:

That is what i thought at first, but lately, i've been following a
thread in a Pro forum that shows some DB, even 1 shot, and i have
to downgrade our cameras even further :-(... Even bayer DBs (or at
least some of them) are much (MUCH) sharper and cleaner, with no
hint on moire when compared to 35mm digital. And i am not talking
about more pixels, i am talking about per/pixel information.

Yes, I do the same, let's call it "full colour" pixels...

I added "practically" to make clear that even pixels containing about 2/3 of the colour information are much better than the current native resolutions "interpolated" from the bayer patterns... of course not perfect either as I agree, 4 of these monochromatic pixels would work out in a better true colour pixel.

Definitely, i am not bashing the D100, i am considering it for
myself vs the S2, and am in a kind of "internal debate" of softness
vs moire.

Maybe I change my mind if you tell me your secret of perfect moiree reduction while maintaining all the detail (-:

regards, A. Schiele

 

[Peter van Vliet]

Tom,

I am also a proud owner of the D100 and I was trying to get my hands on the NC3. You said you have a trial version, do you have by any chance a web site URL where I can find this.

I am also using PS 7.0 so is it then still needed?

I think that the picture is perfect but you have to push it through a lab printer, because then you will be really happy, I was when I saw the result of preproduction picture coming out of there in comparison with Canon.

Peter

How much do the packages cost that you used to enhance the photo?

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

--
Peter van Vliet

 

[Robert Briggs]

Doesn't a trial version of NC3 come with the D100? Look on your disks, pvanvliet.

Bob

I am also a proud owner of the D100 and I was trying to get my
hands on the NC3. You said you have a trial version, do you have by
any chance a web site URL where I can find this.

I am also using PS 7.0 so is it then still needed?

I think that the picture is perfect but you have to push it through
a lab printer, because then you will be really happy, I was when I
saw the result of preproduction picture coming out of there in
comparison with Canon.

Peter

How much do the packages cost that you used to enhance the photo?

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

--
Peter van Vliet

 

[Robert Briggs]

Very nice, Tom. Wonderful macro shot and nice lighting, too. Let the trolls have their fun while the happy D100 owners keep taking nice photos.

[click]

Bob

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

 

[Peter van Vliet]

Nops, nothing there. Nikon D100 is good, but the box is coming rather empty.....

Peter

Bob

I am also a proud owner of the D100 and I was trying to get my
hands on the NC3. You said you have a trial version, do you have by
any chance a web site URL where I can find this.

I am also using PS 7.0 so is it then still needed?

I think that the picture is perfect but you have to push it through
a lab printer, because then you will be really happy, I was when I
saw the result of preproduction picture coming out of there in
comparison with Canon.

Peter

How much do the packages cost that you used to enhance the photo?

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

--
Peter van Vliet

--
Peter van Vliet

 

[alfred_schiele]

Hi

Nops, nothing there. Nikon D100 is good, but the box is coming
rather empty.....

Interesting opinion, what you like to see in the box additionally?

Regards, A. Schiele

 

[Tom899]

Thanks Bob, Yes I am very very happy with my new D100. It is even more than I expected.
Tom B.

[click]

Bob

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

 

[Tom899]

Peter, The advantage of having NC 3 is the ability to manipulate the Raw (NEF) files in 12 bit. You can easily change white balance, exposer (EV), USM, etc.. I really like the trial version and I plan on purchasing it. There are other programs that do somewhat the same such as Bibble and QImage. They may not support the D100 yet, but I'm sure it won't be long. The trial version came with my D100 on a separate CD. It may be US only, I don't know why. I'm sure you will enjoy your D100
Tom B.

I am also a proud owner of the D100 and I was trying to get my
hands on the NC3. You said you have a trial version, do you have by
any chance a web site URL where I can find this.

I am also using PS 7.0 so is it then still needed?

I think that the picture is perfect but you have to push it through
a lab printer, because then you will be really happy, I was when I
saw the result of preproduction picture coming out of there in
comparison with Canon.

Peter

How much do the packages cost that you used to enhance the photo?

I have an old 55mm AF Micro Nikor F2.8 lens. I just brought it out
and shot this CD in NEF. I gave it a little USM in NC3 (40,4,10). I
then saved to tif, opened in Photo Shop, reduced resolution using
Auto, picking 72dpi. This cut the size in half, then saved as jpg
with a setting of 10.
Tom B.

--
Peter van Vliet

 

[Moshe Vainer]

Hi

I enjoy you opinon. It is very close to mine

Thank you

I agree. What is bad in this? Did you ever try to remove colour
moiree??

I routinely do.

How do you manage this without detail loss?

I tried to do this on S1 examples with strange hair moiree and
didn't get any satisfying results at all...

See my answer a little below...

Yes, I do the same, let's call it "full colour" pixels...

I added "practically" to make clear that even pixels containing
about 2/3 of the colour information are much better than the
current native resolutions "interpolated" from the bayer
patterns... of course not perfect either as I agree, 4 of these
monochromatic pixels would work out in a better true colour pixel.

Or one tri-color pixel would be even better, since then you would not loose light for color...

Maybe I change my mind if you tell me your secret of perfect moiree
reduction while maintaining all the detail (-:

Please understand, i am not trying to change your mind, only to put my thoughts forward and in the process make my own decision clearer.

As for the process, it is not one, but rather a combination of techniques that together give me results that i usually find satisfying...

Being an owner of FP6900, i got quite used to get some moire with my pictures, and therefore had to find ways to defeat it.

I am sure you are aware that a general way to reduce moire is to blur the color channel, so i would not go into detail about it. It is the way you selectively apply the blur that is important.

I have noticed several things:

1. Images coming from digital cameras, especially Fuji, are very saturated when you compare them to film, therefor this leaves you with ability to loose some color without loosing too much of a visual image quality.

2. The other, and most important thing that i have noticed is that moire is more often than not appears in areas that are monochromatic in nature - Hair, fabric etc.. Let's take hair for example, it has a basic hue and is not very rich in color detail, it is the luma detail that confuses the debayering and creating the colorfull mosaic we see in hair. What i do is select the affected areas, like hair, usually by restricting the general area with rectangle selection, using color selection to roughly select the area of interest, then expand the selection to include everything i am interested in, feather the selection so that decoloration effects would not be very visible, and then (being of course in lab mode) blur the color channels so that moire is gone. I usually prefer to use smart blur, and not gaussian blur, as this usually gives me better results with less color detail loss.

If you have quantum mechanic, applying it to the selected area also helps.

I repeat this process, varying the selection, degree of blurring and the blurr method depending on the area in question, and the results i find most pleasing.

Usually, the whole process is very quick.

If you are interested, i can take several examples of moire in available Fuji samples, and post the results that i get.

Be your own judge, and let me know what you think.

regards, A. Schiele

regards,
Moshe

 

[alfred_schiele]

Hi

I enjoy you opinon. It is very close to mine

Thank you

How do you manage this without detail loss?

I tried to do this on S1 examples with strange hair moiree and
didn't get any satisfying results at all...

See my answer a little below...

Well, there is a S1 example in phils gallery with a brick wall that containes red-green stripes of "hair-moiree" a lot more than some pixel in size... I tried of course bluring the colour channel but loss in detail was dramatic as colour resolution is already worse from the beginning... Well fiddeling around with selective and colour selections is possible but to me it is too much effort for a standard procedure...

Or one tri-color pixel would be even better, since then you would
not loose light for color...

Yes of course, a good point... but I argue current CCD can not provide this, we have to wait for different technology here...

Maybe I change my mind if you tell me your secret of perfect moiree
reduction while maintaining all the detail (-:

Please understand, i am not trying to change your mind, only to put
my thoughts forward and in the process make my own decision clearer.

Yes, I know this, so I added the smiley (-; ok?

Usually, the whole process is very quick.

? Maybe, but it sounds like a lot of effort to be done to every single image...

If you are interested, i can take several examples of moire in
available Fuji samples, and post the results that i get.

From your explanations I think I will prefer more aggressive LP...

My opinion remains that bayer pattern sensors that produce moiree have a wrong LP filter, maybe to suggest more resolution than theses systems are really able to produce...

Be your own judge, and let me know what you think.

I won't doubt your method works but to me it sounds like lot's of additional work when summing time up for this to be done on every single image.

Regards, A. Schiele