Unable to reproduce DPR Z7 banding

Started 6 months ago | Discussions
JimKasson
JimKasson Forum Pro • Posts: 24,208
Unable to reproduce DPR Z7 banding
20

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

-- hide signature --

Posted as a regular forum member.
https://blog.kasson.com

 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
Nikon Z7
If you believe there are incorrect tags, please send us this post using our feedback form.
ncsakany
ncsakany Senior Member • Posts: 2,214
Re: Unable to reproduce DPR Z7 banding
4

I have, but in a different setting. Even so, they are not visible unless I push exposure all the way and also lift shadows. If you want, I'll gladly place those shots on a link so you can download and examine.

Edit: for the sake of clarification, I am not considering this a real-life issue, it's for technical illustration only.

 ncsakany's gear list:ncsakany's gear list
Nikon Z7 Carl Zeiss Apo Sonnar T* 2/135 Nikon Z 24-70mm F4 Nikon Z 14-30mm F4
RobertBetchel
RobertBetchel Regular Member • Posts: 150
Re: Unable to reproduce DPR Z7 banding
4

Very interesting results and it begs the question of what was distinct in the DPR methodology that led to the visible banding?  Considering the vast number of people who are taking real-world photos and also not seeing the banding, it makes the splashy DPR headlines rather annoying.  Thank you for the diligent testing.

 RobertBetchel's gear list:RobertBetchel's gear list
Nikon D750 Fujifilm X-Pro2 Sony a7R III Nikon Z7 Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +13 more
Horshack Veteran Member • Posts: 7,338
Maybe also try to reproduce Bill's FPN?
1

Bill's single-frame and stacked FPN images here seem to visually match up with the banding in the dpreview's DR studio sample here. Maybe a good next step is to attempt reproducing Bill's findings on your copy to compare the quality and quantity of FPN to his sample, which I assume came from the same copy Rishi used in the DR studio sample (no idea though).

JimKasson
OP JimKasson Forum Pro • Posts: 24,208
Re: Maybe also try to reproduce Bill's FPN?
1

Horshack wrote:

Bill's single-frame and stacked FPN images here seem to visually match up with the banding in the dpreview's DR studio sample here. Maybe a good next step is to attempt reproducing Bill's findings on your copy to compare the quality and quantity of FPN to his sample, which I assume came from the same copy Rishi used in the DR studio sample (no idea though).

I know the noise level varies with row, but the pictures on the DPR site look like the mean is varying with row. I have some old tools I can dust off from the PDAF striping studies, but that's a lot of work to go to if it's not a real-world problem.

Jim

-- hide signature --

Posted as a regular forum member.
https://blog.kasson.com

 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
James Stirling
James Stirling Senior Member • Posts: 4,240
Re: Unable to reproduce DPR Z7 banding

Jim, could it be something specific to DPreviews lighting of their test scene thate exacerbates banding ? According to their update to the new test scene article link below. They use Kino Flo RF55 lamps for the daylight part of their test and a 25W tungsten incandescent light bulb for the low light part

https://www.dpreview.com/articles/4553555371/introducing-the-test-scene

-- hide signature --

Jim Stirling
“Facts are stubborn things; and whatever may be our wishes, our inclinations, or the dictates of our passion, they cannot alter the state of facts and evidence.” John Adams

 James Stirling's gear list:James Stirling's gear list
Panasonic Lumix DMC-GH4 Sony a7R II Panasonic Lumix DMC-GX8 Sony a7R III Panasonic Lumix G Vario 100-300mm F4-5.6 OIS +11 more
Horshack Veteran Member • Posts: 7,338
Another EXIF difference (compression?)
4

I did a deeper dive of the EXIF differences between the raw you supplied me and the dpreview/IR samples that show banding. Another difference I found is that your raw has an NEF linearization table whereas the dpreview/IR samples do not. I think this may relate to a different NEF compression setting but I can't tell for sure because exiftool isn't decoding the compression type for the Z7 raws (yet) like it does for previous Nikon bodies, likely because of the new compression type.

Based on the image sizes the dpreview/IR samples may be lossy whereas yours is lossless. If you shot yours with lossless can you try it lossy instead?

RobertBetchel
RobertBetchel Regular Member • Posts: 150
Re: Unable to reproduce DPR Z7 banding
5

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop. This frame was almost completely black before the exposure lift.

2048 pixel export of -5 EC 1/13th of second ISO 64 shot with 35 Z at F8

100% crop showing subtle banding with 5 EC push.

100% crop with 5 EC push and +40 shadows

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

 RobertBetchel's gear list:RobertBetchel's gear list
Nikon D750 Fujifilm X-Pro2 Sony a7R III Nikon Z7 Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +13 more
Horshack Veteran Member • Posts: 7,338
Re: Unable to reproduce DPR Z7 banding

RobertBetchel wrote:

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop.

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

Thanks. Is it possible for you to share the raw? I'd like to compare it to Jim's raw vs the dpreview/Image Resource samples I have.

Horshack Veteran Member • Posts: 7,338
Banding sample from Imaging Resource
2

100% crop, +5EV exposure adjustment in ACR (NEF available here):

RobertBetchel
RobertBetchel Regular Member • Posts: 150
Re: Unable to reproduce DPR Z7 banding

Horshack wrote:

RobertBetchel wrote:

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop.

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

Thanks. Is it possible for you to share the raw? I'd like to compare it to Jim's raw vs the dpreview/Image Resource samples I have.

https://drive.google.com/drive/folders/1ztLYo4XHwSLbl_AvRhvc8fj29Hk7MC5d?usp=sharing

Shot with compressed raw, opened and edited in Lightroom CC

 RobertBetchel's gear list:RobertBetchel's gear list
Nikon D750 Fujifilm X-Pro2 Sony a7R III Nikon Z7 Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +13 more
Dan Wells Contributing Member • Posts: 571
Re: Unable to reproduce DPR Z7 banding
3

It seems like you have to lift exposure by 5-6 stops before you see it. Even in wide dynamic range landscapes, I can't remember EVER lifting a shadow more than about 3 stops... How many sensors would even let you THINK about a 5 stop push (and, going back to film days, who EVER shot Tri-X at ISO 12,800 and thought push processing would save the day)?

silentstorm Senior Member • Posts: 1,133
A different banding question

Hi Jim, this banding looks different from the A7.3 PDAF type, that is sharp lines. This Z7 banding looks more like the traditional ones. Can you shed some light on this?

Is the Z7 banding due to PDAF or the uniformity of the ADC amp?

Thank you very much

RaLuC
RaLuC Regular Member • Posts: 370
Re: Unable to reproduce DPR Z7 banding

ncsakany wrote:

I have, but in a different setting. Even so, they are not visible unless I push exposure all the way and also lift shadows. If you want, I'll gladly place those shots on a link so you can download and examine.

Edit: for the sake of clarification, I am not considering this a real-life issue, it's for technical illustration only.

Oh yes, it certainly is. I came across some of my shots, landscapes with 24mm focal length. Shot against the indirectly lit sky from the already somewhat vanished sun. I decided to expose directly onto the beautiful yellow red orange coloured clouds in the sky, resulting in strong underexposure of the trees in front. I did that to ensure 100% the atmosphere of that nice sun set.

I didnt search for banding as I was already convinced at that time, that banding is a non issue. If you saw my other comments on that topic in former threads, you might know that...

When back at home I did some PP with these photos, when I suddenly found out, that there was clearly visible banding in some of these files. Only visible after like 4-5 stops push in the shadows. I investgated further and found out very strange things with no logical pattern.

For example there were two pictures, taken with the nearly exact same scene, only couple of meters different standpoint and only like 30s from eachother.

Both with the exact same settings, only the shutter speed was different, ISO 64 but 1/640s vs 1/5000s. And the one with a tree in the ftont and the other without. Other than that the exact same scene.

I guess this is also the reason, some do and some dont find banding, even when doing tests like Jim Kasson. Simply because it is not 100% the same scenery as DPR had it or there might be some very specific conditions, that either is there or not, that does not only have to do with underexposure etc.

Until picture 800 I was also 100% sure that banding is a non issue. But I was wrong, banding IS an issue. But yet we could not identify the definite reason and circumstances when it occurs.

If you now ask me to post NEF, no I wont, I never load up personalized files to the annonymous internet. The only thing I can do is load up some crops of my findings.

 RaLuC's gear list:RaLuC's gear list
Nikon D500 Nikon D850 Nikon Z7 Nikon AF-S Teleconverter TC-17E II Nikon AF-S Nikkor 50mm f/1.8G +4 more
JimKasson
OP JimKasson Forum Pro • Posts: 24,208
Re: A different banding question
1

silentstorm wrote:

Hi Jim, this banding looks different from the A7.3 PDAF type, that is sharp lines.

The a7III has PDAF striping. That is an entirely different phenomenon.

This Z7 banding looks more like the traditional ones. Can you shed some light on this?

At this point, no. I can't even make it happen.

Is the Z7 banding due to PDAF or the uniformity of the ADC amp?

The DPR shots showed banding with the PDAF row spacing, so that is the prime suspect.

Thank you very much

Sorry I can't help much.

Jim

-- hide signature --

Posted as a regular forum member.
https://blog.kasson.com

 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
Iliah Borg Forum Pro • Posts: 25,442
Re: Unable to reproduce DPR Z7 banding
11

RobertBetchel wrote:

Horshack wrote:

RobertBetchel wrote:

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop.

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

Thanks. Is it possible for you to share the raw? I'd like to compare it to Jim's raw vs the dpreview/Image Resource samples I have.

https://drive.google.com/drive/folders/1ztLYo4XHwSLbl_AvRhvc8fj29Hk7MC5d?usp=sharing

Shot with compressed raw, opened and edited in Lightroom CC

It seems Lightroom applied noise reduction to both chroma and luma. Some types of noise reduction may result in posterization and false banding artifacts.

I converted the shot in floating point, all noise reduction switched off, 5 stops push in highlights, 6 stops in shadows, with sharpening and saturation bump, conversion parameters shown on this screenshot:

The banding is pretty much undetectable on the result.

Converter (Mac-only): https://www.raw-photo-processor.com/RPP/RPP64_1864Beta.zip

-- hide signature --
Horshack Veteran Member • Posts: 7,338
Re: Unable to reproduce DPR Z7 banding

Iliah Borg wrote:

RobertBetchel wrote:

Horshack wrote:

RobertBetchel wrote:

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop.

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

Thanks. Is it possible for you to share the raw? I'd like to compare it to Jim's raw vs the dpreview/Image Resource samples I have.

https://drive.google.com/drive/folders/1ztLYo4XHwSLbl_AvRhvc8fj29Hk7MC5d?usp=sharing

Shot with compressed raw, opened and edited in Lightroom CC

It seems Lightroom applied noise reduction to both chroma and luma. Some types of noise reduction may result in posterization and false banding artifacts.

By default ACR is adhering to the NR settings specified in the NEF by Nikon, fields such as "Luminance Smoothing" and "Color Noise Reduction". These don't have a material effect on the ACR conversion wrt to banding in the raws I've examined.

I converted the shot in floating point, all noise reduction switched off, 5 stops push in highlights, 6 stops in shadows, with sharpening and saturation bump, conversion parameters shown on this screenshot:

The banding is pretty much undetectable on the result.

Can you show a 50% or 100% crop?

The banding in his raw is very subtle and light. Here's a conversion in Capture NX-D, Neutral profile, all corrections turned off, NR off, no sharpening, all Nikon picture controls zero'd out.

Iliah Borg Forum Pro • Posts: 25,442
Re: Unable to reproduce DPR Z7 banding
1

Horshack wrote:

Iliah Borg wrote:

RobertBetchel wrote:

Horshack wrote:

RobertBetchel wrote:

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop.

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

Thanks. Is it possible for you to share the raw? I'd like to compare it to Jim's raw vs the dpreview/Image Resource samples I have.

https://drive.google.com/drive/folders/1ztLYo4XHwSLbl_AvRhvc8fj29Hk7MC5d?usp=sharing

Shot with compressed raw, opened and edited in Lightroom CC

It seems Lightroom applied noise reduction to both chroma and luma. Some types of noise reduction may result in posterization and false banding artifacts.

By default ACR is adhering to the NR settings specified in the NEF by Nikon

Supposedly so.

, fields such as "Luminance Smoothing" and "Color Noise Reduction". These don't have a materially effect on the ACR conversion wrt to banding in the raws I've examined.

I converted the shot in floating point, all noise reduction switched off, 5 stops push in highlights, 6 stops in shadows, with sharpening and saturation bump, conversion parameters shown on this screenshot:

The banding is pretty much undetectable on the result.

Can you show a 50% or 100% crop?

Crop: http://s3.amazonaws.com/IliahBorg/RCB_0756_crop.tif (70MB)
Full: http://s3.amazonaws.com/IliahBorg/RCB_0756.tiff (0.5GB)

The banding in his raw is very subtle and light.

That's what I call pretty much undetectable. Printing makes it go away.

Here's a conversion in Capture NX-D, Neutral profile, all corrections turned off, NR off, no sharpening, all Nikon picture controls zero'd out.

-- hide signature --
Horshack Veteran Member • Posts: 7,338
Re: Unable to reproduce DPR Z7 banding
1

Iliah Borg wrote:

Horshack wrote:

Iliah Borg wrote:

RobertBetchel wrote:

Horshack wrote:

RobertBetchel wrote:

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop.

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

Thanks. Is it possible for you to share the raw? I'd like to compare it to Jim's raw vs the dpreview/Image Resource samples I have.

https://drive.google.com/drive/folders/1ztLYo4XHwSLbl_AvRhvc8fj29Hk7MC5d?usp=sharing

Shot with compressed raw, opened and edited in Lightroom CC

It seems Lightroom applied noise reduction to both chroma and luma. Some types of noise reduction may result in posterization and false banding artifacts.

By default ACR is adhering to the NR settings specified in the NEF by Nikon

Supposedly so.

, fields such as "Luminance Smoothing" and "Color Noise Reduction". These don't have a materially effect on the ACR conversion wrt to banding in the raws I've examined.

I converted the shot in floating point, all noise reduction switched off, 5 stops push in highlights, 6 stops in shadows, with sharpening and saturation bump, conversion parameters shown on this screenshot:

The banding is pretty much undetectable on the result.

Can you show a 50% or 100% crop?

Crop: http://s3.amazonaws.com/IliahBorg/RCB_0756_crop.tif (70MB)
Full: http://s3.amazonaws.com/IliahBorg/RCB_0756.tiff (0.5GB)

The banding in his raw is very subtle and light.

That's what I call pretty much undetectable. Printing makes it go away.

Perceptual banding usually survives the resampling done to fit the DPI/dimensions of the page for printing. Sometimes it even exacerbates it. Here are 100% crops at native resolution and then resampled to half the size:

Iliah Borg Forum Pro • Posts: 25,442
Re: Unable to reproduce DPR Z7 banding

Horshack wrote:

Iliah Borg wrote:

Horshack wrote:

Iliah Borg wrote:

RobertBetchel wrote:

Horshack wrote:

RobertBetchel wrote:

JimKasson wrote:

I tried to reproduce the banding observed by DPR in the Z7 tests, and I failed.

I lit a Macbeth chart with a LED ringlight. I set the camera up for a series of exposures that were 6.5 stops under an ETTR exposure, at ISOs 64, 100, and 400, with both 14 and 12-bit precision. I gave all the images a five-stop push in Lightroom.

No banding on any of them.

Details, including 100% and 200% crops of all six images, here:

https://blog.kasson.com/nikon-z6-7/nikon-z7-banding/

Has anyone else reproduced the DPR banding shots?

Jim

Well I was able to reproduce the banding shooting in my dim living room- using the 35 z lens at f8, ISO 64 and underexposed 5 stops. It is not visible at a 2048 pixel output but can be discerned at 100% crop.

OK, so without a doubt banding is real in certain circumstances. Am I worried about it? No, my shooting style does not depend on such extreme exposure adjustments. I bracket if I am trying to capture a wide dynamic range.

Thanks. Is it possible for you to share the raw? I'd like to compare it to Jim's raw vs the dpreview/Image Resource samples I have.

https://drive.google.com/drive/folders/1ztLYo4XHwSLbl_AvRhvc8fj29Hk7MC5d?usp=sharing

Shot with compressed raw, opened and edited in Lightroom CC

It seems Lightroom applied noise reduction to both chroma and luma. Some types of noise reduction may result in posterization and false banding artifacts.

By default ACR is adhering to the NR settings specified in the NEF by Nikon

Supposedly so.

, fields such as "Luminance Smoothing" and "Color Noise Reduction". These don't have a materially effect on the ACR conversion wrt to banding in the raws I've examined.

I converted the shot in floating point, all noise reduction switched off, 5 stops push in highlights, 6 stops in shadows, with sharpening and saturation bump, conversion parameters shown on this screenshot:

The banding is pretty much undetectable on the result.

Can you show a 50% or 100% crop?

Crop: http://s3.amazonaws.com/IliahBorg/RCB_0756_crop.tif (70MB)
Full: http://s3.amazonaws.com/IliahBorg/RCB_0756.tiff (0.5GB)

The banding in his raw is very subtle and light.

That's what I call pretty much undetectable. Printing makes it go away.

Perceptual banding usually survives the resampling

I mean dot gain and other ink effects.

done to fit the DPI/dimensions of the page for printing. Sometimes it even exacerbates it. Here are 100% crops at native resolution and then resampled to half the size:

-- hide signature --
Keyboard shortcuts:
FForum MMy threads