[Qwntm]

Wait, I thought you went all canon, save for K5II and a fisheye?

Well I didn't go "ALL" Canon.


And what Canon I did go, didn't work out so well. Making big prints (30x40, 40x60, cropped 16x48's) kinda shows you the limitations of your system pretty fast.

---------- Post added 10-02-15 at 09:59 AM ----------

Was there any wind? That shot with so much foliage would not normally be suitable for pixel shift if there's a breeze.

I agree. But this wasn't a "grab" shot. I have driven the 250 miles one way three times in the past 6 weeks. The first time was to find the shot on a non windy SUNNY day. I found the shot, but the sun killed it. The next time was on a forecast cloudy day with no wind, and the 6D got the shot as well as it could. Yesterday I went back and left at 6am MDT to be there by 9am PDT because the forecast was for AM clouds, no wind.


So far I have spent 3 days and over $200 in diesel just for this shot. And sometimes that's what it takes. The point I'm trying to make is that what I have realized is that the majority of my best selling landscape shots could have been shot with pixel shift with the proper planning required to get the shot anyway. There are very few spectacular landscape shots by anyone that have been shot in a hurricane! (Or even a slight breeze. )

---------- Post added 10-02-15 at 10:05 AM ----------

One does have to allow for the possibility that the light on the day favoured the Pentax, but WOW. What a difference. How washed out and lifeless that Canon shot looks by comparison.

If the full-frame comes anywhere near this, the landscape photographers are going to be all over it. Come on, Pentax - we really want that DA15-like full-frame prime now.

It's not even that fair: The Canon 6D had a $1000 Canon 24-70 F4.0 L lens on it and the K3 was using a $450 well worn Tamron 17-50 F2.8.


I actually ordered the Pentax 16-50 2.8 SDM last night after seeing the results from the pixel shift. Not because the Tamron did a bad job, but because now that I know this is going to be viable, I also know that the Pentax 16-50 will give my shots that magic Pentax pixie dust color too.


I thought about waiting for the 24-70, but I can always use the 16-50 and k5IIs and K3II for weddings while waiting for the FF to gets to the being sold stage.

[biz-engineer]

Pixel shift is not comparable to a 6D. The 6D does not have the limitation of having to use a tripod on totally still subject. However, when conditions allow pixel shift, pixel shift almost compares to diffraction limited medium format system. The reason is combining 4 shots shifted shots roughly equate to the amount on data of a single frame from an MF sensor at the same ISO, the random noise of the smaller APSC sensor being averaged out in the raw stacking process (deterministic errors remain). For best pixel shift results should are obtained by using a f2.8 lens and stop-it down to f4.0 or F5.6 (that's the reason why Ricoh use the DA*55 f1.4 stopped down a bit).

[shardulm]

Bottom Line: Pixel Shift does expand the capability of the camera's sensor and allows it do much more resolution-wise without having to expand the camera.
Even with its limited tripod laden use it scores on image quality with no extra moolah. Its a no brainer winner.

[Qwntm]

Pixel shift is not comparable to a 6D. The 6D does not have the limitation of having to use a tripod on totally still subject. However, when conditions allow pixel shift, pixel shift almost compares to diffraction limited medium format system. The reason is combining 4 shots shifted shots roughly equate to the amount on data of a single frame from an MF sensor at the same ISO, the random noise of the smaller APSC sensor being averaged out in the raw stacking process (deterministic errors remain). For best pixel shift results should are obtained by using a f2.8 lens and stop-it down to f4.0 or F5.6 (that's the reason why Ricoh use the DA*55 f1.4 stopped down a bit).

OK, if you say so. All I know is that no matter how hard you try, the 6D can't produce the same IQ quality for what I'm shooting and printing with a single press of the shutter button.


So it really isn't fair to compare the 6D as there's no way it could win... (But that maybe was what you were saying... )

---------- Post added 10-02-15 at 04:31 PM ----------

Bottom Line: Pixel Shift does expand the capability of the camera's sensor and allows it do much more resolution-wise without having to expand the camera.
Even with its limited tripod laden use it scores on image quality with no extra moolah. Its a no brainer winner.

For me, I'm 99% of the time on a tripod anyway. It may be a no brainer but it took some thought to realize the potential. And in actual use, it's like taking a time exposure. It's THAT easy and well implemented. Like I said, most of my best selling landscapes could have been shot with pixel shift and would have been better for it.


Moolah, size, weight, complicated interface...

[biz-engineer]

I can see the difference from the images posted here from the K-3II and 6D. The 6D image has a resolution of 800x268 pixels, and the K-3II image has a resolution of 1200x520. Indeed, pixel shift is a winner.

[Rondec]

That is impressive. As you say, regardless of format, most of the best landscape photos come on relatively still days. Anyway, whether pixel shift was used or not, the image is very nice. Beautifully captured and worth the effort, in my opinion.

[radeon7750]

Could you post the original images (full size)?

Another thing about pixel shift, your lens only has to be 24mp sharp, not 50mp sharp, yet it gives you 45mp output. think about it.

This is confusing me. I thought that the file you get using pixelshift is a 24 MP file. 6000x4000. Is that wrong? Are there different settings with the pixelshift or just one?

[zapp]

Another thing about pixel shift, your lens only has to be 24mp sharp, not 50mp sharp, yet it gives you 45mp output. think about it.

Nonsense

[biz-engineer]

From full size shots taken with the 35 F2.8 macro, tripod, still subject, I see only slightly sharper image with pixel shift, but I have to zoom at 100% and look very very carefully. When there is moiré artefacts without pixel shift, the moiré artefacts also end-up in the pixel shift resolution, so regarding resolving power for periodic details, this is not more than 24Mp. What pixel shift does is slightly better demosaicing than bayer's pattern demosaicing, and noise reduction by averaging 4 pixels into 1. To benefit from pixel shift for landscape, diffraction should not kick-in, hence lower f-stop such f4 or f5.6 max...then depth of field should be considered...

Basically, stitching 4 vertical images and downsizing to 6000x4000 deliver more details then pixel shift. But obviously, stitching require more PP work than pixel shift.

[Qwntm]

Nonsense

Nonsense

Can someone with the K-3 II please answer if the pixelshifted image has 6000x4000 pixels or if it has another amount of pixels...

[biz-engineer]

Can someone with the K-3 II please answer if the pixelshifted image has 6000x4000 pixels or if it has another amount of pixels...

The jpeg converted image is 6000x4000, with pixel shift or without pixel shift. But the RAW file (before conversion) contains the information of 4 raw shots of 6000x4000 pixels.
After raw to jpeg conversion, the image is a 24Mpixels image, with eventually slightly more quality (eventually better sharpness and less noise).

The jpeg converted image is 6000x4000, with pixel shift or without pixel shift. But the RAW file (before conversion) contains the information of 4 raw shots of 6000x4000 pixels.
After raw to jpeg conversion, the image is a 24Mpixels image, with eventually slightly more quality (eventually better sharpness and less noise).

Thank you! (I finally found a pixel shifted DNG to check out...)

[biz-engineer]

I'd be interested to see a 100% crop comparison between pixel shift and non pixel shift images taken with zoom lenses and apertures such at f16.