[Senko]

Been using my K-01 recently and started playing around in the in-camera aspect ratios (4:3, 1:1, 16:9) - I am just not clear what is actually happening in-camera.

When shooting 4:3 for example, the specs state it's max 14.2 MP (down from the original 16 MP). Is the sensor simply recording the shot at 3:2 and cropping to 4:3 (1:1, 16:9, etc,.) in the camera giving me a 4/3 image with x2.0 crop factor vs the native 3:2 x1.5 crop?

What I'm getting at is that hypothetically all else being equal, would the image be equivalent to a camera with 14MP m-4/3 sensor?

Thanks

[aslyfox]

didn't find much in

http://www.ricoh-imaging.co.jp/english/support/man-pdf/k-01.pdf

or in depth review

Pentax K-01 Review - Review Introduction | PentaxForums.com Reviews

or in the owners review

Pentax K-01 - Pentax Mirrorless Interchangeable Lens Cameras - Pentax Camera Reviews and Specifications

as a new owner of the K - 01, I'm curious too

[DeadJohn]

Try to forget the x1.5 crop factor when thinking about the aspect ratio settings. The aspect ratio settings merely discard pixels at the edges of the sensor to change the shape of the final image rectangle.

For example, a 1:1 aspect ratio uses the full sensor height, ignoring parts on the left and right sides.

My preference is to always use the camera's native aspect ratio. I can crop with software later.

[aslyfox]

thank you for that clear explanation

[Senko]

Try to forget the x1.5 crop factor when thinking about the aspect ratio settings. The aspect ratio settings merely discard pixels at the edges of the sensor to change the shape of the final image rectangle.

For example, a 1:1 aspect ratio uses the full sensor height, ignoring parts on the left and right sides.

My preference is to always use the camera's native aspect ratio. I can crop with software later.

Thanks for the reply - I couldn't find anything about this.

Are you positive this is how it works? Reason I ask is I thought the same thing and did all my cropping in post until I did some quick in-camera testing.

Playback on images in camera when comparing 4/3, 3/2, 16/9 are all drastically different - when set to the 4/3 aspect ratio looks exactly like 2.0 crop (more zoomed in) vs 1.5 crop...

It's very neat setting, I like that I can see my image immediately and make any framing/composition adjustments on the spot - much prefer it over copping in post.

[pschlute]

Are you positive this is how it works?

Look at the image properties, that will tell you the pixel dimensions.

[Senko]

Look at the image properties, that will tell you the pixel dimensions.

Good call....It's listed in the specs.

4928 x 3264 (16.1 MP, Other),
4224 x 2816 (11.9 MP, 3:2),
3456 x 2304 (8.0 MP, 3:2),
2688 x 1792 (4.8 MP, 3:2),
4352 x 3264 (14.2 MP, 4:3),
3840 x 2880 (11.1 MP, 4:3),
3072 x 2304 (7.1 MP, 4:3),
2304 x 1728 (4.0 MP, 4:3),
4928 x 2776 (13.7 MP, 16:9),
4224 x 2376 (10.0 MP, 16:9),
3456 x 1944 (6.7 MP, 16:9),
2688 x 1512 (4.1 MP, 16:9),
3264 x 3264 (10.7 MP, 1:1),
2880 x 2880 (8.3 MP, 1:1),
2304 x 2304 (5.3 MP, 1:1),
1728 x 1728 (3.0 MP, 1:1)

So by looks of it, 14MP 4/3 aspect ratio out of the K-01 at 4352 x 3264 is equivalent, scale wise, to the 4/3 16MP Olympus OM-D E-M10 sensor which is 4608 x 3456.

[Bob 256]

And that top format ("Other") is actually (really really close to) 3:2 also, but uses the full sensor coverage. The next down (4224x2816 - spot on 3:2) saves on memory space, as do the other 3:2 options. I prefer to shoot full sensor and then post-crop which allows any of the listed formats with the best resolution provided from the camera. The downside is that the crop frame has to be previsualized from the displayed image and more memory is used for each shot.

Note that the corresponding formats with lower pixel counts do use the full sensor (along the horizontal and/or vertical axis) but interpolate the lower pixel values in camera. If pixels were actually masked off (not used), the effective focal length of a given lens would change as a smaller and smaller area of the sensor was used. It's actually a combination of masking (to get an aspect ratio other than 3:2) where pixels of the original sensor are not used, and interpolation of the remaining pixels to get the desired resolution (if a sub-resolution is being used for a given aspect ratio).

[BrianR]

So by looks of it, 14MP 4/3 aspect ratio out of the K-01 at 4352 x 3264 is equivalent, scale wise, to the 4/3 16MP Olympus OM-D E-M10 sensor which is 4608 x 3456.

The k-01 in 14MP 4/3 mode represents a sensor size of 20.9 x 15.7 mm (diagonal 26.1mm). This is larger than the Olympus which is 17.3 x 13 mm (diagonal 21.6 mm).

So the 'crop factor' (relative to 35mm) would be a little less on the k-01, 1.65x compared to the Olympus 2x.

[Bob 256]

The k-01 in 14MP 4/3 mode represents a sensor size of 20.9 x 15.7 mm (diagonal 26.1mm). This is larger than the Olympus which is 17.3 x 13 mm (diagonal 21.6 mm).

So the 'crop factor' (relative to 35mm) would be a little less on the k-01, 1.65x compared to the Olympus 2x.

The actual K-01 sensor is spec'd at 23.7 x 15.7mm (for the full sensor coverage size). This gives 16.1Mp as previously noted but with the discarded pixels the effective resolution & size of the sensor is that stated.

[Senko]

Thanks for the detailed explanations, much appreciated.

You got to the crux of what I was getting at.

If to get the different aspect ratio the camera interpolates to lower pixel value internally, would you see a slight hit to IQ due to interpolation or is it a non-factor?

[BrianR]

If to get the different aspect ratio the camera interpolates to lower pixel value internally, would you see a slight hit to IQ due to interpolation or is it a non-factor?

There's no interpolating at the different aspect settings if you use the highest resolution for that aspect ratio.

For example, the 4352 x 3264 (14.2 MP, 4:3) mode is just a straight up slicing off of the sides to get down to 4:3, no interpolation necessary. If you used 3840 x 2880 (11.1 MP, 4:3), it would slice off the sides and then downsample.

If you're set on shooting at a different aspect ratio at the time of capture, I'd still use the highest resolution available for that aspect ratio and do my downsampling in the computer when I'm exporting the photo to something (print, web, email, whatever). If you aren't doing much post-processing (e.g. additional cropping), I'd doubt there's a huge difference in downsampling in-camera vs on the computer, but storage is cheap and you might decide you needed a higher resolution after the capture.

[Senko]

There's no interpolating at the different aspect settings if you use the highest resolution for that aspect ratio.

For example, the 4352 x 3264 (14.2 MP, 4:3) mode is just a straight up slicing off of the sides to get down to 4:3, no interpolation necessary. If you used 3840 x 2880 (11.1 MP, 4:3), it would slice off the sides and then downsample.

If you're set on shooting at a different aspect ratio at the time of capture, I'd still use the highest resolution available for that aspect ratio and do my downsampling in the computer when I'm exporting the photo to something (print, web, email, whatever). If you aren't doing much post-processing (e.g. additional cropping), I'd doubt there's a huge difference in downsampling in-camera vs on the computer, but storage is cheap and you might decide you needed a higher resolution after the capture.

Makes perfect sense - thank you.

To be honest, never thought to shoot different aspect ratios in camera as I don't believe any of my other bodies offered the feature. Decided to mess with it on the K-01 and found it pretty great. It's actually fun and also nice to be able to see different framing / results on the spot and adjust accordingly.

[08amczb]

It's just a crop, like done in post processing, nothing more... except one thing: it saves less pixels, so your RAW/out of camera jpeg file will be smaller.

I suggest staying with the native 3:2 and do it in post processing.