PDA

View Full Version : The biggest Four Thirds sensor yet


Four Thirds User
6th April 2009, 04:10 PM
Four Thirds User (Fourthirds-user.com (http://fourthirds-user.com)) is a sibling site to the e-group.

Its not widely hailed, but Four Thirds has a new, bigger, image sensor from Panasonic and it made its debut in the new Lumix DMC-GH1. We explore its benefits.

More... (http://fourthirds-user.com/2009/04/the_biggest_four_thirds_sensor_yet.php)

Bill Gordon
6th April 2009, 05:54 PM
If the sensor size is bigger than 4/3rds then how can it be called a 4/3rds sensor??

shenstone
6th April 2009, 07:14 PM
I get the theory from the FTU page, but those that like 4/3 ratio lose out quite a bit in the top and bottom of the sensor frame if I have understood your graphics correctly

I'd like to know whether this can be made to work within the standard as it would give us the best of both worlds

http://e-group.uk.net/gallery/data/500/Drawing1.jpg

In the software you do some form of mapping based on the format you are using and capture only what complies with the standard

Actual sensor size is the blue rectangle which is intended to break the circle - the rest are my too hasty drawing - green being 16:9 and yellow 4:3

Regards
Andy

Ian
6th April 2009, 08:42 PM
If the sensor size is bigger than 4/3rds then how can it be called a 4/3rds sensor??

The Four Thirds rule is that the diagonal of the rectangular sensor area that represents the image frame must be 22.5mm, the diameter of the image circle. Basically, the frame corners must be on the circumference of the image circle.

Until now, all Four Thirds sensors have been 4:3 aspect ratio, 18mm wide and 13.5mm high and 22.5mm diagonally.

As the article points out, the Four Thirds name is nothing to do with the default 4:3 aspect ratio frame we all know and love. It's to do with the sensor carrier size being four thirds inches in size.

Panasonic has created a sensor for the GH1 that is slightly wider than what we are used to, enabling wider aspect ratio frame formats (3:2 and 16:9) to be accommodated with that 22.5mm diagonal.

The perfect utopian Four Thirds sensor would be 22.5mm x 22.5mm (square), so you could shoot any frame aspect ratio you'd like, in vertical or landscape camera orientation. Note, the square frame format would be 15.9mm x 15.9mm, not 22.5 x22.5, as the corners of the frame must be on the circumference of the image circle. Such a square frame format would number 12.7 megapixels at the pixel pitch of this particular sensor.

It does take some getting used to :)

Ian

Ian
6th April 2009, 08:45 PM
I get the theory from the FTU page, but those that like 4/3 ratio lose out quite a bit in the top and bottom of the sensor frame if I have understood your graphics correctly

I'd like to know whether this can be made to work within the standard as it would give us the best of both worlds

http://e-group.uk.net/gallery/data/500/Drawing1.jpg

In the software you do some form of mapping based on the format you are using and capture only what complies with the standard

Actual sensor size is the blue rectangle which is intended to break the circle - the rest are my too hasty drawing - green being 16:9 and yellow 4:3

Regards
Andy

Not sure what you mean Andy :)

The GH1 sensor retains the maximum 4:3 aspect ratio coverage possible.

Ian

Bill Gordon
6th April 2009, 08:59 PM
Thanks Ian fro straightening me out!! I would love a square format, just like my old Hasselblad.

shenstone
6th April 2009, 09:04 PM
Not sure what you mean Andy :)

The GH1 sensor retains the maximum 4:3 aspect ratio coverage possible.

Ian

I didn't think it was clear in your FTU article, but your description above makes sense and I think agrees with what I was asking in that the sensor size can be outside the circle, but your framing must be inside at any specific use

Right?

Regards
Andy

Invicta
6th April 2009, 09:20 PM
Interesting development Ian.

Have I understood this correctly. With the new sensor a 4:3 aspect ratio photo would leave some unused pixels on each side whilst a 16:9 aspect ratio photo would leave some unused pixels at the top and bottom?

Also would all existing E-system lens work OK as they would cover the full image circle? Would the extra pixels at the edges in 16:9 ratio suffer from greater vignette effects?

Ian
6th April 2009, 10:32 PM
I didn't think it was clear in your FTU article, but your description above makes sense and I think agrees with what I was asking in that the sensor size can be outside the circle, but your framing must be inside at any specific use

Right?

Regards
Andy

Yes, that's right :)

Ian

Ian
6th April 2009, 10:38 PM
Interesting development Ian.

Have I understood this correctly. With the new sensor a 4:3 aspect ratio photo would leave some unused pixels on each side whilst a 16:9 aspect ratio photo would leave some unused pixels at the top and bottom?

Also would all existing E-system lens work OK as they would cover the full image circle? Would the extra pixels at the edges in 16:9 ratio suffer from greater vignette effects?

The corners of any aspect ratio frame are always on the periphery of the image circle, so any corner shading would the same. The wider the aspect ratio frame, the closer the sides get to the image circle, so in theory they could be more shaded, but not as much as the corners.

Ian