The # of megapixels of a Red One still extract

[Andrew Rieger]

Pretty self, explanatory. I had this info earlier but can't seem to find it know. I am looking for the megapixel count of a still extracted from R1 footage.

With 6K on the horizon, does anyone have any predictions about what to expect in terms of megapixels from the Scarlet FF35?

I am daydreaming of a camera that does 24 at 24 (24 megapixels at 24 frames per second). Hey, a guy an dream right?

UPDATE

Found This: http://camerarentalz.com/red-sensor-formats/

Does this mean that it is possible to get a 24mp still from the Scarlet FF35's video mode? If I recall correctly, the new DSMC captures the stills by extracting them from video instead of a dedicated stills mode. (Not sure about this, I would love clarification).

 

[Ketch Rossi]

I recall of the S35 sensor been around 12/15MP and the FF35 around 24MP. For either the Scarlet and or Epic, not sure at all about the R1 new M-X sensor.
I think that since was then put as MP count, will one imagine that it was referred so do to Stills, in either way.

I admit I also look forward to some more info on this regard for the Epic-X S35, but don't think I would go to FF35 as I will wait to move to the 645 instead from the Epic-X S35.

 

[Lee Saxon]

Megapixels is calculated by multiplying horizontal pixels by vertical pixels. So logical. I don't understand the use of the "5K" type specs. 5000 x what? 100? 10,000,000?

A great example of this is all the "1080p" cameras that are actually 1440x1080 rather than 1920x1080. They'd hide "1.55 megapixels" on the extended spec sheet somewhere and hope nobody knew what that meant [1920x1080 is 2.07 megapixels].

 

[G Lap]

Cinema v Stills traditional terminology I suppose.

This information is spread across the site in various threads (with the caveat that not all posts are RED official and everything is subject to change......of course).

2/3" - 3072 x 1620 - 3K - (approx 5 mpx) (3.3 micron pixel size)
R1 - 4520 x 2540 - 4K - (approx 12mpx)
S35 - 5120 x 2700 - 5K - (approx 14 mpx) (5.4 micron pixel size)
FF35 - 6000 x 4000 - 6k - (approx 24 mpx)
645 - 9334 x 7000 - 9k - (approx 65 mpx)
617 - 28000 x 9334 -28k - (approx 261 mpx)

 

[Lee Saxon]

If you think about it neither system is perfect for a hybrid still/motion camera.

Where it gets confusing is still photographers aren't going to be using the 2:1 aspect ratio of the S35, they'll cut the width way down and get much less than 5K horizontal pixels. Cinema shooters aren't going to (usually) be using the 3:2 aspect ratio of the FF35, they're going to cut the height way down and get much less than 24 megapixels.

Maybe we should just stick with max available dimensions - "6000x4000"

 

[David Mullen ASC]

The reason we use only the horizontal dimensions like "4K" in a label is that we have more than one aspect ratio in cinema (but not an infinite number either), so the vertical dimension is a variable depending on whether the 4K image is 16x9, 1.85, 2.39, etc.

A megapixel label would be more vague in that regards.

But it's easy enough to take out a calculator and multiply the horizontal by the vertical and get the total megapixel amount.

The max active pixel area that the Red One uses is 4520 x 2540.

Think about it -- if you had two 4K images and one was 4x3 and the other was 16x9, then the 4x3 version would have more total megapixels because it was taller. But would that mean it had more resolution? No.

 

[Andrew M.]

And TV resolution we like to express in vertical count of lines or pixels since it comes from the time when we had analog TV drawing lines on the picture tube where horizontal resolution was limited by the bandwidth of analog video channel, so no distinctive horizontal entities but one could test how many black and white points (vertical lines) you can show on each line of TV signal.
Now days, as David mentioned the horizontal pixel count determine the digital cinema format.

DLSR however are stuck on total MP count per entire frame. I guess for picture taking the total number of pixels is more convenient to use since there is a lot of portrait versus landscape versus full frame atc.
And when we are talking pixel count expressed in MP we mean all the pixels in the Bayer pattern sensor.

When we talk 4K, 2K, 1920, etc. As a cinema format we mean RGB triplets or the pixels that can show all the colors.

Andrew

 

[David Rasberry]

If you think about it neither system is perfect for a hybrid still/motion camera.

Where it gets confusing is still photographers aren't going to be using the 2:1 aspect ratio of the S35, they'll cut the width way down and get much less than 5K horizontal pixels. Cinema shooters aren't going to (usually) be using the 3:2 aspect ratio of the FF35, they're going to cut the height way down and get much less than 24 megapixels.

Maybe we should just stick with max available dimensions - "6000x4000"

R1 has been used to shoot major fashion spreads and cover shots for glossy magazines, so it would seem to be more than adequate as far as resolution is concerned.

Stills or motion one is always cropping to fit some format or another anyway. Unless one is doing very large still fine art prints, 6-8 mega pixels in a cropped image is plenty of resolution for most publication work. For a full bleed glossy 11x14 magazine image, the maximum screen resolution used for color plate separations for offset printing is 150 dpi which works out to about 3.5 mega pixels. Originals with twice this resolution are good enough to avoid issues with moire and aliasing. Newspapers normally print using 60 dpi screens.

 

[Lee Saxon]

Think about it -- if you had two 4K images and one was 4x3 and the other was 16x9, then the 4x3 version would have more total megapixels because it was taller. But would that mean it had more resolution? No.

How is a 4000x3000 image not more information than a 4000x2550 image?

R1 has been used to shoot major fashion spreads and cover shots for glossy magazines, so it would seem to be more than adequate as far as resolution is concerned.

I just meant the terminology, ie "6K" vs "24 megapixels" both have their flaws.

 

[Andrew M.]

I just meant the terminology, ie "6K" vs "24 megapixels" both have their flaws.

The first one is used to tell you how many all-color pixels you have on the horizontal line without telling you how many MP the photo-sensor has that delivers this 6K.
Some cameras will have 12K pixels per 6K horizontal line, or just 8K etc.
More oversampling than better.

The second one (MP) is used to tell you how many photo-sites the sensor has.

So without knowing bit more about the sensor it is difficult to compare

 

[David Mullen ASC]

How is a 4000x3000 image not more information than a 4000x2550 image?

It's more information but it's not necessarily more resolution. Think about it -- if you had a 4K sensor and took a 4x3 image of a subject and another picture that was less tall and thus 16x9 and measured the amount of fine detail in an object in the center of the frame, it would have the same amount of detail, it's just that the taller frame had more picture information above and below the object, hence more total information, but not more resolution devoted to the subject in the frame.

In other words, you can take a single 4x3 picture and crop it vertically to 16x9 and reduce the data, but the amount of detail captured in the subject will be the same in both the cropped and uncropped image.

Having more megapixels doesn't always help you if the aspect ratio of the sensor requires you crop the image to obtain the final aspect ratio and thus lose some of those pixels. So comparing a 10MP camera to a 12MP camera, you'd have to assume the same aspect ratio if you are going to assume that the 12MP camera has more resolution. Otherwise, the 12MP image may have to be cropped to 10MP to match the aspect ratio and thus there is no real difference in resolution.

What matters is the number of pixels you can use (among other things.) If I need a widescreen image then I'm not going to be able to use all of the pixels on a square sensor (unless I use an anamorphic lens.)