Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Understanding Red’s and Canon’s Image sensor sizes..

P

Paul Sutton

Member
Joined
Oct 14, 2009
Messages
19
Reaction score
0
Points
0
Location
California
Sorry, I’m still new to cameras and trying to understand Image sensors, here are my questions and any help would be greatly appreciated..

1. I’m trying to understand Image sensors. Okay so if the 7D shoots 1080 HD video approximately close to a 2k image, is it using all of it’s 18mp for the HD video??

2. and what roughly would a Red one 4k and Epic mx 5k be at as far as MP count?

3. I know that the MX sensor is slightly larger than the Red Ones so the MX sensor will have a larger image or (field of view) but does 4k affect the size of the image or (field of view)? Example, if you set a 7D and a Red One side by side, the 7D being at 1080 and the Red One being at 4k, both with the same focal length, would the image be roughly the same size?? I understand it as the Red Ones 4k footage would just be more pixel dense over the 7Ds 1080p and both would have roughly the same size image ??

4. This is a part of question 3. but from what I understand the images from both cameras should be close in size or (Field of View) with just varying pixel density.. I ask this because if you “Google” 1080p, 2k, 4k and 5k you can find sizing charts that show that 2k is larger than 1080p and that 4k is much larger than 2k and that 5k is larger than them all; they’re like little box diagrams, the biggest box being 5k and the box inside the 5k box is a 4k box and inside the 4k box there will be a 3k or 2k box, all the way down to 1080, 720 and SD.. This is what confuses me, because if the size of an image or (Field of view) is determined by actual sensor size and not pixel count why would these carts show that a 5k image is physically larger then them all??

The box diagrams like I’m talking about are similar to this one at Red.com, it’s the diagram all the way at the bottom on the left, this is the link: http://www.red.com/cameras/technology/ .. see how the 4k box is much larger than the 2k box?? If the Image sensor in my Canon 7D is roughly the same size as the Red Ones Mysterium sensor why would the 2k box be so much smaller??

I’m sorry I feel like the answer to my question is much easier to understand than I’m making it..

Any help with understanding my dilemma would be greatly appreciated. Thank you –Paul an amateur
 
Paul Sutton said:
Sorry, I’m still new to cameras and trying to understand Image sensors, here are my questions and any help would be greatly appreciated..

1. I’m trying to understand Image sensors. Okay so if the 7D shoots 1080 HD video approximately close to a 2k image, is it using all of it’s 18mp for the HD video??

2. and what roughly would a Red one 4k and Epic mx 5k be at as far as MP count?

3. I know that the MX sensor is slightly larger than the Red Ones so the MX sensor will have a larger image or (field of view) but does 4k affect the size of the image or (field of view)? Example, if you set a 7D and a Red One side by side, the 7D being at 1080 and the Red One being at 4k, both with the same focal length, would the image be roughly the same size?? I understand it as the Red Ones 4k footage would just be more pixel dense over the 7Ds 1080p and both would have roughly the same size image ??

4. This is a part of question 3. but from what I understand the images from both cameras should be close in size or (Field of View) with just varying pixel density.. I ask this because if you “Google” 1080p, 2k, 4k and 5k you can find sizing charts that show that 2k is larger than 1080p and that 4k is much larger than 2k and that 5k is larger than them all; they’re like little box diagrams, the biggest box being 5k and the box inside the 5k box is a 4k box and inside the 4k box there will be a 3k or 2k box, all the way down to 1080, 720 and SD.. This is what confuses me, because if the size of an image or (Field of view) is determined by actual sensor size and not pixel count why would these carts show that a 5k image is physically larger then them all??

The box diagrams like I’m talking about are similar to this one at Red.com, it’s the diagram all the way at the bottom on the left, this is the link: http://www.red.com/cameras/technology/ .. see how the 4k box is much larger than the 2k box?? If the Image sensor in my Canon 7D is roughly the same size as the Red Ones Mysterium sensor why would the 2k box be so much smaller??

I’m sorry I feel like the answer to my question is much easier to understand than I’m making it..

Any help with understanding my dilemma would be greatly appreciated. Thank you –Paul an amateur
Click to expand...

You have to divorce yourself from the notion that the amount of pixels has much to do with sensor size. You can have 1920 x 1080 pixels (about 2MP) on a 1/3" sensor or a 35mm sensor, it just depends on the size of the photosites, their spacing, etc. You can have two sensors the same size but one can have 1920 x 1080 pixels, the other can have 4096 x 2304 pixels. And you can have smaller sensors but in a 3-sensor prism block design rather than a single sensor.

So charts that show a 2K box inside of a 4K box are useless unless they are describing the same sensor being cropped to smaller areas and thus lower pixel counts. They are more there for symbolic purposes, to imagine that a 4K image has 4X the data of a 2K image, or can be shown on twice the width screen with the same apparent resolution.

You also have to remember that a single-sensor has a pattern of RGB filters overlaid on an essentially monochrome imaging device, usually in a Bayer mosaic, and from that, separate RGB channels have to be created.

No, the Canon 7D does not use all of its photosites on the sensor to create the HD video image (it does use the overall area of the sensor for the image though). In other words, it does not downscale & convert an 18MP RAW image down to a 2MP RGB image. That would take more processing power than they want to currently put into a Canon still camera. But it also doesn't crop an 18MP sensor area down to a smaller 2MP area either. Ideally, it would use all 18MP to generate the HD image; instead, it uses various schemes (pixel binning, line skipping, etc.) to reduce 18MP way, way down before it does the conversion to HD color. This accounts for the artifacts like color moire that the DSLR video image has.

A 16x9 and a 2:1 image would have a different total pixel count even if they had the same horizontal pixel count due to the aspect ratio. So if you record a 4K 16x9 image on the Red One, that recording is 4096 x 2304 pixels; if you record 4K 2:1, that would be 4096 x 2048 pixels. I don't know what the final recording sizes will be on a Epic, but obviously 5K will be more than 4K.

The 5K MX sensor is something like 30mm x 15mm I think, but currently if you record 4K 16x9 on the Red One with that sensor, you are still recording the same sized area as before on the previous sensor, 22.12mm x 12.44mm. The max size you can record on the Red One is now 4.5K, which would be about 24mm wide -- you can only use more of the 5K sensor when it is in the Epic.

The Red One, when recording 2K, does crop the sensor down to a smaller 2K area, so the used image area is closer to 16mm in size rather than 35mm in size.
 
Last edited:
David said it all. Don't really think there is much more to say.
 
The Abel chart is a bit misleading with regards to EPIC.

Sensor size (recordable area):
EPIC: 27.65mm x 14.58

Resolutions:
EPIC: 5120x2700, 5120x2560, 5120x2160, 4096x2304, 3840x2160, etc

Jim
 
Jannard said:
The Abel chart is a bit misleading with regards to EPIC.

Sensor size (recordable area):
EPIC: 27.65mm x 14.58

Resolutions:
EPIC: 5120x2700, 5120x2560, 5120x2160, 4096x2304, 3840x2160, etc

Jim
Click to expand...

Not meant to be misleading in any way. The chart clearly states that this is for comparing 16x9 image areas only. Several cameras (including film) offer imagers in different aspect ratios. We tried hard for clarity and simplicity in the chart, and 16x9 is by far the most commonplace aspect ratio in use today. No Phantom cameras have 16x9 imagers either, so these had to be cropped as well for purposes of the chart. In the end we felt this was fair because that is how the vast majority of real world clients use the cameras.

The image area used is based on a spec that I got off this forum and confirmed with RED.
 
You must log in or register to reply here.
Back
Top