Could anyone tell me if red is working on a 65mm camera (and while we're at it what I would want in it)?

[fgonz031]

See how the URSA Cine 17K is and if that might work for you, it's supposed to be coming out in the coming months. There's no doubt that RED should be working on the Medium Format field and there is most likely a camera coming to suit this new level of camera but there's been no word yet on when or in what form it will take. The way to do things now is to use a Kipon Medium Format speedbooster and there are models directly compatible with RED cameras and more universal options for RF and PL as well.

Hopefully!!

 

[fgonz031]

Sticking to the DCI 1.9:1/17:9 aspect ratio as Phil suggested makes it easier to see what might be possible with the current sensors.

4096x2160 - 1.9:1/17:9 (4K DCI)
6144x3240 - 1.9:1/17:9 (6K DCI)
8192x4320 - 1.9:1/17:9 (8K DCI)
10240×5400 - 1.9:1/17:9 (10K DCI)
12288x6480 - 1.9:1/17:9 (12K DCI)
14336x7560 - 1.9:1/17:9 (14K DCI)
16384x8640 - 1.9:1/17:9 (16K DCI)

So, using the given pixel sizes per camera -

V-Raptor 8K VV - 5 micron (0.005mm)
Komodo-X - 4.4 micron (0.0044mm)
V-Raptor 8K S35 - 3.2 micron (0.0032mm)

Here's a list of expanded sensor sizes and image circles for the Raptor VV & S35 and Komodo-X -

RED V-Raptor 'VV' - 10K 17:9
Format Size = 51.20 x 27.00mm (57.88mm image circle/diagonal)
Format Resolution = 10240 × 5400 (55.3 megapixels) (+56.3%)
Format Aspect Ratio = 1.896:1
Up to 1250 MB/s (+56.3%)

RED V-Raptor 'VV' - 12K 17:9
Format Size = 61.44 x 32.40mm (69.46mm image circle/diagonal)
Format Resolution = 12288 x 6480 (79.6 megapixels) (+44%)
Format Aspect Ratio = 1.896:1
Up to 1800 MB/s (+44%)

RED V-Raptor 'VV' - 14K 17:9
Format Size = 71.68 x 37.80mm (81.04mm image circle/diagonal)
Format Resolution = 14336 x 7560 (108.4 megapixels) (+36.2%)
Format Aspect Ratio = 1.896:1
Up to 2452 MB/s (+36.2%)

RED V-Raptor 'VV' - 16K 17:9
Format Size = 81.92 x 43.20mm (92.61mm image circle/diagonal)
Format Resolution = 16384 x 8640 (141.6 megapixels) (+30.6%)
Format Aspect Ratio = 1.896:1
Up to 3202 MB/s (+30.6%)

-----------------------------

RED V-Raptor 'S35' - 10K 17:9
Format Size = 32.77mm x 17.28mm (37.05mm image circle/diagonal)
Format Resolution = 10240 × 5400 (55.3 megapixels) (+56.3%)
Format Aspect Ratio = 1.896:1
Up to 1250 MB/s (+56.3%)

RED V-Raptor 'S35' - 12K 17:9
Format Size = 39.32mm x 20.74mm (44.45mm image circle/diagonal)
Format Resolution = 12288 x 6480 (79.6 megapixels) (+44%)
Format Aspect Ratio = 1.896:1
Up to 1800 MB/s (+44%)

RED V-Raptor 'S35' - 14K 17:9
Format Size = 45.88mm x 24.19mm (51.87mm image circle/diagonal)
Format Resolution = 14336 x 7560 (108.4 megapixels) (+36.2%)
Format Aspect Ratio = 1.896:1
Up to 2452 MB/s (+36.2%)

RED V-Raptor 'S35' - 16K 17:9
Format Size = 52.43mm x 27.65mm (59.27mm image circle/diagonal)
Format Resolution = 16384 x 8640 (141.6 megapixels) (+30.6%)
Format Aspect Ratio = 1.896:1
Up to 3202 MB/s (+30.6%)

------------------------------

RED Komodo-X - 8K 17:9
Format Size = 36.05mm x 19.01mm (40.76mm image circle/diagonal)
Format Resolution = 8192 x 4320 (35.4 megapixels) (+78%)
Format Aspect Ratio = 1.896:1
Up to 997 MB/s (+78%)

RED Komodo-X - 10K 17:9
Format Size = 45.06mm x 23.76mm (50.94mm image circle/diagonal)
Format Resolution = 10240 × 5400 (55.3 megapixels) (+56.3%)
Format Aspect Ratio = 1.896:1
Up to 1558 MB/s (+56.3%)

RED Komodo-X - 12K 17:9
Format Size = 54.07mm x 28.51mm (61.13mm image circle/diagonal)
Format Resolution = 12288 x 6480 (79.6 megapixels) (+44%)
Format Aspect Ratio = 1.896:1
Up to 2244 MB/s (+44%)

RED Komodo-X - 14K 17:9
Format Size = 63.08mm x 33.26mm (71.31mm image circle/diagonal)
Format Resolution = 14336 x 7560 (108.4 megapixels) (+36.2%)
Format Aspect Ratio = 1.896:1
Up to 3056 MB/s (+36.2%)

RED Komodo-X - 16K 17:9
Format Size = 72.09mm x 38.02mm (81.50mm image circle/diagonal)
Format Resolution = 16384 x 8640 (141.6 megapixels) (+30.6%)
Format Aspect Ratio = 1.896:1
Up to 3991 MB/s (+30.6%)

The data rates are just extrapolated from the percentage increases in megapixels, I don't know if that's an accurate way of estimating them.

There could be some other mistakes with the numbers or more fundamental problems with the working out, but I just wanted to try and see how the sensor sizes and resolutions relate to each other.

It's all theoretical for me personally, but I wouldn't want to see dynamic range reduced or refresh rates increased to achieve higher resolutions or larger formats, while limiting frame-rates and compression ratio's would be less of an issue. I also think there's more of a limit nowadays to how big a popular camera can be, but that seems to fit with RED's own aims anyway, though I imagine some of those sensor sizes and the amount of internal processing and cooling required would necessitate a definite variation on the current DSCMC3 body design. I also wouldn't have a problem if the data rates needed a proprietary media solution, but I get the impression no-one, including RED, would want to go back to that, so that alone could limit potential advances.

Anyway...

You must send this also to the guys of RED!

 

[M Harvey]

An interesting reference point might be the “Big Sky” camera built to capture imagery for the Sphere venue in Las Vegas. Per the ASC magazine piece on it, it’s a 77.5mm x 75.6mm sensor that captures in 18k (and up to 120, though I don’t know if that’s full resolution or otherwise limited.)

Check out the piece:

Sphere and the Big Sky Camera

An inside look at the creation of Las Vegas' entertainment venue Sphere, and the "Big Sky" camera system custom-built to capture its immersive imagery.

theasc.com

 

[Brian Timmons]

An interesting reference point might be the “Big Sky” camera built to capture imagery for the Sphere venue in Las Vegas. Per the ASC magazine piece on it, it’s a 77.5mm x 75.6mm sensor that captures in 18k (and up to 120, though I don’t know if that’s full resolution or otherwise limited.)

Check out the piece:

Sphere and the Big Sky Camera

An inside look at the creation of Las Vegas' entertainment venue Sphere, and the "Big Sky" camera system custom-built to capture its immersive imagery.

theasc.com

As mentioned in the article, there is a RED connection with the Big Sky camera through Deanan DaSilva who was RED’s CTO in the very early days at RED and worked on the Dalsa Origin (the 1st 4K cinema camera) before RED.
He was also a frequent contributor on this very forum.
He has definitely made his mark pioneering quite a bit in digital cinema.

Brian Timmons
BRITIM/MEDIA

 

[Les Hillis]

You must send this also to the guys of RED!

I imagine they'd already have a list like this somewhere, along with other possible sensors based on different pixel sizes.

Is there something there in particular that you'd like to see in an actual camera?

 

[Edith Blazek]

Sticking to the DCI 1.9:1/17:9 aspect ratio as Phil suggested makes it easier to see what might be possible with the current sensors.

4096x2160 - 1.9:1/17:9 (4K DCI)
6144x3240 - 1.9:1/17:9 (6K DCI)
8192x4320 - 1.9:1/17:9 (8K DCI)
10240×5400 - 1.9:1/17:9 (10K DCI)
12288x6480 - 1.9:1/17:9 (12K DCI)
14336x7560 - 1.9:1/17:9 (14K DCI)
16384x8640 - 1.9:1/17:9 (16K DCI)

So, using the given pixel sizes per camera -

V-Raptor 8K VV - 5 micron (0.005mm)
Komodo-X - 4.4 micron (0.0044mm)
V-Raptor 8K S35 - 3.2 micron (0.0032mm)

Here's a list of expanded sensor sizes and image circles for the Raptor VV & S35 and Komodo-X -

RED V-Raptor 'VV' - 10K 17:9
Format Size = 51.20 x 27.00mm (57.88mm image circle/diagonal)
Format Resolution = 10240 × 5400 (55.3 megapixels) (+56.3%)
Format Aspect Ratio = 1.896:1
Up to 1250 MB/s (+56.3%)

RED V-Raptor 'VV' - 12K 17:9
Format Size = 61.44 x 32.40mm (69.46mm image circle/diagonal)
Format Resolution = 12288 x 6480 (79.6 megapixels) (+44%)
Format Aspect Ratio = 1.896:1
Up to 1800 MB/s (+44%)

RED V-Raptor 'VV' - 14K 17:9
Format Size = 71.68 x 37.80mm (81.04mm image circle/diagonal)
Format Resolution = 14336 x 7560 (108.4 megapixels) (+36.2%)
Format Aspect Ratio = 1.896:1
Up to 2452 MB/s (+36.2%)

RED V-Raptor 'VV' - 16K 17:9
Format Size = 81.92 x 43.20mm (92.61mm image circle/diagonal)
Format Resolution = 16384 x 8640 (141.6 megapixels) (+30.6%)
Format Aspect Ratio = 1.896:1
Up to 3202 MB/s (+30.6%)

-----------------------------

RED V-Raptor 'S35' - 10K 17:9
Format Size = 32.77mm x 17.28mm (37.05mm image circle/diagonal)
Format Resolution = 10240 × 5400 (55.3 megapixels) (+56.3%)
Format Aspect Ratio = 1.896:1
Up to 1250 MB/s (+56.3%)

RED V-Raptor 'S35' - 12K 17:9
Format Size = 39.32mm x 20.74mm (44.45mm image circle/diagonal)
Format Resolution = 12288 x 6480 (79.6 megapixels) (+44%)
Format Aspect Ratio = 1.896:1
Up to 1800 MB/s (+44%)

RED V-Raptor 'S35' - 14K 17:9
Format Size = 45.88mm x 24.19mm (51.87mm image circle/diagonal)
Format Resolution = 14336 x 7560 (108.4 megapixels) (+36.2%)
Format Aspect Ratio = 1.896:1
Up to 2452 MB/s (+36.2%)

RED V-Raptor 'S35' - 16K 17:9
Format Size = 52.43mm x 27.65mm (59.27mm image circle/diagonal)
Format Resolution = 16384 x 8640 (141.6 megapixels) (+30.6%)
Format Aspect Ratio = 1.896:1
Up to 3202 MB/s (+30.6%)

------------------------------

RED Komodo-X - 8K 17:9
Format Size = 36.05mm x 19.01mm (40.76mm image circle/diagonal)
Format Resolution = 8192 x 4320 (35.4 megapixels) (+78%)
Format Aspect Ratio = 1.896:1
Up to 997 MB/s (+78%)

RED Komodo-X - 10K 17:9
Format Size = 45.06mm x 23.76mm (50.94mm image circle/diagonal)
Format Resolution = 10240 × 5400 (55.3 megapixels) (+56.3%)
Format Aspect Ratio = 1.896:1
Up to 1558 MB/s (+56.3%)

RED Komodo-X - 12K 17:9
Format Size = 54.07mm x 28.51mm (61.13mm image circle/diagonal)
Format Resolution = 12288 x 6480 (79.6 megapixels) (+44%)
Format Aspect Ratio = 1.896:1
Up to 2244 MB/s (+44%)

RED Komodo-X - 14K 17:9
Format Size = 63.08mm x 33.26mm (71.31mm image circle/diagonal)
Format Resolution = 14336 x 7560 (108.4 megapixels) (+36.2%)
Format Aspect Ratio = 1.896:1
Up to 3056 MB/s (+36.2%)

RED Komodo-X - 16K 17:9
Format Size = 72.09mm x 38.02mm (81.50mm image circle/diagonal)
Format Resolution = 16384 x 8640 (141.6 megapixels) (+30.6%)
Format Aspect Ratio = 1.896:1
Up to 3991 MB/s (+30.6%)

The data rates are just extrapolated from the percentage increases in megapixels, I don't know if that's an accurate way of estimating them.

There could be some other mistakes with the numbers or more fundamental problems with the working out, but I just wanted to try and see how the sensor sizes and resolutions relate to each other.

It's all theoretical for me personally, but I wouldn't want to see dynamic range reduced or refresh rates increased to achieve higher resolutions or larger formats, while limiting frame-rates and compression ratio's would be less of an issue. I also think there's more of a limit nowadays to how big a popular camera can be, but that seems to fit with RED's own aims anyway, though I imagine some of those sensor sizes and the amount of internal processing and cooling required would necessitate a definite variation on the current DSCMC3 body design. I also wouldn't have a problem if the data rates needed a proprietary media solution, but I get the impression no-one, including RED, would want to go back to that, so that alone could limit potential advances.

Anyway...

Now that I think about it, a 10k sensor based the 8k vv sounds pretty good too, though I think it would be harder to sell than 12k or 16k even in dci format. Also, if we're talking 60p, the data rate in hq would be 1.62gb per second

 

[fgonz031]

I imagine they'd already have a list like this somewhere, along with other possible sensors based on different pixel sizes.

Is there something there in particular that you'd like to see in an actual camera?

I would particularly like that RED/Nikon follow the path of BlackMagic in terms of high resolutions and obvious with a good global shutter sensor and high dynamic range stops, but that advance fast and do not stagnate in the 8K

 

[Les Hillis]

An interesting reference point might be the “Big Sky” camera built to capture imagery for the Sphere venue in Las Vegas. Per the ASC magazine piece on it, it’s a 77.5mm x 75.6mm sensor that captures in 18k (and up to 120, though I don’t know if that’s full resolution or otherwise limited.)

Check out the piece:

Sphere and the Big Sky Camera

An inside look at the creation of Las Vegas' entertainment venue Sphere, and the "Big Sky" camera system custom-built to capture its immersive imagery.

theasc.com

I think it's a good example of some of the technical challenges that need to be to overcome when pushing the limits like that.

Reading between the lines, it seems RED passed on making the camera. If so it would be understandable given the timing of it and the scale of the project, but I wonder if RED had made that camera what they might have been able to learn and develop that could have then been applied to their other cameras.

 

[Les Hillis]

I would particularly like that RED/Nikon follow the path of BlackMagic in terms of high resolutions and obvious with a good global shutter sensor and high dynamic range stops, but that advance fast and do not stagnate in the 8K

I agree, increase the capability without sacrificing the quality.

I think there's a good chance we are going to see that with the future RED and Nikon cameras.

 

[Les Hillis]

Now that I think about it, a 10k sensor based the 8k vv sounds pretty good too, though I think it would be harder to sell than 12k or 16k even in dci format. Also, if we're talking 60p, the data rate in hq would be 1.62gb per second

I think they could sell a DSMC4 camera with a 10K sensor based on the 5-micron pixel size, if it had enough other improvements over the 8K DSMC3 version, beyond the re-framing and downscaling to 8K it would enable by default. But yeah 12K or 16K might make more sense. Even while there are a lot of people who have no want or need for even 8K and who would prefer a more capable (and possibly cheaper) DSMC4 6K camera.

 

[fgonz031]

If you base whatever larger sensor sizes RED might come up with on the 5 micron pixel pitch they've used in their Dragon, Monstro and V-Raptor VV cameras, you only need to go up to between 10-12K to get a sensor size that could fit the sensor dimensions outlined in the original post.

If you also maintain the RED FF 1.90:1/17:9 maximum Aspect Ratio, here's the list of dimensions for the sensor size range mentioned by the OP (52-54mm wide and 28-30mm high) -

52mm wide x 27.30mm high (1.90:1/17:9)
53mm wide x 27.80mm high (1.90:1/17:9)
54mm wide x 28.30mm high (1.90:1/17:9)

28mm high x 53.2mm wide (1.90:1/17:9)
29mm high x 55.1mm wide (1.90:1/17:9)
30mm high x 57mm wide (1.90:1/17:9)

If you take the largest of those combinations (30mm high x 57mm wide) and calculate the resolution using the 5 micron pixel pitch, you get -

RED V-Raptor 11.4K - 17:9 (11400x6000)
Format Size = 57x30mm (64.41mm image circle/diagonal)
Format Resolution = 11400x6000 (68.4 megapixels)
Format Aspect Ratio = 1.90:1

Doesn't seem unreasonable.


If you want to make the sensor an actual 65mm wide, you only need to go up to 13K.

A theoretical RED V-Raptor VV 65mm sensor -

RED V-Raptor 65 - 13K 17:9 (13000x6840)
Format Size = 65x34.20mm (73.45mm image circle/diagonal)
Format Resolution = 13000x6840 (88.92 megapixels)
Format Aspect Ratio = 1.90:1


These calculations are all based on the current RED V-Raptor VV sensor specs -

RED V-Raptor 8K VV - 8K 17:9 (8192x4320)
Format Size = 40.96x21.60mm (46.31mm image circle/diagonal)
Format Resolution = 8192x4320 (35.4 megapixels)
Format Aspect Ratio = 1.90:1


Just to add, it's been mentioned before that at a certain point, it's more of an ecomonic barrier than a technical one that's prevented RED from making larger sensor cameras already. There has to be enough of a market there for it to make sense from a business perspective. That means a high enough amount of buyers, or a more limited amount of buyers prepared to pay more, which could even include another business or venture paying for the initial development of a product that RED can then turn around and sell to others.

So far, it just hasn't happened and RED, perhaps for additional reasons, haven't just gone ahead and taken a build-it-and-they'll-come approach.

The weird thing is, with Nikon on board, there's the opportunity for something developed for the limited cinema camera market to also upend other camera markets as well...

But 12K is not 80Mpx is 100Mpx according to this Of Cine 12K Camera Features

Effective Sensor Size

35.64mm x 23.32mm (Large Format)

Lens Mount

PL mount. Interchangeable with URSA Cine lockable EF mount included.

Lens Control

Active EF and PL mount lens control with compatible lenses.

Dynamic Range

16 Stops

Shooting Resolutions

12,288 x 8040 (12K 3:2 open gate) 12,288 x 6480 (12K 16:9)

12,288 x 6912 (12K 17:9) 12,288 x 5112 (12K 2.4:1)

9648 x 8040 (12K 6:5)

9408 x 6264 (9K 3:2)

8688 x 4896 (9K 16:9)

9312 x 4896 (9K 17:9)

9312 x 3864 (9K 2.4:1)

7680 x 6408 (9K 6:5)

8192 x 5360 (8K 3:2 open gate)

8192 x 4608 (8K 16:9)

8192 x 4320 (8K 17:9)

8192 x 3408 (8K 2.4:1)

6432 x 5360 (8K 6:5)

4096 x 2680 (4K 3:2 open gate)

4096 x 2304 (4K 16:9)

4096 x 2160 (4K 17:9)

4096 x 1704 (4K 2.4:1)

3216 x 2680 (4K 6:5)

Frame Rates

Project frame rates of 23.98, 24, 25, 29.97, 30, 50, 59.94 and 60 fps supported. Maximum sensor frame rate dependent on resolution selected.

High Speed Frame Rates

12K 3:2 open gate up to 80 fps

12K 17:9 up to 100 fps 12K 2.4:1 up to 120 fps

9K 3:2 Super 35 up to 100 fps 8K / 4K 3:2 open gate up to 144 fps 8K / 4K 2.4:1 up to 224 fps

 

[Les Hillis]

Yes, the larger (taller) '12,288 x 8040 12K 3:2 open gate' aspect-ratio adds up to approx 100 megapixels - 12288x8040 = 98,795,520

While the smaller (shorter) 12,288x6480 12K 1.9:1/17:9 (12K DCI) adds up to approx. 80 megapixels - 12288x6480 = 79,626,240


Also, not sure why, but in that list you posted, these aspect ratio's seem to be incorrect (reversed) -

",,,Shooting Resolutions...

12,288 x 6480 (12K 16:9)

12,288 x 6912 (12K 17:9)..."

12,288 x 6480 is actually 1.9:1/17:9 (12K DCI) and 12,288 x 6912 is actually 1.78/16:9.


So many different names and numbers describing the same thing and so many different things being given the same name, with a bunch of odd-ball variations thrown in just to make it all more confusing.

 

[Brian Timmons]

In terms of pixels and resolution this was an interesting article featuring the Achtel 9x7 65 megapixel camera comparing it's MTF to other cinema cameras.

MTF Camera Comparison- How does the Achel 9x7 Digital Cinema Camera compare to high end digital cinema cameras - Newsshooter

Pawel Achtel ACS has conducted an MTF camera comparison test with the help of the Australian Cinematography Society to see how sharp his 9x7

www.newsshooter.com

What's interesting is how the MTF test produced results that created even more questions regarding what's going on under the hood of these cameras to produce the image they do.

How a 4K Alexa LF mini had relatively close MTF results to an 8K RED Raptor.

How the Arri 35 seemed to add sharpening in it's RAW software even when sharpness was turned off.

How Black Magic 12K actually had the lowest MTF and highest amount of artifacts and aliasing.

There is naturally a lot going on beyond the pixel numbers but the quality of the pixels has a big impact on the end result.

Brian Timmons
BRITIM/MEDIA

 

[Zack Birlew]

In terms of pixels and resolution this was an interesting article featuring the Achtel 9x7 65 megapixel camera comparing it's MTF to other cinema cameras.

MTF Camera Comparison- How does the Achel 9x7 Digital Cinema Camera compare to high end digital cinema cameras - Newsshooter

Pawel Achtel ACS has conducted an MTF camera comparison test with the help of the Australian Cinematography Society to see how sharp his 9x7

www.newsshooter.com

What's interesting is how the MTF test produced results that created even more questions regarding what's going on under the hood of these cameras to produce the image they do.

How a 4Kish Alexa LF mini had relatively close MTF results to an 8K RED Raptor.

How the Arri 35 seemed to add sharpening in it's RAW software even when sharpness was turned off.

How Black Magic 12K actually had the lowest MTF and highest amount of artifacts and aliasing.

There is naturally a lot going on beyond the pixel numbers but the quality of the pixels has a big impact on the end result.

Brian Timmons
BRITIM/MEDIA

With that test, I wonder how the results would look after a nice round of Neat Video to denoise and clean up the images. For many, this is the workflow that goes into cinematic feature films on digital and why there are so many digital grain plugin options. On my feature film, "The Diner"(working title), the RED One MX, Blackmagic Pocket 4K, Sony A7S, and GoPro footage all seemed to balance out once Neat Video did its work and the difference between original footage, no grain, and Filmconvert Nitrate on top changes things dramatically the whole way through.

 

[Phil Holland]

With that test, I wonder

Or better yet, have all cameras feature a Low Pass Filter installed or not as that would be a more useful test. The production cameras do have filters installed btw as can be seen by the results.

 

[Karim D. Ghantous]

Speaking of low pass filters... Leica could never make a cinema camera, because they could never bear to include a LPF in it. The last time I checked, cinema cameras absolutely need a LPF, whether we like it or not.

 

[Afshin]

What's interesting is how the MTF test produced results that created even more questions regarding what's going on under the hood of these cameras to produce the image they do.

How a 4K Alexa LF mini had relatively close MTF results to an 8K RED Raptor.

Pixel fill factor and the size of the microlens plays a big role in that. When microlens is smaller than the pixel itself, or the sensitive part of the pixel is smaller than the overall pixel size, you get an image roughly resembling a line skipped video that very common in mirrorles cameras (to lower sampling rate to keep data and heat manageable in those compact bodies). And with that kind of image you get more aliasing, so you have to use stronger OLPF, and hence lower MTF numbers.
Jim Kasson did a great comparison between two medium format Fujis with different fill factor:

On microlens size in the GFX 100 and GFX 50R/S - the last word

This is one in a series of posts on the Fujifilm GFX 100. You should be able to find all the posts about that camera in the Category List on the right sidebar, below the Articles widget. There’s a drop-down menu there that you can use to get to all the posts in this series;

blog.kasson.com

I don't know about RED pixel/microlens design but generally global shutter eats a big chunk of pixel area due to placement of capacitors needed to store the electrons right beside the photo diode.

 

[Karim D. Ghantous]

I don't know about RED pixel/microlens design but generally global shutter eats a big chunk of pixel area due to placement of capacitors needed to store the electrons right beside the photo diode.

Fascinating stuff. From what I've seen so far, I am not happy with the image quality of the Sony A9 III. Having said that, I can say the same for the first two. It didn't get worse, but it didn't get much better IMHO. So many variables...

 

[Brian Timmons]

So many variables...

Yes. This here.
Naturally the vast majority of us are filmmakers and not sensor engineers and I think we tend to see things at a macro level of pixels, resolution, bit depth, etc which is great to know but there are clearly deeper levels to this that influence the end result.

A part of me is not completely convinced that more pixels is necessarily the only way to go with a larger sensor but again I’m not a sensor designer and I’m positive there a varying opinions amongst them.

Hope this discussion generates deeper understanding regarding sensors and larger formats.

Thanks Afshin for the informative link.


Brian Timmons
BRITIM/MEDIA

 

[Les Hillis]

I agree and think it's at the pixel level that the real action happens. The decisions made there inform the basic characteristics of the overall sensor and the fundamental image-quality of the camera.

At some point saying I want 'this' sized sensor or format or resolution is like saying I want a motorcycle or 4-wheel-drive or sports car. Nothing wrong with that, but there are so many design and materials and other decisions that get made, for better or worse, within the broader categories.

I'm always more interested in products where the form follows the function, rather than when the function gets altered to fit within a particular set of constraints serving some other purpose.

With cameras of any sort there are already so many limits put in place by the laws of physics and the fact that whatever device you come up with has to interface with an actual human end-user, but in some ways it's still a pretty big playground to be in. So when it comes to actual pixel and sensor design, I'd rather see someone just hand it over to the engineers and scientists and let them get creative with expanding upon the functionality of the pixels and sensors, allowing the results of that to determine the form the camera ulitmately ends up in.

I suspect there are in fact people at Nikon and RED who could come up with stuff like that. And equally importantly, it seems to me there are also people in those companies who are willing and capable of allowing those crazy solutions to get translated into products the rest of us can actually use.