Steve Yedlin, ASC on Resolution

[Adrian Jebef]

This guy gets it:

http://yedlin.net/ResDemo/#

[Frank Glencairn]

The main takeaway for me is, that compression kills any resolution. Something I was saying for years. Whats the point of having a 6k+ sensor, when your camera is averaging out boxes of 16x16 pixels at the recording level. That's why I prefer an uncompressed raw out of my BM cameras, over any compressed 6k+ material. We also need better delivery codecs, not more pixels.

[Michel Hafner]

Graeme, comments?
Well, some comments from me. Overall a good presentation and I agree mostly with his 3 statements, but...
It is no coincidence that the second part of the demo shows us a close up and examines it in detail. When you see how 2K upsampled looks practically the same
as 6K downsampled it is obvious we have a garbage in garbage out situation. Garbage not meaning crap here but simply an example that has no 4K real world detail and much less 6K detail to start with so the result is a given. What is the message? That normal photography for cinema when doing close ups provides nowhere 4K detail anyway independent of the Ks of your camera? Probably fair enough to claim. But people asking for more Ks are not complaining about close ups usually, but wide shots or other intended uses like reframing.
And yes, even close ups can be very sharp when shot that way as the 4K "Billy Lynn" demonstrates repeatedly. The demo also shows rightly that downsampling algorithms and upscaling algorithms play a huge role in the look of the final result. You can make the 2K upsample look sharper than the 6K downsample, but the looks are deceiving. The 2K looks forced, unnatural and coarse. The 6K effortless, natural and finely resolved. If you start to sharpen the 6K during downsampling you get a 4K that beats the shit out of the 2K upsample long before it gets itself forced, unnatural and coarse as long as we have a picture that actually has visibly more detail to offer than 2K to begin with.
These are not very interesting examples. Interesting examples are wide shots full of 4K and above detail and comparing different compression and noise levels of cameras here.
In the second example he uses, the difference between 2K and 6K in 4K is obvious to me. The edge sharpness is clearly degraded by the 2K. A better example but again not one that pushes the issue.
Also, the 2K he uses is super 2K since it comes from 6K. When using straight 2K the results for 2K are worse. 2K does not hold up on large cinema screens unless you sit sufficiently far away. And that is not the case for a big part of a modern auditorium when your vision is 20/20. Laymen don't care, but these demos are made for professionals with a discerning eye.
He's absolutely right that the Ks are maningless without knowing about the quality of the pixels. And with same quality pixels the quality improvement for more Ks over less Ks is getting smaller and smaller till it becomes invisible for typical footage shot for cinema and typical viewing conditions. The threshold is not at 2K, though.
He says one should not see pixels. Well, pixels can not only be seen the way he shows them in a still. They can also be seen indirectly in the form of aliasing/moiré in motion. And how to fight aliasing best? Optical filtering and oversampling with more K. The best mix of them is not at 2K for sure. He completely ignores this important point.
The halation/grain/sharpening/grading/compression... part is very interesting. The Alexa65 pixels are higher quality than Weapon 6K. But is this due mainly to HW (pixel size) or compression? Is it still true for the 8K and IPP2 pipeline? His compression examples seem to use compression for consumer product, not compression as used by Red cameras. I would like to see this with uncompressed versus 3 times, 5 times, 9 times or so compressed, not 20 or 50 times.
PS Don't show Christopher Nolan...

[Kemalettin Sert]

i was suprised by Alexa65 which looks sharper and better than imax 15perf 65mm.

[Christoffer Glans]

https://theasc.com/articles/a-clear-...-of-resolution

[EMMANUEL DELPECH]

i was suprised by Alexa65 which looks sharper and better than imax 15perf 65mm.

I don't know what to think about his tests, he applies film grain on every single image, which is cheating of sorts , to make them virtually interchangeable, but those are very controlled situations. What I do know is that seeing Dunkirk in laser IMAX (razor sharp in the fifth row), and seeing the War For The Planet Of The Apes trailer beforehand (which is a 2K release), which looked fantastic, the Alexa 65 has nothing on IMAX 70mm & 65mm 5 perf, it's two different worlds imo. Now of course, watching Steve's demo on a much smaller screen is going to make it harder to see which is which. I also do know (even though that is not the point of his demo) that you could shoot La La Land, or The Lost City Of Z or whatever on the Alexa or Alexa65, etc, and it wouldn't look at all the way it does on film.

[Michael Lindsay]

As always Steve Yedlin's dedication and rigure is impressive..

if anyone close to the test could leak what compression ratios the Weapon was shot at? the wide shot suggested to me it was greater than 5to1!

Also in the halation section there seemed to be a bluring (pun intended;-) of target artifacts and lens artifcacts? ..

what would be a little helpful for me in evalutating this would be understadning which camera used which lens and at what stop?

Also it would be great to be able to watch a lees compressed file?

Re lenses

Leica Sumilux
PV Primo
PV Primo 70 (I assume Alexa 65... due to vignetting)
Masterprimes
Zeiss Hasslblad (assuming IMAX?)

any one know what the deal was...

thank you for posting..

Michael

[Kemalettin Sert]

I don't know what to think about his tests, he applies film grain on every single image, which is cheating of sorts , to make them virtually interchangeable, but those are very controlled situations. What I do know is that seeing Dunkirk in laser IMAX (razor sharp in the fifth row), and seeing the War For The Planet Of The Apes trailer beforehand (which is a 2K release), which looked fantastic, the Alexa 65 has nothing on IMAX 70mm & 65mm 5 perf, it's two different worlds imo. Now of course, watching Steve's demo on a much smaller screen is going to make it harder to see which is which. I also do know (even though that is not the point of his demo) that you could shoot La La Land, or The Lost City Of Z or whatever on the Alexa or Alexa65, etc, and it wouldn't look at all the way it does on film.

see pt2 video it shows sharpness comparison as an insanely zoomed in.
Yeah whats up with that? i could have tell difference in those cameras with even only 1 frame but those haliation and grain looks exactly same on every camera.Looks like he applied lut all of it.

[Graeme Nattress]

From an 8k source, top is 2k upsampled to 4k, bottom is 4k downsampled from 8k. I think it's pretty obvious which I prefer and which looks better....

But the argument being made is a straw-man. It's also problematic to think of the audience as the only consumer of the image.

Smaller pixels are noisier - unless they are not. To state that as a general rule would be incorrect. It's also incorrect to assume that people in industry don't test to see the benefits and trade-offs of compression / resolution because I know that they do, and they go off the results of those tests and not headline figures.

Graeme

[Graeme Nattress]

More importantly, even taking the result from above down to 2k, the benefits are still visible.

Graeme