FS7: 5 reasons the URSA Mini is better than the Sony FS7

I think these guys pulling their lenses off and testing white paper is wrong way to go. JB posted an ARRI camera that produced magenta in the same conditions. Having said that my F55 did not however.

But there is for sure a problem because I can spot magenta gradient corners on most people's footage posted online, etc.. Just go watch a few, it's there.

Those guys who are experiencing the problem should at least share info on camera setting, lens used, lens adapter if any, and set aperture. There has to be some commonality.

If it ends up this problem only comes down to some crazy specific combination of lens/camera settings ONLY then I might still consider getting a 4.6K mini. But I'll keep the F55 for real shoots, that's for sure.


[EDIT]

glutton_for_punishment_hat-r26f989937ffb4ad1bfaeb82a8781c695_v9wfy_8byvr_324.jpg
 
Last edited:
But wait, there's more…..

(anyone actually still reading this thread?)

So today I saw a video posted on vimeo which the guy has made un-linkable, but I've included some screenshots below. He shot on the URSA 4.6K 18-35mm sigma lens and as he zoomed out the magenta cornering went away (almost entirely but there's still a hint of it near the end). No idea what aperture he used but this is very very interesting.

The plot thickens:

2qupwle.png


2ikttgi.png
 
If it's shading of the micro lenses on the sensor then. If you imagine light coming in at a low angle (low being horizontal), it hits the lens but the next lens along is in shadow now because the lens itself stops the light. I think the wavelength of light is the colouring reason. However i think there are many other ingredients in play that i don't know about but broadly speaking this kind of effect is something along those lines. In very wide angle lenses on Leicas/A7x series, the light will shade magenta on one side and green on the other - but i suspect that's to do with the physical nature of the lenses.

The key is to remember the micro lenses on top focus the light to the recessed sensor collection material. Different sensors are built in different ways to help this. Sony recently introduced a curved sensor to get around these kinds of issues.

The fact that changing the focal length of the lens removes it seems to point towards a light/sensor/lens interaction.

IMHO, which could be totally wrong

cheers
Paul
 
I am tending to Paul's interpretation. While I still have the Ursa 4.6K, I'm going to try shooting white bead board at varying focal length / fstop / codec variations to see if I can reproduce the magenta.

Overall impression, though of the camera is that is an excellent camera for indie work, interviews and even some doco / doco-drama.

Most detracting feature of camera is the need to open the view screen door to access menu and controls (particularly color temp and ISO which are pretty crucial in my book). With the excellent BMD EVF plus baseplate and handle, very nicely balanced on shoulder. The handle extension needs work but that's where the 3rd party guys will come in (I've replaced my FS7 handle with the Shape).

What really remains to be seen is how BMD can keep up with service as well as ultimate durability of the camera.

I know that BMD is aware of the magenta concerns and they really have a good track record of addressing these things with as prompt a firmware update as possible.

Ned Soltz
 
nsoltz said:
I am tending to Paul's interpretation. While I still have the Ursa 4.6K, I'm going to try shooting white bead board at varying focal length / fstop / codec variations to see if I can reproduce the magentaz
Click to expand...

I would also try different lenses.. I would guess that the more retrofocal ones will suffer less. One 24mm may give different results from another 24mm.

S
 
Regarding angles of light, distance of rear element to sensor, and sensor microlenses, it is hence why I think white paper test is worst test to try on the URSA. People should stick to lenses and look for the commonality.

I've been posting on Facebook and official BMD forums for users to try different lenses, aperture settings and focal length if a zoom, etc.. to see what the connection is and some owners say they will try some tests and report back. Others are returning the camera and want nothing to do it with.

So far out of two videos I watched that were flawless and free from magenta cornering, Canon brand zooms were being used - this is interesting.
 
starcentral said:
So far out of two videos I watched that were flawless and free from magenta cornering, Canon brand zooms were being used - this is interesting.
Click to expand...

I wonder if its the difference between Full Frame lenses and APSC. They would have different image circles and direct light at different angles wouldn't they? It seems a lot of BM users have noticed it with the Sigma 18-35 which is a fantastic but is APSC where most Canon zooms used are full frame.

I have only come across the grid pattern on my sony sensor twice and both were with full frame glass. I had wondered if it had something to do with angles of light let in with FF glass but it was so rare I didn't bother to test the theory further.
 
Last edited:
paulcurtis said:
Me? Off topic? Surely not...

All i meant was that in the case of the Ursa Sensor design started, what a few years ago? Production hitting now.

In the case of the F5 it was released 2012? And the design of that sensor probably started a few years before that.

So the Ursa sensor is perhaps 6 years ahead of the sony? I can't say when the design of it started, but certainly BM tweaks to it are a lot more recent.

And my argument is that in 6 years there is quite a lot of change, perhaps not with the silicon itself but with the rest of it. So to my mind the Ursa sensor could be considered a generation newer? I'm sure there are Sony sensors in R&D right now that would be pretty amazing.

cheers
Paul
Click to expand...


I think you did go off topic. The point is the current industry standard is a basically 10 years old sensor. It has nothing to do with how old the Ursa or F5 sensor is and all to do with, not because it's newer it means it's better. So it doesn't mean because the Ursa sensor is 6 or whatever years newer than the F5's, that it is any better, when a 10 years old sensor is the industry leader. But yes, the Ursa sensor is most likely "a generation" newer than the F5's.
 
FrameFarmMedia said:
I wonder if its the difference between Full Frame lenses and APSC. They would have different image circles and direct light at different angles wouldn't they? It seems a lot of BM users have noticed it with the Sigma 18-35 which is a fantastic but is APSC where most Canon zooms used are full frame.

I have only come across the grid pattern on my sony sensor twice and both were with full frame glass. I had wondered if it had something to do with angles of light let in with FF glass but it was so rare I didn't bother to test the theory further.
Click to expand...


Adam, that is a damn good theory. Really good.


I'm also really trying to argue with these people doing paper tests over the lens mount that its bad because it will just introduce more light scatter (bounce of sensor bay innards, etc.. ) and regardless it's not how the camera is going to be used so what the hell does it prove anyways?

To further my point, I just saw a test where with paper over lens mount there was no magenta at all, then with a lens there was. lol
 
Aaveum said:
I think you did go off topic. The point is the current industry standard is a basically 10 years old sensor. It has nothing to do with how old the Ursa or F5 sensor is and all to do with, not because it's newer it means it's better. So it doesn't mean because the Ursa sensor is 6 or whatever years newer than the F5's, that it is any better, when a 10 years old sensor is the industry leader. But yes, the Ursa sensor is most likely "a generation" newer than the F5's.
Click to expand...


Just because something is the industry standard it doesn't mean it's the best either. It just means a lot of industries are slow to adopt new standards.

NTSC was an industry standard for decades. Rec709 is an industry standard. 23.976fps is in some less enlightened parts of the world an industry standard.
 
starcentral said:
Adam, that is a damn good theory. Really good.


I'm also really trying to argue with these people doing paper tests over the lens mount that its bad because it will just introduce more light scatter (bounce of sensor bay innards, etc.. ) and regardless it's not how the camera is going to be used so what the hell does it prove anyways?

To further my point, I just saw a test where with paper over lens mount there was no magenta at all, then with a lens there was. lol
Click to expand...


As I alluded to a few pages back, the variance in strength of the problem between cameras, and the fact some cameras do not exhibit the problems at all, really point to it being a manufacturing and then QC issue. Which are precisely where you wouldn't be surprised to see a company cut some corners when releasing a feature packed camera for less than half the price of the camera closest to it, going by 'on paper' specs.
 
Aaveum said:
I think you did go off topic. The point is the current industry standard is a basically 10 years old sensor. It has nothing to do with how old the Ursa or F5 sensor is and all to do with, not because it's newer it means it's better. So it doesn't mean because the Ursa sensor is 6 or whatever years newer than the F5's, that it is any better, when a 10 years old sensor is the industry leader. But yes, the Ursa sensor is most likely "a generation" newer than the F5's.
Click to expand...

A valid point, though I'm not sure where you get that the Alexa is 10 years old. It was released almost exactly six years ago. It's the Panavision that was the first S35mm digital cinema camera, and that was released 10 years ago.
 
This threat started as a joke, still smile if I see reactions. I had the ursa in my hands in Amsterdam nab 2015 but the FS7 was available and the URSA wasn't. So there was no much to choose then.

I am used to buy things after reading reviews. Nowadays we buy things that even don't exist yet. Kickstarter and Blackmagic invented a whole new way of selling things. No reviews but paper specs. It is a bit like the stockmarket, you can win or you can loose. The early adapters lost the global shutter. But now have a nice camera.

The thing is, competition is good and I like the cool accent of their makers at BM.
 
Last edited:
I ran a magenta test as follows:

White bead board lit by a single Cineo Matchbook 5600K. I know this produces a pure white light with no spikes.

Ursa 4.6K with Canon 24-105. Camera set at ISO 400 and Iris at f4 for Raw/Log footage and 5.6 for ProRes Rec 709.

Shot full frame 4.6K in Raw and Raw 3:1 at focal lengths on the Canon of roughly 24, 35. 50, 70 105mm. Then shot ProRes 4444 Log (they call it Film) same lengths. Next, ProRes 4444 Rec709 (Video in BMD speak) also at all of these focal lengths.

Bottom line-- I had no magenta vignette or cast as others have experienced.

This does not mean that the problem doesn't exist because we've seen examples of it. I just couldn't reproduce with the combinations used above.

Also shot last week (real footage) in ProRes log and 3:1 RAW mostly with Zeiss Otus 55/1.4 and two Milvus lenses. No magenta on this footage either.

The problem has to be one of specific sensors and the combination of interaction between sensor, lens, focal length and perhaps exposure (as has been discussed here).

I was so hoping to reproduce the magenta but yet another unexpected experimental result.

Ned Soltz
 
Ned, great that you have done this and shared.

Can you use ND filters and shoot wide open on the lens? this has the greatest chance to reproduce it. Particularly on you widest focal length.

Do you have any other Canon zooms you can try or other lenses?

Is your test URSA 4.6K preproduction or production model?

cheers matey
 
I'm curious if this issue arises as well with vintage manual lenses. Like the nikkors, contax zeiss, etc.

Seems like people are seeing it mainly with the modern canon/sigma lenses.
 
You must log in or register to reply here.
Back
Top