URSA Mini 4.6k The Original™ Magenta Cast Issue

Here is my first grade from the 4.6k


The last screenshot is the shot of the sky pic i posted above, corrected..i'm practicing my teal and orange blockbuster look

4.6k grade.png

I definitely do not see any magenta in those sky shots or the corners of the frames on the wall on the tests above

Welcome to the Brooklyn Ursa Mini club. Benton and are the only existing members, so you just increased membership by 50%

Thank you fahnon, happy to be part of this most prestigious club indeed ,Brooklyn Ursa mini owners rule!!!!

Sent from my iPad using Tapatalk

Looking good, like it should.
Cheers

Thanks Denny and rick

Howdy Folks,

What numbers should a UM4.6k start with to be in the second batch... or the batch that has the updated sensor.

Just purchased a new UM4.6k which is coming tomorrow.

If anybody could enlighten me I'd be very grateful.

I also got the Atoch solution... so any thoughts on that would be welcome too

Best,
Johnny

i dont got that answer but i got my 4.6k ef last week serial 3234.. looks clean on my ends. will post dngs..

This is the first example of a telecentric lens on the ursa mini 4.6k that I have seen. It would be interesting to see what the sigma 18-35 set to 35mm and f18 look like on the same camera, just to get an idea how much of the problem is sensor shading.

My 4.6k PL is 3169. Today i did shoot a little more outside ..no magenta here, so i guess there is something to be said about the telecentric lenses. Gene we'll be able to test that when sigma's cine line is available, since its what i believe to be their re-housed still lenses

Testing out that new auto white balance..wow!

Dusk in Brooklyn

Dusk in brooklyn.png (653.51 KiB) Viewed 18755 times

What I've noticed when testing magenta corners by shooting the sky is that unless you get some white cloud in the corners it doesn't show up much at all.

Gene this is a sky shot just for you A little overcast but hope it helps. Shot on the 35mm Schniedar Xenar 3 Pl, 4.6k raw, @ T-16, ISO 800, White balance 5600k No filters. I included the Parade scope of the sky shot as is.

SKY 4.6K RAW.png (838.73 KiB) Viewed 18624 times

SCOPES FOR SKY IMAGE.png (169.12 KiB) Viewed 18624 times

This is some lens test i made today to see how bad was the magenta cast issue in my camera, my veredict: unusable

as shot:

https://vimeo.com/183560108/71834a3660


with official film to video lut applied:

https://vimeo.com/183560107/cf190737ab


I don´t have cfast cards at the moment, so i used a shogun as my main recorder through sdi.

My first 4 digits in my serial number are: 2951


I already had a rma before, i guess i'll have to rma again.

Any opinion on what should i do?

This is some lens test i made today to see how bad was the magenta cast issue in my camera, my veredict: unusable

as shot:

https://vimeo.com/183560108/71834a3660


with official film to video lut applied:

https://vimeo.com/183560107/cf190737ab


I don´t have cfast cards at the moment, so i used a shogun as my main recorder through sdi.

My first 4 digits in my serial number are: 2951


I already had a rma before, i guess i'll have to rma again.

Any opinion on what should i do?

It does look like you'll have to send it in again

Arturo, I'm a hobbyist, so I don't make a living off a camera. When I got my um46k and found it to have strong magenta corners I sent it back for a full refund. I decided to not play the RMA game until I saw more positive reports of cameras with minor to no magenta corner issues. I also suspect, based on this forum and the reviews available on the internet that many others did the same. Unless BM is doing something to fix these cameras I suspect they are just going back out to customers.

That said, I am seeing improvement with the magenta issue, but I'll wait a bit longer until I see more positive reports with lenses that appear to have the biggest problems such as the Sigma 18-35.

Gene this is a sky shot just for you A little overcast but hope it helps. Shot on the 35mm Schniedar Xenar 3 Pl, 4.6k raw, @ T-16, ISO 800, White balance 5600k No filters. I included the Parade scope of the sky shot as is.

SKY 4.6K RAW.png

SCOPES FOR SKY IMAGE.png

Looks good. More indication that the problem is fundamentally a sensor shading issue. It can be corrected to some extent in firmware and software, but the best fix will require a different sensor. I wonder if BM concurs and they are doing just that?

Mine is a 2909 and has been showing signs of improvement. I'm also keeping an eye on progress in eliminating the problems (more problems than just magenta). My warranty expires before NAB 2017, so I'll likely make a decision by February/March when I hope I'll be able to see great images consistently given they're exposed reasonably well.

Although I'm only an enthusiast, I now have an almost weekly assignment covering sports events this year! So I need to hang onto it until at least January 2017. The events are filmed in 2K with the Fujinon B4 so no magenta issue there for me. But I do have a 4.6K project pending...


Sent from my iPad using Tapatalk

Is there some trick to uploading files on this forum. I tried twice to upload a png file here but it gives me a blank page when I press "add the file".

I had done a shot of a white wall.

Johnny - Message boards are notoriously tetchy with image files and bandwith, so if you're going png use something like Imgur. If it's images you want us to look at, you're much better off using DNG files and putting them on a Dropbox.

Johnny, maybe you can upload it somewhere and link it here?

Thanks!

It'd be really cool to have an official response/update. And by cool, I mean essential...

Okay, unfortunately I can't upload anything on this forum because the upload function doesn't seem to work (tried 3 browsers) so here's an Imgur shot. Done some external sky shots etc... I'm very happy with the camera so far.

Hey Johnny,

what lens and focal length was that? And what was the aperture? FOr white wall magenta vignetting tests we generally like 35mm to 50mm at f/8 for worst case.

Can you share a link to a non compressed DNG of 35mm at f/8?

Hi all,
I am currently tracking this thread for a while as I am about to order the mini 4.6.
It has been quiet for months now, and hopefully BM had fixed this issue. I am in Canada so in case of RMA it will not be fun (custom$, transport$, time).
If I do order, I will post a white shot at 35mm/f8 dng file and serial number just for the audit.
Regards,
Harold

Just to follow up with my earlier remark in September: I'm shooting a variety of apertures in various codes both windowed and open gate. Not seeing any magenta now in my images that might concern me from the 25mm, 50mm, and 85mm APO primes. I've never found a white wall to shoot here, but when I first had the camera, magenta was evident in many of my shots with the 50mm APO.


Sent from my iPad using Tapatalk

Hey Rick, have you tried the 16mm, 24mm and the 32mm Zeiss Standard Speeds with the UM46K?

Thanks!

Just my 25mm APO as my current wide. 32mm APO will come in 2017. Sorry not able to comment on those lenses.


Sent from my iPad using Tapatalk

Hi all,
I am currently tracking this thread for a while as I am about to order the mini 4.6.
It has been quiet for months now, and hopefully BM had fixed this issue. I am in Canada so in case of RMA it will not be fun (custom$, transport$, time).
If I do order, I will post a white shot at 35mm/f8 dng file and serial number just for the audit.
Regards,
Harold

Just a heads up: It's been quiet because BMD has stopped responding about this issue, so there's no reason to assume it has been fixed. In fact, the last we heard officially was that it will appear for apertures smaller than f/5.6, so if you test at f/8, you may see it.

I sold my camera, but I was talking to a friend today who just bought one and apparently B&H told him that the magenta issue was fixed with FW4.0, so there's lots of misinformation out there. My friend did some tests and I see magenta corners in his (and he just bought it last week). I don't know what aperture he was shooting with so I'm not sure if it falls within the expected behavior of it appearing at apertures smaller than f/5.6, but it clearly hasn't been eliminated completely.

If I were you, I wouldn't RMA it. If you end up with one, just make sure you get it from someone local with a good return policy and go that route.

Good luck!

Thank you for this last update. I will take precautions.

Hello from Québec.

I have received mine (3355xxx) and the first thing in my list was verified the magenta corner cast.
Not having a white wall, I shot the snow.
The screen shot from resolve speak for itself.

4608x2592
FW4.0
raw 4:1
Sigma 18-35mm f/1.8 DC HSM
iso 200
k5600
tint 10

f5.2
35mm

f5.2-35

f5.2-35.png (997.81 KiB) Viewed 17403 times

f8.0
35mm

f8-35

f8-35.png (996.49 KiB) Viewed 17403 times

f8
24mm
saturation 100%

22-f8-sat100

22-f8-sat100.png (1016.85 KiB) Viewed 17403 times

Oh, unhappy day, Harold! Seriously, looks very good albeit not perfect when you use 100% saturation. But have you seen anything troubling when you shoot a real subject of interest in a motion picture?


Sent from my iPad using Tapatalk

Hello from Québec.

I have received mine (3355xxx) and the first thing in my list was verified the magenta corner cast.
Not having a white wall, I shot the snow.
The screen shot from resolve speak for itself.

4608x2592
FW4.0
raw 4:1
Sigma 18-35mm f/1.8 DC HSM
iso 200
k5600
tint 10

f5.2
35mm

f5.2-35.png

f8.0
35mm

f8-35.png

f8
24mm
saturation 100%

22-f8-sat100.png

Unfortunately it's in these examples. Not nearly as bad as I've seen, but still there.

Hello from Québec.

I have received mine (3355xxx) and the first thing in my list was verified the magenta corner cast.
Not having a white wall, I shot the snow.
The screen shot from resolve speak for itself.

4608x2592
FW4.0
raw 4:1
Sigma 18-35mm f/1.8 DC HSM
iso 200
k5600
tint 10

f5.2
35mm

f5.2-35.png

f8.0
35mm

f8-35.png

f8
24mm
saturation 100%

22-f8-sat100.png

Unfortunately it's in these examples. Not nearly as bad as I've seen, but still there.

Thx for looking at.
I am considering returning it, but will check what bmd has to say with my dng I just sent. They responded fast. In winter we receive a lot of light in day time, f8 and up are sometimes useful for moving targets.

Have you noticed any crosshatching? In my unscientific observation (simply noticed after working with 3 cameras), the ones with stronger magenta corners have almost no crosshatch. The ones without corners have very strong crosshatch.

Have you noticed any crosshatching? In my unscientific observation (simply noticed after working with 3 cameras), the ones with stronger magenta corners have almost no crosshatch. The ones without corners have very strong crosshatch.

No, I will have to hurry to check that before I decided to return it.

I will read how to perform a descent test of the crosshatch.

Confirmed mild magenta in mine. No crosshatch noticed so far. Serial start: 3298

My TL;DR for new/prospective 4.6K owners:
- The magenta issue is not fixed, just improved by firmware 4.x.
- If you can live with it in certain situations, you simply won't find a better image for your money (or even arguably a better camera) at this price point. For me, I wasn't willing to spend 2-4x for an FS7, C300, or RED.
- You may be able to solve the magenta vignette in certain cases with an HLS Secondary (Lumetri) in Premiere CC. I successfully did this on a white cyc wall shoot.

Side note: Firmware 4.x is absolutely awesome. Kicks the crap out of Sony and Canon menus IMO, but don't be surprised if you have to roll back to Firmware 3.1 then update to 4.x due to failed updates.

I have received the RMA form from BMD.
Sorry if redundant, but did anyone get back with a rectified unit after RMA ? Just reading the threads and I have not noticed many that succeed.

Harold, you got an RMA for magenta corners? Thats new...

I havent seen a UM4.6k camera yet that has less magenta in the corners than this one does at these settings.

Thx Peter and Kyle, your comments are encouraging me to persist looking at way to use it under certain restrictions. and sometimes correct in post-prod.

It's too bad that after many months of this camera being released, it still has the magenta corners issue on new cameras. There are obviously plenty of cameras out there without this problem, or at least so minimal that it's not much of an issue, but it seems that being prepared for repeated RMA rounds is part of the process to expect and factor in when buying this camera. I had to go through 4 before I got one reasonably clean. And even then it's not perfect, but for the money, it's freaking amazing and well worth the initial hassle.
The main thing is don't let BMD be the judge on wether a camera is good or not. The second camera I received from them had been fully tested out and I was told it was a great one. Yet it still had just as much magenta and I sent that back too. I actually think that part of the reason this camera is priced so relatively low for the potential of what it can deliver is that no intensive quality control is put in beyond just the basics. More through quality control is left up to the consumer so they can actually deliver a camera at that price. Higher end cameras reject a very high number of chips that actually make it into their cameras. It's likely that if the same level of scrutiny was put in on the UM4.6 chip, the cost of each camera would skyrocket. It's just a guess, but there must be some reason that different performing UM4.6k cameras are out there. It's really hard to imagine that BMD really doesn't know what's causing the issue.

At this point I'm on the sidelines with this camera. I'm hoping that for the next generation they adopt a back illuminated sensor.

Shift your tint from 10 to zero and retest. I've never taken my tint off of zero and always do a custom white balance and don't seem to ever see it in real world shooting.


Sent from my iPhone using Tapatalk

Shift your tint from 10 to zero and retest. I've never taken my tint off of zero and always do a custom white balance and don't seem to ever see it in real world shooting.

Ok, thx, I will retest tomorrow. I tested the cross hatching and will publish tomorrow.
Bye

Hi,

In 50 winter I had never seen magenta snow like this before..
In desperation, I had tried the 4k and UHD window on, and magenta vignette is still there.
I am sure I can do a test very quick now, why not just drive to NY, get friend with someone at B&H, gently open a box of UM46, insert the card and the lens, shoot a white board and check the dng.
It is a dream but the reality is I will have to return it from Canada, pay custom$ and transport$ etc etc. I don't know if I will take chance for replacement. I am not a casino player.

Regards,


4608x2592
FW4.0
raw: lossless
Sigma 18-35mm f/1.8 DC HSM
iso 200
k5600
tint 0

35mm F4.8

sigma35f4.8.jpg (573.6 KiB) Viewed 17089 times

35mm F8

sigma35f8.jpg (260.36 KiB) Viewed 17089 times

You don't pay customs and taxes on a warranty repair, but I just learned a few days ago that your shipping information that you send must clearly state it is a RETURN for REPAIR UNDER WARRANTY. If you just said RETURN FOR REPAIR as I did, Canada Customs will charge duty!


Sent from my iPad using Tapatalk

You don't pay customs and taxes on a warranty repair, but I just learned a few days ago that your shipping information that you send must clearly state it is a RETURN for REPAIR UNDER WARRANTY. If you just said RETURN FOR REPAIR as I did, Canada Customs will charge duty!


Sent from my iPad using Tapatalk

Ho, thx Rick.
I sure will word it like this.
Withs customs we always have to mesure each words and align them in the right order. Not to few not too mutch. It is an art.

And it's a black art that you only learn by experience, no one has a clear set of instructions!


Sent from my iPad using Tapatalk

Have you noticed any crosshatching? In my unscientific observation (simply noticed after working with 3 cameras), the ones with stronger magenta corners have almost no crosshatch. The ones without corners have very strong crosshatch.

Hi Stephen,

I did not find crosshatching on the unit I have right now, but I am not sure if I tested it right.

It's impossible to say anything definitive from a single JPEG image.

Read through Gary Yost posts in the thread linked below for an clearly detailed explanation of how to test for the issue:

https://forum.blackmagicdesign.com/viewtopic.php?f=2&t=50175&p=290768&hilit=yost#p290768

So I have updated the firmware to 4.2 and done a Black Shading/Sensor Calibration and there seems to be an improvement to the Italian flag(asymmetrical shading) effect on my camera.
So then I did an auto white balance and shot a white corflute board, in raw 4.6K with my Sigma 18-35 at 35mm and f8. Lighting is via a warm white led on the ceiling of my living room.
At this low lighting level, I increased shutter to 360 and ISO to 1600.

UM4.6K fw 4.2.jpg (930.77 KiB) Viewed 10427 times

I flipped the camera 180 degrees but it made no difference to the waveform.
No adjustments in Resolve.

This seems to me to be a vast improvement and I am assuming it is due to the new Black Shading feature.
The shading on the right is so negligible that it does not concern me.
Link to DNG attached.
https://drive.google.com/open?id=0ByCkZ_iRaK0_Skdyck9RcFJ3WW8

Hey Paul Thanks for sharing the DNG.

Can I ask the first four digits of your serial number? Thanks!

Also just to confirm, that DNG was taken with the camera upside down?