V-Raptor Sensor Issues

[Patrick Tresch]

I just think anyone that's using this camera on a professional level should know if they put something like a Tribe 7 Blackwing on and shoot some flares for a music video it's very likely to show the sensor corruption.

So you think it's on all Raptors due to design flaw or just some faulty ones? I think there where a lot of bodies sold and I see only your sensor that has this problem.

 

[Mike P.]

I think that's because RUser is quiet these days (pretty sure Facebook groups have more)... Also Vicky started another thread showing it too. The YM article isn't based off their units, either (unless they work for BP40?).

Also Phil confirmed their looking into it, as did Jarred (which I presume means they could replicate it, or no?)

 

[Phil Holland]

So you think it's on all Raptors due to design flaw or just some faulty ones? I think there where a lot of bodies sold and I see only your sensor that has this problem.

I think that's because RUser is quiet these days (pretty sure Facebook groups have more)... Also Vicky started another thread showing it too. The YM article isn't based off their units, either (unless they work for BP40?).

Also Phil confirmed their looking into it, as did Jarred (which I presume means they could replicate it, or no?)

There are 3 threads here on REDuser, 3 posts as far as I know on Facebook. The difficulty is tracking down all of the information and where it's coming from and responding accordingly. Especially digging through posts everywhere to find any information. When it was just REDUSER or even just here and the RED Camera Facebook group, things were a lot, lot easier to find useful information quickly. Not complaining about that, it's just a side effect of the expansive market, audience, and internet/social media growth.

It's being looked into. That's the last official word I have from RED. I still have my camera that exhibits this, awaiting their findings really. I reported it to RED pretty early on. I know it was difficult to recreate on some cameras and there do seem to be cameras that don't exhibit this artifact. Without me testing every single camera in the world it would be impossible to really know. All components are tracked, batches and such too in my previous experiences with RED, ARRI, and other camera companies.

I shoot flares and into hot sources a lot, so it is concerning to me, but I have also been filming with the camera. I suspect we'll get official word soon if somebody has sent their camera in. That to me sounds like they now what to do to fix it.

The second I know anything or likely anybody from RED does I'll post, maybe Jarred will post, fireworks will go off, and the world will spin again.

 

[Charles Bergquist]

So you think it's on all Raptors due to design flaw or just some faulty ones? I think there where a lot of bodies sold and I see only your sensor that has this problem.

Hard to tell at all, you can only go by what others report. If it varies from camera to camera, it’ll be something that affects the perception of V-Raptor on set. I’m hoping that mine is just a mismatched sensor or a bad calibration, mine will go in next week for service with Red. The others that are out there, could just be the calibration as well.

Monstro and LF do utilize the same technique though no one has seen it yet or posted images, so it could just be a calibration issue.

 

[Michael Lindsay]

doesn't seem to happen (at all or as much) with lenses that are more image plane side tell-centric...

 

[Kemalettin Sert]

I ordered mine last month and it's probably gonna get delivered in 2 weeks.I trust RED and i'm sure they'll do their best to fix this issue.

 

[Chris McKechnie]

It’s definitely going to be interesting to see how RED handles this debacle. They do things their own way…so only time will tell.

When I had issues with previous red cams..I was sort of kicked to the curb and said it is what it is, and you’re seeing more than you should because the sensors are better. I’ve owned almost every single red cam…and in my opinion, they just aren’t made like they used to be. R1 and the Epic MX/dragon cams are still my fav reds ever.

I know they’ve sold a crap ton of these raptor cams…so they are gonna enter another s-storm if they don’t handle it right. Maybe they rushed it too quickly. I highly doubt it’s a software fix though. Definitely won’t be renting one anytime soon though for any project until it gets all sorted out.

 

[Charles Bergquist]

I’ve owned 6 Reds, AKS and a Monstro. Definitely felt like the first time I brought it up I was brushed off. I think they’re taking a different approach now that people see it and it’s circulating in the public. I just hope it confined to only a batch of cameras.

Every project I’m doing now I try to find a way to work in some fun flares, especially with the BlackWing X-Tunes. We’ll see, definitely hopeful at the moment. I do think the Red community needs to follow Red a little less blindly and make sure they get what they pay for.

The unfortunate thing about all this, is the V-Raptor is an incredible camera…they just have to sort out this sensor issue.

 

[Thomas Winston]

Charles - I looked at your samples. The issue is obvious.

Am I correct that the easiest way to show the line is with no lens on camera and pointed at a light source.

That would mean any talk of lens adapters or specific lenses is mute.

Correct?

Or am I missing something?

I'd hope RED clarifies that a redirect to lens adapters or lenses is not appropriate.

Thanks!

Thomas Winston
www.grizzlycreekfilms.com

 

[Phil Holland]

Am I correct that the easiest way to show the line is with no lens on camera and pointed at a light source.

Yes Thomas, that is the easiest way to see the seam on all the cameras that have seams. You take a flashlight, ideally a dimmed one btw, and graze the surface of the sensor with light and look in monitor. But that wouldn't be a the best way to determine if this is an issue however due to optics and light rays. Meaning you can produce the seam without a lens and never see it with a lens.

The issue in this case is for some the seam can be seen with under certain lighting conditions when a lens is attached. Another added variable is the lens itself.

My sensor does exhibit this, but not on every lens with the same controlled testing source. That to me provides some useful hints on what's going on btw. This is likely why RED initially stated that it could be tied the lenses and/or lens adapters (tolerances are a sticky widget and the importance of correct FFD can also be important).

If I had to theorize, it's either due to something pertaining to the sensor, sensor coverglass, or filter itself. I've produced test results that make me think it might be more of a filter/OLPF related issue, but it's very hard to tell. I have forwarded all of this to RED weeks ago now. The sensor itself? I'd say that's oddly the least likely culprit.

There also appears to be cameras out there that don't exhibit the artifact at all. Which if it is the filter/OLPF that could equate to just a batch issue. But again, all theory until an official words is given.


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

Making a clear separator here because I want to point out something else that should be of use to you all and isn't tied to this. But in an effort not to complicate things I have made a separator.

Regarding some adapters and interchangeable mounts from 3rd parties, and this is something that expands across all cameras, so be on the lookout. Some adapters as a whole do have "unique optical properties", but that's not related to this. But the 3rd party adapters and even interchangeable mounts can produce an interesting landscape of variables such as potentially added light bouncing around inside the sensor chamber. Have this issue on both an E-Mount to EF and E-Mount to PL, their issue mainly manifests as a very dark ring which may initially could be shrugged off as vignette, but you can see a shape to it and it's not a smooth gradient. Some adapter have increased vignetting, an impact on bokeh, a subtle loss in light transmission, and other oddities. Which is why you occasionally get the "this never happened on camera X with this lens" sort of thing.

I don't want to set the world on fire with this, but take into account light rays, various lens design related things.

I also have a host of theories and side by sides that make me think it has something to do more with the ID of the mount itself, but it's also more likely specific adapter or mount related. I have nearly every E-Mount to whatever mount and RF-Mount to whatever you can think of here.

Some numbers to keep in mind.

Inner Diameter of native lens mounts (this is not the Flange Focal Distance, but rather the diameter of the "hole" light travels through):
- Arri PL Mount = 50mm
- Arri LPL Mount = 62mm (between this and the FFD of 44mm, this is why Arri made this mount btw)
- Canon EF = 54mm (most don't adapt PL directly to this, but rather make a bolt on PL adapter if they need to use an EF cam with PL)
- Canon RF = 54mm
- Nikon Z = 55mm
- Sony E-Mount = 46.1mm
- Leica M-Mount = 40mm
- Leica L-Mount = 51.6mm
- Fujifilm G = 65mm

Consider how and where light rays project out of the rear of a lens for a given focal length and specific lens design, two of those numbers is very different than the other in potentially "there might be an issue" sort of way. Not something that impacts native mount glass, but something to look out for particularly when going from PL to any of this. Particularly when it comes to using VV or LF PL glass on these cameras that feature imagers of this approximate format size as that's where I've seen gremlins hiding in the brush. I'd rather see E-Mount be between a 54-56mm ID. I suspect this has a lot to do with the timeline of when these were invented and how or what the purpose of the mount itself is. The other variable is obviously testing all lenses from other mounts, which isn't exactly the focus when coming up with a new proprietary mount you want people to buy your glass with.

FFD versus ID likely also plays a role here, i.e. where the "hole" is from the imager as well as how much extra metal sits within the ID of the hole.

For instance, I love the look of my Zeiss Otus on the Sony A1, but I will avoid backlit conditions with 85 and 55 at deeper stops as that can induce an artifact. A subtle one, but an annoying one.

In a perfect world for manufacturers, you just use their native mount lenses and nothing else because that is a controlled ecosystem. But for creative people who want to use a lot of glass from a lot of eras and mounts, this is something to be on the lookout for.

 

[Charles Bergquist]

...trying to get as much proof out in the world as possible. Here's an example from a Mini LF with the LPL Mount fully removed, bare sensor. You can see the sensor split here, so it exists on Mini LF with the LPL mount OFF ( Bare Sensor ). The interesting thing is with the LPL mount ON there is absolutely no center line split. WIth he LPL with PL adapter, there is NO center line sensor split. So, it appears that by Arri controlling the mounts, they've eliminated the sensor split issue. Again, the easiest way to see the split on a camera is no lens and no mount. When you put a mount on and it does or doesn't exhibit it, that's when I think we need to be concerned. Have not seen this on Mini LF or Monstro with a lens mount on, only Raptor. It could be the RF mount and built in OLPF and it's depth causing the sensor split issues. I'm not an engineer in the slightest bit. So, conclusion is, it exists with a Mini LF with NO LENS MOUNT, with Lens Mount of any kind and with a lens, it doesn't present itself. This could be due to a whole much of stuff, Mini LFs have ND / different OLPFs, etc. With this in mind, I think it's a fixable thing if Red is able to rework their OLPF or Mount System for existing V-Raptor customers. RF mount might need to go and we might need to go to Red Approved only PL / EF / LPL mounts, which I'm all for if it works.

 

[Charles Bergquist]

Here's an example with the LPL Mount with no lens under the exact same conditions. So, we're starting to see a picture here that the Mini LF with their LPL Mount or even PL Mount has controlled the light and deflected it from making a sensor split happen. This doesn't happen under the V-Raptor V1 ( ST / BLK ) with the Canon EF Mount and the Wooden Camera PL or with lenses on either. So, it might be a thing where it takes some time to come up with a way to find a solution, but I would imagine it's either a Re-Mounting thing for V-Raptors or Red will make their own approved PL / EF or LPL...which I'm all for.

 

[Charles Bergquist]

For comparison, here's V-Raptor with RF-EF Adapter (Canon ) with a 135 F2 @ 2. Screen grabbed so everyone can see the differences.

 

[Phil Holland]

Sharing some of my notes here with you Charles on the LF and Mini LF. Briefly pushing Raptor aside here and that is something that needs looking into for reasons shown and discussed.

Arri LF/MLF - Wide through 85 primes, glass for FF35 and larger formats, and particularly anamorphic and flare conditions, i.e. flashlight in frame ideally a small point source..

With no mount on camera light rays are coming in from lots of angles potentially, with a lens mount on (you've also created an aperture, but that's not a factor), but flocked off stray rays and angles of light. i.e. Same conditions as shown, but light rays are entering in "not the same" fashion. However lenses protrude into the mount and their exit pupil exists within, which is where you may or may not see this on LF.

Not exactly the most visible thing on Signature's due to them being fairly telocentric, but take something a bit less so and things can show up.

I've tested lenses physically the most on Dragon, Monstro, recently V-Raptor, Arri LF (full sized), and Venice the most. In the case of RED and ARRI I'd say it's been the most aggressive testing by a mile. All the glass I own, some I don't, and pretty much a whole inventory of a rental house nearby have been on RED and ARRI cams.

Hope that helps a bit in the sleuthing. Arri calls it out in their manual for exactly the reasons I'm describing really, i.e. "rare situations". I also think the variable of back focus on some lenses could play a factor making it about 1000X harder to realistically come up with playbook.

One fun thing to examine is if and when you do see it, add some diffusion in the mix, like 1/8th Promist or similar. See if that source and lens combo are still creating the issue.

I feel Raptor's issue is "not of the same cloth" and is a tech issue, likely optically, somewhere in the chain which I posted above.

 

[Charles Bergquist]

Phil Holland , I think you’re probably onto something with it being a sensor cover glass or OLPF stack issue. I actually shot my Monstro with no OLPF for IR work a lot and flares are out of control without it. In all the footage I never noticed a sensor split. Obviously, my Mini LFs in use I’ve never seen it until I fully removed the mount.

As much as I like the 3rd party stuff, I think this is an issue with it possibly. The Wooden Camera PL and the EF Canon are approved but still can create a sensor tear / glitch or image corruption, however it wants to be said.

Overall, I think everyone needs to remember, Red is fallible just as Arri or Sony. I think it’s smart to keep a critical eye on everything so not only do we as cinematographers or camera owners not get screwed but the manufacturers see the issue and rework it to a better camera. If there’s a V-Raptor recall for a OLPF / cover glass which did happen with Dragon OLPF, it’s for those that opt in. Red fixed that problem and actually made it a feature people used ( removable OLPFs ). Oddly, my most reliable cameras are my Vision Research Phantoms.

 

[Phil Holland]

fallible[/QUOTE]

I agree. And truthfully with the top 4 companies (Panasonic was good at this a while back too, maybe 5 then), if you can present something in the right way (and sometimes to the right people, very hard to do this with overseas companies really) in order for them to respond accordingly they do listen. I don't know how many bugs or feature requests I've submitted over the years, but I do know my feedback has been well received and has changed some aspects of the cameras we use to make a living and even far lesser cameras.

Right now it's more of a waiting game to see what the actual answer here is for V-Raptor owners experiencing this. I am in that category. I'm curious what is updated on your camera. I'll be sending mine in when I can for sure.

Too early for me to call it on what the most reliable camera is for me (some stuff is too new), but I have been around the globe a few times with DSMC2 bodies in absolutely horrific conditions at times. But that's years of fimmaking adding up. To that point, in my mind that puts the bar for DSMC3 really, really high. But I'm also aware that early adoption of pretty much every top tier system comes with sharp edges.

It'll get there. Raptor as a concept and platform is a hell of thing. IQ outside of the seam is alarmingly high. Sort this little bit up and there's nothing close to it on the market. Not for a long while either.

 

[Mattia Palombi]

For comparison, here's V-Raptor with RF-EF Adapter (Canon ) with a 135 F2 @ 2. Screen grabbed so everyone can see the differences.

Thanks for sharing. The imperfection it's quite obvious. Incredible RED didn't test the sensor properly before releasing the new camera. Hopefully this will get fixed asap.
Please keep us posted. Ciao.

 

[Matt Maniego]

Not to beat a dead horse, but I did a test myself today and found that my Raptor is exhibiting the same thing. Using a DZO RF to PL adapter and a 32mm ARRI Ultra Prime. Test was done on my dining table and used a maglite flashlight to shine into the lens haha Just wanted to see if i could reproduce the artifact easily, and I gotta say it was pretty easy to reproduce. Will be reaching out to RED first thing in the AM.

 

[Thomas Winston]

Phil and Charles,

This is all super helpful and informative.

Thanks so much for taking the time!

-Tom

Thomas Winston
www.grizzlycreekfilms.com

 

[Phil Holland]

So um..... I have an idea.

If you do have this issue, please let me and us know what lenses you are using and VERY specifically what mount adapter you are using.

Did an aggressive amount of testing last night. And I have a lot of mounts here, noticed something rather specific. Too early to tell, but the direction I'm pursuing here is stray light bouncing around within some adapters. Which would also somewhat echo my LF experiences to a degree.

I'm sure RED is doing a lot of this, but we have numbers and variables that their engineers might not be able to recreate.

So if you can, these things are the questions I have:

- Did you experience this with a native RF lens?
- Did you experience this with a mount adapter?
- If on a mount adapter what lens?

Thank you in advance. Post this here so people who need to see it can see it FYI.