V-Raptor Sensor Issues

[Ryan Chernin]

Dude….. yes!

I knew your name looked familiar. Small world, let’s chat offline. What’s your IG?

Wild!! Sent you a PM!

 

[Mike P.]

Oououf. Yeah, it's like a dead pixel: you might not notice in passing but as soon as you do see it, you can't *un*see it (especially on that clip; it's basically centre face, which is where an audience's eyes would go to by default).

Do we think this is going to get resolved or will RED brush it under the rug? I love most things about my raptor but have seen this in a few shoots I've had at this point. It's tough for me to bring this on set with this current flaw.

Tough to say given RED's track record... I mean, REDusers have been dealing with aperture reflection "red dots" at medium stops (as low as T4 on some of my lenses) with literally every RED camera/sensor to date (except Raptor) and RED's response was always 'just don't shoot like that'. Similarly, they ignored/denied that their ISO was off by a full stop for almost a decade, and when they did fix it they left out older cameras (even though they weren't officially EOL yet). That said, when the OLPF fiasco happened with DSMC1 they did have an optional recall to get it the interchangeable system installed.

Something tells me if it can't be reduced via firmware/debayer tweaks in the SDK or user installable flocking/gasket, they'll do the 'it doesn't happen often/don't shoot like that/real pros use a matte box and control flares' until their next hardware launch (kinda like how they handled Dragon's penchant for CMOS Smear... or timecode issues in general).

 

[Ryan Chernin]

Oououf. Yeah, it's like a dead pixel: you might not notice in passing but as soon as you do see it, you can't *un*see it (especially on that clip; it's basically centre face, which is where an audience's eyes would go to by default).



Tough to say given RED's track record... I mean, REDusers have been dealing with aperture reflection "red dots" at medium stops (as low as T4 on some of my lenses) with literally every RED camera/sensor to date (except Raptor) and RED's response was always 'just don't shoot like that'. Similarly, they ignored/denied that their ISO was off by a full stop for almost a decade, and when they did fix it they left out older cameras (even though they weren't officially EOL yet). That said, when the OLPF fiasco happened with DSMC1 they did have an optional recall to get it the interchangeable system installed.

Something tells me if it can't be reduced via firmware/debayer tweaks in the SDK or user installable flocking/gasket, they'll do the 'it doesn't happen often/don't shoot like that/real pros use a matte box and control flares' until their next hardware launch (kinda like how they handled Dragon's penchant for CMOS Smear... or timecode issues in general).

If it is an adapter related thing, as Jarred seems to be suggesting, then they might abdicate responsibility for the artifact altogether, passing the buck on to the third party manufacturers of lens adapters. If this winds up being the case, then they're probably using this time to fine tune their own EVND PL adapter, which they'll then market as a surefire way to avoid the artifact. And with the XL using bolt on adapters, it'll likely avoid this issue altogether (and if it is showing the artifact now, then Red will make sure it doesn't by the XL's release date). That's just my somewhat pessimistic prediction based on the trends of capitalism.

But.. didn't someone earlier in the thread mention that they saw the artifact with an RF mount lens?

 

[Steven Schloss]

Trying to be optimistic that there will be a new mount solution, either aftermarket or from RED themselves that will address it. I agree that I do not see RED taking back cameras to do sensor repairs/replacements

Looking at the new Strata Mount by Kippertie... seems like because it's build is a "bolt in design" it may help with the issue if it is caused by light bouncing around in the adaptor. There is also a video of how the adaptor works and it seems to mimic the DSMC2 mount design.

In an article they write "In order to achieve high optical performance, the inside of the adapter features optical baffles and flare traps that should minimize internal reflections."

Full article below:
https://www.cined.com/kippertie-stra...-red-v-raptor/

Facebook video of mount in works:
https://www.facebook.com/watch/?v=329441629069297

Response from Kipper tie
"we’ve not been able to provoke this issue with the Strata mount. It is very hard to prove a negative however, so there could still be specific circumstances in which that may occur."

Any one else seeing this thinking the same thing? Since they are in cahoots with RED this may be something that has been in the works behind the scenes.

Thoughts??

 

[Michael Lindsay]

...I mean, REDusers have been dealing with aperture reflection "red dots" at medium stops (as low as T4 on some of my lenses) with literally every RED camera/sensor to date (except Raptor) and RED's response was always 'just don't shoot like that'.

i thought the responce was shoot skin tone? ... that does work!

Something tells me if it can't be reduced via firmware/debayer tweaks in the SDK or user installable flocking/gasket, they'll do the 'it doesn't happen often/don't shoot like that/real pros use a matte box and control flares' until their next hardware launch (kinda like how they handled Dragon's penchant for CMOS Smear... or timecode issues in general).

i hope you are wrong! I feel they need to make this better by any means... Monstro and Dragon W didn’t have these issues (ok I have heard one example but I have never been able to replicate myself... )

i don’t actually need it to be 100% perfect... just nearly as good as a Monstro which also has a stitched sensor. Ryan’s clip above was one of the worst example for me as it was very far from a synthetic test and it was with a signature lens which is pretty telicentric. (Though I do feel the Wooden camera or c7 mounts are far from ideal... maybe Red should make these?

Lots of us shoot into big back lights all the time... the love of flares isn’t going away soon SO I don’t think it would be acceptable from a Red POV.. to leave as is.

On a more hopefully note has anyone diy’d the mount into a much better state?

I have some new flocking material coming this week and will repost back here if it works..

 

[Ryan Chernin]

i don’t actually need it to be 100% perfect... just nearly as good as a Monstro which also has a stitched sensor. Ryan’s clip above was one of the worst example for me as it was very far from a synthetic test and it was with a signature lens which is pretty telicentric. (Though I do feel the Wooden camera or c7 mounts are far from ideal... maybe Red should make these?

Lots of us shoot into big back lights all the time... the love of flares isn’t going away soon SO I don’t think it would be acceptable from a Red POV.. to leave as is.

On a more hopefully note has anyone diy’d the mount into a much better state?

I have some new flocking material coming this week and will repost back here if it works..

Yeah it's pretty annoying. I wound up cutting around a lot of the affected footage. In hindsight, it's not bad enough to sell or stop using the camera. But now it's something we'll always have to be conscious of. And it's something I'll always notice when it shows up.

I showed the artifact to a director and he didn't care at all. I think it'll only be any issue on a case by case basis. Some directors will be more meticulous. Owner/operators will notice, and it'll be a tiny cloud hanging over their creative process. But not a complete "dealbreaker."

In my opinion, Raptor isn't priced "premium" enough for Red to address the artifact in the form of a recall. Like everyone's saying, it'll probably be a "don't shoot that way" situation. They'll make note of it going forward, and future products won't show the artifact. At this quality/price ratio, everything's a beta until it happens to work well enough. I wish it wasn't that way, because 25K is still certainly a stupidly large chunk of change for a tool. But in the world of 90K Alexas, Raptor IQ holds a strong candle to the LF, but it's everything else about the LF that warrants the price tag. I haven't checked, but I have doubts that Alexa owners are on their own forum stressing over possible quality control issues, haha. That being said, most of us aren't in $90K+ of debt just to own a high quality cinema camera. It's a give and take.

 

[Mike P.]

i thought the responce was shoot skin tone? ... that does work!

Pretty sure the only reason that helped was because STH was noisy at even mid-ISOs and took a stop from the bottom and put it on top... Meaning you didn't need to stop down as much in bright scenes. But yeah, it still happened on STH far more (and at far lower apertures) than Alexa or any segment Sony/Canon/BMD (Personally, I've never been able to duplicate it on any of those cameras).

i hope you are wrong! I feel they need to make this better by any means... Monstro and Dragon W didn’t have these issues (ok I have heard one example but I have never been able to replicate myself... )
...
I have some new flocking material coming this week and will repost back here if it works..

Yeah, I hope so too. The problem is it's been months and aside from acknowledging it, mums the word. If it's not fixed soon, the damage will be done (in the same way that everyone thought R1s still overheated long after it was resolved on Che). I think everyone is pointing to the mount because that's the major difference between Monstro and Raptor. My concern would be that it's the gap between the OLPF and the sensor on DSMC2 (and presumably LF, since I think its mount-platform "flange-distance" is similar) that makes the split almost impossible to replicate on those cameras. Conversely earlier in the thread, someone pointed out it's very easy to replicate on *any* stitched sensor -- LF, Monstro, Raptor -- if you don't have a lens on them.

Looking at the new Strata Mount by Kippertie... seems like because it's build is a "bolt in design" it may help with the issue if it is caused by light bouncing around in the adaptor. There is also a video of how the adaptor works and it seems to mimic the DSMC2 mount design.

In an article they write "In order to achieve high optical performance, the inside of the adapter features optical baffles and flare traps that should minimize internal reflections."

Response from Kipper tie
"we’ve not been able to provoke this issue with the Strata mount. It is very hard to prove a negative however, so there could still be specific circumstances in which that may occur."

Any one else seeing this thinking the same thing? Since they are in cahoots with RED this may be something that has been in the works behind the scenes.

Thoughts??

I don't think they're in cahoots or anything, but I also have a hard time taking the use of 'you mileage may vary' type language without a huge grain salt/skepticism... Like does 'we haven't been able to *provoke* it' mean they haven't been able to duplicate it at all with their mount, or that it's *no worse* than stock-mount Raptor? If it was just flocking/mount~adapter material, you'd think RED would've solved it by now.

On the filpside, if John/KipperTie confirms their mount doesn't exhibit it, he's not lying/exaggerating. Every interaction I've had with him over the years (general RED questions, questions about his products, or purchases) have been super smooth. He is insanely transparent, straightforward, and has excellent communication (Apparently he doesn't sleep either because he would respond to msgs [posts, emails, or DMs] within ~30mins regardless of the time of day, and in spite of being in a completely different time zone/country)... All that is to say, I don't think he would sugarcoat it.

...Plus that mount looks solid af compared to any adapter that uses the RF mount... Shimmble too!

 

[James Darren]

It seems strange that by using a RF adapting mount is the cause of the sensor split issue. To me the direction of the light hitting the sensor & lens tele-centricity are the main causes.

Regarding the KipperTie ND Filter mount, if this mount doesn't show the issue, I wonder if the glass filter is somehow "filtering" the issue. Hard to imagine but it's a possibility.

 

[Phil Holland]

It seems strange that by using a RF adapting mount is the cause of the sensor split issue. To me the direction of the light hitting the sensor & lens tele-centricity are the main causes.

Regarding the KipperTie ND Filter mount, if this mount doesn't show the issue, I wonder if the glass filter is somehow "filtering" the issue. Hard to imagine but it's a possibility.

My Strata arrived. I'll be testing it next week. Built like a tank!

I'm not speaking for RED, but I've done a lot of testing. My solution was to flock the mount and/or build a light shield as the issue pertains to "stray light rays", which is why the mount can play a roll. Each mount is designed different and you do have a fair bit of space between the PL flange and the sensor. And yes, this is why I've been asking what mounts and lenses people are using to exhibit the artifact. I've purchased literally all of them to figure out what works. And in a few cases made suggestions to improve the mount.

It can still occur in other rare circumstances. But certainly less so. Which is why you see me blasting light into my V-Raptor's eye with confidence.

 

[Ryan Chernin]

My solution was to flock the mount and/or build a light shield as the issue pertains to "stray light rays", which is why the mount can play a roll.

If I say that you hold zero responsibility for any damage I might do, any chance I could beg you to share your method for this? Wondering if I could improve the Wooden Camera RF to LPL

 

[Ryan Chernin]

Phil Holland might something like this do the trick? https://www.amazon.com/dp/B087PDD53...t_i_EJ7WCWVC497XDGSV19A3?_encoding=UTF8&psc=1

Or best to use adhesive and flocking powder?

 

[Ryan Chernin]

Phil Holland might something like this do the trick? https://www.amazon.com/dp/B087PDD53...t_i_EJ7WCWVC497XDGSV19A3?_encoding=UTF8&psc=1

Or best to use adhesive and flocking powder?

Ok folks, so, update (since Phil is either busy, can't comment, or hates me, haha). The Amazon linked adhesive flocking tape seems to be helping. Got a shoot tonight in the same location as the vimeo clip, so I'll get a chance to test the exact lighting conditions with the added flocking.

 

[Mike P.]

It’ll be interesting to see if the flocking reduces flares in general (and that’s why it “improves” the split from being visible), or if it actually just reduces the visibility of the split while retaining the flares.

Anyway, keep us posted!

 

[Ryan Chernin]

It’ll be interesting to see if the flocking reduces flares in general (and that’s why it “improves” the split from being visible), or if it actually just reduces the visibility of the split while retaining the flares.

Anyway, keep us posted!

Will do. When I did a preliminary test this afternoon, I was still getting some pretty gorgeous flares shooting direct sunlight through a window. In fact it may have aided contrast while flaring. Hard to say without a control group. But I believe fewer internal reflections can have that effect.

 

[James Darren]

Got a shoot tonight in the same location as the vimeo clip, so I'll get a chance to test the exact lighting conditions with the added flocking.

It should be pretty easy to test & recreate without being on an actual shoot. Just get a fresnel style or telescopic spot light beam (doesn't need to be even that powerful) and put it on a stand.

Use a neutral colored backdrop, underexpose a bit and pan through the light source.

 

[Ryan Chernin]

It should be pretty easy to test & recreate without being on an actual shoot. Just get a fresnel style or telescopic spot light beam (doesn't need to be even that powerful) and put it on a stand.

Use a neutral colored backdrop, underexpose a bit and pan through the light source.

I know. I've tested it under such circumstances when the artifact first became known, but I'm a little more concerned with "does it show in my real shooting scenarios" at the moment (until there's word on an official fix from Red).

I've had three shoots since flocking, lots of center frame flares that should have triggered the artifact, but everything looks good. New problem: some vignetting at 8K. Will have to try the adhesive and flocking powder method to see if it makes a difference.

 

[Mattia Palombi]

Alright, so I tempted fate by joining the "this is hard to replicate" side. Finally happened In a real shooting scenario on set. Pretty much present throughout an entire take while light flared the lens.

https://vimeo.com/673066563/45c75fcefc

Wooden Camera LPL to RF adapter and a 40mm Signature Prime


*Edit: upon further inspection, you can see the artifact anytime there's a substantial lens flare passing through center frame. James is right!

Heartbreaking. Thanks for sharing. It's so frustrating but the sensor doesn't seem reliable at this stage. I'm not surprised to see V-Raptor owners selling off. I truly hope RED is going to address and fix this issue sooner than later. My biggest question is why this is not happening on Monstro or any other FF Sony or Canon cameras. Please let us know if you will be able to send your V-Raptor back for "repair". Cheers.

 

[Ryan Chernin]

Heartbreaking. Thanks for sharing. It's so frustrating but the sensor doesn't seem reliable at this stage. I'm not surprised to see V-Raptor owners selling off. I truly hope RED is going to address and fix this issue sooner than later. My biggest question is why this is not happening on Monstro or any other FF Sony or Canon cameras. Please let us know if you will be able to send your V-Raptor back for "repair". Cheers.

I think we all have a tendency on here to get amped up over these QC issues. Partly because we know Red's listening, we think that if enough of us make a ruckus, only then will Red address the issue.

But I'm very happy with the camera, and while I considered selling in a moment of frustration, personally I don't really see any benefit at the moment. It's performing beyond my expectations, and it just, well, makes me happy. All of the footage I've shot on it puts a smile on my face. That is by no means a scientific metric! Everyone's personal experience and expectations will vary.

If and when Red announces some sort of repair option for the sensor split artifact, I'll send it in.

 

[Charles Bergquist]

Now that more people are seeing it on important work, I feel like Red has to address this. Which I feel is sensor glass redesign and recall. I’m going to send mine in finally, but they’ve said there is no solution though it was said they were working on it. I feel like now is the time to be transparent. Also, the option is just sell it all and stick with Arri. I started on Red built all the way up to a Monstro for a while. I’m just disappointed with this reaction by Red. My take, you may not see it now but you’ll see it on something, just hope it’s not important if Red doesn’t offer a fix here immediately.

 

[Audy Erel]

I'll be ready to purchase my next cinema camera gear by the end of this year, I was on the fence about buying the V-Raptor up until this sensor issue thing surfaced.. Now I'll wait and see how RED handle this situation before making any decision. Curious about the upcoming ARRI offering but I doubt the budget will be friendly to me.