Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

OLPF v1 and v2

Thanks Roger, (In these examples is v1 olpf not during daytime whereas version 2 is during daytime? Trying to get a handle on that window light in v2 vs. no window light in v1).)

The difference in low end info is immense. If you're shooting in dimly lit conditions, version 1 is for you. Remember to work around the orbs. The low end color on the "lowlight olpf" is far superior to the low end on the "highlight olpf." The highlight olpf version has a green tint in the low end where the lowlight olpf maintains a neutral color. As we all know, lowlight olpf is far superior for lowlight (just watch out for the artifacts), and the highlight olpf is king for daylight and bright settings. The in-between olpf is where it's at and most likely the sweet spot tradeoff between the two.

(right click, open in a new tab to view larger version)
OLPF1vs2.jpg
 
Hi Matt,

I bet you could make a very acceptable image out of both of those r3d's without much effort. :)

Even with the moderate lowlight advantage, Whenever I ever think about old OLPF I just look at this test that Phil did. For me, and the situations I encounter, avoiding artifacts like this is far more important. But I appreciate that folks would rather stick with the older one for their style of shooting.

Battistella


phfx_redDragon_NewOLPF_FlareSun.jpg
 
David Battistella said:
Hi Matt,

I bet you could make a very acceptable image out of both of those r3d's without much effort. :)

Even with the moderate lowlight advantage, Whenever I ever think about old OLPF I just look at this test that Phil did. For me, and the situations I encounter, avoiding artifacts like this is far more important. But I appreciate that folks would rather stick with the older one for their style of shooting.

Battistella


phfx_redDragon_NewOLPF_FlareSun.jpg
Click to expand...

Thanks David ;)

I agree, the artifacts killed the lowlight olpf for me. But if it didn't have those artifacts, look out. We wouldn't even be having this discussion.

Even with the lowlight olpf (referred to as the old olpf or olpf 1) the comparisons between Dragon and MX showed Dragon had more DR on both ends. I think for the majority of users that was the best case scenario. Now with the highlight olpf (also referred to as the new olpf or olpf 2) we gain a couple stops over MX in the highlights but lose a bit to the MX on the low end. Not to mention the increase in the red noise. While the increase in highlights is desirable for some (due to the fact it's got a more film like rolloff) to be honest I didn't think the roloff on the old olpf was that much worse. It was better than MX so that was good.

Anyways, we all could go back and forth till the end of time. Shoot what you want based on what you are looking for. I believe that is the key. I personally look forward to the inbetween olpf so we get back some low end and keep some extra stuff up top without the artifacts of either new or old olpf (I consider the excessive red noise with olpf 2 to be an artifact). If it has no red noise, no sensor pattern flare, no orb flares, and retains some more low end info, we will be set.
 
Matt,

I think you nailed it. Shoot what you like. Everyone gets to choose and that is a good thing. I like that RED is responding by giving people a choice.

Your tests show that you can get very decent results with both and that bit of green in the V2 could easily have been graded out. You posted exactly what I did when I saw those two R3D's I loaded them both up and went through the range, up to about 3200ISO or 6400ISO (I might never push anything that much) they were about the same. When you get into that territory with any system you need to be considering NR anyway. So for me there is no huge advantage to V1.

What I like most about V2 is that I do not have to be SO concerned about highlights (the way I was on MX) and that the clipping point is "further away" than it used to be. You can open up a fair bit before the highlights "bite you" and when they did on MX it hurt. Now that rolloff in the top is smoother and it seems I want that filmic feel in many more shooting situations.

Battistella
 
David Battistella said:
Hi Matt,

I bet you could make a very acceptable image out of both of those r3d's without much effort. :)

Even with the moderate lowlight advantage, Whenever I ever think about old OLPF I just look at this test that Phil did. For me, and the situations I encounter, avoiding artifacts like this is far more important. But I appreciate that folks would rather stick with the older one for their style of shooting.

Battistella


phfx_redDragon_NewOLPF_FlareSun.jpg
Click to expand...

While this is awful, I'm not really concerned about the pattern thing on V1, because from what I hear it only happens at high f-stops, which I never use. I think the orb issue would rear its face more often.
 
I am buying two Dragons. Would it make sense to get an old and a new? And could someone clarify if these will be usable in the new changeable mount?
 
Agreed, Jake, you can lose those dots easily by using the tools available properly, and not shooting at T16?
Fighting low light levels with OLPF2 after the sun's gone might not be so easy.

The conditions in the stress tests seen in this thread that make the the OLPF1 produce artifacts/issues are not back-against-the-wall conditions (e.g. using T16, which is not ideal for almost any lens anyway; or adding 9 stops of ND to effect vulnerable fabrics).
Phil, unless I'm mistaken, your ND in this test is non-IR. Why would you ever need to shoot like that if IR glass would help avoid the pollution issue?
However, quickly tripling light levels when there's not enough light on set, to make OLPF2's density not adversely effect the shadows in a shot, is not always possible or desirable.
How do you justify choosing an OLPF that produces an 320 ISO camera, when a sudden creative thought comes to shoot at 120fps, and you have to say "No- we can't do that"?
 
Mark Pugh said:
Agreed, Jake, you can lose those dots easily by using the tools available properly, and not shooting at T16?
Fighting low light levels with OLPF2 after the sun's gone might not be so easy.

The conditions in the stress tests seen in this thread that make the the OLPF1 produce artifacts/issues are not back-against-the-wall conditions (e.g. using T16, which is not ideal for almost any lens anyway; or adding 9 stops of ND to effect vulnerable fabrics).
Phil, unless I'm mistaken, your ND in this test is non-IR. Why would you ever need to shoot like that if IR glass would help avoid the pollution issue?
However, quickly tripling light levels when there's not enough light on set, to make OLPF2's density not adversely effect the shadows in a shot, is not always possible or desirable.
How do you justify choosing an OLPF that produces an 320 ISO camera, when a sudden creative thought comes to shoot at 120fps, and you have to say "No- we can't do that"?
Click to expand...

I am very interested in knowing too how much better off you would be with the old if you plan on doing a lot of 120fps
 
Hey Mark,

I get what you are saying and your points are all valid. It's indeed a pretty lively discussion and has been for a while. The other side of this, of course, is when the producer (or someone eventually comes on here and says that the producer turned to them) and says, "the Alexa has better color, skin tone and more room in the highlights".

This is what came out of the "London tests" on the V1 OLPF.

So, the detractors would then be saying, "shit, we waited for a camera that does not compete with the Alexa on skin tone and in the highlights."

This discussion is a no win for RED. But, to their credit, they are developing a flexible solution in an effort to satisfy all their customers. There will also be those who say, "well, that is not good enough either because you (red) should build the OLPF that satisfies ever single possible situation. Of course that is what we all would like.

Making that happen maybe is not as simple as saying it.

As for getting after Phil (who has given an arm for the community). Maybe he shot that at T16 (which he exposed equally and fairly for both camera's) to show worst case scenario, highlight the old problem and show the improvement. I think that is fair.

Now, let me take your exact words to Phil and apply it to Roger's two R3D's posted here.

ORIGINAL "The conditions in the stress tests seen in this thread that make the the OLPF1 produce artifacts/issues are avoidable (e.g. using T16, which is not ideal for almost any lens anyway; or adding 9 stops of ND to effect vulnerable fabrics)."

SIMILAR "The conditions in the stress tests seen in this thread or Rogers "test" that make the OLPF2 produce artifacts/issues are avoidable (e.g. placing a camera on the floor of my office with no lighting, which is not ideal for almost any lens anyway; or shooting tests that aren't actually side by side)."

Anyway, you can see how this gets into semantics.

Bottom line is that people do get to choose their OLPF.

Battistella
 
I think I'm going with V1. But the more I think about this, I LOVE the switchable OLPF idea. It's decided.
 
David Battistella said:
Hey Mark,
The other side of this, of course, is when the producer (or someone eventually comes on here and says that the producer turned to them) and says, "the Alexa has better color, skin tone and more room in the highlights".

This is what came out of the "London tests" on the V1 OLPF.
Click to expand...

Can you point to the London test for me? I'd like to see it. I can't recall a big storm about the results.
The dragon was generally gaining momentum with a lot of praise before the issues of OLPF came out.

I've had real world shoots using Dragon with OLPF1 next to the Alexa, and found it pretty difficult to pick a difference.
I've never had a producer or anyone else complain about a problem induced by OLPF1, and it's the knly one that I've been shooting with, all year. However, Rogers's low light test shows results that might make one complain about OLPF2, for sure.

It's fine to talk about future solutions, but they are not here now. I'm talking about practical shooting today.
 
No I'm definitely not saying that OLPF2 is unusable. For day exteriors it needs a little less wrangling to produce a great image. I'm saying that not having enough light when you really need it is a bigger drama than opening a lens a little to avoid dots around the sun. I haven't had highlight issues with OLPF1, as I'm happy creating a curve in camera that lifts shadows a little (noIse-free) which means that I can stop down a little.
 
I was kinda happy to see those legacy MX things disappear with dragon v2. I can't speak for Phil but I doubt he'd be shooting much at t16 in bright sun.

Battistella
 
There are folks who really wanted clean 2000ISO imagery without an additional processing step and feel ripped off that V2 is noisier than V1 in low light. Fair enough. That said, let's not rewrite reality and ignore the reasons RED - and the majority of the user base - canned it for V2.

V1 had both the obvious red dot issue AND subtle magenta blooming around in frame light sources. Moreover, I had a few shots at wider apertures with diffuse red orbs that weren't nearly as obvious, but still made global (primary) grading problematic.

In a Bayer CFA based CMOS camera the optical path is intimately co-dependent with the decoding algorithm. RED has publicly stated that further development of their Dragon based systems will be V2 only (at least until there is a swappable OLPF system extant). My analytical left brain tells me that's the horse to bet on, as further optimization - based on my experience with RED cams and other similar tech - will yield benefits in performance (like DEB) over time that will ameliorate many of the current issues. My right brain just looks at the color rendition of V2 and it's game over.

I realize it's not viable for everyone, but with a noise reduction pass in finishing I would feel confident I could kill all but the most extreme noise issues. YMMV.

Cheers - #19
 
Thanks Matt. These examples were taken in different days and time. Each OLPF version offers specific attributes that I wish to be combined into a single new OLPF V3, avoiding the swapping technique, specially in not optimal environmental situations.
 
Although of course i'd love OLPf1 to have the benefits oh OLPF2, I'm interested in what's the best general purpose solution for shooting today. Not so much about future development priorities, future modular OLPFs, and relying on someone to do noise reduction in an online ( which won't get back shadows that are gone)...

OLPF1's "obvious red dot issue" is only an issue if you don't know what you're doing.
Driving a sports car really fast is only a problem if you're doing it despite the road conditions. Just because driving around everywhere at 140MPH will eventually lead to a crash doesn't mean that that the car is a lemon. It means that you need to know how to drive, and to use your head in the conditions.

Happy shooting, everyone.
 
You must log in or register to reply here.
Back
Top