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 :)

First Look: Interchangeable OLPF on Dragon

Killing it Phil, thanks for these tests.

Any noticeable improvement on the Magenta Streak with this OLPF?

I do think giving people what they originally signed up for Dragon for, the 2000ISO, is going to calm a lot of Dragon owners.
 
Matt Ryan said:
Phil, you are the man! Thanks so much for the flashlight test!!! A HUGE help!
Click to expand...

Heh. That one was for you Matt. I nearly dressed up like Brad, but that would have been too dead on.


David Battistella said:
You know, many "working DPs " don't share or show their tests to anyone.

Phil is doing this community a service on his dime and is presenting it in RED's house. If I read another post suggesting he is "selling" or "spinning", not disclosing or implying that he is working for RED or a fanboy or doesn't have post knowledge, then I'm going to act on that that feeling of wanting to barf that I get in my stomach when I read that stuff.
Click to expand...

Thank you David.

For anybody who wants to learn about what I've been up to for about 16 years:

http://www.imdb.com/name/nm0390802/

I'm not credited on a about 80 or so productions as well. This also doesn't list things like the ride film I just shot or any of my video game credits :)

Seriously though. I've been swimming in a world of film, digital cinema, production, pre-pro, VFX, color insanity, post pipelines, and pretty much everything else you can imagine for a long time and for a good chunk of that I was working non stop. Took two 3 month sabbaticals while working at my old studio and guess what, I worked for other studios in that time. Don't worry, I took a week off in there!

I've been very, very fortunate to learn from and work with some of the best in the industry.

Much of my film background is what eventually lead me to be a Red shooter and owner. I'm also very grateful for Red in general and Jarred for trusting in me to occasionally test out new gear to see if I can break it :)


Charles Bergquist said:
Killing it Phil, thanks for these tests.

Any noticeable improvement on the Magenta Streak with this OLPF?

I do think giving people what they originally signed up for Dragon for, the 2000ISO, is going to calm a lot of Dragon owners.
Click to expand...


It's in there in similar over exposure and underexposure high contrast ratios which can be revealed more by shooting higher ISOs. This artifact really comes down that. Here's something outside of the recommended ISO range at ISO 6400 to fully show the CMOS smear.




This is a tricky artifact that is fairly common among a few other CMOS sensors, though it pops up at different sensitivities. The good news is when it does reveal itself is pretty consistent across multiple systems on where it shows up. Once you get solid on that, it's easier to avoid. This is a CMOS limitation in this case, though I can think of a couple of things that Red can do to help minimize it. I'm sure there's talks behind closed doors. I've certainly sent some fairly interesting test images their way.
 
Phil is putting a lot of time in this, I honestly think that did not go to sleep at night.
I hope people now realize how Dragon is something amazing.
Is getting better with each passing month :)


Thanks Phil for these shares :)


I can not wait to throw a OLPF low contrast in front of the beast :)
 
Phil, your photo from the two people crossing the street in Hollywood, is just the look that “Michael Mann” was looking for in “Collateral”, a real night environment as we see-it, not as in most other cameras see-it.

http://boxofficemojo.com/movies/?id=collateral.htm and http://en.wikipedia.org/wiki/Collateral_(film)

From Wikipedia, the free encyclopedia; “Collateral is a 2004 American crime thriller film directed by Michael Mann from a screenplay written by Stuart Beattie, and starring Tom Cruise as a contract killer and Jamie Foxx as a taxi driver who finds himself his hostage. The film is set in Los Angeles, California in January 2004, and the supporting cast includes Jada Pinkett Smith and Mark Ruffalo. Foxx and Cruise's performances were widely praised, with Foxx being nominated for the Academy Award for Best Supporting Actor.”

Great Job on that image and all the other you have posted, I can’t say enough “Good Things” about them.

Humberto Rivera

It's in there in similar over exposure and underexposure high contrast ratios which can be revealed more by shooting higher ISOs. This artifact really comes down that. Here's something outside of the recommended ISO range at ISO 6400 to fully show the CMOS smear.




This is a tricky artifact that is fairly common among a few other CMOS sensors, though it pops up at different sensitivities. The good news is when it does reveal itself is pretty consistent across multiple systems on where it shows up. Once you get solid on that, it's easier to avoid. This is a CMOS limitation in this case, though I can think of a couple of things that Red can do to help minimize it. I'm sure there's talks behind closed doors. I've certainly sent some fairly interesting test images their way.
Click to expand...
 
All well and good, and im convinced, not to mention eager, so when does this start to transpire as far as sending the cameras in? whats the turnover etc... are Olpf in plentitude (please say yes) and how long till we start seeing new fangled artsy ones rollout? Im in I love the idea of a few new stocks every quarter to learn test play with to have signature looks that make our craft ours again...
 
We should start calling Phil "Prof. Holland". I can now check the "post-graduate education" on the census form. Thanks for all the sharing, Phil.
 
Ryan Purcell said:
Phil - could you please just every once in a while SMILE in one of your RED selfies.
Click to expand...

One day I squeak one out :)


Just a quicky with more info on the CMOS Steak/Smear and it not showing up.

If you take a look at this shot that I pushed up to ISO 4000, the RAW data is clipping both the sign and the headlights, however due to it being within a more "wrangled" contrast ratio there are no streaks to be seen.




One thing I'll also say about this shot, is while I'm fairly adamant about using Dragon within the "Red recommended" ISO range of 250-2000, at ISO 4000 this shot is actually rather usable. It was rated at ISO 800 and lifted up to 4000. On the Skin Tone - Highlight OLPF this wouldn't even look close to this at ISO 4000.

Here's a 350MB ProRes HQ of that shot in motion at high ISO:

http://www.artbyphil.com/temp/reduser/phfx_redDragon_OLPF_LLO_nightHighISO.zip


Also, I've reviewed all of my orb tests and night shots, this is the worst Orb I was able to make happen:





Also that particular shot would be the type where it's not as clean at ISO 4000 because of where the sky values land:




However, it does show just how much sensitivity Dragon has way back there. This is actually fairly cool if you think from a capture and post perspective. If you were knocking out a night shot like this with a subtle low contrast grade (flatter than RG4 basically) from the base exposure you would have really nice rich detailed shadow areas. I can't even see those trees against the night sky at ISO 800 DC/RG4, but the information is there if you want it to be.


Right now things I'm seeing overall with the Low Light Optimized OLPF are:

- Highlight Orb Artifact is greatly reduced since the "V1 OLPF".
- Low Light/High ISO performance is improved over the Skin Tone - Highlight OLPF
- The Small Aperture IR Dot Reflection Grid Pattern still exists and is something to be mindful of.
- Color on the LLO OLPF is good, Color on the STH OLPF is better than good.
- IR Containment is better on the STH OLPF, but it's not bad on the LLO OLPF. You shouldn't need an IR Cut on your NDs (tested up to 1.8 ND so far in broad daylight)
- Highlight Roll-Off is very good on both OLPFs and doesn't "feel" digital. This is very important for me.


It's a difficult thing to truly describe but it's more of a combination of the OLPF and what it's layers/coatings block or allow through in combination with the accompanying calibration that can produce cleaner ISO shooting across the board.

I think we'll find two different perspectives here, but for the moment I'm keeping the LLO OLPF in my camera, but I will likely go back to shooting the STH OLPF just because the color has that extra juice I like and the two orb and dot artifacts are nonexistent.

For those who have been looking for that cleaner ISO performance across the board but especially from ISO 800-2000 the LLO OLPF is what you want.

For the film people who mainly are mainly exposing/rating for the ISO 250-1280 range I'm thinking the STH OLPF is still what you want.
 
sergio arguello said:
All well and good, and im convinced, not to mention eager, so when does this start to transpire as far as sending the cameras in? whats the turnover etc... are Olpf in plentitude (please say yes) and how long till we start seeing new fangled artsy ones rollout? Im in I love the idea of a few new stocks every quarter to learn test play with to have signature looks that make our craft ours again...
Click to expand...

Sounds like end of the month/October for the LLO and STH OLPFs:

http://www.reduser.net/forum/showthread.php?121658-Interchangable-OLPFs

I was fortunate to take early delivery of one, don't even have the proper case yet. I'm using a small lab specimen container to haul my OLPFs around in at the moment.

There hasn't been official word on the release date of the artsy, more content specific, or 3rd Party OLPFs yet. The calibration/profiling will take a moment as well as making the specific OLPFs. I'm actually curious what sort of container Red will come up with.

One thing I should say is since I've been handling these and have now swapped OLPF about 35 times (which I honestly think is more than what many hard core shooters will do in a year even) is the OLPFs have a front face metal frame that's pretty sturdy and easy to grab. There is a rubber gasket that presses into the sensor box area to block out light and all that. When inspecting your OLPF before insertion make sure you hit that with the flashlight and hand blower if there are any stray particles.

I haven't been experiencing anything I couldn't remove with the hand blower and I'm also swapping the OLPF in rather not-so-ideal locations. Like outside or in the middle of Hollywood and Highland, so I have been inspecting and blowing out the little dust bunnies when they show up.

For most of you I recommend taking a trip to the camera cart or more likely just a cleaner prep area prior to shooting. I've been doing it on the tripod on location and often in the dark to do these A/B comparisons, be mindful of those tiny bolts. Two finger pinch action and you can grab them nice and easy.
 
Phil, Can you confirm that LLO is actually different in some unspecified way than the V1 OLPF when dragon was first released ? I had thought they were identical.
 
Dave Blackham said:
Phil, Can you confirm that LLO is actually different in some unspecified way than the V1 OLPF when dragon was first released ? I had thought they were identical.
Click to expand...

Jarred has confirmed they tweaked it. The post is in this thread.

Jarred Land said:
Once we committed to the swappable OLPF it gave us the opportunity to dial each OLPF into getting the best image from a specific use case.


Things have become alot more interesting ..and some of the 3rd party OLPF modules that 3rd parties are working on are pretty incredible. From the clinical all the way to the lost-your-mind artistic.


And Phil.. you are one badass motherfucker. Thank's for putting in the effort here.
Click to expand...
 
clipped from Phil...
What the heck does this mean to me? Well the LLO OLPF is sort of surprising me now, which at the beginning of these tests and before I knew that there was a revision in it's design I wasn't expecting. It's giving me cleaner higher ISO material. Like ISO 800-2000 feels groovy. There's texture at ISO 2000, but nothing killing me.

Basically I'm going to shoot with it a bit and experiment more because it's unlocked some curiosities and grabbed my attention in a different way. The color is good and the dynamic range is there too. Highlight is still a pleasing roll-off. It's not as "banging" as the STH OLPF in some ways, but I'm pleasantly surprised by what I'm getting here.
Click to expand...

Thanks for sharing your hard work Phil! Much appreciated. I shoot a lot of 90 degree shutter and high framerates - the improved LLO OLPF is going to be like having a new camera. Get some rest and thanks again!
 
Phil Holland said:
At this point I am woefully confused by the concept of me "selling" anything to anybody. I am spending a rather large amount of my time to literally to dig out these gremlins and artifacts hoping to reveal what this system can do for the working DP.

You've brought it up a few times and I have a thick skin, but I'm telling you man, stings a bit when I'm up at 3am processing test footage. Especially when I'm giving you specifically things you're looking for. I'm testing this stuff hard because I use these cameras to make a living and I understand the gravity of others in the same position.


Anyways. Gate Flare. Hard topic and a difficult one to give a "whole" answer to because every lens is different. Right now, this is what my 35mm T2.1 is doing:




This is improved at the top and sides of the frame, which is nice. It does depend on a million factors though. Lens design, mount, source distance, angle of incidence, etc..... So in this particular PL mount case it's a bit better than it was at 6K FF.
Click to expand...

I never said you personally were selling anything, I was talking about a general principle. If anything, I was encouraging you to share remaining problems you may find as completely as improvements, but this without (sorry if I gave another impression) accusing you of not intending to do so.

I do, however, sometimes feel that your honest enthusiasms gets the better of you, but by no means does that make you an even slightly bad person in my book.

Look at it from my perspective - I am not friends with anybody except my clients. I am just mercilessly, coldly objective, as best I can.

Indeed all issues, from your very, very useful tests seem to show huge improvement. Is there a "New OLPF 01" Dragon in Montreal or Toronto? If there is I will do my own tests, and even publish them here.

If I, a much derided "complainer" who does not mince his words (not a "Red Supporter", or an any-brand supporter more than any other), am happy by what I now see, and can show my torture tests designed to find problems, imagine the value that would have.

I have my flaws, sure (just ask my wife…) but certainly no one has ever accused me of being over enthusiastic about Dragon, or afraid of being critical of Red, or any other companies. I am nobody's fanboy, and I think that is bluntly clear to all.

So whenever you can hook me up with a Dragon with these new OLPF 01 improvements, which indeed do look very, very promising, let me know and I'll make the time.
 
Rob.
This is actually a warning. :) Take it as an official moderator moderation...

And this is why:

You imply repeatedly that your scepticism is objective and everyone else is wrong.
So far, I at least don't have many issues with that.
In principle.

BUT
As you never document anything and don't do much else than crapping all over what is easily the current biggest single resource for this forum, who doesn't in any way HAVE to share his findings, it's getting close to stupid.

So.
If you have issues with Phils findings (or mine, Or Johns or Davids or whoever you can think of as biased stupid non-profesionals who don't think of their clients best), post an image and a R3D, and explain what you think is wrong.

Same goes for Dragon vs Epic. I have both cameras and I cannot on any way recognize your observations in tests or other images.
So again.
I'll accept it. But document it.

If not, you're just trolling.
And there's nothing objective or really interesting in that.

Makes sense?

Cheers!
 
Jake Wilganowski said:
My rep told me the filter is exactly the same, it is the firmware that makes if different
Click to expand...

My hunch is that Phils hunch, and Jarreds kinda confirmative hunch, might be a better hunch.
 
Last edited:
Gunleik Groven said:
Rob.
This is actually a warning. :) Take it as an official moderator moderation...

And this is why:

You imply repeatedly that your scepticism is objective and everyone else is wrong.
So far, I at least don't have many issues with that.
In principle.

BUT
As you never document anything and don't do much else than crapping all over what is easily the current biggest single resource for this forum, who doesn't in any way HAVE to share his findings, it's getting close to stupid.

So.
If you have issues with Phils findings (or mine, Or Johns or Davids or whoever you can think of as biased stupid non-profesionals who don't think of their clients best), post an image and a R3D, and explain what you think is wrong.

Same goes for Dragon vs Epic. I have both cameras and I cannot on any way recognize your observations in tests or other images.
So again.
I'll accept it. But document it.

If not, you're just trolling.
And there's nothing objective or really interesting in that.

Makes sense?

Cheers!
Click to expand...

With all due respect,

I asked if there was a Dragon in Montreal anyone knew of with new OLPF 01 that I could do tests on, and side by sides with MX, and offered to do the work, and publish my results, at my expense I just offered to document it, in the very post you are replying to by saying I should do that. I also said the new OLPF 01 showed real improvement. I'm sorry if I fail to understand how that constitutes trolling.

I could not publish the tests I did a few weeks ago because I did not pay for them or own them, and the rental company who sponsored them, and as such owns the rights, decided is was not in their business interest to publish images from my findings, and anyway, apparently now things have improved, so new tests are in order.

To my credit, I am not alone in my findings. Many DPs are not Reduser members, or choose to no longer post here.

However, you can read about their very similar results in other forums, or ask them in person.
 
Gunleik Groven said:
My hunch is that Phils hunch, and Jarreds kinda confirmative hunch, might be a better hunch.
Click to expand...

You may be right, but my rep went and talked to someone who would know (he said) and that's what he told me. I even referenced this thread and Jarred's comment. Sigh, I wish there wasn't conflicting info.
 
You must log in or register to reply here.
Back
Top