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

Hey guys.

I think phil probably just got through the first round of his pretty excellent testing procedures. He just is now uploading a big file for all of us to look at.

Not to be a Debbie downer, but could we maybe pile through all of that before asking him to do a round of specific tests?

I mean, he has been pretty frickin' generous to take a few days to do this for all of us.

just sayin'


battistella
 
David Battistella said:
Hey guys.

I think phil probably just got through the first round of his pretty excellent testing procedures. He just is now uploading a big file for all of us to look at.

Not to be a Debbie downer, but could we maybe pile through all of that before asking him to do a round of specific tests?

I mean, he has been pretty frickin' generous to take a few days to do this for all of us.

just sayin'


battistella
Click to expand...

He's uploading a "texture" test. I asked for that shot a couple days ago. I don't think asking for specific tests is a bad thing considering he got an early look and started this thread to show off the results... WHich I'm thankful for.
 
Matt Ryan said:
He's uploading a "texture" test. I asked for that shot a couple days ago. I don't think asking for specific tests is a bad thing considering he got an early look and started this thread to show off the results... WHich I'm thankful for.
Click to expand...

Less panic. I'm shooting night material tonight. Need to make a glass decision.

Sorry about any delays. I'm orchestrating 3 upcoming shoots, scheduling a few important meetings, prepping some expansion documents, and working on 2 trips out of town for other things coming up. I'm fighting a bag of alligators bare handed out here outside of doing camera tests in my "free" time.

Most of my tests thus far have been targeted towards "certain" other people out here and overseas to demonstrate the differences between the two OLPFs and what they actually can do.
 
Sean Cruser said:
Phil these tests are great. thanks for all the time and effort. i've only ever used the STH OLPF so maybe i'm just biased, but i think that's still the winner for me.
Click to expand...

Same here, although I did use the v1 OLPF for a short time before release of the v2.
 
Thanksssss Phil ;)

yep same here, still in V2 i love this:)

Jarred, when ready the low con?

Give us a good news:)
 
Phil Holland said:
Heh. I will always have my work gloves on even in that sexy garb.


Alright, so here's sort of what everybody wants to see. An ISO Rating Texture comparison between the Low Light Optimized OLPF and the Skin Tone - Highlight OLPF across the ISO range. I left D.E.B. out of this party to keep things "vanilla", but D.E.B. can be utilized to effect things slightly of course. And that goes for A.D.D. as well.

First handful of shots are from the Skin Tone - Highlight OLPF graded at ISO 500. After that everything is labelled. You'll see two of the same ISO and properly exposed shots next to each other with the OLPF labelled in the corner.

This is a 1GB download of an H.264 file, which isn't the best quality thing out there, but the 4K ProRes HQ is um..... bigger than what I'm putting up on my web server at 21GB.





http://www.artbyphil.com/temp/reduser/phfx_redDragonOLPF_compareLLOSTH_HD1080.zip

Enjoy.

I'm currently uploading the 4K ProRes file to YouTube. It's gonna be a while.
Click to expand...

Phil great test. I have to say, the new Lowlight OLPF is looking GOOD. At 2000 ISO I thought it was holding up REALLY well.

And the "orbs" are pretty much gone. What's left looks like very very mild bloom.

Thanks again so so much for taking the time to do these tests. They are...in a word...invaluable.

best
 
Robert Ruffo New said:
In this video Magenta orbs are still a real problem. Better sure, but far from acceptable in my book. If they were white, like on Alexa, that would be fine, but they are Magenta, which has very specific psychological associations, and would be a very tricky thing to remove in post, especially if the thing ever creeped onto someone's face, when maybe you could not fix it at all. Also, that light is very far from bright, it is dimmed way down. If you had something really kicking to over-exposure in that shot you would see huge orbs, me thinks, much closer to how they were before. We need the Goldilocks OLPF - and this ain't it yet.
Click to expand...

White bloom is certainly more convenient. Would be interesting to see how the Lowlight OLPF holds up under brighter sources, for sure!

best
 
Robert Ruffo New said:
If you had something really kicking to over-exposure in that shot you would see huge orbs, me thinks, much closer to how they were before. We need the Goldilocks OLPF - and this ain't it yet.
Click to expand...

This is Phil broadcasting from the dangerous streets of Hollywood. Like literally in the street.

Im still out shooting, but I can initially confirm that the orbs are far, far lessened. I dont know if that will turn your frown upsidedown, but these car headlights arent doing anything near what they used to.

Jarred confirmed via email and this thread that theres been a bit of a revised idea od what the V1 OLPF did and I can confirm that the Low Light Optimized OLPF is different.

Images soon. Trying not to get run over. And yes, I shot some flashlight stuff.
 
Phil Holland said:
This is Phil broadcasting from the dangerous streets of Hollywood. Like literally in the street.

Im still out shooting, but I can initially confirm that the orbs are far, far lessened. I dont know if that will turn your frown upsidedown, but these car headlights arent doing anything near what they used to.

Jarred confirmed via email and this thread that theres been a bit of a revised idea od what the V1 OLPF did and I can confirm that the Low Light Optimized OLPF is different.

Images soon. Trying not to get run over. And yes, I shot some flashlight stuff.
Click to expand...

Thanks Phil! Your tireless efforts are so appreciated!!
 
Nick Morrison said:
Thanks Phil! Your tireless efforts are so appreciated!!
Click to expand...

Thank you Nick. Much appreciated for sure. Here's a post 1am post just to emphasize that.

So I went out tonight and did some ISO 800-2000 shooting mainly looking at headlights and flashlights with the Low Light Optimized OLPF. Basically the long and the short of it is I was "orb hunting".

First when I got the LLO OLPF I felt it was a bit different than what I was seeing on the old V1 OLPF and Jarred has since confirmed that some tweaking has gone down. Which makes me feel less crazy :)

Good news? The orbs are to me vastly improved, though they do occur mainly when a source is pointed straight into the lens or just real hot.

So here's some rather stress test inspired stills:













This next one is ISO 2000.




This one is ISO 800.





So I've got lots of thoughts now that I've done these tests. A few things to note. The Low Light Optimized OLPF still produces the Red Dot Pattern when shooting at smaller apertures with a hot source in frame. However, for many of us with Mysterium-X prior to Dragon we have dealt with that for a while. The LLO OLPF however has fairly noticeably reduced the highlight orb effect. Like a whole lot. What can still happen? Well we you get that lens flare cooking at higher ISOs you'll be certainly lifting the shadows a bit and you may experience additional texture in those areas, which is pretty much par for the course in regards to how that reveals itself with other systems and "makes sense". The LLO OLPF clearly has less layers or coatings than the STH OLPF as you can get a lot more of the reddish IR reflection, glare, flare in comparison and in extreme cases you can even get light to reflect off the back side of the PL mount of the lens depending on a variety of factors. Again, extreme thing here.

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.


Oh. And the Texture Comparison Test has uploaded. Looks like I'm having a YouTube/ProRes gamma issue I'll look into, but this is a still a decent A/B comparison between the two where you can see the LLO OLPF providing a cleaner higher ISO image over the STH, though I am having a hard time now figuring out what I prefer.

 
Thanks Phill, love your new findings. Looking forward to sending my Epic for Dragon upgrade and enjoying the luxury of choosing OLPS for whatever suits a shoot best! Shooting RED just keeps getting better!
 
Fantastic tests Phil.

Thanks a million!
 
Phil (amazing testing effort sir!) & RED,

More questions. :)

1) In the colour comparison between LLO and SKO - is it possible to see if the only difference is Saturation - or is there a color shift too?
If Saturation only - could a one-to-one mapping LUT turn the LLO image into a STH (where IR is not contaminating the lighting). If so - if your studios beauty shots used super high CRIs LED lighting ... you could use the LLO without any loss in fidelity.

2) How many stops does an IR mirror add? Would and IR mirror + LLO = the IQ of the STH?

3) Would any concerns of IR/UV colour polution be reduced if the colour chart of LLO, LLO with IR, LLO with UV, and STH be shown together?

4) [RED] Random thought for your next sensor. If the STH sensor has 1.3dB less R + G + B light cause by more exacting IR filtration, and the R channel is both more sensitive and more noisy, and the G channel now > 3 dB less noisy than the R channel ... is there any mileage in considering a rethink of the Bayer pattern having 2 x G pixels?
-> Maybe one of Green pixels should be turned to frequencies of Yellow / Orange?
-> Or even tune one of the G pixels to be IR, and remove the high-pass filtering of the R pixel

5) [RED] If LED Efficacy is jumping up hugely through external quantum dots reducing intenal reflection, I would posit that the reverse would be true in collection of light (ie better absorption,less reflection with quantum dots on the sensor surface)

http://phys.org/news/2014-09-nanotechnology-cheaper.html

AJ
 
Thanks!

Thanks!

I am shooting Dragon with OLPF 2 and, unlike some others, have been knocked out. I don't post much, but I have to say a big thanks, Phil. These tests are legit, side by side, real world, and very helpful. Thanks for the extra time and energy you have taken to share. Great stuff.

Tim Eaton
 
Antony Newman said:
Phil (amazing testing effort sir!) & RED,

More questions. :)

1) In the colour comparison between LLO and SKO - is it possible to see if the only difference is Saturation - or is there a color shift too?
If Saturation only - could a one-to-one mapping LUT turn the LLO image into a STH (where IR is not contaminating the lighting). If so - if your studios beauty shots used super high CRIs LED lighting ... you could use the LLO without any loss in fidelity.

2) How many stops does an IR mirror add? Would and IR mirror + LLO = the IQ of the STH?

3) Would any concerns of IR/UV colour polution be reduced if the colour chart of LLO, LLO with IR, LLO with UV, and STH be shown together?

4) [RED] Random thought for your next sensor. If the STH sensor has 1.3dB less R + G + B light cause by more exacting IR filtration, and the R channel is both more sensitive and more noisy, and the G channel now > 3 dB less noisy than the R channel ... is there any mileage in considering a rethink of the Bayer pattern having 2 x G pixels?
-> Maybe one of Green pixels should be turned to frequencies of Yellow / Orange?
-> Or even tune one of the G pixels to be IR, and remove the high-pass filtering of the R pixel

5) [RED] If LED Efficacy is jumping up hugely through external quantum dots reducing intenal reflection, I would posit that the reverse would be true in collection of light (ie better absorption,less reflection with quantum dots on the sensor surface)

http://phys.org/news/2014-09-nanotechnology-cheaper.html

AJ
Click to expand...

Hi Antony, I just have quick thoughts here, but....

1. I do see a slight color shift overall between these two OLPFs, it's not a major shift in reality and is slight. I do believe it can matched via LUT, but I haven't shot in every single type of lighting condition yet where color, tonal, light intensity, and IR could come into play. The table top setup I utilized featured full spectrum CFL and Tungsten bulbs and not LEDs.

2. It's a bit more complicated than just adding an IR cut as there's certainly a lack or weaker strength anti-reflective layer or coating in the LLO.

3. The STH functions the best without IR filtration at higher ND values, which is more or less to be expected. I don't think we're dealing with UV effecting the image when comparing these two OLPFs. It's calibration/profile, IR, and AR related.
 
You must log in or register to reply here.
Back
Top