Welcome to our community

Be a part of something great, join today!

Register 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 :)

The toe and the shoulder Dragon DR

Dragon definitely displays better light sensitivity than MX on the low end. I spent 2 weeks A/Bing these two in lowlight, controlled lighting environments (unfortunately can't post examples, the footage doesn't belong to me), and Dragon is definitely superior. We shot a ton of lowlight footage using artificial light between 320-400 ISO, and the Dragon looks noticeably cleaner and also allows you to push exposure much harder, while remaining clean. If you light at 320 or one of the lower ISOs properly, you can push the image to 800 and beyond, and it will remain clean. That's not the case with the MX, if you light at 320, you typically can push it to around 640 or 800 but you will introduce noise around 800.

Which is why I've stopped bitching. Having shot these side by side, it's no contest. Dragon rules MX. I am very happy with this upgrade. Initially, I did not like what I was seeing in the shadows at 800, just with initial experiments of shooting in the darker corners of my house. Seeing more noise than with MX at 800 disturbed me - but no more, not when I am seeing that this is the one and only condition under which MX will outperform the Dragon.

The one - and only one - condition where MX out-performs the Dragon and that is when you shoot in an uncontrolled lighting environment and you are trying to push the ISO because you're already shooting wide open and there isn't enough practical or natural light in the scene to get a decent histogram. Dragon will be noisier at 800 and above than MX when you try to use the ISO to push your exposure, once you've exhausted other available tools. Avoid it. Or throw some light in there!

While this doesn't seem "right," necessarily, based on the original lowlight performance projections for Dragon - you get a banquet of other enhancements (no reason to list these yet again...) in addition to this lowlight sensitivity enhancement when the sensor is lit.

I definitely no longer think that MX outperforms Dragon - it simply doesn't. They behave differently in this very narrow band of circumstances, but that is the band of circumstances which has incited the riot. If you need to shoot under these circumstances frequently, then you should consider a different option or opt for V1.

But for the shooting I do, these are atypical circumstances - these conditions definitely come up, but not frequently. And when this condition does arise, I'll know to fight for light.

I have definitely fallen for Dragon V2 after shooting it side-by-side with MX for 5 solid weeks. MX is great techology. Dragon is better. This isn't fan-ism speaking either. This is based on experience shooting them side by side. We shot 3 weeks outdoors and the latitude in Dragon is dreamy for outdoors. We shot set pieces in lowlight for 12 days, and Dragon's better lowlight sensitivity - when properly lit - kicked MX's booty.

The band of conditions where MX will outperform is very narrow indeed, and V1 remains an option and interchangeable OLPFs will be available, if that's the band of circumstances in which you most often shoot.

I would not have expected MX to outperform Dragon under any circumstances at all - and I think that's why people are unhappy. I don't think anyone expected it. But I'm ratcheting in those expectations because everything else is sooooo good! After an initial period of worry, I think my money was well spent on the Dragon. I would do it all over again.
 
Meryem Ersoz said:
Dragon definitely displays better light sensitivity than MX on the low end. I spent 2 weeks A/Bing these two in lowlight, controlled lighting environments (unfortunately can't post examples, the footage doesn't belong to me), and Dragon is definitely superior. We shot a ton of lowlight footage using artificial light between 320-400 ISO, and the Dragon looks noticeably cleaner and also allows you to push exposure much harder, while remaining clean. If you light at 320 or one of the lower ISOs properly, you can push the image to 800 and beyond, and it will remain clean. That's not the case with the MX, if you light at 320, you typically can push it to around 640 or 800 but you will introduce noise around 800.

Which is why I've stopped bitching. Having shot these side by side, it's no contest. Dragon rules MX. I am very happy with this upgrade. Initially, I did not like what I was seeing in the shadows at 800, just with initial experiments of shooting in the darker corners of my house. Seeing more noise than with MX at 800 disturbed me - but no more, not when I am seeing that this is the one and only condition under which MX will outperform the Dragon.

The one - and only one - condition where MX out-performs the Dragon and that is when you shoot in an uncontrolled lighting environment and you are trying to push the ISO because you're already shooting wide open and there isn't enough practical or natural light in the scene to get a decent histogram. Dragon will be noisier at 800 and above than MX when you try to use the ISO to push your exposure, once you've exhausted other available tools. Avoid it. Or throw some light in there!

While this doesn't seem "right," necessarily, based on the original lowlight performance projections for Dragon - you get a banquet of other enhancements (no reason to list these yet again...) in addition to this lowlight sensitivity enhancement when the sensor is lit.

I definitely no longer think that MX outperforms Dragon - it simply doesn't. They behave differently in this very narrow band of circumstances, but that is the band of circumstances which has incited the riot. If you need to shoot under these circumstances frequently, then you should consider a different option or opt for V1.

But for the shooting I do, these are atypical circumstances - these conditions definitely come up, but not frequently. And when this condition does arise, I'll know to fight for light.

I have definitely fallen for Dragon V2 after shooting it side-by-side with MX for 5 solid weeks. MX is great techology. Dragon is better. This isn't fan-ism speaking either. This is based on experience shooting them side by side. We shot 3 weeks outdoors and the latitude in Dragon is dreamy for outdoors. We shot set pieces in lowlight for 12 days, and Dragon's better lowlight sensitivity - when properly lit - kicked MX's booty.

The band of conditions where MX will outperform is very narrow indeed, and V1 remains an option and interchangeable OLPFs will be available, if that's the band of circumstances in which you most often shoot.

I would not have expected MX to outperform Dragon under any circumstances at all - and I think that's why people are unhappy. I don't think anyone expected it. But I'm ratcheting in those expectations because everything else is sooooo good! After an initial period of worry, I think my money was well spent on the Dragon. I would do it all over again.
Click to expand...

my head hurts reading your post. ok what you are trying to say is light for 320 iso ? 800 iso will look good in 320 iso light (what does that mean?)
 
Bob Gundu said:
I also feel the sensitivity on the low end are equal or slightly better than MX. Top end... Forget about it... Dragon rules.
Click to expand...

yes, but I still want the lowlight preformance of the olpf1 it was fantastic. And I refuse to think that there is no possible solution to make a olpf that has the lowlight end but not the red orbs. I'm sure red or some 3rd part will make it happen quite soon, especially if the swappable olpf comes then people can actually experiment with it and I think that will bring up some great solutions.

If there was a olpf with the sensitivity of the olpf1 and the clarity of the olpf2.... then nobody in their right mind would choose the olpf2 option. As it's quite simple, the better light sensitivity you got the better, you can always use an ND, but can not always bring up the light levels of a scene.
 
charles lim yi yong said:
my head hurts reading your post. ok what you are trying to say is light for 320 iso ? 800 iso will look good in 320 iso light (what does that mean?)
Click to expand...
Thank you:) I thought it was just me.
Meryem, since you shot these side by side for weeks, there should be a raw frame from each camera available for all of us to see?
 
Gunleik Groven said:
HM.

I haven't A/B'ed v2 vs MX, but v1 was more sensitive downwards AND upwards than Epic
The way it feels (and that suits the 1.3 stop move upwards number being thrown about) Dragon should be about EQUAL to the MX in the bottom and gain 1.3 more stops in the top.

NOW I haven't tested for this, but the theory suits what it feels like.

Have anyone really tested this in a controlled manner?

Have Dragon and Epic here and have a real hard time seeing when I'd pick the Epic if the Dragon is free to use...
Click to expand...
Can you post a frame of each at cleanest setting? Same scene, does not have to be anything fancy, just a simple low ISO, even ( or what you would consider best) exposure for each camera. I can't believe this is not available to show the advantages of dragon. It is so pointless to hear the same pro and con again and again when two frames can settle this.
 
I did this ppretty thoroughly with v1.

I'll see what I can rack up.

"best case" isn't really as informative as "same scene, same light" at different exposures.
 
Meryem Ersoz said:
The one - and only one - condition where MX out-performs the Dragon and that is when you shoot in an uncontrolled lighting environment and you are trying to push the ISO because you're already shooting wide open and there isn't enough practical or natural light in the scene to get a decent histogram. Dragon will be noisier at 800 and above than MX when you try to use the ISO to push your exposure, once you've exhausted other available tools. Avoid it. Or throw some light in there!
Click to expand...

So Dragon is noiser than MX at 800 ISO? That somehow contradicts your previous argument (Dragon is cleaner and allows to push harder).

The "throw some light in there" argument isn't valid either. Using more light is not always possible, or desirable, even if you have the money and equipment to do that (someone mentioned he shots night aerial stuff, for instance). Even though I preferred to shoot nights at around 400 ISO (overexpose) when I was using MX (Red One and Epic), I have also shot many times at 800-1000 ISO normal exposure (for instance, when using anamorphic lenses, as it's better to stop down them a bit, and 800 ISO with a healthy exposed histogram used to be pretty clean) and found the results 100% acceptable. That's not the case with Dragon. I would shoot 800 ISO again, but only if I really have to, and prepare myself to enter the world of D.E.B. and/or some other noise-reduction software that remains unknown to me up to this day.

The other day, people at CML were posting about the latest Blackmagic camera (URSA). Some folks said that in 2014, it wasn't acceptable for what it supposed to be a professional camera to only have only a 12 stop dynamic range. I feel pretty much the same about Dragon OLPF v2 and its low-light capabilities. It performs great at about 250-400 ISO (middle gray metered by lightmeter, not camera settings) and outperforms RED's previous models. But if you push it, not to far, just to a light level where MX was confortable, it gets really noisy in the red channel, and thus, outperformed by the previous camera model. It's not only that we were promised another thing (we were), it's just that it's a big step back compared to MX, which wasn't and isn't the most light sensitive sensor out there. I just hope that Dragon's poor low-light capabilities can be fixed somehow, via swap OLPF or a whole new OLPF, firmware upgrades or calibration, and not introducing more post-production complexities to an already complex process.
 
Meanwhile, just shot jobs at ISO 500, 800, and 1000 on the V2 OLPF. Tuesday shooting at 500.
 
Christof Paul said:
Thank you:) I thought it was just me.
Meryem, since you shot these side by side for weeks, there should be a raw frame from each camera available for all of us to see?
Click to expand...

I don't own the footage. It belongs to the production, and I was only a humble camera operator. But I was also the rental house, so both cams (1 MX, 1 Dragon) were mine as well, so I feel pretty comfortable with my assessment, after working side by side with both in real-world conditions, on a real movie with real name talent.

I'll look into obtaining a still for comparison, once the producer gets back from his trip to LA.

Charles, I hope this makes your head hurt less....

1) any properly lit footage, shot at the same exposure on both MX and Dragon, will look noticeably superior on Dragon. Dragon for the win.

2) that same footage will be more flexible on the Dragon (you can, for example, push properly lit footage at 320 on both - but MX will limit you around 800, Dragon will let you push past that limitation and still look cleaner than MX because, properly lit, it offers greater image flexibility than MX). Dragon for the win.

3) shooting both MX and Dragon around ISO 800, in lowlight, using the ISO to add light to the image, once you have exhausted other resources like aperture or available light, is the only situation in which MX will actually outperform Dragon, in terms of noise reduction - when you push the sensors without lighting, MX will be a bit cleaner at 800 than Dragon. Dragon will still crave light in some circumstances, where MX may not. In this situation - find light! MX for the win - unless you light. If this is your primary lighting environment, as Ignacio claims - then you might be better off staying with your MX or else taking V1 for a spin and seeing what it can do in this situation. Or exercising patience on the swappable OLPF....

4) Dragon destroys MX in any daylight or high-key light situation, hands down, end of discussion. The latitude is mind-boggling. Dragon for the win.


Does that break it down?
 
Gunleik Groven said:
I did this ppretty thoroughly with v1.

I'll see what I can rack up.

"best case" isn't really as informative as "same scene, same light" at different exposures.
Click to expand...
yes, it would have to be same scene, same light. i meant best possible settings in terms of compression and iso. i can't get a clean file out of my epic at 6:1 regardless of iso, so i almost always shoot 5:1, which still allows for 50fps (at 5KHD), i am not sure how that works with Dragon?
 
and Ignacio - Gunleik asks a good question. why don't you do your own side by side comparisons, since you own both sensors? There's really no substitute for shooting your own stuff.

There's a lot of stuff that gets tossed around forums which bears testing.
 
Phil Holland said:
Meanwhile, just shot jobs at ISO 500, 800, and 1000 on the V2 OLPF. Tuesday shooting at 500.
Click to expand...


But you can shoot any ISO cleanly with proper lighting - the noise issues with Dragon only manifest when you try to shoot in low light using only limited ambient lighting. People want to be able to shoot Dragon at 800 at night (or on dark sets) with limited ambient lighting and are finding that side-by-side, MX outperforms Dragon if the cameras need ISO pushed higher as the vehicle for adding light to the image.

My point is that those concerns are legit but are outweighed.

I think that is the issue that people worried about upgrading are trying to get their arms around.

I think this issue has not been addressed with specificity. There are very imited and singular conditions where MX can outperform Dragon. This is an extremely narrow band, however. But I think it should still be part of the discussion - and hasn't been, so far. Or not in a very specific way.
 
Gunleik Groven said:
Have you shot them side by side, Ignacio?
(Seeing from your sig that you have them.)
Click to expand...

Yes, I have done several tests, which were sent to RED a few months ago. But I don't own an MX anymore, it's in the process of being upgraded to Dragon, even it may be Dragon as we speak (for those who may ask why I've upgraded the second camera after being dissapointed by certain aspects of Dragon, the answer is very simple, I need two identical brains for many jobs, both personal and for rental, as top clients always ask for the newer stuff).

This tests between Dragon OLPF v2 and Epic-MX showed clear differences in the noise floor of the two cameras. With both correctly black shaded, with the lens cap on, or just shooting the same environment (our rental facility, which exposes correctly at T/4 at 800 ISO and even T/5.6 in at least another camera system...), the Dragon showed much grainer images at high ISO (800-4000 ISO) with the lens cap on. And in the darker areas of our place, well exposed at ISO 800, red dots were clearly visible on RCX and put me on the track that maybe there was something wrong with the Dragon. With the latest firmware, things seem to have improved a bit, and D.E.B. hides the color of the noise, but it's still a bit far from what used to be our Epic Mysterium-X and adds complexity to the process.

Anyway, I'm really happy to read that most people out there doesn't share my own, personal opinion, as I've invested a lot of money in these two cameras, and I shoot and rent them (among other equipment) for a living. Hope my clients, directors and producers I work for will find Dragon's sensitivity OK as well.
 
Ignacio Aguilar said:
Yes, I have done several tests, which were sent to RED a few months ago. But I don't own an MX anymore, it's in the process of being upgraded to Dragon, even it may be Dragon as we speak (for those who may ask why I've upgraded the second camera after being dissapointed by certain aspects of Dragon, the answer is very simple, I need two identical brains for many jobs, both personal and for rental, as top clients always ask for the newer stuff).

This tests between Dragon OLPF v2 and Epic-MX showed clear differences in the noise floor of the two cameras. With both correctly black shaded, with the lens cap on, or just shooting the same environment (our rental facility, which exposes correctly at T/4 at 800 ISO and even T/5.6 in at least another camera system...), the Dragon showed much grainer images at high ISO (800-4000 ISO) with the lens cap on. And in the darker areas of our place, well exposed at ISO 800, red dots were clearly visible on RCX and put me on the track that maybe there was something wrong with the Dragon. With the latest firmware, things seem to have improved a bit, and D.E.B. hides the color of the noise, but it's still a bit far from what used to be our Epic Mysterium-X and adds complexity to the process.

Anyway, I'm really happy to read that most people out there doesn't share my own, personal opinion, as I've invested a lot of money in these two cameras, and I shoot and rent them (among other equipment) for a living. Hope my clients, directors and producers I work for will find Dragon's sensitivity OK as well.
Click to expand...
that is so sad. thank you for being so candid.
 
Meryem Ersoz said:
But you can shoot any ISO cleanly with proper lighting - the noise issues with Dragon only manifest when you try to shoot in low light using only limited ambient lighting. People want to be able to shoot Dragon at 800 at night (or on dark sets) with limited ambient lighting and are finding that side-by-side, MX outperforms Dragon if the cameras need ISO pushed higher as the vehicle for adding light to the image.

My point is that those concerns are legit but are outweighed.

I think that is the issue that people worried about upgrading are trying to get their arms around.

I think this issue has not been addressed with specificity. There are very imited and singular conditions where MX can outperform Dragon. This is an extremely narrow band, however. But I think it should still be part of the discussion - and hasn't been, so far. Or not in a very specific way.
Click to expand...


I have lots of thoughts on this matter, more in the relationship between the desires of a digital cinema camera versus a video camera in regards to an image that has texture or not, but I'll push those to the side.

If low light/no light is the only goal there's realities that need to be faced. If the world your capturing lands in the ISO 800-6400 realm and the given exposure reflects that that's where things need to be "worried about" in terms of noise reduction. ISO 800-2000 in terms of a properly metered and exposed shot in relationship to the noise floor looks pretty cool when done correctly.

There's a lot of discussion of what Mysterium-X can do versus what Dragon can do, most seem to be focusing on the small box in the corner rather than the cargo container holding everything in place and that's where Dragon outshines Mysterium-X in a multitude of ways. The V1 Dragon OLPF does a lot of Mysterium-X does and is notably cleaner at equivalent ISO ratings. If that's the focus, that's the answer. If you want the extra goodness found in the V2 Dragon OLPF you know where the goodness is being extracted from. That's as specific as it gets really.

If you want listed benefits and trade offs:

Dragon V1 OLPF
- Similar feel to Mysterium-X OLPF with similar visible artifacts
- Small aperture reflected dot pattern can occur with hot source in frame
- Floating orbs around source highlights can occur in frame
- Sees 1.3 stops more into the shadows (actually midtones through shadows) compared to the V2
- Good color
- No need for IR filtration at various degrees of ND filtration.
- More suited for shooting low light material and rating at higher ISOs

Dragon V2 OLPF
- Doesn't exhibit the reflected dot pattern at small apertures
- Doesn't exhibit the floating highlight orbs.
- 1.3 stops placed in favor of midtones through highlights
- Superior color compared to the V1 OLPF. This is slight to some, more to others.
- No need for IR filtration at various degrees of ND filtration.
- More suited for general shooting conditions


In regards to the OLPF Module, V1, V2, and whatever future OLPFs land. I see concern from some that this is becoming "too complicated". We need to get away from that thinking. User replaceable OLPF tech is optional. Also, on most jobs the concept of changing OLPFs in the field several times on a given shooting day would be largely rare. I would more or less expect many to choose and option and run with it for a given setup or shooting day. Options are good. It's okay for a top tier camera to be a bit more advanced then a point and shoot for a multitude of reasons.

*edit*

Just to add, remember the recommended usable ISO range on Dragon is clearly published as cinema ready ISO 250-2000. Beyond that you're looking at new territory outside of what the system is totally optimized for and NR techniques would be ideal.
 
Phil Holland said:
If you want listed benefits and trade offs:

Dragon V1 OLPF
- Similar feel to Mysterium-X OLPF with similar visible artifacts
- Small aperture reflected dot pattern can occur with hot source in frame
- Floating orbs around source highlights can occur in frame
- Sees 1.3 stops more into the shadows (actually midtones through shadows) compared to the V2
- Good color
- No need for IR filtration at various degrees of ND filtration.
- More suited for shooting low light material and rating at higher ISOs

Dragon V2 OLPF
- Doesn't exhibit the reflected dot pattern at small apertures
- Doesn't exhibit the floating highlight orbs.
- 1.3 stops placed in favor of midtones through highlights
- Superior color compared to the V1 OLPF. This is slight to some, more to others.
- No need for IR filtration at various degrees of ND filtration.
- More suited for general shooting conditions
Click to expand...

how can a camera with the drawbacks of OLPF V1 be considered for a production environment? dot pattern or floating orbs might be fixable with a feature movie post budget but for normal people that shot is just ruined?
and doesn't OLPF V2 require new post processing only available in RCX as of right now to get rid of red noise?
and what about the magenta line issues mentioned in several posts here?
and aren't all these issues not really ISO (high or low) related?
 
Phil Holland said:
In regards to the OLPF Module, V1, V2, and whatever future OLPFs land. I see concern from some that this is becoming "too complicated". We need to get away from that thinking.
Click to expand...

Well, it does make specifying a rental more difficult for someone not intimately involved with REDUSER. It's not like RED has clear instructions on their website about OLPFs or anything.

And it also adds complications / labor for anyone renting cameras out if every second client requests a different OLPF and they keep having to swap out / risk damage.

So that's why I'm wondering... couldn't RED split out the OLPF functions so that instead of being a color filter AND an anti-moire filter, it only does anti-moire? Eg it doesn't do spectral stuff?

Then put the color filter of your choice (OLPF v1, OLPF v2, IR only, etc) in front of the lens?

OR, put in OLPF v1... but add an optional color filter that us crazy color nuts can put in front of the lens to get similar to OLPF v2 spectral performance?

http://www.reduser.net/forum/showth...LPF-quot-color-filter-as-a-4x5-65-or-screw-on

Of course, personally, like you Phil, I am a color nut and would rather they just go OLPF v2 as standard and then work on improving sensor signal to noise ratio (through either building better photo-sites / electronics or a larger sensor or both).

Bruce Allen
www.boacinema.com
 
Phil Holland said:
If you want listed benefits and trade offs:

Dragon V1 OLPF
- Similar feel to Mysterium-X OLPF with similar visible artifacts
- Small aperture reflected dot pattern can occur with hot source in frame
- Floating orbs around source highlights can occur in frame
- Sees 1.3 stops more into the shadows (actually midtones through shadows) compared to the V2
- Good color
- No need for IR filtration at various degrees of ND filtration.
- More suited for shooting low light material and rating at higher ISOs

Dragon V2 OLPF
- Doesn't exhibit the reflected dot pattern at small apertures
- Doesn't exhibit the floating highlight orbs.
- 1.3 stops placed in favor of midtones through highlights
- Superior color compared to the V1 OLPF. This is slight to some, more to others.
- No need for IR filtration at various degrees of ND filtration.
- More suited for general shooting conditions
Click to expand...

Swap OLPFs maybe great, if it can be done safely on set, and it doesn't bring another additional problem to the equation. It reminds me of the time when we were using several film stocks for our shoots or still cameras depending on their ASA rating. Even the late Harris Savides, ASC, once said that he would have loved to have different versions of the same HD camera, one adapted to highlights and the other, to low-light situations, when he did "Zodiac" using the Thomson Viper in the mid 2000's.

But having to swap OLPFs it's still a compromise, and not exactly what was announced. I mean, Dragon characteristics were supposed to be the best of both OLPFs, without all this swap thing. I really hope RED will keep improving the camera in that direction.
 
Last edited:
You must log in or register to reply here.
Back
Top