Dragon??? So after nearly 10months what do we think? NATIVE 2000ASA no 18stops no ..

[Bohdan Klos]

You're mischaracterizing OLPF 1's color performance as "reduced color fidelity." Saying the OLPF 2 has improved color fidelity is NOT the same as saying OLPF 1's fidelity is reduced.

Split hairs how you want - but the V1 colour is closer to MX than it is to V2 OLPF. My point is there is a big difference in colour fidelity between the OLPFs as detailed by Jarred.

skintones .... isn't as good as the new OLPF,

the new OLPF ... the vastly improved color science ...

after testing both extensively over and over and over again. It is the smart man's choice.

 

[Sergio Perez]

I hope the RED TEAM is monitoring this thread. Personal jabs aside, great info and opinions here. Red always listens. I think its time to take all this in and come up with the usual RED kick on the b@lls shocker solution that blows us all out of the water. We are All red users here and I believe- diverging opinions withstanding- everyone just wants the best for the system they support/use/invested/love. We are all in this together. There's something certainly worth discussing and improving here. Lets do this.

 

[Isaac Marchionna]

I hope the RED TEAM is monitoring this thread. Personal jabs aside, great info and opinions here. Red always listens. I think its time to take all this in and come up with the usual RED kick on the b@lls shocker solution that blows us all out of the water. We are All red users here and I believe- diverging opinions withstanding- everyone just wants the best for the system they support/use/invested/love. We are all in this together. There's something certainly worth discussing and improving here. Lets do this.

I certainly hope so, and I hope the solution is measured and restores confidence to those of us for whom a 9500 dollar purchase is and was a major chunk of change. Specifically this new OLPF system, personally this should be a free retroactive upgrade. Some of us literally can't afford to keep chasing the dragon if it's fix after fix.

 

[luigivaltulini]

It's nice that you work with producers who are willing to pay extra to shoot now at 250 ISO, when they got used to shooting at 800 with MX and Alexa. I don't know anyone like that, and frankly, my clients were shocked to learn that Dragon, supposedly an upgrade, not a side-grade new option, was less sensitive than the older MX and required more light budget, which means not only more light but more electricity, more crew at a certain point, more everything. If Dragon had been at least as good as MX in all ways, then there would be no problem, but I for one am not going to blow thousands per day on more lighting just to get slightly improved skintones (that I can quite easily match in the grade anyway) and a mostly useless, excessive overkill of resolution, that is much more a hard drive clogger and a card transfer bottleneck than in any real way an advantage for non-theatrical (and even then…) production.

People keep talking about film, but many things were not possible on film, like shooting a cityscape at night, and only lighting your actors, not the distant buildings in the background. Do you know how much it costs to light a skyscraper in a city, just in permits and logistics? What are we supposed to do, stop doing shots like that now, because Dragon skintones are slightly better? Because on film they weren't possible? I think not.

This is the first time that a new camera model is not only better than an old camera model, of any brand that I can remember.

It's not just that there is more noise in the darks, it's that the curve to black is too steep - that darks actually hard clip and there is very little information to get from down there. This has been shown many times. My own tests and Ryan Walters lead to the same conclusions.

That's not something noise reduction could ever fix, and now with Dragon many shots that were possible with MX no longer are, and even when they are possible, require much more time and expense to create without problems, as one must shoot at 250 ISO. 250 ISO (or even 500) means that a lot of the time, the camera just can't make the shot I want, can't do what my Epic MX does with ease. Even when money is unlimited, actor comfort is a factor, and actor eye fatigue, and some places just won't let you whip out 10Ks and shine them in people's windows, no matter how much cash you have.

I mean Ryan Walters (and I agree with him) came out and said that certain lighting ratios that were fine on MX don't look good on Dragon. So now I can't do certain styles of lighting anymore? Really? I;m supposed to just cross styles of set-ups I very much like to do to "adapt" to Dragon? Sounds like the tail wagging the dog to me - a camera is supposed to get out of your way, not dictate how you shoot.

But hard-clipping blacks and noise are not the only problems, we have:

CMOS smear (magenta lines around bright lights, running all the way across the frame). Yes this occurs on all CMOS cameras, but on most it's so subtle you barely see it, or not at all. On Dragon it can seriously ruin a shot, much too often, way more often than I have ever seen on any CMOS device. It's like skew - sure all CMOS camera have it, but not all are as bad as a 5D Mark II. So some here say "Don't shoot bright lights in frame" Hmmm… That's one more type of shot a Dragon can't do, that is no problem on my Epic MX, let me get out my notebook, the list is getting long….

Flare shadow. I agree with my fellow Canadian Mr Steinberg. This is a real problem. I like in-camera flares. On MX I never worry about them. Good thing I didn't put away my notebook yet.

Oh, and I'm told that if I don't want to be restricted to super-low, much more expensive to light ISOs then I should stop complaining and just use the old OLPF, which has all kinds of flaring weirdness and can't really be trusted. Gee, what an amazing option! As yet another wonderful alternative, I can spend tons of precious set time futzing with tiny screws and switching OLPFs on set! Really!?

I for one am skipping the Dragon upgrade. Even if it were free, I've come to the conclusion that the MX is actually a better camera, all things weighed together, as it is significantly more versatile and less likely to let me down, and some producers and other DPs I talk to are, somewhat in shock, slowly because it's kind of hard to believe, coming to the same conclusion after a slew of bad experiences. Slightly better skin tones that you can equal in a good color grade anyway, and highlight retention I never clip on MX anyway (because hey, I can expose for 800ISO, where this is almost never a real-world problem…) does not outweigh all the negatives and serious issues, at least not for how we and our clients shoot. Again, this even if the upgrade cost $0.

At some level it could even be argued, by reasonable people, that a Red One MX is more versatile and can guarantee more shots being acceptable under a greater variety of circumstances than a Dragon. But its' not free, and what with all the problems circulating around the grapevine (people see the problems on set and in post, not just on these forums, so no, it's not because people have been complaining here too much) it's a very, very hard sell at this point, to producers and clients, yes, but even to myself.

There is no way I would ever amortize back that $10 000 since I would need to be renting MX Epics often just to keep getting many types of shots I like in the can, and the whole fiasco list above has meant that the Dragon rental market is not exactly hot at this stage, to say the least. Elsie may not understand this, as he does not actually work on paid sets, by his own admission, but the most important thing for a camera to be is predictable and reliable - no hidden bad surprises - that's more important, in fact, than potential performance capabilities. An Epic MX is exactly that now. A Dragon? No.

Now maybe DEB or some other advance will make this completely different. For now, saying no (I have the money set aside, it's not the issue, at all) seems by far the most sensible thing to do.

After all, some of the most beautiful films ever made were shot on Red One MX and Epic MX, and we have both fully kitted here.

If you can't make beautiful, perfect even, images on any MX camera then you just suck and should quit now.

Robert, please let're a very intelligent person and character.
Please do not overdo it as they are doing all of them.


1. The magenta line for Dragon I do not see differences with MX, this also applies Shot on Alexa.
I'd rather have the lines with DRagon that with MX, especially if you use light to tungsten.
So Dragon is the same as the MX, and like all CMOS sensors, more than those who less.


2. Deuce to say that Dragon be worse than MX 800 iso, is absolutely not true, as many Phil, John, Jan, Toia etc. etc. I have shown examples where esatamente be the opposite. I have posted video clips, commercials filmed in 1600/2000 iso very good.
So with the new Dragon OLPF is better than MX 800 iso the answer is YES, both externally and internal light Tungsteno.Ma know that's nice to have 2 or 3 more stops in the high lights, they are more calm on the set with MX, If you want then as it was promised in 2000 iso clean and miss a stop put back the old OLPF. But I assure you that once porvate DRagon with the new OLPF you will not go back.


3. Updates of this money worth them all, especially for the color and texture that can give this sensor.


Robert many tests are not part of real life on the set, I sincerely hope that all this fuss is calm, but it will happen only when the Dragon will be many out there and people will realize how good this camera. For now we sistuazione mixed.


As much as I like the MX, I can not think of going back just for a moment, as I all my clients for many years where I was bitten by Alexa now are all over the Dragon and I'm very happy. For me this' status confirmation.


I personally agree on only one factor that unfortunately Dragon is not clean in 2000 as it was with the old OLPF, for the rest I confirm what I said.


my thoughts, sorry for my bad English Robert: (

These are lines with MX (without correction I took the first pictures from the repertoire), are better than ... DRAGON?!?!?!?!

[/URL][/IMG]

[/URL][/IMG]

 

[Jan Reiff]

Robert, please let're a very intelligent person and character.
Please do not overdo it as they are doing all of them.


1. The magenta line for Dragon I do not see differences with MX, this also applies Shot on Alexa.
I'd rather have the lines with DRagon that with MX, especially if you use light to tungsten.
So Dragon is the same as the MX, and like all CMOS sensors, more than those who less.


2. Deuce to say that Dragon be worse than MX 800 iso, is absolutely not true, as many Phil, John, Jan, Toia etc. etc. I have shown examples where esatamente be the opposite. I have posted video clips, commercials filmed in 1600/2000 iso very good.
So with the new Dragon OLPF is better than MX 800 iso the answer is YES, both externally and internal light Tungsteno.Ma know that's nice to have 2 or 3 more stops in the high lights, they are more calm on the set with MX, If you want then as it was promised in 2000 iso clean and miss a stop put back the old OLPF. But I assure you that once porvate DRagon with the new OLPF you will not go back.


3. Updates of this money worth them all, especially for the color and texture that can give this sensor.


Robert many tests are not part of real life on the set, I sincerely hope that all this fuss is calm, but it will happen only when the Dragon will be many out there and people will realize how good this camera. For now we sistuazione mixed.


As much as I like the MX, I can not think of going back just for a moment, as I all my clients for many years where I was bitten by Alexa now are all over the Dragon and I'm very happy. For me this' status confirmation.


I personally agree on only one factor that unfortunately Dragon is not clean in 2000 as it was with the old OLPF, for the rest I confirm what I said.


my thoughts, sorry for my bad English Robert: (

These are lines with MX (without correction I took the first pictures from the repertoire), are better than ... DRAGON?!?!?!?!

[/URL][/IMG]

[/URL][/IMG]

+1
i think the reason why some think MX might be cleaner is, that the Dragon "grain" is much finer and compared to Dragon MX is much more muddy in the shadows like a NR. Dragon keeps more Details in "bad light", and so the grain is sharper and more visible. that´s what i am watching when i compare. i am bit speechless about the whole discussion ... maybe because i did compare some weeks ago to other named cameras and there are not really "better". i do not want to change the Dragon. Expose it right and throw it on a 4K screen. That´s entertaining for the whole family i tell you.

 

[Brian Boyer]

Split hairs how you want - but the V1 colour is closer to MX than it is to V2 OLPF. My point is there is a big difference in colour fidelity between the OLPFs as detailed by Jarred.

You know exactly what you're doing when you word things a certain way. You're trying to make the case that RED offering the OLPF 1 as an option is bogus because, by their own "admission," it's not as good as the new one and therefore must necessarily be unusable.

Not only is that simplistic but you know that's not really the case and you're purposely trying to manipulate the conversation. You're applying straw man logic, which serves no one.

You have some legitimate points but you can't help trying to distort the facts in order to make them.

 

[Jeff Kilgroe]

Jeff

A very well put point. I remember being told during my cinematography studies that according to Kodak a release print probably averaged about 1.4K of resolution, which may have been an optimistic guess.

It is an analogue medium, after all.

JM

That is actually a different way of saying the same thing.

When I was using the reference of 800 lines of resolution, I was speaking of a vertically measured quantity, as that is how such things are typically quantified. When we enter the digital realm of greater than HD, resolutions are usually quantified in horizontal pixel measurements by thousands (K) or overall pixel quantity (MegaPixels). 1.4K from a release print would actually be pretty close to the mark as they tend to hover right around HD resolution in terms of actual resolved detail. And by HD, I mean 1200 to 1600 ticks measurable across the horizontal axis. 1.4K is smack dab in the middle of that range.

 

[Jan Reiff]

Lucky Arri Company that no one complains about the fact that a camera that costs just 1/3 of an Alexa (C300) is better in low light - maybe the Alexa clients know, that other image aspects are more important in a professional set ?



http://www.creativeplanetnetwork.com/news/news-articles/alexa-vs-c300-vs-f55-low-light/407658

 

[Mike P.]

I haven't used V1 OLPF, only V2, but from what I gathered from the V1 .r3ds I downloaded, the only real issue with V1 was the magnets-orbs around blown highlights (somewhat insignificant compared to the noise/ISO250-limit of the V2, in my opinion). People talked about colour fidelity being way better on V2, but I just (maybe wrongfully) figured those things could be tweaked in the grade being that its 16bit raw from the exact same sensor.

Conversely, the 'taking from the bottom and putting it on top' that the V2 OLPF does, should, in theory, be easily reproducible by rating the V1 OLPF at a 1-2stop higher ISO (ISO3200?), forcing the user to expose darker and protecting the highlights as a result (basically what we've been doing with MX by rating it at 800 in exteriors). That would give you the 1-2stops more in the highlights, whilst avoiding V2s haggard noise and very shallow shadows. Or is this bad logic?

The irony is, I do recall people (Boyle/CML specifically) saying that Dragon with OLPFV1 was 'only ~14 stops or so', and everyone here claiming 'the new OLPF will give you an additional 1.5 stops of DR'.... Fast forward to now, and it seems to be the same 14 stops, just taking from the bottom and putting on top, but at a much lower ISO.

 

[Bohdan Klos]

You have some legitimate points but you can't help trying to distort the facts in order to make them.

Legitimate points made from (alleged) distorted facts is still....legitimate points.
I think you are saying we agree. Great.

This thread is full of valuable opinions.

 

[Matt Ryan]

The irony is, I do recall people (Boyle/CML specifically) saying that Dragon with OLPFV1 was 'only ~14 stops or so', and everyone here claiming 'the new OLPF will give you an additional 1.5 stops of DR'.... Fast forward to now, and it seems to be the same 14 stops, just taking from the bottom and putting on top, but at a much lower ISO.

That is exactly what is happening. Taking from the low end and putting it up top. It's funny when people claim the new OLPF increased the DR. FACT: it did not. All it did was shift the calibration up improving the high end and taking away from the low end.

Remember in that same test where Geoff Boyle also said Dragon was way cleaner than Alexa. Now it's noisier than the Alexa...

While I enjoyed the old OLPF better, those orb flares were disgusting and if it didn't have them then that would have been a great camera.

 

[paul schefz]

It's nice that you work with producers who are willing to pay extra to shoot now at 250 ISO, when they got used to shooting at 800 with MX and Alexa. I don't know anyone like that, and frankly, my clients were shocked to learn that Dragon, supposedly an upgrade, not a side-grade new option, was less sensitive than the older MX and required more light budget, which means not only more light but more electricity, more crew at a certain point, more everything. If Dragon had been at least as good as MX in all ways, then there would be no problem, but I for one am not going to blow thousands per day on more lighting just to get slightly improved skintones (that I can quite easily match in the grade anyway) and a mostly useless, excessive overkill of resolution, that is much more a hard drive clogger and a card transfer bottleneck than in any real way an advantage for non-theatrical (and even then…) production.

People keep talking about film, but many things were not possible on film, like shooting a cityscape at night, and only lighting your actors, not the distant buildings in the background. Do you know how much it costs to light a skyscraper in a city, just in permits and logistics? What are we supposed to do, stop doing shots like that now, because Dragon skintones are slightly better? Because on film they weren't possible? I think not.

This is the first time that a new camera model is not only better than an old camera model, of any brand that I can remember.

It's not just that there is more noise in the darks, it's that the curve to black is too steep - that darks actually hard clip and there is very little information to get from down there. This has been shown many times. My own tests and Ryan Walters lead to the same conclusions.

That's not something noise reduction could ever fix, and now with Dragon many shots that were possible with MX no longer are, and even when they are possible, require much more time and expense to create without problems, as one must shoot at 250 ISO. 250 ISO (or even 500) means that a lot of the time, the camera just can't make the shot I want, can't do what my Epic MX does with ease. Even when money is unlimited, actor comfort is a factor, and actor eye fatigue, and some places just won't let you whip out 10Ks and shine them in people's windows, no matter how much cash you have.

I mean Ryan Walters (and I agree with him) came out and said that certain lighting ratios that were fine on MX don't look good on Dragon. So now I can't do certain styles of lighting anymore? Really? I;m supposed to just cross styles of set-ups I very much like to do to "adapt" to Dragon? Sounds like the tail wagging the dog to me - a camera is supposed to get out of your way, not dictate how you shoot.

But hard-clipping blacks and noise are not the only problems, we have:

CMOS smear (magenta lines around bright lights, running all the way across the frame). Yes this occurs on all CMOS cameras, but on most it's so subtle you barely see it, or not at all. On Dragon it can seriously ruin a shot, much too often, way more often than I have ever seen on any CMOS device. It's like skew - sure all CMOS camera have it, but not all are as bad as a 5D Mark II. So some here say "Don't shoot bright lights in frame" Hmmm… That's one more type of shot a Dragon can't do, that is no problem on my Epic MX, let me get out my notebook, the list is getting long….

Flare shadow. I agree with my fellow Canadian Mr Steinberg. This is a real problem. I like in-camera flares. On MX I never worry about them. Good thing I didn't put away my notebook yet.

Oh, and I'm told that if I don't want to be restricted to super-low, much more expensive to light ISOs then I should stop complaining and just use the old OLPF, which has all kinds of flaring weirdness and can't really be trusted. Gee, what an amazing option! As yet another wonderful alternative, I can spend tons of precious set time futzing with tiny screws and switching OLPFs on set! Really!?

I for one am skipping the Dragon upgrade. Even if it were free, I've come to the conclusion that the MX is actually a better camera, all things weighed together, as it is significantly more versatile and less likely to let me down, and some producers and other DPs I talk to are, somewhat in shock, slowly because it's kind of hard to believe, coming to the same conclusion after a slew of bad experiences. Slightly better skin tones that you can equal in a good color grade anyway, and highlight retention I never clip on MX anyway (because hey, I can expose for 800ISO, where this is almost never a real-world problem…) does not outweigh all the negatives and serious issues, at least not for how we and our clients shoot. Again, this even if the upgrade cost $0.

At some level it could even be argued, by reasonable people, that a Red One MX is more versatile and can guarantee more shots being acceptable under a greater variety of circumstances than a Dragon. But its' not free, and what with all the problems circulating around the grapevine (people see the problems on set and in post, not just on these forums, so no, it's not because people have been complaining here too much) it's a very, very hard sell at this point, to producers and clients, yes, but even to myself.

There is no way I would ever amortize back that $10 000 since I would need to be renting MX Epics often just to keep getting many types of shots I like in the can, and the whole fiasco list above has meant that the Dragon rental market is not exactly hot at this stage, to say the least. Elsie may not understand this, as he does not actually work on paid sets, by his own admission, but the most important thing for a camera to be is predictable and reliable - no hidden bad surprises - that's more important, in fact, than potential performance capabilities. An Epic MX is exactly that now. A Dragon? No.

Now maybe DEB or some other advance will make this completely different. For now, saying no (I have the money set aside, it's not the issue, at all) seems by far the most sensible thing to do.

After all, some of the most beautiful films ever made were shot on Red One MX and Epic MX, and we have both fully kitted here.

If you can't make beautiful, perfect even, images on any MX camera then you just suck and should quit now.

thanks for that....you are pissing in the wind....it has become impossible to have a discussion about this here...unfortunately...every thread like this turns into the same couple of people making the same excuses over and over...
the problems are real, very real and MX is the better camera for now....my biggest concern is that this (dragon) was supposed to be the answer and even if the issues are ironed out to get anywhere near the promised specs, in reality RED should be working on the next chip right now behind the scenes...instead of constantly having to meddle with a now already 2 year old one that just won't do exactly what it should be doing....
this is not about hating or bashing or the more and more desperate "don't worry, they will fix it"....it is my concern as a customer and professional who still believes that this is the best system for me right now out there....unfortunatley that system is still MX and that is a little strange and hard to swallow......

 

[paul schefz]

That is exactly what is happening. Taking from the low end and putting it up top. It's funny when people claim the new OLPF increased the DR. FACT: it did not. All it did was shift the calibration up improving the high end and taking away from the low end.

no problem! red will come up with TTS two track shooting....one with new olpf, one with old....problem solved....ahm wait a minute?

 

[Martin Stevens]

To hear people say that they view MX as being better than Dragon is perplexing.

For me the DRAGON outshines MX.

 

[David Battistella]

Dragon was indeed tested in February with original OLFP, to a chorus of it's not the latest OLFP which is much better.

The latest OLPF is actually much better.

 

[Elsie N]

To hear people say that they view MX as being better than Dragon is perplexing.

For me the DRAGON outshines MX.

You do understand that you've just spewed blasphemy to those who are perpetual fault-finders?

 

[Martin Stevens]

You do understand that you've just spewed blasphemy to those who are perpetual fault-finders?

Yes, I guess so. And since I love how film handles highlights the new OLPF really shines.

 

[David Battistella]

I'd like someone to post some graded shots from films or from magazines (and I don't mean shots from real estate magazines) with 16 or even 14+ stops of DR.

I know that people know that Redlog film is not intended to be the final image, it's there to show you all the DR captured. I'm trying to get my head around what people really expect(ed) from this sensor. I'm being sincere.

Battistella

 

[Brian Boyer]

Legitimate points made from (alleged) distorted facts is still....legitimate points.

It does show poor logic skills, though. There is data, and then there is interpretation of that data. Coming to the right conclusion based on flawed logic and incorrect assumptions is generally frowned upon in the scientific/academic community. But hey, some people are doing the best they can with what they have.

To the extent that manipulating the facts doesn't bother you, it tends to hurt your argument, not bolster it.

I think you are saying we agree. Great.

You would, considering how prone you are to distort things.

I said "some." You think the red speck noise is an issue. I agree. You think OLPF 1 isn't production ready, despite entire projects being shot on it. I disagree. See how that works?

 

[Martin Stevens]

It think that because there is 16+ stops of DR, people are finally seeing too much info and it confuses their eye.

Just like if you showed the uninitiated a scan of a film negative they would complain and complain about all sorts of things
until you told them that, hey, this won't be what is projected in the theater.