My Dragon Thoughts

[Christopher Probst]

The cameras ARE what they ARE. We can either choose to understand their performance characteristics or bury our heads in the sand. Now this test was a very quick, spur of the moment idea designed specifically to see where the imagers fell apart and what that actually looked like. I was not interested in comparing different manufacturers in the same test conditions.

If you are unsatisfied with my sloppy test, by all means, conduct your own when the time comes. I drew what information I needed from them and then thought to share some of that information... Perhaps I will just keep my observations to myself in the future and hoard a competitive edge.

Cheers!

 

[Curran Giddens]

Perhaps I will just keep my observations to myself in the future and hoard a competitive edge.

Noooo!!! No need for that. I'm very interested in learning exactly what advantages come with Dragon, no matter how big or small. Thank you for sharing your tests with us!

 

[David Mullen ASC]

The only tests that are really useful are the ones that show failure points, so thanks Chris! Ultimately it doesn't help to live in a self-imposed bubble where facts are bent to fit preconceived notions or are only allowed when they reinforce those notions, hidden otherwise. I like seeing tests where you can study artifacts at the extreme ends of exposure and color, though it is always important to keep in mind the bulk of most images and their details are not at those extreme ends.

Yes, perhaps the test just shows us what we'd expect, that the newer sensor beats the older sensor... ultimately I'm more interested in how a Dragon Epic compares to Alexa and Sony F65.

 

[Mark L. Pederson]

The only tests that are really useful are the ones that show failure points, so thanks Chris!

Amen.

 

[Gunleik Groven]

The cameras ARE what they ARE. We can either choose to understand their performance characteristics or bury our heads in the sand. Now this test was a very quick, spur of the moment idea designed specifically to see where the imagers fell apart and what that actually looked like. I was not interested in comparing different manufacturers in the same test conditions.

If you are unsatisfied with my sloppy test, by all means, conduct your own when the time comes. I drew what information I needed from them and then thought to share some of that information... Perhaps I will just keep my observations to myself in the future and hoard a competitive edge.

Cheers!

The only tests that are really useful are the ones that show failure points, so thanks Chris! Ultimately it doesn't help to live in a self-imposed bubble where facts are bent to fit preconceived notions or are only allowed when they reinforce those notions, hidden otherwise. I like seeing tests where you can study artifacts at the extreme ends of exposure and color, though it is always important to keep in mind the bulk of most images and their details are not at those extreme ends.

Yes, perhaps the test just shows us what we'd expect, that the newer sensor beats the older sensor... ultimately I'm more interested in how a Dragon Epic compares to Alexa and Sony F65.

Smart men....

When you know the limits of your tool, you can maneuver within them.

If you do not recognize those limits, you will end up f***ed in an unfitting and embarrassing way sooner or later.

 

[Eric Santiago]

Well, i don't get it...

The MX sensor is not broken and does the soul good

Whats not to get, its a comparison test.

We all went through this during RED ONE to EPIC changeover.

Just more options thats all and I love that

Love live my Scarlet!!

 

[Daniel P Rudin]

How does that contradict what i have written?

Nobody denies that wrongly exposed MX Footage falls apart..

But imho there is no need to make the MX bad to make Dragon shine.. this isnt a RED Users only forum.. other places started to take notice..

Yes, perhaps the test just shows us what we'd expect, that the newer sensor beats the older sensor... ultimately I'm more interested in how a Dragon Epic compares to Alexa and Sony F65.

totally agree here..

 

[Zeb B]

The only tests that are really useful are the ones that show failure points, so thanks Chris! Ultimately it doesn't help to live in a self-imposed bubble where facts are bent to fit preconceived notions or are only allowed when they reinforce those notions, hidden otherwise. I like seeing tests where you can study artifacts at the extreme ends of exposure and color, though it is always important to keep in mind the bulk of most images and their details are not at those extreme ends.

Yes, perhaps the test just shows us what we'd expect, that the newer sensor beats the older sensor... ultimately I'm more interested in how a Dragon Epic compares to Alexa and Sony F65.

Damn . . . I love this man!

 

[Christopher Probst]

How does that contradict what i have written?

Nobody denies that wrongly exposed MX Footage falls apart..

But imho there is no need to make the MX bad to make Dragon shine.. this isnt a RED Users only forum.. other places started to take notice..



totally agree here..

Reread David's response and if you still don't see the merits of what I shot, I cannot help you.

 

[Daniel P Rudin]

Reread David's response and if you still don't see the merits of what I shot, I cannot help you.

I see where the merits of your tests are.. Dragon can be pushed far and it behaves gracefully.. i like that too, no question about that..
But, the big but.. a shoot against Alexa or F65 etc. should be publicated in my opinion.. that is a fair comparisation..

Because the lack of Dragon shoots againt other Camera Manufacurers, every Dragon footage posted at the moment is being evaluated from different circles.. for different purposes..
you and me are using these results to know exactly how far we can go and what happens.. others may say.. Dragon looks like Alexa.. and MX like shit..

thats the reason i'm being protective of MX.. people tend to take results out of context..

just some food for tought..

 

[Brian F Kobylarz]

Chris,

Great information - really looking forward to the call to send my EPIC in for the upgrade. Everyone that has had a chance to work with the Dragon has indicated that the improvements are obvious out of the gate.
Asking a gut opinion only - are the differences apparent enough that matching an X and Dragon sensor together on set will be workable?

 

[David Mullen ASC]

People always take tests out of context, it's as inevitable as the rising of the sun, the only solution would be to ban all tests, or at least, never show them to anyone.

I don't think it is reasonable for Red to not promote -- and demonstrate -- that Dragon is an improvement over the MX sensor just to keep MX owners happy. When Kodak introduced a new film stock, they always showed a split-screen or other test showing the improvements over the old stock, even if that old stock was on the market. They weren't trying to make the people who used the older stock feel bad about their purchases.

Every time Canon or Nikon puts out a new digital still camera, everyone wants to know what the improvements are over the old camera. That's just commonplace in this day and age. Doesn't stop people with the former model from showing the world that they can still take gorgeous images with it.

As for producers wanting to rent a Red with the newest sensor in it, that's also inevitable, it happened when the MX sensor came out. Everyone saw plenty of demonstrations that the MX sensor was an improvement over the M sensor.

 

[Michael Epple]

Great test. Finally seeing a comparison. I know it's not a definitive test, but it's looking like less than 3 stops of latitude improvement. Maybe there is more that can be pulled out, but I'm sure that's true of both. The increased contrast is a nice surprise, as well as the color fidelity.

 

[Neil W. Smith]

the Pros and Cons of camera tests ...

the Pros and Cons of camera tests ...

People always take tests out of context, it's as inevitable as the rising of the sun, the only solution would be to ban all tests, or at least, never show them to anyone.

I don't think it is reasonable for Red to not promote -- and demonstrate -- that Dragon is an improvement over the MX sensor just to keep MX owners happy. When Kodak introduced a new film stock, they always showed a split-screen or other test showing the improvements over the old stock, even if that old stock was on the market. They weren't trying to make the people who used the older stock feel bad about their purchases.

Every time Canon or Nikon puts out a new digital still camera, everyone wants to know what the improvements are over the old camera. That's just commonplace in this day and age. Doesn't stop people with the former model from showing the world that they can still take gorgeous images with it.

As for producers wanting to rent a Red with the newest sensor in it, that's also inevitable, it happened when the MX sensor came out. Everyone saw plenty of demonstrations that the MX sensor was an improvement over the M sensor.

Agree wholeheartedly, David ... having been involved in numerous ASC, CAS, Zacuto and Geoff Boyle's camera tests I can attest to the problematic and controversial nature of trying to compare camera's characteristics - there will always be someone with an axe to grind and an agenda to push, which is why I haven't gone into too much pixel peeping with the DRAGON footage we have. The only thing I would say is that DRAGON is definitely a step up from the M-X sensor in every sense of the word ... dynamic range, resolution and overall image quality ... it seems that Jarred and his team have managed to combine some new technologies and some secret sauce to make a whole new kind of camera.

The DRAGON is not an upgrade to the M-X in the same sense that the M-X was an upgrade to the Red One M series .. the form factors look similar but under the hood this is a different camera all together ... the smart thing that RED have done for all current RED cameras owners is to make the "upgrade" path both painless and relatively inexpensive and keep all the accessories the same.

The entire RED DRAGON ecosystem is evolving at a rapid rate .... it will surely have an impact on perceptions of the relative merits of RED M-X versus DRAGON and there is no escaping the fact that the new camera is better in almost every way imaginable than the current RED M-X ... the price of progress I'm afraid.

We had the pleasure of grading the very first images off the original RED One way back in the summer of 2006 and have worked on all the upgrades since then ... have to say that if RED had delivered the DRAGON a few years back, the landscape of cinematography and camera choice would be quite different to how it is now.

We put together some of the DRAGON footage that Ketch shot into a fun, one minute sizzle reel just to prove the point that you can make all kinds of content from the DRAGON .. from music videos to serious drama ... this is not meant for pixel peeping - see Phil's, John's and Christopher's illuminating threads if you want that level of detail:

http://www.lumaforge.com/FLOworks/index.html

We'll also be showing the DRAGON footage with FULL debayer using the Red Rocket-X at the CreateaSphere ETE conference in Burbank next week ... we're sandwiched in between the Sony and Canon booths so we should create a bit of excitement to say the least.

Hope to see some of you at ETE next week - it's on 6th and 7th at the Burbank Marriott Hotel near the airport ... as well as our Booth #201, I'll be also leading a presentation on 4K workflow in Theater #1 on Thursday afternoon at 2pm ... a couple of ASC DPs and I will be showing footage from EPIC, Sony F55, Canon C500 and 1DC ... oh, I wonder if there might be a little bit of DRAGON footage shown at the end just for good measure ;-)

http://createasphere.com/event/speakers/entertainment-technology-expo-2013

Look out for the LumaFLOW banner on the ETE booth ...

Cheers,
Neil

 

[paul schefz]

a serious test vs alexa and sony (and incl both dragon and epic) would be great and i am sure will happen but until then a well exposed side by side dragon vs epic would be more interesting to me....using the same lens, the same "safe" base iso for both, same compression...and then maybe get creative with it (iso, compression, trying to align what the lens sees on both cameras,...)
i looked at the files again and there is more noise in the shadows with the dragon files and the strangest thing to me is that i really can't tell which file is which when i am looking for resolution or detail....14mpix vs 19 should be a bigger difference?
the only way i would be able how much DR one has over the other is a well exposed image or in this case probably f11...and look which camera still holds the sky without drowning out the shadows....

 

[Jason L]

People always take tests out of context, it's as inevitable as the rising of the sun, the only solution would be to ban all tests, or at least, never show them to anyone.

Tests are good when used to see how a camera performs, but I think they are very flawed too. A film is never just random people standing in a room walking around turning lights on and off, or staring right into a camera. Camera tests should be actual films, because the camera is only a small part of what the viewer actually takes in.

 

[Christopher Probst]

a serious test vs alexa and sony (and incl both dragon and epic) would be great and i am sure will happen but until then a well exposed side by side dragon vs epic would be more interesting to me....using the same lens, the same "safe" base iso for both, same compression...and then maybe get creative with it (iso, compression, trying to align what the lens sees on both cameras,...)
i looked at the files again and there is more noise in the shadows with the dragon files and the strangest thing to me is that i really can't tell which file is which when i am looking for resolution or detail....14mpix vs 19 should be a bigger difference?
the only way i would be able how much DR one has over the other is a well exposed image or in this case probably f11...and look which camera still holds the sky without drowning out the shadows....

Jarred has already announced a significant improvement in calibration on both a hardware and software level since I shot those clips...

Dragon still is using the same old RC3/RG3 profile while the tuning of the camera continues. I'm already quite happy with what I'm seeing. And it's only going to get better. I had not one firmware issue on my shoot at all, traveling to 3 countries, swapping out motion mounts... Ext.

Even though the Dragon is effectively a brand new camera, it's not starting from groung zero. It's got miles of a heads tart compaired to when Epic was first released.

 

[David Mullen ASC]

Tests are good when used to see how a camera performs, but I think they are very flawed too. A film is never just random people standing in a room walking around turning lights on and off, or staring right into a camera. Camera tests should be actual films, because the camera is only a small part of what the viewer actually takes in.

There are ALL KINDS OF TESTS. One type isn't necessarily more valid than another. A test where the camera points at line resolution charts, dynamic range charts, color charts, etc. is just as valid as a test of a typical narrative set-up that the movie is likely to shoot. So is a test that just shows under and over exposures of a face.

Personally most of the high-end pro cameras do fine in a nicely-lit scene, putting a pretty actress in a nice dress on a nice set with some pleasant lighting. Most look fine outside in daylight. When I shoot a test, I need to find out what happens when mistakes are made, when I encounter an extreme contrast situation, when I have to really push the camera in low-light. I need to know where it falls apart so I can stop myself one step before that point is reached. When a lot of people post tests, i usually just go to the extreme stuff, I figure the camera can handle the easy stuff just fine or else the manufacturer probably wouldn't have released it.

With the internet, the day when you could hide the flaws of some technology are pretty much behind us -- the truth always comes out as soon as the technology is available to the public. And ultimately the camera will be used both by people who know what they are doing and people who don't.

Sure, there is such a thing as first impressions, that you can only make one once, so it behooves a camera manufacturer to release something that is ready for public scrutiny and not too much in beta. But once it's out there, the product has to stand on its own.

My belief is that the more tests, the better. The more footage gets released, the more you can start to see trends and average the results. Even if people make the same mistakes shooting their tests, that also tells you something. It's always in the early days when there are only a few tests to see that people get overly sensitive.

 

[Robert Ruffo New]

Tests are good when used to see how a camera performs, but I think they are very flawed too. A film is never just random people standing in a room walking around turning lights on and off, or staring right into a camera. Camera tests should be actual films, because the camera is only a small part of what the viewer actually takes in.

No they shouldn't because those kinds of tests do not tell you anything about where the camera falls down - they only show you ideal conditions, and even my iPhone has ideal conditions where it can look OK. You learn almost nothing from camera "sizzle reels' unless they are showing you specific tricks the camera can perform at a technical level.

Rigorous test shots designed as a tests - not designed to be "good" shots, is where you learn what a camera can or cannot do, where you determine whether it will save you money in your lighting package (lower wattages at night) allow for more aggressive secondary grading, etc. Watching someone's completed film is utterly useless for any of this.

EDIT: OOPS above David was posting some of the same things as me at the same time, but as usual beautifully expressed.

 

[Jason L]

Miss post.