Dragon 5k vs. Dragon 6k

[Alex Behrens]

Hi there!

I was wondering if both sensors are visually the same in terms of cinematic feel? It is just that I believe to see a difference between both footages especially in motion interpretation of the sensor (did not have a one on one comparison though). The 6k feels more like a blockbuster movie and the 5k a little bit less and towards DSLR-ish more fitting for a music or event video. It is my subjective view and I would like to hear from professionals.

What do you think?

 

[Akin A]

"Cinematic feel" is subjective, but technically:

Scarlet-W Dragon 5K has double the data rate and lower compression than Scarlet Dragon 5K.
Epic Dragon 6K improves on data rate/compression over the Scarlets and Weapon Dragon 6K improves on Epic Dragon 6K.

Using the same lens and aperture to get the same shot, 6K image from the larger sized sensor will have a more shallow depth of field than the 5K image, since you would have to move closer to get the same framing.

 

[Aaron Lochert]

The 6k feels more like a blockbuster movie and the 5k a little bit less and towards DSLR-ish more fitting for a music or event video. It is my subjective view and I would like to hear from professionals.

You might just be seeing a correlation between lower budges and higher budgets. The difference in purchasing a RED camera that has a 5K dragon sensor and one that has a 6K dragon sensor is currently $37,500.

The Scarlet-W is targeted toward that music or event video crowd, so it's natural you'll see more of that kind of work. I'm of the belief that you could get a just as capable and "cinematic" (whatever that means) image out of the Dragon 5K if you took as much care as a large production would in crafting the visuals.

Edit: Exhibit A: See Daniel's post below.

 

[Daniel Stilling]

I always like to point out that "Eyes In The Sky", British movie with Helen Mirren was shot on a Dragon, and they only used a 4K extraction. It's not about if you use 4, 5 or 6K out of the Dragon. It's how you light it, block it, production design it, act it... You get the point...

https://www.youtube.com/watch?v=JpenAqCdkh0

 

[Audy Erel]

Sensor size wise, Dragon 5K has the same dimension as Super 35mm 3-perf celluloid film. I think that is the format used by many 'cinematic' works in Hollywood. http://www.red.com/tools/crop-factor

 

[Alex Behrens]

Thanks guys, you’re totally right about the whole production value thing, lighting, setting and so on. And it’s not even the compression or the sensor size what I mean (you can shoot raw with a blackmagic cinema camera too) but even if you put a Blackmagic on a tripod next to a ARRI Alexa, shoot with both cameras, match the colors perfectly and would make two versions of the same movie I think people would feel that something is off with the Blackmagic version, don’t you think? Maybe it’s rolling shutter or how the camera catches the movement. I “guess” I could see the difference between a Scarlet-W Dragon and a Epic Dragon… But I don’t know since there are no such tests so far.

 

[Akin A]

Thanks guys, you’re totally right about the whole production value thing, lighting, setting and so on. And it’s not even the compression or the sensor size what I mean (you can shoot raw with a blackmagic cinema camera too) but even if you put a Blackmagic on a tripod next to a ARRI Alexa, shoot with both cameras, match the colors perfectly and would make two versions of the same movie I think people would feel that something is off with the Blackmagic version, don’t you think? Maybe it’s rolling shutter or how the camera catches the movement. I “guess” I could see the difference between a Scarlet-W Dragon and a Epic Dragon… But I don’t know since there are no such tests so far.

A big part of the reason I switched from BMCC to Scarlet Dragon a few years ago was the motion.
We can see stuff like that because we trained our eyes for it and we geek out on details, but the average person wouldn't see a difference between a mobile phone video and Monstro/Alexa 65, lol.

 

[Misha Engel]

Thanks guys, you’re totally right about the whole production value thing, lighting, setting and so on. And it’s not even the compression or the sensor size what I mean (you can shoot raw with a blackmagic cinema camera too) but even if you put a Blackmagic on a tripod next to a ARRI Alexa, shoot with both cameras, match the colors perfectly and would make two versions of the same movie I think people would feel that something is off with the Blackmagic version, don’t you think? Maybe it’s rolling shutter or how the camera catches the movement. I “guess” I could see the difference between a Scarlet-W Dragon and a Epic Dragon… But I don’t know since there are no such tests so far.

Have a chat with Frank Glencairn and/or John Brawley, they are both heavy Alexa and UMP shooters (one of the reasons might be that these two camera's are easy to match).

 

[Mike P.]

It's a tough one and likely psychosomatic; sometimes I too think there's a difference to the cadence of the capture when comparing full sensor vs. cropped, but realistically it's probably just the post/finishing done on the film that's shooting with the highest-end versions of the camera.

Case in point, Dan's Eyes in the Sky 4k extraction example. Fincher also always shoots with an extraction/crop in mind, including all his R1 and MX features (GWTDT was shot 4.5k WS with 4kWS extraction on R1mx, which taps out at 7.5:1 redcode), and they all look amazing.

Curious though; aside from data rate, do Scarlets actually have a slower sensor readout/rolling shutter than R1s/Epics/Weapons?

 

[Alex Behrens]

It's a tough one and likely psychosomatic; […]

Curious though; aside from data rate, do Scarlets actually have a slower sensor readout/rolling shutter than R1s/Epics/Weapons?

Hey there Just wanted to ask again, if somebody knows if there is a sensor readout difference between these REDs. Thank you in advance!

 

[Björn Benckert]

6k dragon gives more fine noise structure. 5k dragon gives the opertunity to use lenses with slightly smaller image circles = faster lenses or the same lenses with less corner softness, vinjetting, etc. Also 5k capture on weapon dragon allows more data per pixel. So lots of factors in play and difficult to say, I think, that one option is better than the other. Cropping into the sensor does not only have drawbacks, its all depending what else you have in play.

 

[Zack Birlew]

I made posts about this before during our initial search for a digital cinema camera for our first feature film at Babs Do Productions. It was a couple years ago now so we were looking at an Ursa Major, Sony FS5/7, and RED. At the show we saw all of those plus Kinefinity and the Canon C500. Basically we saw similar things with motion in that the non-digital cinema cameras like the Sonys, Panasonics, and Canons all looked a little video like with their overall motion as things looked just too smooth. Then the digital cinema cameras looked absolutely filmic. Then at Blackmagic they had a development camera that seemed to have all the settings off so I was going through to see what it looked like with the settings I would use and I started adjusting what I thought was the shutter speed and it turned out that, after speaking with a developer, what is was adjusting was the sensor refresh rate and for Blackmagic they keep that in sync with the frame rate but that made me realize that the video cameras must have their refresh rates synced with 60 or 50fps at all times and only adjust the frame rate setting. Because when the Blackmagic was set to 60 refresh rate it looked like what was happening with the video cameras but when it was set to 24 it looked like proper film movement. I would think that RED wouldn’t make that mistake but that may be what you’re seeing with other non-digital cinema cameras.

 

[Mike P.]

I made posts about this before during our initial search for a digital cinema camera for our first feature film at Babs Do Productions. It was a couple years ago now so we were looking at an Ursa Major, Sony FS5/7, and RED. At the show we saw all of those plus Kinefinity and the Canon C500. Basically we saw similar things with motion in that the non-digital cinema cameras like the Sonys, Panasonics, and Canons all looked a little video like with their overall motion as things looked just too smooth. Then the digital cinema cameras looked absolutely filmic. Then at Blackmagic they had a development camera that seemed to have all the settings off so I was going through to see what it looked like with the settings I would use and I started adjusting what I thought was the shutter speed and it turned out that, after speaking with a developer, what is was adjusting was the sensor refresh rate and for Blackmagic they keep that in sync with the frame rate but that made me realize that the video cameras must have their refresh rates synced with 60 or 50fps at all times and only adjust the frame rate setting. Because when the Blackmagic was set to 60 refresh rate it looked like what was happening with the video cameras but when it was set to 24 it looked like proper film movement. I would think that RED wouldn’t make that mistake but that may be what you’re seeing with other non-digital cinema cameras.

Wasn't that only on the monitor path, or you mean with the actual footage files the BMD was making? With RED you can set the monitor path to 24, 30, 50 or 60hz (and/or "open gate"), but that's only for the camera's monitor path/playback. I would presume it's the same with BMD... or no?

 

[Michael Lindsay]

I have seen dragon 5k look better than dragon 6k... very unusual situation and it was to do with ridicules amount of detail in the shot and that switching to 5k allowed me less compression. Yes 6k is typically a tiny bit cleaner but well shot 5k dragon is fantastic!

I have a Monstro here but I still think the scarlet W is the best value camera we ever owned.... looking forward to having our next 5k camera (Gemini)... 5k is really not a issue..