Alexa vs. Dragon vs. MX

[JanneJansson]

All hail A7! This feels like an ad paid by sony.

this is pretty interesting.....http://www.cinema5d.com/dynamic-ran...ira-canon-c300-5d-mark-iii-1dc-panasonic-gh4/
if sony decides to put that sensor in their next cine body.....

 

[Matthew Scott]

Then why do tests at all?

Exactly.

...you are also in some way saying that the upgrade between Epic and Dragon isn't worth it, since a final production won't be much of a difference.

Exactly.

I have been able to pinpoint what was shot on Red and what was shot on Alexa when watching movies and television. There are differences.

Don't fool yourself. If the same DP, on the same film, on the same set, with the same actors read the same script, directed by the same guy/girl, and the film was then finished in the same post house, you would have NO CLUE if it were RED, or Alexa. Especially if both cameras were used and shots were intercut, like a "real world" test.
So, yes, you are right, there are differences but they are so barely visible (especially once processed) that they aren't worth worrying about. It's useful to know these differences though, which is why I enjoy these tests, but only for the sake that it's important to know the strengths and weaknesses of any camera you are shooting with/thinking about purchasing. What get's me going is how the Dragon is being pushed as breath-takingly different, light-years better and a drastic improvement over MX, when really, it's a 1.5 stop increase in DR (highlights only) with slightly better colour. I should also mention that it has better compression which is an excellent advancement IMHO, but hey...image wise, I'm not blown away like I'm lead to believe around here. I know responses like these don't win friends, but I'm not about that. I love shooting RED and I love where RED is taking camera technology. I'm all for any sort of discussion that can empower the end user to be a better cinematographer.

 

[Bruce Allen]

OK... so is it just me or does it seem to show:

1. Dragon's color is looking nice now
2. Alexa has slightly more range than Dragon in the highlights at most ISOs
3. Alexa is less noisy than Dragon in the shadows at high ISOs

It's nice to see someone do what everyone on REDUSER seems to tiptoe around. Eg when you want to compare Alexa / Epic / Dragon, suddenly you get an absence of clear tests and pages and pages of doublespeak. This does not exactly inspire confidence because you know that if Dragon comprehensively beat Alexa in DR or low noise or whatever then people would be going nuts and posting tons of stuff. Instead we get a bunch of non-useful tests and very little real meat. The average guy like me says "hmm, this Dragon obviously falls short or else why would the owners be so darn defensive?"

I do think this shows that if you took whatever testing methodology RED and others got to get 16.5+ stops out of Dragon and applied that to the Alexa, you'd get even MOAR. And I don't really consider the Alexa a 17-stop camera. So, yeah, this keeps reminding me to keep adding a few grains of salt to those sources' words because they are obviously prioritizing their own agendas over objectivity. That's silly because if you look at Dragon dead-on honestly, it's a great camera. I must say that ARRI are coming out really well here because it actually appears that they UNDER-promised on their sensor's specs. That's how you build (and keep) brand confidence. Just as RED's policy of updates that add stuff to their cameras beyond what they originally promised are probably the best thing about the company (well, apart from the insane price/performance ratio).

That said, clear tests like these actually make me want to rent Dragon MORE. It's cheap compared to Alexa, this test shows that it looks comparatively good when downsampled to HD (if you give it enough light), and it has a bunch of advantages for special shoot needs (high-res, high-FPS, etc). I never liked the Mysterium MX sensor's colors (and took a lot of flak for it on REDUSER) but this Dragon sensor is a big improvement. Yay. Oh yeah... and the cherry on the top is I have a bunch of friends who own Dragons (and none with Alexas) so that's the decision-clincher for me Of course, most of the time, for workflow reasons and client reasons it's Alexa, sadly. But that's another thread.

Bruce Allen
www.boacinema.com

 

[Matthew Scott]

OK... so is it just me or does it seem to show:

1. Dragon's color is looking nice now
2. Alexa has slightly more range than Dragon in the highlights at most ISOs
3. Alexa is less noisy than Dragon in the shadows at high ISOs

It's nice to see someone do what everyone on REDUSER seems to tiptoe around. Eg when you want to compare Alexa / Epic / Dragon, suddenly you get an absence of clear tests and pages and pages of doublespeak. This does not exactly inspire confidence because you know that if Dragon comprehensively beat Alexa in DR or low noise or whatever then people would be going nuts and posting tons of stuff. Instead we get a bunch of non-useful tests and very little real meat. The average guy like me says "hmm, this Dragon obviously falls short or else why would the owners be so darn defensive?"

I do think this shows that if you took whatever testing methodology RED and others got to get 16.5+ stops out of Dragon and applied that to the Alexa, you'd get even MOAR. And I don't really consider the Alexa a 17-stop camera. So, yeah, this keeps reminding me to keep adding a few grains of salt to those sources' words because they are obviously prioritizing their own agendas over objectivity. That's silly because if you look at Dragon dead-on honestly, it's a great camera. I must say that ARRI are coming out really well here because it actually appears that they UNDER-promised on their sensor's specs. That's how you build (and keep) brand confidence. Just as RED's policy of updates that add stuff to their cameras beyond what they originally promised are probably the best thing about the company (well, apart from the insane price/performance ratio).

That said, clear tests like these actually make me want to rent Dragon MORE. It's cheap compared to Alexa, this test shows that it looks comparatively good when downsampled to HD (if you give it enough light), and it has a bunch of advantages for special shoot needs (high-res, high-FPS, etc). I never liked the Mysterium MX sensor's colors (and took a lot of flak for it on REDUSER) but this Dragon sensor is a big improvement. Yay. Oh yeah... and the cherry on the top is I have a bunch of friends who own Dragons (and none with Alexas) so that's the decision-clincher for me Of course, most of the time, for workflow reasons and client reasons it's Alexa, sadly. But that's another thread.

Bruce Allen
www.boacinema.com

:yesnod:

 

[NOSA OBAYIUWANA]

Bruce you have made some really salient points. As I said before, in my own experience, I did perceive a 2 stops difference between the Red Scarlet and the F5. Interestingly, I thought, when testing, that the sensitivity of the Scarlet and the C300 where about the same, although the scarlet was a slightly bit noisier. I really don't think Red needs to be over generous on its DR marketing hype. People working with the cameras professionally will discover the "usable" DR. I know all this about the Reds, and I'm still ordering the Epic Dragon anyway. They are great cameras for the price. The Alexa's are great cameras as well, but they are overpriced. There definitely isn't a 30k difference in quality here, so Red is doing a great job.

At the end of the day so many other things determine the quality of the final film/product that the differences between Red's, Alexa's and F55's when shooting 3K, 4K or even 6K become a very small fraction of what contributes to great filmmaking (particularly in narrative filmmaking).

 

[NOSA OBAYIUWANA]

Also, we are not married to any of these cameras... so we can be promiscuous. And you know what? having knowledge of, and using, different camera systems can only be an advantage. I am tempted to say that getting completely locked into any one company's ecosystem is never a great thing. So as people always say on here - use the right camera for the job.

I've recently become a bit "underwhelmed" with Canons DSLR offerings, so now I'm off to sony A7S'. If Red does not give me what I want for the price I can afford, I will be off to Canon, Sony, or renting Alexas.

That's just it, I cavort with whatever camera will leave me satisfied and not feeling cheated. At the moment thats Red (higher end), Sony (mid range and DSLR or mirrorless) and to an extent Canon.

 

[Matthew J]

Also note that this doesn't say anything on how the footage was developed. For example, if you shoot ISO 800 outside, you shouldn't process the footage at 800, but at 320 because otherwise you lose 2 stops in the top end and needlessly brought up the low end. Because of the RAW VS ProRes of the Alexa and low noise low end of the Epic you have to process the footage differently. WYSIWYG is not terribly fair for the Epic, especially now with the dragon.

 

[Paul E. McCarthy]

I do not understand the 800 ASA section of this test. Either the Alexa rates 800 asa 400 or the dragon rates 800 asa at 1600.

Nobody else sees this as odd?

 

[Jon Carr]

I do not understand the 800 ASA section of this test. Either the Alexa rates 800 asa 400 or the dragon rates 800 asa at 1600.

Nobody else sees this as odd?

Yea thats the first thing I noticed as well. Looked like the doubled the clip twice.

 

[Bruce Allen]

I do not understand the 800 ASA section of this test. Either the Alexa rates 800 asa 400 or the dragon rates 800 asa at 1600.

Nobody else sees this as odd?

Based on my own error rate when doing similar tests, I figured that that particular one was a mistake Doesn't really hurt our ability to get an overall impression of performance, though.

Bruce Allen
www.boacinema.com

 

[David Litchfield]

Considering the original Stouffer chart that got shown, the Dragon sensor has taken an almighty tumble in DR and High ISO performance. It's like its a different sensor! But hey, its sold a lot and the loyal followers won't complain too much, so I guess it's all good for Red.

I doubt the Dragon will make too many inroads against the 4 year old Alexa's grip on the industry.

 

[Timur Civan]

Know whats crazy, 90% of the time i pick Alexa out of a line up, cause it looks awesome. This is the firs time, i've ever gone against the German monster. Picked Dragon every time. Looked rounder, fuller, better color and more natural. The alexa for some reason looked flat, and grey.

 

[Ryan Hamblin]

They all look so close it really feels like we need to give up the pixel peeping and say if you got any of these three in your hand you better tell a good story or your doing a disservice to your camera

 

[Nick Morrison]

Know whats crazy, 90% of the time i pick Alexa out of a line up, cause it looks awesome. This is the firs time, i've ever gone against the German monster. Picked Dragon every time. Looked rounder, fuller, better color and more natural. The alexa for some reason looked flat, and grey.

I was thinking much the same...Loved the rich dragon tones...

 

[Paul E. McCarthy]

I think Dragon looked the best also. I wouldn't tell a client that its dramatically better than the Alexa straight off the mag. I would describe it as marginally better. I hope they can unlock some more lift DR. I hate to say it but, if I were RED right now I would price drop the RR X and the Broadcast Module real quick before that AJA camera comes out.

 

[Clayton Burkhart]

The rolling shutter on this camera is pretty horrible. I'm too used to global shutter now, although for certain uses this cam is tempting.

Sony is working on a 4 line readout solution for the A7S. This means that that the rolling shutter will be reduced to somewhere around one quarter of what it currently is right now. It is the only bugaboo currently causing me to hesitate, so I will wait. I own an F3, and I have to say, that I am STILL happy with that camera. It just looks great. Low noise, very organic, great roll-off, not much rolling shutter or artifacts. Frankly, the most underrated camera out there for Full HD. So the little A7S would be a nice complement with the extra frame rates, the occasional need for 4K, SLOG2, etc.

By the way concerning DR, I always preferred the F3 to the Epic/Scarlet until Dragon. Especially for skintones. Basically the F3 has 13 stops of DR, maybe a little more. It definitely had more DR than the Epic, so I think 12 stops was a good estimation for the latter. The Dragon, Alexa and F5/55 all appear to have about 14 stops by comparison when I compare them during camera tests. So honestly I wouldn't be surprised if the little A7S falls into the same range. That's just technology and time advancing.

Frankly, there is a lot of snobbism out there. The only thing that keeps me from showing up on certain jobs with the F3 are people's attitudes, not what the camera is capable of. Aside from framerate concerns (admittedly very important) the vast majority of the jobs I see could just as easily be shot on this 3-4 year old camera rather than any of the newer ones. 4K mostly gets scaled down to Full HD anyhow.

So really, I think the real question is how much more incrementally one is willing to spend to massage some client's ego and whether it would not be wiser to rent in terms of cost savings at that point.

 

[Scott C]

Sony is working on a 4 line readout solution for the A7S. This means that that the rolling shutter will be reduced to somewhere around one quarter of what it currently is right now.

Do you have a source for that? I recall reading that Sony was working on this but never saw that they planned to apply a fix for the A7s.