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 :)

Dragon Vs Alexa Low Light Tests - Someone?

Mark Toia said:
Fact. But at 320 iso (my chosen ISO) The MX sings in resoltion with no noise at all.
Click to expand...
I have a question
I thought the mx sensor always shoots at 800iso (because its the base iso). and their are no advantages at shooting at "lower" iso settings on camera.
 
harjotsingh said:
I have a question
I thought the mx sensor always shoots at 800iso (because its the base iso). and their are no advantages at shooting at "lower" iso settings on camera.
Click to expand...

You definitely want to stay below 800 if possible. You'll avoid a lot of noise. That's not to say that ISO 800 won't produce a usable image, but your shadows will look way better at 320
 
Gunleik Groven said:
There are some tradeoffs both ways. 800 is the safer. 320 needs a bit more attention.
Click to expand...

Exactly.

Just watch that histogram doesn't clip. And you safe as houses with an amazing amount of detail.
I only push to 800 ISO at night. I never go over that with the current MX sensor. I just use extremely bright glass to make sure I never end up underexposed.
Dragon however is a different kettle of fish. I try never go over 2000 ISO with that camera and shoot it mainly 250 ISO for the cleanest images. But if you need to push into 3200 ISO territory (@6k) you'll still safe .
 
harjotsingh said:
I have a question
I thought the mx sensor always shoots at 800iso (because its the base iso). and their are no advantages at shooting at "lower" iso settings on camera.
Click to expand...

Red prefers you shoot at 800 because it's like giving you a stop up your sleeve for over exposure. In other words it saves you if you are a bit loose with your exposures. Saves clipping of highlights in other words.
I shoot exclusively at 320 and on the Dragon 250 . But mind you I do have one eye firmly planted on the histogram making sure I never clip. Sometimes I do because I pan into the sun or do something stupid like that.
 
Gunleik the one interesting comparison I would love to see is the MX vs DRAGON Low light underexposed performance noise. I know and learned from you guys a lot about properly exposure avoids noise even at high ISO testing and that Dragon's noise is more organic and probably the equivalent of MX 800 at 2000 in Dragon Noise grain. But a lot of times there are just not enough lights on set to fill and the scene is shot underexposed even @ f1.4 or f1.2 for low budget reasons, or because it is a run gun gorilla scene in the middle of nowhere.

That would be sweet to see!


 
Alain M said:
Gunleik the one interesting comparison I would love to see is the MX vs DRAGON Low light underexposed performance noise. I know and learned from you guys a lot about properly exposure avoids noise even at high ISO testing and that Dragon's noise is more organic and probably the equivalent of MX 800 at 2000 in Dragon Noise grain. But a lot of times there are just not enough lights on set to fill and the scene is shot underexposed even @ f1.4 or f1.2 for low budget reasons, or because it is a run gun gorilla scene in the middle of nowhere.

That would be sweet to see!
Click to expand...

I agree that we need some more sensor comparison tests. But there has to be some caution when doing this as I learned because my entire thread was deleted. It's probably because it was a beta camera and I showed a zoomed in picture of the grain.
So some examples of real-world photography with some outdoor scenes and some shadows would be great. I think people are a little gun shy about doing it ,quite frankly.
Posting Dragon captures of pretty girls and beautiful scenery doesn't do much for me, I really need that frame of reference with the MX capture as well. The images can be boring, but a lot can be learned from the exercise .
 
Mark Toia said:
There is alot of low light shots in this DRAGON video. Sadly I did not do side by side at 4000asa, only because I new the MX wouldn't get near the dragon at those sort of exposure settings.

https://vimeo.com/71666317
Click to expand...

And this was on a BETA camera. I wonder if current Dragons would look even better?
 
Phil Holland said:
While clean "out of camera" it is actually murdering the detail found in the 4K image and creating a plasticy look that people tend to try and avoid.
Click to expand...

Looking at Toias post with the blowup of the faces, I'm going to file this under "written in haste".

I realize that a Canon DSLR will have few supporters here and the OP has explicitly requested to not deviate from the topic; I'll respect that unless questions arise.

That said, I'm sure the Dragon will be hard to beat in terms of absolute IQ. Only question mark so far is the haloing also seen in the Epic here with the candle. Many times showing up as pink. Do you all agree that this is a lens thing? Seems like a far fetch to me that the exact same thing shows up here, in the Dragon match stick footage and also with the Audi headlights?
 
One thing I think that fails to be mentioned is the distribution of the latitude over the dynamic range.
My impression of the Epic vs the Alexa for instance was that there was more information in the highlights in the latter.
I believe this is because the Epic had more info below 50% grey and the Alexa had more above it.
So it is not enough to say that one camera has more dynamic range than another, it is also about how much of it falls above or below middle grey. This of course also changes depending on the iso implemented and the color temperature of the light source. Tungsten does not perform like daylight in this regard. So simplistic statements about more or less total DR are not adequate.
Thus it is entirely possible to have a high dynamic range sensor and still have halos around a candle flame for this reason.
 
While I agree that distribution is a critical thing, I cannot follow the rest of the reasoning, I am afraid... Have a look at the images Mark Toia posted. I think they tell a story.

I loved and love the Epic for a great many reasons. One actually being that I think that I had a nicer overexposure RESULT with that, than the Alexa, even though the Alexa accepts more light and is nicer UNTIL it clips (but goes quite yellow at clip)

But Marks example shows the difference in the bottom in a way that corresponds well with what I have seen.

Measurable latitude in the bottom may be similarish, but the practical image from the Alexa has... different qualities to the Epic as it goes low.
 
AndreeMarkefors said:
Looking at Toias post with the blowup of the faces, I'm going to file this under "written in haste".

I realize that a Canon DSLR will have few supporters here and the OP has explicitly requested to not deviate from the topic; I'll respect that unless questions arise.

That said, I'm sure the Dragon will be hard to beat in terms of absolute IQ. Only question mark so far is the haloing also seen in the Epic here with the candle. Many times showing up as pink. Do you all agree that this is a lens thing? Seems like a far fetch to me that the exact same thing shows up here, in the Dragon match stick footage and also with the Audi headlights?
Click to expand...

It wasn't written in haste. I'm actually still a Canon shooter for stills and a previous CPS member. Love a lot of their glass too. I like the concept behind the 1DC even, just loathe that codec in reality. I'm just calling it how it is. The 1DC's noise reduction is strong at this level. If you look at those sample clips that Mark posted which are available from CML for download you can see how much detail is lost in that noise reduction. For instance, if you look at the 100% crops of the "3200ASA" text, the skin detail on the man, or the man's wardrobe you can see what's being effected by the on camera noise reduction. I'm fine with that, it's what it is and works as expected. Heck I've even done a similar test with all camera bodies shown and I can tell you directly it effects the resolving power and it varies at different ISOs on the 1DC, which is either a good or bad thing depending on how you look at it.

What I am saying though is processing any one of these other cameras with a smart noise reduction method can yield similar and in a couple cases better results.

This is where sometimes a side by side camera comparison becomes somewhat odd because to get the most out of each camera you may have a different workflow to achieve the highest quality possible. In this case cameras going up against cameras that have no on board noise reduction or slightly different methods of noise reduction yielding different results.

For me it's all about the end results and how to get there. The fundamentals behind REDCODE RAW is it's more or less up to the end user what to do with the material. Red's been pretty clear on that. In Redcine-X Pro you have a bit of control over noise (heck it wouldn't shock me if there was more NR advancements coming in RCX either for those moments when you dip down and pull up from the darkness). If you want more advanced noise reduction there's options outside of RCX of course.

Myself and others have posted 4K samples of Red material from Mysterium-X at ISO 6400 and 12800 processed that looks good with a bit of post NR done. More or less the same technique I'd do on most of the other cameras shown in these samples at this particular ISO.


In terms of the OLPF reflected glare, there's many here who want to see something done about that. That's something I agree with and for the most part is a quirk that I lived with on Mysterium-X and now live with a slightly different incarnation of that aberration on Dragon. There are solutions. None of which are easy to execute given the design. It's a tough one.
 
Phil Holland said:
In terms of the OLPF reflected glare, there's many here who want to see something done about that. That's something I agree with and for the most part is a quirk that I lived with on Mysterium-X and now live with a slightly different incarnation of that aberration on Dragon. There are solutions. None of which are easy to execute given the design. It's a tough one.
Click to expand...

I hope that Jim as a master of optical design took that Dragon OLPF candle flare as one of the highest priorities to be soon fixed on Dragon before NAB 2014.
 
Phil Holland said:
It wasn't written in haste.
Click to expand...

Fair enough. I am looking primarily at the beard. To say that the 1Dc murders detail and let the Epic off the hook is a stretch to my eyes. Doesn't matter of course: the images do their own talking and it's up to everyone to decide what fits into the workflow.

And by looking at them I'd say the 1Dc was left at the standard NR option which is "standard" (the second highest). You can opt to set it to "high", "low" or "off", should you so desire. Each lower setting introducing more grain, as expected. Anyway… back to Dragon!
 
AndreeMarkefors said:
Fair enough. I am looking primarily at the beard. To say that the 1Dc murders detail and let the Epic off the hook is a stretch to my eyes. Doesn't matter of course: the images do their own talking and it's up to everyone to decide what fits into the workflow.

And by looking at them I'd say the 1Dc was left at the standard NR option which is "standard" (the second highest). You can opt to set it to "high", "low" or "off", should you so desire. Each lower setting introducing more grain, as expected. Anyway… back to Dragon!
Click to expand...

The F65 would clean up easily if you wanted to get rid of what I think is some nice film like texture. The Alexa would still have a good amount of detail after NR, but IMHO there wouldn't be much left to the Epic MX image after try to remove that amount noise. How it looks in motion is what really matters though. The 1Dc will probably have that large blocky over compressed and over NR'd feel in motion, but will the average viewer notice? From what I've seen the Dragon is going to have a very nice texture similar to the F65 at higher ISO and when underexposed.
 
AndreeMarkefors said:
Looking at Toias post with the blowup of the faces, I'm going to file this under "written in haste".

I realize that a Canon DSLR will have few supporters here and the OP has explicitly requested to not deviate from the topic; I'll respect that unless questions arise.

That said, I'm sure the Dragon will be hard to beat in terms of absolute IQ. Only question mark so far is the haloing also seen in the Epic here with the candle. Many times showing up as pink. Do you all agree that this is a lens thing? Seems like a far fetch to me that the exact same thing shows up here, in the Dragon match stick footage and also with the Audi headlights?
Click to expand...

Haste... ? I'm not one to ponder if my grammar or spelling is up to spec for a Blog post. I just type quickly and move on... Sorry.

Nick Morrison said:
And this was on a BETA camera. I wonder if current Dragons would look even better?
Click to expand...

It is. And I know for sure it will be even better again when REDCOLOR 4 hits town.
AndreeMarkefors said:
Only question mark so far is the haloing also seen in the Epic here with the candle. Many times showing up as pink. Do you all agree that this is a lens thing? Seems like a far fetch to me that the exact same thing shows up here, in the Dragon match stick footage and also with the Audi headlights?
Click to expand...

I've seen pink in EPIC M & MX images from day dot.. you are correct. Some lenses (lens flares) make it more apparent than others. But it's definitely there in the high lights. BUT... It can be graded out quite easily.
OLPF reflections are a different issue all together and needs to be solved.
 
Last edited:
You must log in or register to reply here.
Back
Top