Welcome to our community

Be a part of something great, join today!

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 :)

All the Epic/Dragon Vs. Alexa/Sony threads are belong to us...

I set up the camera and hit the record button. any operator mistakes are mine.

As Dragon is the camera I have to work with, it is kinda obvious that I would not try to harass it...

We agreed on 5.5k hd for the above reasons.

Geoff and I (and probably anyone who download the RAW files) will approach the analysis in different ways.
I do not think Geoff is off-base from what many will do, though.
Unfortunately I cannot post anything until I get home, as the disk won't mount on the computer I have here. (nothing wrong with the disk, BTW)
I have not more than confirmed that the high and low exposures are in.

As I have mentioned, I think new calibration may affect existing files to some degree.

Dragon was blackshaded and shot at the temperature/shutter it was calibrated for for the general test. I added 2 stops of "even lower" exposure with the shutter, just for kicks on the Dragon, the F55 and the Alexa. (not on the rest of the cameras.)

I had a look at those two exposures from the Dragon, as I still had them on a card, and have some thoughts about that, but I will not comment on any results before I have the relevant files at hand.
For time and practicality, to me that means Tuesday.

Actually I do not think the test is good for measuring exact total DR for a number of reasons. It is not bad for seeing how the cams perform at similar exposures, though.
I do not think Geoff has said that the test measures exact DR either, but it is good for seeing tendencies.

Phils and Johns ways of measuring DR (as in total DR) are good and the way to go.

To me, what I saw in the Phantom Flex and a few other cameras was much more interesting.

The Dragon holds color at insane underexposures. Which is no surprise, but cool to see.

The biggest eye-opener for me (on all cameras) was how they behaved when a scuba-guy filtered for "underwater". I have no idea how good that test emulated that, but if it was anything close to real life, orange filtering for underwater shooting sounds like a really good idea on digital cameras.

There was no intent to break any of the cameras.

And lastly: All tests are faulty...

Unfortunately the F65 did not get released from a gig. THAT was the camera I was most curious about.
 
Last edited:
This will be messy. Geoff seemed a bit positive to start with but his test being invalidated even before he could post something seems a bit in haste.

"I am totally aware of the new OLPF that is becoming available but it doesn't invalidate my tests.The tests are of what was available to "normal" people on the 6th & 7th of February this year.I could have waited for the 6k Alexa or the 8k Canon or the 4k Panasonic or the ??k Panavision.I'm not waiting for any manufacturer, that was what was available on the day and thats what we have to work with, not jam tomorrow!"
Cheers
Geoff Boyle FBKSCinematographerEU Basedgeoff@cinematography.netSkype geoff.boylewww.gboyle.co.ukVia iPad and fat fingers
 
Why the "HD" in 5.5K naming convention ?
 
Ugh...

1) Redundant
2) Misleading

16:9 is frame ratio and HD refers to picture resolution. With 3840x2160 it makes more sense with regards to exactly 4xHD resolution or QuadHD. Above that "HD" tag is completely unnecessary and potentially confusing, "5.5K 16:9" would be unmistakeable.

Also, "5.5K HD" all by itself (without additional explaining) is an oxymoron.



A: Here is 5.5K HD footage displayed on UHD screen.

B: So it's HD but upscaled ?

A: No it's UHD as that 5.5K tag refers to horizontal resolution, but it is shot in 16:9

B: So it's UHD although it says HD ?

A: Yes.

B: Aaand... that HD doesn't stand for High Definition but 16:9 ?

A: Correct.

B: So why doesn't it just say 5.5K 16:9?

A: /crickets/


I suggest ditching that. Yesterday.
 
for "5k hd" it is as you suggest.

for 5.5k, you choose aspect as 16:9 in the "advanced" tab.

methinks it would be good to have the 5.5k 16:9 in the "standard" tab, as that is the "normal" (or vanilla, as phil dubbed it) S35 FOV many DOPs would expect to see when they mount an S35 lens on their cams.

Less confusion. it could even be called S35 and then aspect.
 
Hi Gunleik
thanks for all the hard work. I am an underwater shooter and am very interested as your "underwater" results. Will you be elaborating on these in the published results as well as your thoughts on orange filtration.
All the best
 
Sulekh Suman said:
This will be messy. Geoff seemed a bit positive to start with but his test being invalidated even before he could post something seems a bit in haste.

"I am totally aware of the new OLPF that is becoming available but it doesn't invalidate my tests.The tests are of what was available to "normal" people on the 6th & 7th of February this year.I could have waited for the 6k Alexa or the 8k Canon or the 4k Panasonic or the ??k Panavision.I'm not waiting for any manufacturer, that was what was available on the day and thats what we have to work with, not jam tomorrow!"
Cheers
Geoff Boyle FBKSCinematographerEU Basedgeoff@cinematography.netSkype geoff.boylewww.gboyle.co.ukVia iPad and fat fingers
Click to expand...

It's not messy man, it's how you're taking it. He is right, it doesn't invalidate anything. That is the camera that is available today and what has been sold. So in that respect the results are as valid as anything else. But, when the new OLPF comes out, the results will be invalidated in regards to the updated camera. Then a new test is needed to update the results for accuracy since the previous test used the previously current OLPF before the update. But this test will tell us exactly what we get with the currently available system. Anyone who argues against that needs to check their brain. No one should tell Geoff his test is invalidated. To me that's harsh to say to someone who just put in a lot of time and effort to do a proper test with the gear available to him at the time.

Let's not blow this up bigger and get negative....
 
I think the point is this test will not apply to the majority of Dragon cameras going forward. Good info currently but these tests have a tendency to make impressions that last longer than their relavence to the available technology.
 
Evin Grant said:
I think the point is this test will not apply to the majority of Dragon cameras going forward. Good info currently but these tests have a tendency to make impressions that last longer than their relavence to the available technology.
Click to expand...

That's understandable but you have to look at what we currently have available to test. It's not like he didn't want to make Dragon look good. He's just testing what is available. Maybe RED should have sent a camera with the new olpf then...
 
Red and Geoff have a rough history, and as I've stated before these tests can never really be done fairly for each camera.
My suggestion is always to test yourself to determine what works for your specific projects needs.
 
evin, i agree.

and still took part.

two reasons

1. these tests take place and are published, like it or not. i figured that it was better to have me around than not. Your milage may vary on that statement, but after being there, i still think it was not a bad idea. (and that does not imply foul play of any sort, just that it was good to have someone who knew that camera around, too. Vision Research was there with a rep and two phantom 4ks... and there were others in place setting up the other cameras.)

2. gee.... i am as curious as the next man... nothing that i have seen from the test has changed my impression on Dragon.

i know of course about the history of red and geoff, but (which i think also is reflected in geoffs quote) the dragon as it is, is not suddenly "bad" because of the new improvements, it will only get better. as other cameras have gotten better over time.

i did consider to pull my cam when the new olpf came out with a timeframe, but one out of two would probably have happened.
a not insignificant conspiracy-speculating noise
or they would just get another. + the above. and on a few marks it was good that i was there with it.

the spirit of the test and all the people there was:
let's have a look at this, not: let us break these cams, and i suggested and performed a few extra tests, that geoff commented "i dare not" :-)

geoff was pretty cautios in advance and offered me anonymity. i would not have any of that. too much bad stuff in the history already.
after accepting, i informed red that i would take part, in case that would be taken badly, and i also posted it here, for people to have opinions on that, if they felt like having opinions... :-)

of course red has not approved, that is not what i imply, just that i think the procedure has been pretty open.

and geoff took "what is to come" into his initial remarks.

raw files will be released. what you can read out of those is simply "what happens if you do what we did".

there are a few reasons i think "exact DR" is not one thing we really tested for. what we did test for, was what happens if we shoot these cameras side by side over a range of about 14 stops (+2 for alexa, dragon and f55) in 1/2 stop increments. from overexposure to underexposure. i also checked for some color-consistancy at different exposures, which has come to interest me, and a few other things, like resolution (at full-res), two outdoor scenes, hdr and the mentioned "underwater emulation".

as far as i can judge, nothing was done to harm any cam, and i look forward to the new olpf and color-scence as much as anyone. but i have stated that i have liked the dragon since the prototype, i have seen it improve since then and i am sure it will get even better. somehow i do not want and cannot see why we should in any way make the Dragon sound "bad" just because it gets "better"... :-)
it is not bad. it is the coolest cam out there, remember!

cheers

g
 
Last edited:
Mark van Coller said:
Hi Gunleik
thanks for all the hard work. I am an underwater shooter and am very interested as your "underwater" results. Will you be elaborating on these in the published results as well as your thoughts on orange filtration.
All the best
Click to expand...

I have no problem publishing those, but I have less background for evaluating the relevance of that test, than the others we did. We blue-gelled a 2k HMI and lit a scene with that, and took a few exposures.

I can describe what I initially see, and that is basically that the blue seems to cancel out a lot of the other colors, as opposed to when you shoot tungsten. I would have expected an "inverted tungsten" result, but on the cameras we shot, we basically had "no colors" (that is abit exaggerated, but it shows a tendency. These were Alexa, Dragon and F55). If that reflects with your observations, (the scuba-dude said the results fitted his experience), the only practical way I can see one can counter that, would be with correcting the optical signal before it hits the sensor(s), and that would presumably be with some kind of orange filter, to get all channels saturated.

Does this sound half sensible?
 
Shooting Dragon, I would definitly go the 1600 iso route and have +7 & -7 stops DR with a total of 14 stops in my findings on Goef's test. Full RES and max Redcine-x denoise will smoothen the remaining small amount of noise.

To bad they didn't push further than -8 as I still see some black separation in the shadows. Here a test of how clipping is made in the highlight colors.

Pat
 

Attachments

  • 6.0000000.jpg
    6.0000000.jpg
    9 KB · Views: 0
  • 6.5000000.jpg
    6.5000000.jpg
    8.7 KB · Views: 0
  • 7.0000000.jpg
    7.0000000.jpg
    8.6 KB · Views: 0
  • 5.5000000.jpg
    5.5000000.jpg
    9.3 KB · Views: 0
You must log in or register to reply here.
Back
Top