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 :)

Dragon vs Epic - underexposure/ overexposure test

M

Michael H Booth

Member
Joined
Nov 19, 2013
Messages
15
Reaction score
0
Points
0
I had the chance to get an Epic and Epic Dragon side by side to test underexposure and overexposure in each camera. The specs are included in the video. One thing I forgot to leave out is that both cameras compression ratios were set to 8:1. I'll try to get a 4444 link up here soon.

https://vimeo.com/86458847

Cheers
 
Hey, Michael. Thanks for sharing this test. A couple of thoughts:

1. Because you rated the two cameras at different ISOs (800 for MX, 2000 for Dragon), you were effectively underexposing the Dragon by 1.5 stops more than the MX. Had you rated them both at 800, the differences in the two cameras would have been more apparent.

2. It would be really informative to see the under and over-exposed images pulled back to normal in Redcine so we can see how much information is recoverable in deep shadows and blown highlights.

Again, thanks so much for posting this test. It's always generous when people share their findings here, and even more generous when they put up with a slew of additional requests :)

Julio
 
Thank you for doing it. Would love to see them both set to 800 same F/T-stop and pushed to extremes. If you would be so kind as to post some R3D's so we could push the ISO around after the fact that would be amazing :p
 
I think they are trying to match the over/under distribution of DR on both cameras by setting different EIs. This gets into why testing cameras against each other is so problematic.
If they we're set the same you'd feel less difference in the highlights but much more in the shadows because of the distribution of tonality. However, by setting to EI 2000 you only really feel 1.5 Stops on each end of the DR, but that is indeed 3 stops more total DR regardless wether that's enough for Zak:sifone:

More importantly to me is how much better the color and character of the blown out areas are, much cleaner (no magenta cast) with a much more natural rolloff.
 
The darks look similar to me (except the skin is brighter) and the highlights look waaaay better to me. Great job Red and thanks for the test.
 
i don't get it....for me the iso is a very basic guideline and really has not much to do with correct exposure....either way epic at 800 looks exactly the way my epic would look if i metered at 800 and set exposure accordingly....but that is at least 1/2 stop under.....
i also have to say that i shoot outside all the time in LA and yet have to experience the pink highlight problem.....actually not true, i did once when an assistant knocked the f-stop off and we over exposed about 3 stops.....but why would anyone do that on purpose?
judging from this i would have to be with zak....not nearly enough to justify 10k....
phil's test with the new olfp and color are much more convincing, i am assuming this dragon did not have the latest upgrades/updates?
regardless: raw files are the only way to make a call....would love to get a chance to play.....
 
Zakaree Sandberg said:
not enough of a difference to justify for me
Click to expand...

Are you thinking of not upgrading? Are are you saying that just as a comment on what this one video shows. Full disclosure I think we will upgrade, but it isn't the kind of almost no-brainer easy decision that upgrading to Mx was, and upgrading to Epic was. There the advantages were extreme and very numerous.

Now I find the Epic is really great, especially if you take the time to grade it properly. I'm not left wanting for a better camera, except when I have to hand in ungraded footage that will likely never be graded with any care, and I hope to do fewer of those kinds of gigs anyway.

Would love to hear your further thoughts on this.
 
I agree. I think of ISO on my Epic almost like gain. If you think about it, ISO in a digital world is typically only used when you "need to" as it can introduce noise and other undesirable effects. I use ISO on the Epic as a tool to either protect highlights (800 on a bright day on Epic) or "turn it off" to see what the sensor is seeing (320 on a dark scene) If these images were shot at 800 on both cameras, it would give me a chance to see how much I could push the iso on the Dragon and also compare the noise. 800 seems like a reasonably fair baseline as well. Id also like to see a couple R3D's where the Epic is pushed just to the edge of clipping highlights (maybe some pink highlights) and then see the same shot with the Dragon at 800 ISO at the same aperture then opened up one more stop. I'd do it if I had access to the both cameras :)


paul schefz said:
i don't get it....for me the iso is a very basic guideline and really has not much to do with correct exposure....either way epic at 800 looks exactly the way my epic would look if i metered at 800 and set exposure accordingly....but that is at least 1/2 stop under.....
i also have to say that i shoot outside all the time in LA and yet have to experience the pink highlight problem.....actually not true, i did once when an assistant knocked the f-stop off and we over exposed about 3 stops.....but why would anyone do that on purpose?
judging from this i would have to be with zak....not nearly enough to justify 10k....
phil's test with the new olfp and color are much more convincing, i am assuming this dragon did not have the latest upgrades/updates?
regardless: raw files are the only way to make a call....would love to get a chance to play.....
Click to expand...
 
Last edited:
David Battistella said:
here. It has been done...and done very well I think.. :)

http://www.reduser.net/forum/showthread.php?109565-Dragon-Epic-comparison&highlight=


Battistella
Click to expand...


Good point. I'll take more examples if anyone wants to shoot them? !!! :)



Not sure if the R3D' download links were there when I saw that the first time or if I just missed them… (referring to the very first images in the thread)
I just downloaded them and tried to recover both clips by reducing ISO / Exposure etc…. Truly amazing how much more workable the Dragon footage is. Amazing really.
Surprised that nobody posted stills showing how much could be recovered in the Dragon clip. Maybe it's down the thread a bit...
 
This is actually a very nice test Michael. Shows a couple of things clearly, though on the older calibration and OLPF.

Things that are very impressive to me:

- Baseline normal exposure outside.
- The +2 exposure, both in the highlight retention and shadow density. (This particular image shows off the dynamic range very well)
- Overall color, especially the skin and sky.
- Overall tonal separation.
 
I like the color separation on the outdoor shots, watch it paying attention to the color chips and you can see the separation there. I like the sky too, less "dingy". Nice blue sky in the future.

its funny Phil is calling it the "older" color science and OLPF, (it's what? Two months old?six?) it's funny how fast things become old in the red world.

Battistella
 
Thanks for putting the test together Michael.
 
Phil Holland said:
- Overall color, especially the skin and sky.
Click to expand...

+1 This was my favorite part about the test. Right out the gate I thought colors and skintones on Dragon were superb. Looking forward to similar tests with new OLPF.

David Battistella said:
its funny Phil is calling it the "older" color science and OLPF, (it's what? Two months old?six?) it's funny how fast things become old in the red world.

Battistella
Click to expand...

Haha. You have a point. The old OLPF does feel like "ancient history" (even though most people still have it in their cams!)
 
The shots of this test are not transcoded to log but use a Rec709 compliant gamma that chops off tons of information in both, blacks and whites. In this test the DR of MX seems to be almost on par to Dragon. Partly because of the heavy gamma losses, partly because of the 1+ stop ISO settings of Dragon in the interior scene, stressing the sensor particularly in the blacks. in the daylight scene Dragon is half a stop or just about one stop overexposed compared to MX, the shot is visibly brighter. The Dragon scene should have been developed with ISO 800. The overexposure cuts off valuable DR in the whites which are again already badly stressed by the gamma. On the Vimeo video one cannot judge the noise but clearly one can state that for web videos Dragon has probably more headroom regarding sensitivity, see the interior scene. Lastly it shows that colour rendition with Dragon is more neutral and not prone to a magenta cast.

Ok. Have seen better tests. But nonetheless it's always welcome to see side by side shots. For me the obvious better colour rendition is reason enough to upgrade. But then other test show that Dragon has about 1+ stop more DR in the whites and at least 1, perhaps 2 stops more depth in the blacks when both are set to ISO 800. The most obvious advantage of Dragon is the better colour science and the richness of colour rendition in the shadows. Here Dragon excels, probably even the Alexa. Plus there is plenty of resolution added for further manipulation or UHD projects. Dragon is a no-brainer.

Thanks,

Hans
 
David Battistella said:
its funny Phil is calling it the "older" color science and OLPF, (it's what? Two months old?six?) it's funny how fast things become old in the red world.
Click to expand...

Nick Morrison said:
Haha. You have a point. The old OLPF does feel like "ancient history" (even though most people still have it in their cams!)
Click to expand...


Heh. Starting alarmingly soon the new OLPF and Sensor Calibration will be available in all new bodies and upgraded bodies. I only say old because there is a new cool :)

The interesting thing is if you have a Dragon it is indeed optional to send it in for the new tech, which Red is installing for free. Though, I don't know how many people will hold onto those potential uglies the older OLPF produces (magenta highlight orbs, IR glare/reflection, potential lifted shadow noise due to IR contaminatio). Plus you get more color stability and highlight retention at the top end while maintaining the same midtone exposure and base shadow density with the New Calibration.
 
You must log in or register to reply here.
Back
Top