Dragon/Epic comparison

[Brian Boyer]

I think the best way can be R3ds, but I am a bit unsure.

First two examples updated with r3ds

12 frames from each cam...


G

I tried twice and got error messages opening the zips (Mac OS 10.8.5). Other .zip files open just fine. Might there be a problem with the files?

 

[Gunleik Groven]

reuploading...

 

[Stephen Franchek]

There are some great responses on this thread! And after talking with my bomb squad rep I feel a bit safer with my purchase. I will, however, need to see an apples to apples comparison before making my final judgement. That's a reasonable request.
However, I will tell you that as a scarlet owner, I'm already past the point of being sick of losing rentals to Epics when the client doesn't need one.
"I can rent you a scarlet for this price"
"Oh but we must have the EPIC!"
"Are you shooting slomo?"
"No..."
"Do you need 5k or better compression ratios?"
"What are those?"
"Then you should save a few bucks and rent my Scarlet."
"...Oh but we MUST have the Epic!"

For the sake of avoiding that (abridged) conversation, upgrading to Epic Dragon will be well worth it.
What I don't understand is some of the anger on the forum. We're consumers and as such it's our job question companies. That's what pushes the market forward and why we're using these amazing digital cameras instead of being stuck with film (not knocking the great quality of film).

 

[Gunleik Groven]

ok now...

 

[Daniel Soderberg]

Awesome, thanks John!

I've shot Xyla charts on Dragon, but they're from the very early days so I won't be sharing them. I could see 16+ stops in them as advertised, but things have moved on and they're no longer representative IMHO.

When our Dragon comes back from its current job and / or our Carbon arrives I'll try and grab the 21 stop Xyla again to share the results.

 

[Martin Stevens]

Thanks Gunleik and all.

Great thread.

 

[Michael Ou]

All I can say is I can't wait to get my hands on it! Oh and thanks for the tease, to all those in the know and sharing! :thumbup1:

 

[Matt Ryan]

If you push both of these files you will notice not only how color response is better on Dragon, Dragon images are sharper at 6K than MX at 5K (not to mention when 6K dragon is down-sampled to match 5K MX, even more sharp), and also noticeable is Dragons better handling of IR. At 12,800ISO IR is obvious in a few spots of the black shirt on MX footage, not even present on Dragon.

Thanks for these G! You should maybe reconsider posting the super underexposed shot. If for anything it will help show the tremendous amount of latitude available vs. the current MX. To be able to see that one lacks an image while the other has one would be HUGE. If people don't get that, they shouldn't be allowed to use a camera...

 

[Gunleik Groven]

If you push both of these files you will notice not only how color response is better on Dragon, Dragon images are sharper at 6K than MX at 5K (not to mention when 6K dragon is down-sampled to match 5K MX, even more sharp), and also noticeable is Dragons better handling of IR. At 12,800ISO IR is obvious in a few spots of the black shirt on MX footage, not even present on Dragon.

Thanks for these G! You should maybe reconsider posting the super underexposed shot. If for anything it will help show the tremendous amount of latitude available vs. the current MX. To be able to see that one lacks an image while the other has one would be HUGE. If people don't get that, they shouldn't be allowed to use a camera...

I am glad someone see what I see...

I am reluctant, because of course the image is noisy like hell and pretty unuseable.

But there should really not be any image data in the parts where the first exposure has image data.

And there is both image and colordata.

Blew my mind.

But judging from some of the feedback to this thread, I fear that people will make noise about seeing noise.

Obviously there is noise. Should really be nothing but noise. Still there is more than a stop of info there...

EDIT

I think I stand by not showing it....

It shows those interested that there is a discernible color-signal 16 stops bellow that first exposure (which is the important thing), but as it seems many people think that "each stop should be equal", and I do not want to spend a lot of time explaining the difference between sensors and eyes, I will just leave it out.

To me it is part of the explanation on how extremely flexible the Dragon is over anything else I have used.
But I see that the data can be misinterpreted.

It has been very good to show people who are interested in the Dragon, though.

That the Dragon has a lot more in the top and a lot more in the bottom than the Epic.

 

[Matthew Scott]

Yes, an awesome test/comparison indeed, and very much appreciated, thank you Gunleik! Any "noise" I've made about noise, was simply an observation that at 1:1 the noise seems virtually same on both the MX and Dragon...which was surprising. That's not a terrible thing to say, is it? Maybe it is.

 

[Will Keir]

Thanks Gunleik for taking the time to shoot these tests, and to your friend, Espen who is now famous!

 

[Robert Ruffo New]

There are some great responses on this thread! And after talking with my bomb squad rep I feel a bit safer with my purchase. I will, however, need to see an apples to apples comparison before making my final judgement. That's a reasonable request.
However, I will tell you that as a scarlet owner, I'm already past the point of being sick of losing rentals to Epics when the client doesn't need one.
"I can rent you a scarlet for this price"
"Oh but we must have the EPIC!"
"Are you shooting slomo?"
"No..."
"Do you need 5k or better compression ratios?"
"What are those?"
"Then you should save a few bucks and rent my Scarlet."
"...Oh but we MUST have the Epic!"

For the sake of avoiding that (abridged) conversation, upgrading to Epic Dragon will be well worth it.
What I don't understand is some of the anger on the forum. We're consumers and as such it's our job question companies. That's what pushes the market forward and why we're using these amazing digital cameras instead of being stuck with film (not knocking the great quality of film).

The fact is Epic when used at 5K looks noticeably better for a 1080p master - it has a wider field of view, we are more zoomed out from the noise. It looks sharper, smoother. Lower compression also simply looks better and can be graded very aggressively without issue. Epics also tend to have less breakdown issues and some tests have shown that Epic has faster read-out and thus less CMOS skew (something like half as much)

SO whether the producer knows the tech reasons for this or not, the producer you mention is making the same choice I would make, and the questions you are asking don't cover the real and always meaningful differences between an Epic and a Scarlet. Simply put: every shot will look better shot on epic in 5K at 3:1 than anything Scarlet is capable of shooting (if aimed at the same set-up with the same lenses and filters), even if you are delivering 1080p or even 720p.

This same producer will demand Dragon, yes, and again I will agree with him/her (within reason of cost difference, right now at 2K a day it's too much for many prods).

 

[James_Mills]

Yes the MX magenta blowouts is nothing new and not much to worry about on the MX those magenta patches has nothing to hide than white. Simple Colour correction and its out.
I think this post is great and it really shows the DR of the dragon and if people do not see the difference then so be it, Latitude looks amazing.

Can you help me find the differences in latitude? The only noticeable differences I see out of all the samples is the way we can drop the ISO to 50 and still have a somewhat useable image in the over exposed one. Does this suggest an increase in latitude or just light sensitivity? If we're getting the exact same image when exposed properly, what is the real advantage of this? I thought extra stops would mean extra details in shadows and highlights in our properly exposed images as well. I don't see this, could someone help me find spots where there is some sort of difference in detail at the proper exposures?

I do notice a difference in the skintone and tint of the image, but I'm not entirely sure I prefer the greener / creamier look to be honest.

 

[James_Mills]

The fact is Epic when used at 5K looks noticeably better for a 1080p master - it has a wider field of view, we are more zoomed out from the noise. It looks sharper, smoother.

It isn't a wider field of view, a Scarlet's 4k is the same crop as Epic's in terms of FOV (unlike the Dragon size to Epic size).

 

[Robert Ruffo New]

Can you help me find the differences in latitude? The only noticeable differences I see out of all the samples is the way we can drop the ISO to 50 and still have a somewhat useable image in the over exposed one. Does this suggest an increase in latitude or just light sensitivity? If we're getting the exact same image when exposed properly, what is the real advantage of this? I thought extra stops would mean extra details in shadows and highlights in our properly exposed images as well. I don't see this, could someone help me find spots where there is some sort of difference in detail at the proper exposures?

I do notice a difference in the skintone and tint of the image, but I'm not entirely sure I prefer the greener / creamier look to be honest.

if you can reduce ISO and see new information then that is latitude. It means the highlights are not clipped and there is color information in pixels that received a whole heck of a lot of light - more so than would be the case in MX when looking at such light-drenched pixels. Not rocket science.

 

[Robert Ruffo New]

It isn't a wider field of view, a Scarlet's 4k is the same crop as Epic's in terms of FOV (unlike the Dragon size to Epic size).

Sigh… Yes, but only an idiot would shoot Epic at 4K (not to say some people are not that stupid- I have seen this choice made on sets by others, and it's the kind of thing that makes me feel better about charging as much as I do as a DP) so who cares - why would you do this unless your prod is trying to save $15 of hard drive space. You shoot at 5K, you downsample to your delivery format, it looks better. Again, not rocket science.

 

[Phil Holland]

Sigh… Yes, but only an idiot would shoot Epic at 4K (not to say some people are not that stupid- I have seen this choice made on sets by others, and it's the kind of thing that makes me feel better about charging as much as I do as a DP) so who cares - why would you do this unless your prod is trying to save $15 of hard drive space. You shoot at 5K, you downsample to your delivery format, it looks better. Again, not rocket science.

I've shot on three shows that chose to shoot 4K for the aesthetic it provides as well as a few other reasons. Lots of people have done it around here. Also, on Mysterium-X 4K falls within the optimal coverage of most Super 35mm designed lenses. Some people are picky about that. Really picky.

Personally I prefer to shoot 5K on MX for the various gains and the larger format of course. Dragon is the interesting multi-format tool among the bunch because 5K now lands in the S35 sweet spot, 5.5K is essentially Mysterium-X Full, and 6K is it's own bigger playground. It's been interesting so far to see who is using what format and how with Dragon so far. There isn't one answer really and all resolutions from 4K and up make for a good 4K finish.

 

[James_Mills]

Sigh… Yes, but only an idiot would shoot Epic at 4K (not to say some people are not that stupid- I have seen this choice made on sets by others, and it's the kind of thing that makes me feel better about charging as much as I do as a DP) so who cares - why would you do this unless your prod is trying to save $15 of hard drive space. You shoot at 5K, you downsample to your delivery format, it looks better. Again, not rocket science.

I wasn't talking about shooting at 4K on an Epic (I would never do this either and have urged directors not to when they asked me if we could on my Epic). I meant that, as far as I know, the actual 5K image from the Epic is the same FOV as the Scarlet's 4K. You said that it has a wider field of view and that we are thus zoomed out from the noise, but I don't believe this is true?

if you can reduce ISO and see new information then that is latitude. It means the highlights are not clipped and there is color information in pixels that received a whole heck of a lot of light - more so than would be the case in MX when looking at such light-drenched pixels. Not rocket science.

That's what I figured. Unfortunately I don't see the advantage of this in these shots when exposed properly. Perhaps this just isn't the best test to display this because of the lack of high contrasty shadows, but I was asking for examples of this DR increase in the images that are exposed properly.

A better side by side comparison for latitude would be shooting directly into a sunset-backlit face or something, then we can see how much useable DR is really availed. I think some of the earlier Dragon tests showcased this increase, but I'm really interested in a side by side comparison with the MX.

 

[Gunleik Groven]

Sigh… Yes, but only an idiot would shoot Epic at 4K (not to say some people are not that stupid- I have seen this choice made on sets by others, and it's the kind of thing that makes me feel better about charging as much as I do as a DP) so who cares - why would you do this unless your prod is trying to save $15 of hard drive space. You shoot at 5K, you downsample to your delivery format, it looks better. Again, not rocket science.

4kHD is pretty much "standard" for a lot of TV production.
I am not discussing that 5k is better, just stating that I know that it is.

It will be interesting to see if 5kHD can be established in similar Dragon shows.

 

[Gunleik Groven]

James... Give me a sec... I think I can explain pretty simple