A Dragon assesment...

[Martin Stevens]

Less speculation is really needed about this. How many do you see here:

Myself, Gavin, and Stacey have all posted individual measurements of various things. Xyla-21, individual measured patches, individual calibrated illuminated sources, etc......

The next question to ask is what are you looking at it? REDgamma4? REDlogFilm Flat? Curve? The next question is how are you looking at it? Color Space? Viewing Device?

Long and the short of it Red's description of 16+ is accurate. How much of that Dynamic Range at any given point in time will depend on your ISO rating and what sort of viewing Gamma/Curve you are utilizing.

Thanks Phil...

Yes, but this does not really answer the question. On your chart it does not say which chip is middle gray... unless I am blind.

I would prefer a direct answer as I just received my Dragon and something is not right, unless is is supposed to be this way. ???

Also, I use REDLogFilm Flat and use RedCine's wave form tools etc.

 

[Timur Civan]

I'm feeling significantly more reach in the highlights on new olpf. Stop less on the bottom due to noise, but 2 up top. About 14+ overall.

Edit: @800

 

[Phil Holland]

Thanks Phil...

Yes, but this does not really answer the question. On your chart it does not say which chip is middle gray... unless I am blind.

I would prefer a direct answer as I just received my Dragon and something is not right, unless is is supposed to be this way. ???

Also, I use REDLogFilm Flat and use RedCines wave form tools etc.

I'll attempt to give you the most concise answer ever on this topic.

The direct answer I can give you is based on my own and Stacey's/Gavin's measured data, which is thankfully identical.

Where is 18% gray on that chart with within measured patches?

- At ISO 0050 18% gray at IRE 41 lands on patch 4 via REDlogFilm Flat with no Curve.
- At ISO 0100 18% gray at IRE 41 lands on patch 5 via REDlogFilm Flat with no Curve.
- At ISO 0200 18% gray at IRE 41 lands on patch 6 via REDlogFilm Flat with no Curve.
- At ISO 0400 18% gray at IRE 41 lands on patch 7 via REDlogFilm Flat with no Curve.
- At ISO 0800 18% gray at IRE 41 lands on patch 8 via REDlogFilm Flat with no Curve.
- At ISO 1600 18% gray at IRE 41 lands on patch 9 via REDlogFilm Flat with no Curve.
- At ISO 3200 18% gray at IRE 41 lands on patch 10 via REDlogFilm Flat with no Curve.
- At ISO 6400 18% gray at IRE 41 lands on patch 11 via REDlogFilm Flat with no Curve.

Additional Information:

- This chart contains 17 patches measured with the new OLPF. Patch 1 is the moment just before clipping and contains textured detail.
- With the new OLPF I have measured beyond 16 stops of what the camera can "see". I can actually see/measure 19 stops throughout the available ISO range.
- That 19th stop is dark and has a good bit of noise before tonal detail is "washed into the abyss", but there is separation and Dragon can see it amazingly.
- For fun additional info I have never seen anything beyond the 15th patch with a film stock. Any film stock. Commonly it tops out at 13 stops in reality.
- I would generally describe the common usable range as a 16-18 stops of Dynamic Range. The description of 16+ actually makes sense.
- Remember that Red recommends ISO 250-2000 as "cinema ready" usable ISO ratings.
- Some of these measurements are outside of this range because you have to test everything and look at what the camera can see.
- Patch 8 or 9 is really what you are looking for.
- Depending on who you are and how you "feel" about those stops that means ISO 800-1600 is the ideal range where you have equal stops above and below.
- RAW View on Dragon is currently a REDcolor2/REDlogFilm (FLAT) image at ISO 800 for this very reason.
- Personally for that reason ISO 800-1280 is likely the number you are looking for.

Film Logic and Other Tid Bits:

- If you want to maintain the most film like response you'll likely be shooting between ISO 250-800, I'll add a personal note for ISO 1000/1280 because of texture as well.
- I really dig what ISO 320-500 can do in reality. Especially in relationship to what film's tonal response is. Thought Dragon is much cleaner at those sensitivities in comparison.
- ISO 800 is essentially that middle point you are looking for above and below 18% gray. I provides a lightly textured image.

One Last Thought/Thing:

- One other note that falls out of a laboratory test such as this. Often real world lighting conditions and subject matter aren't going to fill up all of the Dynamic Range. However, having a lot of Dynamic Range captured does indeed effect the Workable Latitude of the image after the fact. An interesting thing I've found is what information is kept outside of the "recommended cinema ready ISO 250-2000" use. This equates overall to an image, a REDCODE RAW image, that can be manipulated by a Digital Colorist in powerful ways. I am nearly 100% of the time working with Dragon like I'm working with film. Thought I will say grading a clip it's interesting going beyond standard 1 stop pushes and pulls. More or less depending on how you shoot a very friendly approximate 3.3 stops across the tonal range without hurting the image quality. This is impressive actually and not something other cameras do very well.

 

[Medavoym]

Long and the short of it Red's description of 16+ is accurate. How much of that Dynamic Range at any given point in time will depend on your ISO rating and what sort of viewing Gamma/Curve you are utilizing.

Hi Phil,

If you measure the MX like this - counting the stops in RedLogFilm and NOT in any Red Gammas or without any curve applied - how many stops do you think you will see?

Probably 14.

Is that an accurate statement to say that the Epic MX has a 14-stop range?

Moreover - how many stops does the Alexa show in LOG (if you test it similarly)? Probably 18. Is that a fair assessment to say that the Alexa is an 18-stop camera?

Thanks!

 

[Nick Morrison]

- RAW View on Dragon is currently a REDcolor2/REDlogFilm (FLAT) image at ISO 800 for this very reason.

Awesome info. Thanks for sharing.

 

[Phil Holland]

Hi Phil,

If you measure the MX like this - counting the stops in RedLogFilm and NOT in any Red Gammas or without any curve applied - how many stops do you think you will see?

Probably 14.

Is that an accurate statement to say that the Epic MX has a 14-stop range?

Moreover - how many stops does the Alexa show in LOG (if you test it similarly)? Probably 18. Is that a fair assessment to say that the Alexa is an 18-stop camera?

Thanks!

Measured Dynamic Range tested under identical circumstances and methods:

Red Epic Dragon = 16+ for the reasons described above.
Arri Alexa = 14+ (due to the 14.1-14.25 nominal results as there is a hint of detail measured but it's not a clear stop of information)
Sony F65 = 14
Sony F55 = 14
Red Mysterium-X = 13.5
Phantom Flex 4K = 12
Canon C500 = 12

I need to stress here that I'm not merely looking at an image at one particular ISO. I am measuring what each system can "see" as well. Meaning where detail can be resolved, measured, and seen above and below the line. You have to look deeper into each "camera manufacturer X's" version of log. Things need to be measured with the image data and precisely to the intensity of the sample target. My methods involve both illuminated and reflected targets for a reason as it reveals truly who's "adding a bit more" into the optical and image processing mix. Each camera takes a full day to accurately test in reality.

An interesting thing really here is those last possible stops of visibility, meaning textured and resolved detail. Not pure image noise.

Everybody is keen on comparing Dragon to the Alexa, but in reality it's real competition is the Sony F65. At least in my mind. And that's where the discussion actually fun because then we get to talk color again.

I'll add one other thing, you may notice the Alexa is the only camera on this chart not suitable for 4K finishes. That should inform some people's opinions about a lot of things.

I'll add one other, other thing. To get the most out of each of these cameras you got to shoot them very right. I can make all of them look like junk if shot wrecklessly. However, when used for their strengths that's where interesting things really happen and pretty images are born.

 

[Martin Stevens]

Thanks again Phil,

Now that I know where the IRE chips fall (in your nice chart), I would like to make sure my deductions are correct.

So, are you saying that with the New Dragon OLPF that if you place middle gray at 41 IRE when at ISO 800, that there are
8 stops until clip? I feel like this is what your chart is showing, but I keep feeling like the answer is no.

And if there are 8 Stops from 41 IRE to clip with the new OLPF, then how many stops from 41 IRE to clip is there
with the earlier (low light) OLPF?

I just received my Dragon Upgrade and feel like I do not have the correct (new) OLPF installed as I requested, but
I need to do more tests as I had been using 48 IRE instead of 41 for middle gray measurements.

Also, do you think that there is a one stop difference between 41 and 48 IRE?

 

[Brian Boyer]

So, are you saying that with the New Dragon OLPF that if you place middle gray at 41 IRE when at ISO 800, that there are
8 stops until clip? I feel like this is what your chart is showing, but I keep feeling like the answer is no.

I think you're making the mistake of counting the chips when you should be counting the movement between chips. In other words, the jump from chip 8 to chip 7 is 1 stop. The jump from 7 to 6 is another stop, etc.

At 800 ISO it's 7 stops.

 

[Phil Holland]

Thanks again Phil,

Now that I know where the IRE chips fall (in your nice chart), I would like to make sure my deductions are correct.

So, are you saying that with the New Dragon OLPF that if you place middle gray at 41 IRE when at ISO 800, that there are
8 stops until clip? I feel like this is what your chart is showing, but I keep feeling like the answer is no.

And if there are 8 Stops from 41 IRE to clip with the new OLPF, then how many stops from 41 IRE to clip is there
with the earlier (low light) OLPF?

I just received my Dragon Upgrade and feel like I do not have the correct (new) OLPF installed as I requested, but
I need to do more tests as I had been using 48 IRE instead of 41 for middle gray measurements.

Also, do you think that there is a one stop difference between 41 and 48 IRE?

Yes. What I'm saying is that at ISO 800 you have 8 Stops (including that patch) from IRE 41 before clipping occurs.

18% Gray has essentially an IRE range for 41-48. The patches and data was measured precisely at IRE 41 for 18% Gray.

One full stop using REDlogFilm without a curve should give you approximately IRE 48-50.

 

[Brian Boyer]

Yes. What I'm saying is that at ISO 800 you have 8 Stops from IRE 41 before clipping occurs.

18% Gray has essentially an IRE range for 41-48. The patches and data was measured precisely at IRE 41 for 18% Gray.

One full stop using REDlogFilm without a curve should give you approximately IRE 48-50.

Well damn, I stand corrected. Now, I'm confused.

 

[Phil Holland]

Well damn, I stand corrected.

No, you're right. It depends basically if you start counting the next step up or start counting on the patch itself. Since this patch sequence is shot just at the moment before clipping that is indeed a stop of light and not full clip. Which is why some people may start on 0 or 1 respectively.

 

[Brian Boyer]

No, you're right. It depends basically if you start counting the next step up or start counting on the patch itself. Since this patch sequence is shot just at the moment before clipping that is indeed a stop of light and not full clip. Which is why some people may start on 0 or 1 respectively.

Okay. Thanks for the clarification.

Phew! I've been on a losing streak around here lately and I needed a win.

 

[Martin Stevens]

No, you're right. It depends basically if you start counting the next step up or start counting on the patch itself. Since this patch sequence is shot just at the moment before clipping that is indeed a stop of light and not full clip. Which is why some people may start on 0 or 1 respectively.

Woah, so if he's right, its 7, but you said you were right at 8.

When I count I do not count the chip I start on.

So, I started on chip 8 and then counted 8 chips to the left ending on chip 0 which is not there but is the real clip CHIP
because chip one still has info according to you.

 

[Brian Boyer]

Woah, so if he's right, its 7, but you said you were right at 8.

When I count I do not count the chip I start on.

So, I started on chip 8 and then counted 8 chips to the left ending on chip 0 which is not there but is the real clip CHIP
because chip one still has info according to you.

Yeah, if I understand him correctly, the white of the chip isn't the "whitest" you could go with the sensor because it's not clipping. Therefore, some people would consider the clip as another stop.

How'd we do, Phil?

 

[Martin Stevens]

If I understand him correctly, the white of the chip isn't the "whitest" you could go with the sensor because it's not clipping. Therefore, some people would consider the clip as another stop.

How'd I do?

Great, but I think Phil is saying 8 Stops from 41 IRE to full no detail clip if you do not count the 41 IRE chip, but do count
the pure white, no-info chip which is not on his chart.

Also............... "And if there are 8 Stops from 41 IRE to clip with the new OLPF, then how many stops from 41 IRE to clip is there
with the earlier (low light) OLPF?" ........ ???? Or, MX?

 

[Brian Boyer]

Great, but I think Phil is saying 8 Stops from 41 IRE to full no detail clip if you do not count the 41 IRE chip, but do count
the pure white, no-info chip which is not on his chart.

I should probably just let him answer but I can't help myself. I thought he was saying:

8 stops if you start from 0, which is full clip, to the 41 IRE chip, inclusive.
7 stops if you start from 1, which is just before clipping, to the 41 IRE chip, also inclusive.

...not counting the chip we start on, like on a board game.

 

[Martin Stevens]

I should probably just let him answer but I can't help myself. I thought he was saying:

8 stops if you start from 0, which is full clip, to the 41 IRE chip, inclusive.
7 stops if you start from 1, which is just before clipping, to the 41 IRE chip, also inclusive.

This is not counting the chip we start on, like on a board game.

Thanks Brian,

I just wish his chart had a zero chip so there was perhaps less confusion.
This way you could start on the zero chip, but not count it.

 

[Phil Holland]

Woah, so if he's right, its 7, but you said you were right at 8.

When I count I do not count the chip I start on.

So, I started on chip 8 and then counted 8 chips to the left ending on chip 0 which is not there but is the real clip CHIP
because chip one still has info according to you.

Thanks Brian,

I just wish his chart had a zero chip so there was perhaps less confusion.
This way you could start on the zero chip, but not count it.

Well that's where it's sticky. Starting from "white" which is within exposure if you count from there to 18% gray you land on patch number 8. Since that's a value it is a captured stop. You either start at 0 or 1 depending on who you are. In measuring however total stops you do obviously count what the camera reads as a stop.

Also, that's why I made that crazy note about ISO 800-1280. You decide where you like it The camera has enough to support any of those as a true middle in my mind.

The reason ISO 800 is considered that base is exactly that as it's measured there.

When you count your fingers you don't call your pinky zero. Them are fight'n words and pinky doesn't like to be left out.

 

[Martin Stevens]

When you count your fingers you don't call your pinky zero. Them are fight'n words and pinky doesn't like to be left out.

I love this line. Great point, but I'm trying to find out what my camera is supposed to deliver as stops ABOVE middle gray,
and not including it. heh heh heh.

When I get back to my camera I will test using 41 IRE as the mid gray base and see.

THANKS PHIL.... Let me ask this whole thing in another way. How many more stops above middle gray
does the new Dragon OLPF have over MX?

 

[Phil Holland]

I love this line. Great point, but I'm trying to find out what my camera is supposed to deliver as stops ABOVE middle gray,
and not including it. heh heh heh.

When I get back to my camera I will test using 41 IRE as the mid gray base and see.

THANKS PHIL.... Let me ask this whole thing in another way. How many more stops above middle gray
does the new Dragon OLPF have over MX?

Dragon with the new OLPF retains about 2 stops in highlights at ISO 800 compared to Mysterium-X at the same rating. The rest is found on the shadow side.

**edit, I'll add the highlight roll-off to clip is much, much nicer too.