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

[Blair S. Paulsen]

I clearly stipulated that my comments were predicated on particular use cases to AVOID the kind of distortions you claim I am creating.

In any case, I really don't think I claimed there were magic beans in RAW data. In fact, my intent was to cite the interdependency of the decoding step with measured performance of other parts of the signal chain upstream from user access.

FWIW the more pixels/compression (RED), less pixels/compression (ARRI), stratagems each have upsides - IMO, they also add a subtle character to the imagery as they impact the precise response characteristics of the MTF curve.

I am fascinated by many subtle aspects of how the technology impacts imaging, as are many of us. That said, I agree that the Epic and Alexa have much more in common than they do meaningful differences from a broader viewpoint. Sold.

I suppose Photoshop is more important on some level but isn't the whole point what we can with such tools working together? Am I missing something?

I've learned plenty from both Jake and Adrian on RU and respect their contribution to the community. I certainly did not wish to mislead fellow RedUsers with my comments, perhaps my phrasing was poor.

Cheers - #19

 

[Adrian Jebef]

I'm not trying to get on anyone's nerves, I just enjoy a good discussion.

The thing I take issue with is how a raw recording seems to be held up as a holy grail on these boards. The idea that because the EPIC or DRAGON records a raw R3D file it is somehow innately superior because of it. There is constant opaque reference to the incredible amount of dynamic range, color gamut, shadow detail, highlight retention, etc, etc. that can somehow be "found" in the raw R3D file. All of these attributes are a result of the sensor and recording path as Blair pointed out. Though they do indeed "exist" in the R3D they are only made real when developed in the software. There is no real magic behind developing raw files. If you are not getting all you think you should be getting out of an R3D it is either a limitation of the sensor or a limitation of the post software.

The question concerning a 5D frame vs Photoshop was rhetorical. They are equally important. It takes both to get an image. The reliance on the idea of shooting raw first and foremost often overshadows the reality of developing it into a usable image. IMO a camera system should be able to show you on-set what you'll be able to get in Post.

 

[luigivaltulini]

I am much more simple,
It is true that out by Alexa is better than EPIC, never denied the opposite, but now Dragon is better than Alexa.
It is true Alexa still has a small advantage in DR, but not in noise, color and resolution and flexibility in post production, this applies to ProRes and RAW.
At the end of the day what is best today?
This month I had 5 productions that after seeing Dragon vs. Alexa only monitor on the set have decided to switch all to Dragon, whereas before they were convinced on Alexa, this is a great reference for me and a big change.
So what?
Simple time Dragon is better in both cases, ProRes and RAW and other formats.
Why make a thousand questions when the answer is staring at the monitor in front of you.

Dragon is better and is growing now ..

 

[Gunleik Groven]

Luigi... That sounds a bit like the "confidence building" quality Dragon has.

It is quite easy for a DP (no matter what can be done in post) and look at two images and think:
"I take the nice one"...

 

[D Fuller]

Here's a question: What is more important, a Canon raw frame from a 5D or Adobe Photoshop?

Isn't that a little like asking "Which is more important, a fish or fire?"

It all depends how hungry you are?

 

[Adrian Jebef]

Isn't that a little like asking "Which is more important, a fish or fire?"

It all depends how hungry you are?

It's rhetorical.

The fish/fire example suggests you can just eat the fish without cooking it. Unfortunately that's not how raw files work. They gotta get baked one way or the other.

 

[D Fuller]

It's rhetorical.

The fish/fire example suggests you can just eat the fish without cooking it. Unfortunately that's not how raw files work. They gotta get baked one way or the other.

Yeah, I understand. (The essential weakness of analogy.) But I can see a 5D image with Preview if I'm in a hurry.

You're right, the images have to get baked, and I believe the development of the digital footage is very important to the craft of cinematography in digital cinema. But I don't agree that a camera should be able to show you on-set everything you can get in post. For it to be able to do that would be to deny they incredible power of a grading suite. I do think that the camera should be able to show you the outlines -- a good idea of what you'll be able to get, that gives you confidence that you got what you came for. But I think part of the essential value propositions of digital cinema (and the thing that makes it not like video) is that you get to leverage all the computer power that sits back in the grading suite to finish the image. And you get to spend time that is not as expensive as on-set time to do it.

I think an awful lot of stuff is getting pushed on to the set that gets in the way of performance. And I don't mean just the actors' performance. I include the DP and director. Committee meetings around the monitor are stifling.

 

[Brian Boyer]

Here's a question: What is more important, a Canon raw frame from a 5D or Adobe Photoshop?

Isn't that a little like asking "Which is more important, a fish or fire?"

It's rhetorical.

The fish/fire example suggests you can just eat the fish without cooking it. Unfortunately that's not how raw files work. They gotta get baked one way or the other.

Yeah, I understand. (The essential weakness of analogy.)

Don't be too hard on yourself, David. The question isn't as clearly rhetorical as it could be.

As asked, I'd say the answer is Photoshop because it's used for a great deal more than JUST reading raw files from cameras.

I realize Photoshop's ubiquity in doing such tasks is supposed to be understood but, if you're a graphic designer, reading raw camera files might not be the first thing that comes to mind.

I think a less ambiguous way to ask the question would be, "What is more important, a Canon raw frame from a 5D or Adobe Camera Raw?"

 

[luigivaltulini]

as expected ... the more things change ......
ADD has arrived and is making great strides ..

[/URL][/IMG]

 

[Paulo Emílio]

Luigi, based on your images it's the old is better than new?

 

[D Fuller]

Don't be too hard on yourself, David. The question isn't as clearly rhetorical as it could be.

As asked, I'd say the answer is Photoshop because it's used for a great deal more than JUST reading raw files from cameras.

I realize Photoshop's ubiquity in doing such tasks is supposed to be understood but, if you're a graphic designer, reading raw camera files might not be the first thing that comes to mind.

I think a less ambiguous way to ask the question would be, "What is more important, a Canon raw frame from a 5D or Adobe Camera Raw?"

Or in the case of the camera this forum is built around, "What is more important, an r3d or the tool you use to develop it?"

The key idea being that the development of an r3d is an integral part of the exposure decision. This relates to a discussion we had in real depth a week or two ago in the "Exposure and Native ISO" thread. I think Adrian and I understand the role of development in an r3d image very similarly, but we have different views on just how that should inform the design of a shooting set.

If I understand him correctly, Adrian would like the image out of the camera to be pretty much finished. I'm happier when it's more like what we used to see in dailies: an image that tells you what you captured, but not necessarily an image will look like in post. (Or maybe I'm mis-interpreting him -- Adrian, please chime in here if that's the case.)

Back at the beginning of time, i spent about as much time in a darkroom as I did shooting, and I came to understand the relationship between exposure and development very intimately. It took me a long time to realize that the same relationship exists in digital cinema. And for me, that's the primary differentiator between digital cinema and video.

 

[luigivaltulini]

Luigi, based on your images it's the old is better than new?

ciao Paulo,
old is better only in low condition (noise only) vs. new, new is more better in highlights vs old.
But do not worry adjustments will come and everything will be even better.
These are just examples of how RED advances every month with tools and calibrations.
A.D.D is a great weapon when it will be implemented for the movie.

 

[David Battistella]

David,

I could not agree more. The advancements of RED and the debayering of r3ds gets better over time and we do need to know how much light is hitting the film plane/sensor to discover the best way to develop based on the creative image you are going for. That thread was a good discussion. I'm more like you in that I do not need a pretty image out of the box, but I do want a good idea of how that sensor is going to respond to the amount of light it is fed.

RAW, is an important consideration for the future of the image you captured. The container, the r3d file, just houses data that needs to be extracted and wee have already seen with r3d that this improves over time.

The he reason I say it's important is that the raw image you process today will not yield as good an image as it will some time in the future as reds decoding of that image improves.

These sensor advancements equal about have the equation as the way the image gets decoded is the other half. All my 2008 footage is far better now than it was. This why I am so glad it was never baked in.

Pretty images out of the box used to be more for consumers more than professional image makers. I prefer the two step process over a WYSiWYG image, even if it's raw.

battistella

 

[James_Mills]

ciao Paulo,
old is better only in low condition (noise only) vs. new, new is more better in highlights vs old.
But do not worry adjustments will come and everything will be even better.
These are just examples of how RED advances every month with tools and calibrations.
A.D.D is a great weapon when it will be implemented for the movie.

I don't see any blatant difference in highlights between the two in your newest test, but the new OLPF with A.D.D. appears to have twice the noise of the old. And there's still color tints everywhere. To my eyes this test suggests that the old OLPF is much better at this point.

So far all that has been proved is that RED's claim of 16+ stops is specious at best, and the justifications from mods thus far on here have been extremely equivocal and unconvincing. I'm also disappointed in the noise found at 2500 iso here in Bjork's test and don't find this "useable": https://vimeo.com/87804971

I'm curious now as to how we had ever justified the MX's claim to having 13.5 stops of DR? The Dragon appears to maybe have 13.5 stops now yet everyone is so sure it outperforms the MX by leaps and bounds (and many of the tests posted thus far seem to support this theory). I'm in this camp:

Adrian Jebef[/URL]]IMO a camera system should be able to show you on-set what you'll be able to get in Post.

 

[luigivaltulini]

James,
be for example here you can see more than 2-3 stops in mx in high light, not to mention color etc. etc., this is a significant improvement.
http://www.reduser.net/forum/showth...Vs-Alexa-Sony-threads-are-belong-to-us/page51
So we come back to say that Dragon has 16.5 stops usable total.
Perhaps it is wrong to show these tests are interpreted in the wrong way.
Why are you disappointed with the noise found? What would you like to 80000iso clean a camera?
I honestly do not understand you, I do not know where you're going?

just my thought.

 

[Brian Boyer]

I don't see any blatant difference in highlights between the two in your newest test...

Really? You don't see the 1+ stop of difference between the old OLPF and new OLPF with calibration in these two images? I think it's significant, but I tend to favor highlight retention over virtually noise-free shadow areas, especially when there are products like Neat Video.

These are Luigi's overexposed T2.8 shots, with nothing but identical lift/gamma/gain adjustments made to both.


Old OLPF



NewOLPF

Or, the 2 stop or so difference between the new OLPF with calibration and the MX sensor in Scarlet?

just for fun Dragon and scarlet :

https://copy.com/p82Gg3MmKNDx

This amounts to 7+ stops above middle gray compared to MX's 5. Yes, the lows are noisier now, but that could change in the future and A.D.D. seems to help on that front as well.

 

[luigivaltulini]

thanks Brian ,
in fact, the comparison between Scarlet and DRagon was still Old OLPF, with the new OLPF you need to add 1.3 stops.

 

[Brian Boyer]

Or in the case of the camera this forum is built around, "What is more important, an r3d or the tool you use to develop it?"

I thought about that but Adrian put forth an analogy which, as a matter of course, uses examples other than the ones being discussed (R3D & SDK, for example), so I left it alone.

But, I agree, it's still a more pointed question than his original one, IMO.

And, in case anyone is wondering, I apparently don't have anything better to do on a Saturday morning than nitpick semantics.

 

[Brian Boyer]

thanks Brian ,
in fact, the comparison between Scarlet and DRagon was still Old OLPF, with the new OLPF you need to add 1.3 stops.

Wow. I missed the old OLPF part vs. Scarlet. Thanks for pointing that out, Luigi.

That's even better.

 

[James_Mills]

James,
be for example here you can see more than 2-3 stops in mx in high light, not to mention color etc. etc., this is a significant improvement.
http://www.reduser.net/forum/showth...Vs-Alexa-Sony-threads-are-belong-to-us/page51
So we come back to say that Dragon has 16.5 stops usable total.
Perhaps it is wrong to show these tests are interpreted in the wrong way.
Why are you disappointed with the noise found? What would you like to 80000iso clean a camera?
I honestly do not understand you, I do not know where you're going?

just my thought.

I know it has more useable stops than the MX, but I'm saying if teh Dragon can't even argue a legitimate 14 stops in comparison to the Alexa's 14 (by their measure), then the MX claiming to have had 13.5 is ridiculous. It's a testament to how far beyond the MX Alexa really is in terms of DR, and it is thus of little surprise why it is used 10 times as often in recent features.

And no, I'm not expecting a camera that can have clean 8000 iso, but there was claims as lofty as having useable 6400 ISO early on (though ones to this extreme were probably made by mods and early users as opposed to actual RED staff), so I was just expecting less noise at 2500. The video Bjorn posted at 2500 looks easily as noisy if not noisier than shots I've done at 1600 on MX.

Really? You don't see the 1+ stop of difference between the old OLPF and new OLPF with calibration in these two images? I think it's significant, but I tend to favor highlight retention over virtually noise-free shadow areas, especially when there are products like Neat Video.

I was referring to Luigi's most recent test in which the difference in highlights aren't that apparent whereas the alarming amount of noise in the new OLPF suggest the OLD OLPF looks a lot better imo:

as expected ... the more things change ......
ADD has arrived and is making great strides ..

[/URL][/IMG]