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 :)

HDRx Overview...

Here we go:

Standard DR:

- Canon 5DMK2 11.86 stops DR @ 1080p
- Aaton Penelope ∆ 13 stops DR @ 4K
- RED1 MX 13.5 stops DR @ 4.5K
- ARRI Alexa 14 stops DR @ 1080p


HDR:

- RED Epic HDRx™ 18 stops DR @ 5K
- University of Warwick HDR camera 20 stops DR @ 1080p
- SpheroCam HDR 26 stops DR @ Full spherical: up to 50 megapixel (5.300 x 10.600 pixel), Panoramic: up to 100.000 pixels horizontal
 
Sanjin Jukic said:
Here we go:

Standard DR:

- Aaton Penelope ∆ 13 stops DR @ 4K
- RED1 MX 13.5 stops DR @ 4.5K
- Canon 5DMK2 14 stops DR @ 1080p
- ARRI Alexa 14 stops DR @ 1080p


HDR:

- RED Epic HDRx™ 18 stops DR @ 5K
- University of Warwick HDR camera 20 stops DR @ 1080p
- SpheroCam HDR 26 stops DR @ Full spherical: up to 50 megapixel (5.300 x 10.600 pixel), Panoramic: up to 100.000 pixels horizontal
Click to expand...

No argument with any of this except the 5D MKII. Not close.

How about some other comparative specs and prices?

Jim
 
Sanjin, let's put it this way:

if I had to guide a DP on different sides of these cameras and tell the up- and downsides, I would definitely get fired and be out of further gigs after a day or two if he or she took this advice serious...

Lists like these can be very counterproductive and indeed harmfull if you cannot sustain your claims in a usefull way...

In addition to that, you could actually risk harming the brands you mention, if they in some way don't live up to the numbers you're posting...
At least if people tend to use you as any kind of guide on these subjects...
 
Correction:

By DxOMark website Canon 5DMK2 DR was rated @ 11.86 Ev.

See the attachment or visit DxOMark website >>>
 

Attachments

  • 5DMK2_DR.jpg
    5DMK2_DR.jpg
    87.9 KB · Views: 0
If one are to publish stuff like this, one has to show how one came to the results, too...

Graeme loves his stouffer charts.

There are other methodologies, but it needs to be repeatable.

A number is NOT repeatable, and they can be pretty bewildering.

"Stops" does for example NOT tell us anything about "which stop" - in other words: At what light intensity range, the the effective DR holds...

I now the "base ISO value" should be a hint, but in my experience... it isn't...
 
Jannard said:
Heh heh...that's better.

Now we should talk about resolution, size, price and support. Then I would be happy.

Jim
Click to expand...

+ overall shootingflexibility, image flexibility and post flexibility...
There are more things to this equation than pure stops or pure resolution.

That's what will be cool with the Epic.

But I'd be surprised if we didn't see different performance of the MX in the Epic than in the R1 MX...
 
There is also a big difference between a "science project" and a deliverable camera.

Jim
 
Jannard said:
Heh heh...that's better.

Now we should talk about resolution, size, price and support. Then I would be happy.

Jim
Click to expand...

Jim,

you are right but in this case you also maybe forgot colors.

Actually the color accuracy of skin tone, sky, landscape, @ low light, etc,... should be the most important parameter of any image capture system.

That what I think and also not only one...

By my experience with using and owing RED1, RED1 MX, Canon 7D & 5D, Nikon D7000, Panasonic GH2 and Olympus EP2

the last one that is Olympus has the most accurate and pleasant image (stills also video) but has doomed 30p f***ed up video codec.

I would say in the near future high end digital cine cameras could be at the same (4K+ resolution)

but what will be difference between them is IQ (Image Quality) that means color accuracy, etc,...
 
Skin tone.

This one to me is often kind of full of it. Here is why I am begining to take this view.

- I think it is more important that cameras reproduce color accurately. Skin tone is only ONE factor.
- if you walk around and really look at people. Really look at their skin you will notice that few, if anyone has a similar skin tone. A roomfull of ten people, if you look them in the face will have different skin tones.

- I agree that a camera needs to produce the "texture" of the skin accurately. This is something film attempted to have a claim on for years.

- video destroyed skin tone. Panasonic made everyone pink and SONY made everyone green.

-red captures the texture of the skin and realistic natural color of the skin. For truly balanced skin tones, everyone needs makeup and that is also going to have to evolve with RED.
 
Sanjin Jukic said:
Correction:

By DxOMark website Canon 5DMK2 DR was rated @ 11.86 Ev.
Click to expand...

The 5D2 has nowhere *near* 11.86 stops of dynamic range. DxOMark only measures random noise. If all the 5D2 had was random noise, then we could shoot 12 stops on the 5D2 no problem (stills). But the reality is that the 5D2 is plagued by some of the worst pattern noise (AKA "banding" or "temporal FPN") on the face of the planet. It reduces the dynamic range to
a maximum of maybe 10 stops -- and that's only in good (balanced) light. If you are shooting in tungsten, deep shade, or any other imbalanced light, the usable dynamic range falls another 2-3 stops to 7 or 8 stops.
 
Daniel Browning said:
The 5D2 has nowhere *near* 11.86 stops of dynamic range. DxOMark only measures random noise. If all the 5D2 had was random noise, then we could shoot 12 stops on the 5D2 no problem (stills). But the reality is that the 5D2 is plagued by some of the worst pattern noise (AKA "banding" or "temporal FPN") on the face of the planet. It reduces the dynamic range to
a maximum of maybe 10 stops -- and that's only in good (balanced) light. If you are shooting in tungsten, deep shade, or any other imbalanced light, the usable dynamic range falls another 2-3 stops to 7 or 8 stops.
Click to expand...

This corresponds well with what I have seen...
 
Probably already asked,

Will there be a provided way to combine the two tracks of data into scene-referred HDR vales? I noticed that the info given in the first post says that a EXR output will spit out a multi-view EXR, not a single image float-point EXR with corresponding 'above 1.0' value highlight data. Will the track-a and track-x combining process create HDR images that have been camera-response curve compinsated? Just curious. Thanks.

-Eric
 
Sanjin Jukic said:
Here we go:

Standard DR:

- Canon 5DMK2 11.86 stops DR @ 1080p
- Aaton Penelope ∆ 13 stops DR @ 4K
- RED1 MX 13.5 stops DR @ 4.5K
- ARRI Alexa 14 stops DR @ 1080p


HDR:

- RED Epic HDRx™™™ 18 stops DR @ 5K
- University of Warwick HDR camera 20 stops DR @ 1080p
- SpheroCam HDR 26 stops DR @ Full spherical: up to 50 megapixel (5.300 x 10.600 pixel), Panoramic: up to 100.000 pixels horizontal
Click to expand...

The Spheron that I have is only a 1 pixel wide sensor by 5K high. It takes almost 11 minutes to do a full 360 degree pass at the highest resolution, but you can get reasonable 6-7K wide panoramas in 5-6 minutes (for one still). If the scene has any movement, you'll get temporal shifts in the RGB channels that shows up as striping. And while the 26 stops of range is very impressive, the shadows are very noisy, meaning the bottom 2-4 stops aren't useful to us in a production environment. Mine is several years old, so they may have released something faster, but its hard to tell from their website.

Now, if we can just get a some software for a Talon head to automatically do the panning/tiles with an Epic on board....
 
David Battistella said:
Skin tone.

This one to me is often kind of full of it. Here is why I am beginning to take this view.

- I think it is more important that cameras reproduce color accurately. Skin tone is only ONE factor.
- if you walk around and really look at people. Really look at their skin you will notice that few, if anyone has a similar skin tone. A roomful of ten people, if you look them in the face will have different skin tones.

- I agree that a camera needs to produce the "texture" of the skin accurately. This is something film attempted to have a claim on for years.

- Video destroyed skin tone. Panasonic made everyone pink and SONY made everyone green.

-Red captures the texture of the skin and realistic natural color of the skin. For truly balanced skin tones, everyone needs makeup and that is also going to have to evolve with RED.
Click to expand...


FIlm chemists spent years fine tuning emulsions to suite audience and Hollywood desire for gloss and glamour of golden and warm skin tones.
Scientists tell us that the human eye is particularly sensitive to skin tone, something to do with being able to perceive a threat revealed by a rise in blood pressure in an adversary.



Although a video camera may reproduce a test chart perfectly, there is not enough bits in the (10 bit video) system allocated to yellows and red to help record and later fine tune skin tone to the same degree as film.

I've suggested targeting bits to skin tone before now to manufacturers, who do make subtle changes to the AD. However, since we are moving to the combination of 14 bit AD and 14 bit recording I'm not sure if the concept of allocating more bits in a particular zone to create greater subtly in tone and colour would make a lot of difference?



Mike Brennan
 
Michael Brennan said:
Although a video camera may reproduce a test chart perfectly, there is not enough bits in the (10 bit video) system allocated to yellows and red to help record and later fine tune skin tone to the same degree as film.
Click to expand...

The difference in skin tone rendition between video and film has nothing to do with the limitations of 10 bit log recording formats. It's because of the differing spectral response. Digital camera manufacturers that use color filters (e.g. Bayer) tend to be going for "accurate" more than "film-like", but it's theoretically possible to build one that matches the color of a given film stock *exactly*. But if they did that, they'd have to take the bad (film's very inaccurate color) with the good (film's nice skin tone rendition).

Personally, I don't think film's skin tone rendition is any better than RED or even a DSLR. I think we can and do get ideal and finely tuned skin tones with digital CFA (color filter array) cameras. But if you still prefer film, then the only option is to suggest that manufacturers change the spectral response of their color filters.
 
You must log in or register to reply here.
Back
Top