Jump to content

MX v Alexa


Stephen Williams

Recommended Posts

  • Replies 54
  • Created
  • Last Reply

Top Posters In This Topic

  • Premium Member

Seems like the Red camp are not happy, another 'ambush' test that knew nothing about..........apparently it was Ted's camera one of the first to be upgraded...... firmware not the latest......sounds like a repeat of the Red One debut and every subsequent test everyone ever did.

Link to comment
Share on other sites

Seems like the Red camp are not happy, another 'ambush' test that knew nothing about..........apparently it was Ted's camera one of the first to be upgraded...... firmware not the latest......sounds like a repeat of the Red One debut and every subsequent test everyone ever did.

 

Not unhappy at all... just posting the facts while answering the question. We could have easily brought a camera with the production OLPF and new firmware. No worries though... we are happy to be included with the Alexa as one of the two future cameras.

 

Why the need to stir it up so much? :-)

 

Jim

Edited by Jim Jannard
Link to comment
Share on other sites

They lost me when they said film is dead....but then again...it is a story by HDusermagazine.

 

 

 

same here, can't bring myself to finish reading. It's one thing to talk about HD cameras and tests but another to come out and attack film. This guy is an butt. Anyone care to give the Cliff notes?

Link to comment
Share on other sites

  • Premium Member

Same here. Though I assume it says something along the lines of "the red MX and the alexa perform similarly with a lower noise floor on the red, but more pleasing renditions on the arri" or swap the two.

Both cameras are a moving target, so tests are moot a bit until such time as you're ready to use one, then you should test qll systems your considering out to see which one works best for your specific situation.

Link to comment
Share on other sites

  • Premium Member
Not unhappy at all... just posting the facts while answering the question. We could have easily brought a camera with the production OLPF and new firmware. No worries though... we are happy to be included with the Alexa as one of the two future cameras.

 

Why the need to stir it up so much? :-)

 

Jim

 

Your post yesterday obviously pushed a button

 

"Stephen... 1st off, you didn't get banned. You got a few hour "timeout". 2nd... it wasn't for questioning if the specs would be met.

 

Jim"

 

Edit,

 

I was truly hoping that future products from Red would work out of the box, and that the weekly upgrade programme & beta issues would be a thing of the past. As a prospective customer I am a little disappointed.

Link to comment
Share on other sites

  • Premium Member
Not unhappy at all... just posting the facts while answering the question. We could have easily brought a camera with the production OLPF and new firmware. No worries though... we are happy to be included with the Alexa as one of the two future cameras.

 

Why the need to stir it up so much? :-)

 

Jim

 

Haha,because we enjoy pressing the buttons,they always work.

 

I also read that piece,perhaps a little premature in sounding the death bells for film.

I know it has largely happened in stills,particularly with commercial work.

I for one still shoot occasionally on 10x8,although it's getting harder to get it processed.

 

But seriously,I haven't had my Red upgraded to MX yet,but recently I did get the opportunity to have a play with a pre-production Alexa at Arri UK,and I hope to shoot with it soon.Seriously impressive camera,especially from an operators point of view.

It looks like an ordinary shoe box,rather than the post-apocalyptic shoe box that the Red is.

However I also had a chance to hold Ted's Epic prototype at the London Red day.

I must say that Red have learned an enormous amount since the early days,regarding ergonomics.(Spec's influence perhaps?)

If you can deliver on this one Jim,with real 4K,100fps & 15stops DR you'll have a camera to be very proud of.

 

Tom.

Link to comment
Share on other sites

Haha,because we enjoy pressing the buttons,they always work.

 

I also read that piece,perhaps a little premature in sounding the death bells for film.

I know it has largely happened in stills,particularly with commercial work.

I for one still shoot occasionally on 10x8,although it's getting harder to get it processed.

 

But seriously,I haven't had my Red upgraded to MX yet,but recently I did get the opportunity to have a play with a pre-production Alexa at Arri UK,and I hope to shoot with it soon.Seriously impressive camera,especially from an operators point of view.

It looks like an ordinary shoe box,rather than the post-apocalyptic shoe box that the Red is.

However I also had a chance to hold Ted's Epic prototype at the London Red day.

I must say that Red have learned an enormous amount since the early days,regarding ergonomics.(Spec's influence perhaps?)

If you can deliver on this one Jim,with real 4K,100fps & 15stops DR you'll have a camera to be very proud of.

 

Tom.

 

Tom... we are proud of EPIC but need to deliver it. Images from "Blackie" (1st working prototype) will be shown soon. It is measured 4K (from 5K Bayer), 100fps and 13.5 stops DR (according to the Arri test setup). 15 stops of DR won't happen until the next sensor due late this year.

 

Jim

Link to comment
Share on other sites

  • Premium Member
I also read that piece,perhaps a little premature in sounding the death bells for film.

I know it has largely happened in stills,particularly with commercial work.

 

 

The problem for me is that anyone who says "film is dead" is just buying into a ridiculous gear pissing match over numbers and statistics.

 

Even in stills, film is actually not dead. Sure, not many people choose to use it, but there are plenty who still do. And critically perhaps in this analogy, it's fine artists rather than commercial photographers.

 

It's like saying LP's are dead. Well they are not are they ? Diehards still use them. Yes they make up a small minority of the total music market.

 

I'm realistic and would say most of my recent work hasn't been shot on film.

 

This ridiculous clamouring to pronounce film being dead speaks volumes about the person making such pronouncements. It says they care more about specs than using their eyes. It says that they care more about the emotional choice to NOT shoot film and desperately want the reassurance that their chosen alternative is as good as film.

 

Meanwhile, good DP's who care about telling stories visually will simply choose the best tool for the job, based on the script, directors intent and available budget.

 

jb

Link to comment
Share on other sites

  • Premium Member

"I'm realistic and would say most of my recent work hasn't been shot on film."

 

That's true for me also,but I do still shoot on film,not the smaller formats,but 10x8 & 5x4.

All my cine work is digital,although I keep on trying to convince clients.

I've got a job coming up next month where film is really the best choice,and I've been able to convince the client that film is the best choice,but only for the stills part of the job.

It wasn't an easy job to convince them,I can tell you.

Yes,film will always be here,and as long as it is I'll continue to use it if I can.

However I won't use or recommend film,if digital would be better for what I want to achieve.

Choose the best tool for the job.

 

Tom.

Link to comment
Share on other sites

  • Premium Member
15 stops of DR won't happen until the next sensor due late this year.

???

I'm confused.

You're 8-9 months away from a sensor with 15 stops of DR?!

An extraordinary claim.

A camera like that would wipe the floor with the competition.

So, why are you bothering with the M-X?

 

Also what do you mean by "15 stops"?

The Dalsa has/had a 16-bit ADC but it had only about a 12 stop dynamic range.

 

According to your competition, what you are suggesting appears to be theoretically impossible, particularly with a CMOS sensor.

The military applications alone would probably be worth quite a few orders of magnitude than your present target market.

And they'd probably let you play with lots of neato stuff :lol:

Link to comment
Share on other sites

???

I'm confused.

You're 8-9 months away from a sensor with 15 stops of DR?!

An extraordinary claim.

A camera like that would wipe the floor with the competition.

So, why are you bothering with the M-X?

 

Also what do you mean by "15 stops"?

The Dalsa has/had a 16-bit ADC but it had only about a 12 stop dynamic range.

 

According to your competition, what you are suggesting appears to be theoretically impossible, particularly with a CMOS sensor.

The military applications alone would probably be worth quite a few orders of magnitude than your present target market.

And they'd probably let you play with lots of neato stuff :lol:

 

These specs have been out for awhile... the M-X sensor has been measured in the last month by many people and in several different ways... including on the Arri DRTC (13.5 stops). Monstro has greater than 1.5 stops less noise than the M-X.

 

Late this year is our target... but we currently don't have a history of being on time. Hopefully we will do better with the Monstro sensor. 1st application of that sensor is 24x36mm at 100 fps...

 

Jim

Link to comment
Share on other sites

  • Premium Member
... we are happy to be included with the Alexa as one of the two future cameras.

 

Third out of the gate will be Sony's Q67 chip -- Basically it looks like a Bayer pattern rotated 45 degrees, with smaller photosites. It'll be the next generation in the F-35 line.

 

 

 

 

-- J.S.

Link to comment
Share on other sites

Third out of the gate will be Sony's Q67 chip -- Basically it looks like a Bayer pattern rotated 45 degrees, with smaller photosites. It'll be the next generation in the F-35 line.

-- J.S.

 

Which is amusing, because according to Sony, Bayer patterns are bad :-) I guess their new mantra will be Bayer patterns are bad unless they're rotated 45°. Perhaps they've not read page 124-125 of Alan Robert's excellent "Circles of Confusion" book.

 

Graeme

Link to comment
Share on other sites

  • Site Sponsor

As someone on CML said " you will know when film is dead when you cannot buy it anymore" I would go one further and say you will know it is dead when you cannot buy or make it anymore.

 

These "film is Dead" statements are just juvenile and seem like some kind of insecurity to me. The painting world went through something like this years ago with the introduction of Acrylics and now years later you can get both Acrylic paint and Oil paint. I don't see any reason why emulsion and electronic imaging cannot co exist.

 

Enthusiastic DIY types are already making their own emulsion and with material becoming much more malleable I see no reason why a small plant could not use nanotech to formulate any emulsion recipe you would want.

 

Ten years ago there was just a glimmer of possibility that a small shop like Red or Arri could get it's own sensor built and make a camera around it but now it is possible...It is not rocket science anymore and we have small shops doing that too these days.

 

-Rob-

Link to comment
Share on other sites

  • Premium Member
I guess their new mantra will be Bayer patterns are bad unless they're rotated 45°.

 

The rotation is a good idea wrt aliasing on strong horizontal and vertical patterns, of which there are plenty in the real world. Then again, we have a pilot shooting now where dang near every location includes some chain link fence..... ;-) (They're on D-21's btw.)

 

 

 

 

 

-- J.S.

Link to comment
Share on other sites

  • Premium Member

I couldn't get past this howler:

 

To be honest, I never thought the dynamic range discussion was a valid reason. Just look at the final result. It’s hardly a reason to baulk at two stops less than film, if a theatrical print made in a high-speed printing process has at best seven stops dynamic range, and your TV (that goes for DVD and Blu-Ray as well, of course) only displays 8-bit colour. But anyway, if that’s your only reason, it’s gone. The dynamic range of these two cameras blew me away. ARRI has measured 13.5 stops. What we found seemed to support that. Which brings me finally to our test.

 

 

The reality is, you don't need anything like the dynamic range of the original scene to produce a satisfactory picture.

The average sunny outdoor scene can have a total range that easily exceeds 20 stops, which means that no known capture medium can capture everything.

However once it is captured (as well as the medium will allow), the difference in brightness betweent he brightest and the darkest parts of the reproduced image can be massively reduced, and you eye still interprets it as an acceptable picture. Look at these images

 

8stopto1.jpg

 

 

The first one has the full 256 possible levels of an 8-bit JPEG image. That is, the brightest pixels are at level 255, while the darkest pixels are at zero. Since we need at least one step in the 255-step "staircase" to produce any sort of image, it follows that there is a 2^8 or an 8-stop brightness range in this image.

Now, by fiddling with the photoshop "curves" function, we can lift the darkest pixels away from the zero line, reducing the contrast.

The second picture has the blacks lifted up to a level of 32, so there is now only a 256/32 = 8 times difference betrween the brightest and darkest pixels. 8 = 2^3, so that's only three stops.

The third picture has the blacks lifted up to a level of 64, which means only 2 stops.

The last one has its blacks lifted up to a level of 128, so there is now only a 1-stop difference between the brightest and darkest pixels.

OK a trifle dull and washed-out looking, but you can still clearly make out the model's face, (or so they tell me :rolleyes:... )

 

The important thing to understand is that, while you can have a vastly smaller difference in the brightness of adjacent features in a reproduced image, and still get perfectly satisfactory results, you still have to be able to capture the original, sometimes huge, differences in the first place, so as to have something to reduce! The more dynamic capture range the medium has, the better it can do this.

 

It is routinely stated here and in other places that film has a dynamic range of "about 14 stops". I would dearly love to know where people get this figure from.

Link to comment
Share on other sites

  • Premium Member
you still have to be able to capture the original, sometimes huge, differences in the first place, so as to have something to reduce!

 

What we do with those sometimes huge differences is reduce them in the real world to fit within the dynamic range of the system we're shooting with. Them's fancy words for what everyone here knows as "fill light". ;-)

 

 

 

 

 

-- J.S.

Link to comment
Share on other sites

  • Premium Member
What we do with those sometimes huge differences is reduce them in the real world to fit within the dynamic range of the system we're shooting with. Them's fancy words for what everyone here knows as "fill light". ;-)

-- J.S.

 

 

Quite so. Dead easy in an enclosed TV studio or sound stage.

 

"reduce them in the real world to fit within the dynamic range of the system we're shooting with"

 

Or try to at any rate.

 

Let's see you fill-light the Grand Canyon, or the streets of New York. :lol:

 

What film effectively does is put a gigantic black tent over the outside world, and supply you with millions little Pixies all running around carrying lights and who know exactly how to light it to your personal taste, and can do it in the time it takes you to set up your camera

 

What is then recorded on the emulsion can be easily imaged by an electronic camera.

 

Film may be dying, but it's not not from natural causes.

Link to comment
Share on other sites

I couldn't get past this howler:

 

To be honest, I never thought the dynamic range discussion was a valid reason. Just look at the final result. It’s hardly a reason to baulk at two stops less than film, if a theatrical print made in a high-speed printing process has at best seven stops dynamic range, and your TV (that goes for DVD and Blu-Ray as well, of course) only displays 8-bit colour. But anyway, if that’s your only reason, it’s gone. The dynamic range of these two cameras blew me away. ARRI has measured 13.5 stops. What we found seemed to support that. Which brings me finally to our test.

 

 

The reality is, you don't need anything like the dynamic range of the original scene to produce a satisfactory picture.

The average sunny outdoor scene can have a total range that easily exceeds 20 stops, which means that no known capture medium can capture everything.

However once it is captured (as well as the medium will allow), the difference in brightness betweent he brightest and the darkest parts of the reproduced image can be massively reduced, and you eye still interprets it as an acceptable picture. Look at these images

 

8stopto1.jpg

 

 

The first one has the full 256 possible levels of an 8-bit JPEG image. That is, the brightest pixels are at level 255, while the darkest pixels are at zero. Since we need at least one step in the 255-step "staircase" to produce any sort of image, it follows that there is a 2^8 or an 8-stop brightness range in this image.

Now, by fiddling with the photoshop "curves" function, we can lift the darkest pixels away from the zero line, reducing the contrast.

The second picture has the blacks lifted up to a level of 32, so there is now only a 256/32 = 8 times difference betrween the brightest and darkest pixels. 8 = 2^3, so that's only three stops.

The third picture has the blacks lifted up to a level of 64, which means only 2 stops.

The last one has its blacks lifted up to a level of 128, so there is now only a 1-stop difference between the brightest and darkest pixels.

OK a trifle dull and washed-out looking, but you can still clearly make out the model's face, (or so they tell me :rolleyes:... )

 

The important thing to understand is that, while you can have a vastly smaller difference in the brightness of adjacent features in a reproduced image, and still get perfectly satisfactory results, you still have to be able to capture the original, sometimes huge, differences in the first place, so as to have something to reduce! The more dynamic capture range the medium has, the better it can do this.

 

It is routinely stated here and in other places that film has a dynamic range of "about 14 stops". I would dearly love to know where people get this figure from.

 

Unfortunately, there are many ways to measure DR and most results are subject to some interpretation.

 

The engineering definition is signal/noise. Typically, most understand it as holding detail in the brightest and darkest chips in a chart, like a Stouffer. The discrepancy usually comes on the dark end of the chart and how much noise is acceptable where there still is detail. Which is why DR measurements are not nearly as relevant, except for marketing, as useable range... what can you see in range while accepting the amount of noise. Then there is NR.

 

Our sensor program is designed to continue to lower the noise floor. The M-X sensor is certainly proof that we have greater than 2 stops less noise than our original sensor. Our next sensor will have an additional more than 1.5 stops less noise than the M-X. This translates to greater DR and more useable range. The primary difference with the coming sensor is better highlight protection as well as less noise in the shadows.

 

Digital usually falls down in the highlights... film in the shadows. This is one of the reasons that people still believe there is more DR in film. They are used to seeing better highlight control. The primary evidence for digital over film as it relates to DR is now found in the shadows. You can't shoot ISO 2000 and above and get descent results with film. You can with both our current sensor and the Alexa. All high ISOs with digital are just RAW plus a curve. The only way to achieve such great high ISO results with digital is if the noise floor is incredibly low.

 

Sensor design... or more succinctly pixel design, is attacking both ends of the range. But it is still a work in progress/process. It seems clear to us that the day is near where a digital sensor will outperform film on every level... except grain. That can always be added.

 

Film will never be dead. The projects shot on film will live forever (or at least as long as man keeps records). But it is dying in the respect that it's use will diminish relative to digital jobs. If that is the case... someone has to be respectful to the capabilities of film. Resolution, color, range and "feel". We are trying our best.

 

Jim

Link to comment
Share on other sites

  • Premium Member

I dunno Jim, I've seen '19 from Kodak pushed 2 stops which it still a-ok in my book (on 35). . . . Depends, as mentioned on what one can accept and most importantly, scene content! I've seen film and digital look "noisiest," in lower contras scenes where the eye is more prone to "notice," the differences in luminance that grain and noise introduce...

I'd also say that while Digital sensors are getting better, I seriously doubt they'll outperform the blanket term of "film." Will they be better than 500asa film--- some of them probably already are. But, will they be better, which is in and of itself an arbitrary term, than 50D film at the same time, probably not. That's film's biggest strength, the ability to choose speed relative to what you're shooting. Regardless of what smart things you can do with data off of a chip, the fact remains that it is still balanced for some speed and some color. With a film camera you just swap out the stock for what's appropriate in that scene as opposed to doing so electronically. The electronic method can often yield very nice results, but the more you finagle the data, the more you tend to loose and the more artifacts you're prone to introduce. Perhaps one day we'll have "field swappable sensors," where you can load a 50D sensor into a camera for that shot in the desert at noon, and then swap out to a 3200T sensor for a night exterior in a hostile location where you can't rig. But, until that day comes I wholly believe that film and digital will coexist as production formats much as they do now chosen for budgetary(sad)/physical reasons as well aesthetics.

 

p.s. if you do invent "field swappable sensors" I request 1 free camera of the current generation for thinking it up!

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Forum Sponsors

Metropolis Post

New Pro Video - New and Used Equipment

Gamma Ray Digital Inc

Broadcast Solutions Inc

Visual Products

Film Gears

CINELEASE

BOKEH RENTALS

CineLab

Cinematography Books and Gear



×
×
  • Create New...