The best $5,750 you can spend [updated]

Brook Willard · 01-30-2010, 02:01 PM

Hey folks,

For the past few weeks, I've been the DIT on a pilot J. Michael Muro was shooting [we wrapped last night]. A few weeks ago, Jarred and Keslow Camera arranged for RED to send us Jarred's own personal MX camera to our location days in Chicago for some key night exteriors.

Now over the past few years, I've spent way more than my fair share of days shooting with the RED ONE. I also spend a lot of time [more than might be suggested by how RED-centric I am online] working as a DIT for traditional HD camera systems.

Shooting with the new MX sensor in Chicago impressed me massively. I was hoping for a camera with more dynamic range and a lower noise floor and RED has delivered. The night exterior city footage from Jarred's preproduction firmware MX camera looked better at T4 than the RED ONE M sensor looked at T2.2. Now don't take that as me saying it's "X" stops more sensitive or "X" times better, but for this particular setup [overlooking the river], it was terrific. We ended up shooting the shot at T2.2 on the MX anyways... just to expose to the right as much as possible.

Those of you that went to RED DAY at RED Studios Hollywood would've seen a few of the shots from our project. I wasn't at RSH myself [was teaching the Data Handling training for local 600 in NY], but I'm told that [among others], a shot of Emmy Rossum on the phone was played. I want to give a little information about that particular shot as well.

It was our last shot before switching to night mode, so the whole conversation happens at sunset. The shot starts front-lit, follows Emmy around a patch of hot sky [and snow...] before landing with the sun setting behind her head. There was no fill light, only a ball as an eye light. Those that saw the shot know what they need to know, but to those that didn't, the entire sky [save the sun itself] was not clipped, but there was enough information left in Emmy's face to bring it up without introducing unacceptable noise. It's pretty cool.

Just as with the last example, don't take this the wrong way. It doesn't mean you don't need fill light, it doesn't mean you can expose for the sun and let the rest fall into place, etcetera. But for this particular shot, the exposure that we frankly guessed on seemed to do the trick.

When I'm looking at the MX, I don't think of it as a tool that's going to change the way we shoot movies. I'm not approaching it as something that will stop us from lighting or let us see in the dark - that's just not what it's about to me. To me it's about making what we already do just that much better. We're always going to light, we're always going to give our assistants a decent stop, etcetera. Yes, there are situations where you will be able to get a shot with this camera that you couldn't get with the M sensor... just as there are shots that play at a T1.3 instead of T2.8. There are entire productions that the sensor's sensitivity will enable. But for most productions, a stop or two of dynamic range won't reinvent the wheel.

No matter what tool one uses, there is always a solution. I've not kept my thoughts on certain elements of the RED ONE a secret over the years. Things like sensitivity, ergonomics, etcetera. But like everything else, there is a solution. And for the past two and a half years we've all been shooting with the camera and having a blast. Camera doesn't balance on your shoulder? Invent a way to make it balance. Camera looks noisy if you light it like 500T? Find a new way to light it to communicate your intended look through the scope of what the camera can see. Camera takes 90 seconds to power up? Find a way to keep it running all day. Solving problems is what we do.

So with that in mind, briefly try not to look at cameras in terms of "best" or "worst" - something that many people get a little too caught up on. Every system has pros and cons and every system can have solutions built for its problems. Nobody asks DPs whether they're "converting" over to chromes when they shoot a movie on reversal... why should we talk about "converting" to RED after shooting a movie digitally?

Panavision rocks. Arri rocks. Canon rocks. Sony rocks. Grass Valley rocks. Photosonics rocks. Vision Research rocks. Moviecam rocks. Aaton rocks. RED rocks.

So yeah, the MX is better than the M. Way better. It has a lot more dynamic range and a lower noise floor. Is it "better" than the F35? Does it have more DR? I dunno. I'm not one for OMG BEST CAMERA EVER type statements.

The MX has a few more stops of measurable DR than the M. It also has a lower noise floor, meaning that more stops are usable at lower light levels.

But if we were already lighting for the M sensor and fitting our world into the limitations of that sensor's world, doesn't that just steal all its thunder? Sure, it might mean we see a little further into the highlights or shadows... or maybe that we see a little more of that sunset than we used to... but does that really change anything? If you treat the MX exactly the same way as you treat the M, why upgrade? A lot of people have been asking themselves this question. I don't need the extra dynamic range or sensitivity, I'm happy with the M.

So what's the big deal about this sensor? I didn't really know until last night when I got to spend some time with a production MX camera in a situation I'm familiar with. Shooting off of a roof in Chicago at night or shooting unlit into a sunset are special situations that require special attention. But we have all been in hundreds of night exteriors and day interiors and know what to expect from them. Using the MX camera in a night exterior alongside a few M cameras really made the difference clear to me for the first time.

The MX sensor handles color in a way that I've never seen before with a digital camera. It's miles beyond that the M sensor and RedSpace are capable of. Highlights [brake lights, street lights, etc] maintain their color without pushing to white in a way that the M sensor never did. Shadows maintain color accuracy and saturation [and resolution, due to the noise characteristics] in ways that really impressed me. The subtle shifts of color, light and shadow I was seeing in our talent's faces was fabulous. Putting the T3.5 MX footage up against the T2.6 M footage was almost comical. Like any new tool or film stock, you will need to light differently with this sensor, both in terms of value and character.

RED has delivered something tremendously impressive here. If you own a RED ONE, you would be absolutely crazy not to upgrade. I haven't had enough time with the MX to form real, balanced opinions and experiences... but my initial impression is immensely positive.

I'll be posting more direct experience and footage here very soon. Stay tuned.

Brook Willard · 01-30-2010, 11:54 PM

Here are the links to the uncompressed TIFFs. Please only download them once... I only have so much bandwidth. It's hosted on my personal website, so if I run out of bandwidth... I become harder to find for the rest of February!

Tungsten and daylight.

Here's the process:

I lit the charts to F8 at 500 ASA. They were just a touch hotter/specular on the left - less than a tenth of a stop variation in incident. The chart is a Gamma & Density 709 chart.

The lens is an Optimo 28-76mm at 76mm.

I exposed from T2.8 [not WFO on the lens] to T22 on the lens. When I got to T22, I added ND [Format, including IR .9 and 1.2] until I had the 1.2 and .9 stacked in the camera. I then started adding wire to the light until I ran out [2 stops on the Tungsten light, 3 stops on the HMI]. I finally held in the .3 and .6 [Format non IR] in front of the matte box.

The M camera [Evin Grant's 199 - thanks, Evin!] is running the release build 21. The MX camera is running the release build of 30.

Since build 96 of RCX doesn't have 64 ASA as an option, I set the ASA to 80 and the FLUT™ to -.3 for the T2.8/+3 exposures. Everything else I balanced using ASA [125, 250, 500, 1000, 2000, 4000] in full stop increments. After that, I added FLUT™ in full stop increments. When I maxed out FLUT™ [8], I maxed the ASA to 6400 and processed the remaining images as such.

The M footage was processed RedSpace/RedSpace, the MX footage using RedGamma/RedColor. I changed no settings [sharpness, NR, etc.] beyond what I've discussed above.

I do not know if this is the best way to process MX footage. I haven't had time to experiment enough with footage yet to really understand the relationship between ASA and FLUT with the new sensor. I am looking forward to more clarification on the subject as I experiment.

The TIFFs are 100% crop uncompressed of the grey scale on the chart. It is a just-left-of-center cut from the sensor. I will try to do full frame [scaled] charts soon.

Brook Willard · 01-30-2010, 11:54 PM

This is a 70% quality "save for web" JPEG from Photoshop.

Here's a 640x360 ProRes LT file of that clip.

Please download sparingly.

I've purposefully selected a crap shot in a crap situation to show what it can pull out. When you actually take some time with this footage, it's a fabulous tool. I've been shooting between 1,000 and 1,600 ASA for several days with brilliant results that I'd be happy to put on TV.

Steve Freebairn · 01-31-2010, 12:06 AM

Brook is right, for those that haven't gotten a chance to see footage or better yet, play with a M-X camera, you're in for a serious, serious treat.

Jarred Land · 01-31-2010, 12:35 AM

Thanks Brook for doing what you do best on this shoot.... from what we have seen the images look incredible.

Jannard · 01-31-2010, 12:38 AM

Quote:

Originally Posted by Jarred Land

Thanks Brook for doing what you do best on this shoot.... from what we have seen the images look incredible.

Brook... thanks for your observations. Everyone here knows you speak your mind both ways so you carry a lot of credibility. You are not alone in your thoughts.

Jim

JanneJansson · 01-31-2010, 01:10 AM

The chance to upgrade the R1 with the latest sensor tech will make it survive longer as a trusty tool. R1 is a solid camera and a pleasure to work with. I feel safe that it always will do a great job and that my clients will be happy, and in the end of the day bring in some $. My R1 had a near-death-experience the other day, but survived with just broken LCD. The R1 have been very good for me as a small entrepreneur. A big leap of faith for me when I spent a year of salary from a steady job and going freelance with just a R1. When#4744 comes up for MX upgrade I'll be glad

Best

Sanjin Jukic · 01-31-2010, 01:10 AM

Brook,

thanks for a great M-X report.

Waiting for footage...

Brook Willard · 01-31-2010, 01:13 AM

Thanks to Jarred and Jim for helping us out on Shameless - it really made a huge impression on everybody involved.

I'll post some footage/stills tomorrow, I'm a bit exhausted tonight. What sort of stuff do you guys want to see?

Sanjin Jukic · 01-31-2010, 01:19 AM

Quote:

Originally Posted by Brook Willard

What sort of stuff do you guys want to see?

More different shots, available light, talent's skin tones, sunset colors and flares, car lights, skin tones at night car drive, street lights, neon, night sky, moon light, etc,...

01-30-2010, 02:01 PM	#1
Brook Willard Moderator Join Date: Dec 2006 Location: Burbank, CA Posts: 4,453	The best $5,750 you can spend [updated] Hey folks, For the past few weeks, I've been the DIT on a pilot J. Michael Muro was shooting [we wrapped last night]. A few weeks ago, Jarred and Keslow Camera arranged for RED to send us Jarred's own personal MX camera to our location days in Chicago for some key night exteriors. Now over the past few years, I've spent way more than my fair share of days shooting with the RED ONE. I also spend a lot of time [more than might be suggested by how RED-centric I am online] working as a DIT for traditional HD camera systems. Shooting with the new MX sensor in Chicago impressed me massively. I was hoping for a camera with more dynamic range and a lower noise floor and RED has delivered. The night exterior city footage from Jarred's preproduction firmware MX camera looked better at T4 than the RED ONE M sensor looked at T2.2. Now don't take that as me saying it's "X" stops more sensitive or "X" times better, but for this particular setup [overlooking the river], it was terrific. We ended up shooting the shot at T2.2 on the MX anyways... just to expose to the right as much as possible. Those of you that went to RED DAY at RED Studios Hollywood would've seen a few of the shots from our project. I wasn't at RSH myself [was teaching the Data Handling training for local 600 in NY], but I'm told that [among others], a shot of Emmy Rossum on the phone was played. I want to give a little information about that particular shot as well. It was our last shot before switching to night mode, so the whole conversation happens at sunset. The shot starts front-lit, follows Emmy around a patch of hot sky [and snow...] before landing with the sun setting behind her head. There was no fill light, only a ball as an eye light. Those that saw the shot know what they need to know, but to those that didn't, the entire sky [save the sun itself] was not clipped, but there was enough information left in Emmy's face to bring it up without introducing unacceptable noise. It's pretty cool. Just as with the last example, don't take this the wrong way. It doesn't mean you don't need fill light, it doesn't mean you can expose for the sun and let the rest fall into place, etcetera. But for this particular shot, the exposure that we frankly guessed on seemed to do the trick. When I'm looking at the MX, I don't think of it as a tool that's going to change the way we shoot movies. I'm not approaching it as something that will stop us from lighting or let us see in the dark - that's just not what it's about to me. To me it's about making what we already do just that much better. We're always going to light, we're always going to give our assistants a decent stop, etcetera. Yes, there are situations where you will be able to get a shot with this camera that you couldn't get with the M sensor... just as there are shots that play at a T1.3 instead of T2.8. There are entire productions that the sensor's sensitivity will enable. But for most productions, a stop or two of dynamic range won't reinvent the wheel. No matter what tool one uses, there is always a solution. I've not kept my thoughts on certain elements of the RED ONE a secret over the years. Things like sensitivity, ergonomics, etcetera. But like everything else, there is a solution. And for the past two and a half years we've all been shooting with the camera and having a blast. Camera doesn't balance on your shoulder? Invent a way to make it balance. Camera looks noisy if you light it like 500T? Find a new way to light it to communicate your intended look through the scope of what the camera can see. Camera takes 90 seconds to power up? Find a way to keep it running all day. Solving problems is what we do. So with that in mind, briefly try not to look at cameras in terms of "best" or "worst" - something that many people get a little too caught up on. Every system has pros and cons and every system can have solutions built for its problems. Nobody asks DPs whether they're "converting" over to chromes when they shoot a movie on reversal... why should we talk about "converting" to RED after shooting a movie digitally? Panavision rocks. Arri rocks. Canon rocks. Sony rocks. Grass Valley rocks. Photosonics rocks. Vision Research rocks. Moviecam rocks. Aaton rocks. RED rocks. So yeah, the MX is better than the M. Way better. It has a lot more dynamic range and a lower noise floor. Is it "better" than the F35? Does it have more DR? I dunno. I'm not one for OMG BEST CAMERA EVER type statements. The MX has a few more stops of measurable DR than the M. It also has a lower noise floor, meaning that more stops are usable at lower light levels. But if we were already lighting for the M sensor and fitting our world into the limitations of that sensor's world, doesn't that just steal all its thunder? Sure, it might mean we see a little further into the highlights or shadows... or maybe that we see a little more of that sunset than we used to... but does that really change anything? If you treat the MX exactly the same way as you treat the M, why upgrade? A lot of people have been asking themselves this question. I don't need the extra dynamic range or sensitivity, I'm happy with the M. So what's the big deal about this sensor? I didn't really know until last night when I got to spend some time with a production MX camera in a situation I'm familiar with. Shooting off of a roof in Chicago at night or shooting unlit into a sunset are special situations that require special attention. But we have all been in hundreds of night exteriors and day interiors and know what to expect from them. Using the MX camera in a night exterior alongside a few M cameras really made the difference clear to me for the first time. The MX sensor handles color in a way that I've never seen before with a digital camera. It's miles beyond that the M sensor and RedSpace are capable of. Highlights [brake lights, street lights, etc] maintain their color without pushing to white in a way that the M sensor never did. Shadows maintain color accuracy and saturation [and resolution, due to the noise characteristics] in ways that really impressed me. The subtle shifts of color, light and shadow I was seeing in our talent's faces was fabulous. Putting the T3.5 MX footage up against the T2.6 M footage was almost comical. Like any new tool or film stock, you will need to light differently with this sensor, both in terms of value and character. RED has delivered something tremendously impressive here. If you own a RED ONE, you would be absolutely crazy not to upgrade. I haven't had enough time with the MX to form real, balanced opinions and experiences... but my initial impression is immensely positive. I'll be posting more direct experience and footage here very soon. Stay tuned. __________________ Brook Willard IATSE Local 600 DIT Click through for contact information. Click and follow the world's most boring production blog...

01-30-2010, 11:54 PM	#2
Brook Willard Moderator Join Date: Dec 2006 Location: Burbank, CA Posts: 4,453	Here are the links to the uncompressed TIFFs. Please only download them once... I only have so much bandwidth. It's hosted on my personal website, so if I run out of bandwidth... I become harder to find for the rest of February! Tungsten and daylight. Here's the process: I lit the charts to F8 at 500 ASA. They were just a touch hotter/specular on the left - less than a tenth of a stop variation in incident. The chart is a Gamma & Density 709 chart. The lens is an Optimo 28-76mm at 76mm. I exposed from T2.8 [not WFO on the lens] to T22 on the lens. When I got to T22, I added ND [Format, including IR .9 and 1.2] until I had the 1.2 and .9 stacked in the camera. I then started adding wire to the light until I ran out [2 stops on the Tungsten light, 3 stops on the HMI]. I finally held in the .3 and .6 [Format non IR] in front of the matte box. The M camera [Evin Grant's 199 - thanks, Evin!] is running the release build 21. The MX camera is running the release build of 30. Since build 96 of RCX doesn't have 64 ASA as an option, I set the ASA to 80 and the FLUT™ to -.3 for the T2.8/+3 exposures. Everything else I balanced using ASA [125, 250, 500, 1000, 2000, 4000] in full stop increments. After that, I added FLUT™ in full stop increments. When I maxed out FLUT™ [8], I maxed the ASA to 6400 and processed the remaining images as such. The M footage was processed RedSpace/RedSpace, the MX footage using RedGamma/RedColor. I changed no settings [sharpness, NR, etc.] beyond what I've discussed above. I do not know if this is the best way to process MX footage. I haven't had time to experiment enough with footage yet to really understand the relationship between ASA and FLUT with the new sensor. I am looking forward to more clarification on the subject as I experiment. The TIFFs are 100% crop uncompressed of the grey scale on the chart. It is a just-left-of-center cut from the sensor. I will try to do full frame [scaled] charts soon. __________________ Brook Willard IATSE Local 600 DIT Click through for contact information. Click and follow the world's most boring production blog... Last edited by Brook Willard; 01-31-2010 at 09:06 PM.

01-30-2010, 11:54 PM	#3
Brook Willard Moderator Join Date: Dec 2006 Location: Burbank, CA Posts: 4,453	This is a 70% quality "save for web" JPEG from Photoshop. Here's a 640x360 ProRes LT file of that clip. Please download sparingly. I've purposefully selected a crap shot in a crap situation to show what it can pull out. When you actually take some time with this footage, it's a fabulous tool. I've been shooting between 1,000 and 1,600 ASA for several days with brilliant results that I'd be happy to put on TV. __________________ Brook Willard IATSE Local 600 DIT Click through for contact information. Click and follow the world's most boring production blog... Last edited by Brook Willard; 02-09-2010 at 05:12 PM.

01-31-2010, 01:10 AM	#7
JanneJansson Senior Member Join Date: Aug 2008 Location: Sweden Posts: 692	The chance to upgrade the R1 with the latest sensor tech will make it survive longer as a trusty tool. R1 is a solid camera and a pleasure to work with. I feel safe that it always will do a great job and that my clients will be happy, and in the end of the day bring in some $. My R1 had a near-death-experience the other day, but survived with just broken LCD. The R1 have been very good for me as a small entrepreneur. A big leap of faith for me when I spent a year of salary from a steady job and going freelance with just a R1. When#4744 comes up for MX upgrade I'll be glad Best __________________ JJ

01-31-2010, 01:10 AM	#8
Sanjin Jukic Senior Member Join Date: Dec 2006 Posts: 6,735	Brook, thanks for a great M-X report. Waiting for footage... __________________ Sanjin Jukic - RED1 Camera Experimentalist "There is no point in having sharp images when you've fuzzy ideas." Jean-Luc Godard. Digital Cinema Production Vienna-Austria

01-31-2010, 12:06 AM	#4
Steve Freebairn Senior Member Join Date: Jan 2007 Posts: 998	Brook is right, for those that haven't gotten a chance to see footage or better yet, play with a M-X camera, you're in for a serious, serious treat. __________________ IATSE Local 600 DIT www.ipaddailies.com/

01-31-2010, 12:35 AM	#5
Jarred Land Fire Chief Join Date: Dec 2006 Posts: 4,444	Thanks Brook for doing what you do best on this shoot.... from what we have seen the images look incredible.

01-31-2010, 01:13 AM	#9
Brook Willard Moderator Join Date: Dec 2006 Location: Burbank, CA Posts: 4,453	Thanks to Jarred and Jim for helping us out on Shameless - it really made a huge impression on everybody involved. I'll post some footage/stills tomorrow, I'm a bit exhausted tonight. What sort of stuff do you guys want to see? __________________ Brook Willard IATSE Local 600 DIT Click through for contact information. Click and follow the world's most boring production blog...

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode