ProVideo Coalition.com: FRESHDV by Matthew Jeppsen & Kendal Miller

While Red was making new product announcements in November 2008, I was busy AC’ing a 10-day Red One shoot. Between takes the crew discussed the new camera and sensor options announced by Jim Jannard and Co. The ability to mix and match sensor sizes for maximum flexibility whets my appetite as a filmmaker. However, for myself and others I’ve spoken with, that excitement is tempered with a bit of cold reality. The general concensus is, “Finish the camera you promised at NAB 2007 before you deliver another line of products.” Leading up the camera department on this project and others has really driven that point home to me. Donning my flame-retardant gear, I’m going to take a few minutes and get some of these issues off my chest. Read on…

I’ll start with the positive aspects of a Red production; the sensor captures beautiful filmic video, and the camera is a relatively well-designed system that makes great images day in and day out. In general, the system is solid and well-designed. However, the problems creep into play with the details. Unfortunately, these small wrinkles tend to accumulate, and they’ll drive you mad if you are the one responsible for the camera’s daily operation.

On our cold November mornings in Chicago, most days the camera refused to boot without being reset 2-3 times. I recall one particular morning that required us to hold vehicle shots while trying to start the camera. It was the first shot of the day, and we had built up the camera and rigged the vehicle we were shooting from before booting the Red. Meanwhile, crew is standing by and the client is waiting. It was a tense moment for me and my 2nd as we reset the Red multiple times, waiting for that one magic moment when it would boot without freezing. Multiply that anxiety by the necessary 6-10 battery changes in a full shooting day (all requiring shutting down the camera), and you begin to understand why some users are frustrated by the bugginess. I know that I am. Prior to the start of principal photography, the Red was updated to Build 17, so at that time we were at the latest stable version of firmware. As I write this in mid-March 2009, the Red One stable firmware is still at Build 17. Build 18 is in Beta, and I assume that fixes are coming. And inevitably, new bugs. It is software, after all.

For us in November, Build 17 introduced another wrinkle… 17 had a bug that makes the ability to toggle in and out of View Raw impossible after a random period of time. This was problematic for us, as the DoP had designed a preview look that closely approximated what the final film grade was to be. Toggling View Raw enabled us to quickly see what the look would be, but still monitor the Raw data being recorded at the touch of a button. This greatly simplified exposure decisions on-set. Unfortunately, Build 17 would “forget” the View Raw toggle by noon, and we’d be guessing for the rest of the day. The only workaround we were aware of at the time was to do a complete Build 17 re-upgrade on the camera. This “fix” lasted several hours before the option vanished again. Since then, I learned that there is a simpler workaround that involves re-applying the View Raw option in the menus. We had no way of knowing that at the time. However, even that is unacceptable for a camera that shipped 18 months ago. These things should work out of the box on a camera that is supposed to be production-ready. Furthermore, that brings up another side-concern…there is no one location that contains a list of Known Issues for the camera or the various firmware builds. With people being brought on to crew a Red shoot that have never worked with a Red, it’s a herculean task to get up to speed on what works and doesn’t work with the camera, and what limitations and gotchas exist. Sure there is information scattered all over the Reduser forums, but keeping on top of it all is nearly impossible.

Another bug I discovered on a recent project relates to reference audio. If you are shooting 100% sync sound and not capturing reference audio on the Red, you’ve probably never seen this error. On a recent shoot with a lot of interviews, we needed to capture reference audio to the Red. It worked fine, except that on playback most clips would fail to play, freezing with an audio error on screen. The workaround was to exit playback, and try to play the clip again. Sometimes we would do that as many as 8-10 times before the clip would finally play without error. We learned very quickly to disable audio capture when it wasn’t entirely necessary, so we could actually play back the footage for the director. This camera had not yet been sent in for the audio board upgrade, so I don’t know if that was related. But I do know that it was a very irritating, very time-consuming issue that sapped my confidence and slowed down our already compressed shooting schedule. These sort of simple details need to work every day, all day. Oh, and while we are on the subject of playback…can we have playback thumbnails already? Hitting Next-Next-Next through 50 clips so the director can look at one detail from a shot early in the day is time-consuming and feels 100% unecessary on a camera that already has a joystick thumb control.

Working with the Red on a daily basis also revealed some other fit and finish issues. At first glance, the Lemo connectors for drives, EVF, and LCD look like a great idea. They are compact and appear solid. Unfortunately, they are clunky and unreliable in actual use. The slide lock design is difficult to disengage, and our EVF plug in particular had been tweaked on a previous shoot and was notoriously unreliable. The display would turn red from time to time, requiring a repatch of the EVF cable to fix it. The LCD plug was loose on the back as well, it would wiggle about 1/4-inch when you touched it. I noted this on the November shoot. By the time the March shoot rolled around, the LCD’s plug was dying a slow death, sometimes turning green or red in the middle of a take while I’m using it to pull focus, and other times just blinking to black. The EVF wasn’t doing well in March either, sometimes it showed around 15 vertical “jailbars” across the frame, and no amount of wiggling or jiggling cords would change that. Very distracting when monitoring or rolling, though ironically somewhat helpful for checking verticals and leveling the camera. Connectors that are hardier and can handle the constant stresses of production should be standard. I shouldn’t have to buy a third-party breakout box from Element Technica just to get an HD-SDI plug that doesn’t feel like it’s going to fall apart the first time I tweak it. Because it WILL get tweaked. It’s not a matter of if, it’s a matter of when.

Other issues we ran into, and some of these are perhaps “normal” wear-and-tear related; when configuring the camera for a handheld shot, the Red Handle locking screw broke off while tightening the rig down. Not exactly confidence-inspiring when you are about to shoulder a 25lb rig. Thankfully the rental package included all the necessary extra screws and replacement items that made the handle repair quick and easy. The 18-50mm Red Zoom also decided to stop zooming during the November shoot. Wouldn’t push past 25mm. Thankfully we had a set of Superspeeds that basically covered that zoom range. On the March shoot the batteries stopped reporting percentages to the camera, the display would just read “DC” instead of a number showing what was left in the batt. We were able to get a good read using the hard button and 5-light LED display on the battery itself, but sometimes you’d forget to check (since it’s not in front of you on the display), and then realize right before a long take that you were at 1 or 2 bars and risking losing the camera mid-take. I don’t know if this was related to our battery plate, or what, but I do know that it was a major irritation. That sort of thing should work on a production-ready rig.

Another issue we ran into in March related to image shear. I don’t really know if that’s the right term, but that’s how I describe it. From time to time, the image would shear in half horizontally on fast pans. Like the top and bottom halves of the sensor were split apart and not in sync. This was easy to catch on motion, but sometimes surprised us on very subtle moving shots like interview setups. The fix? Reboot the camera. Again, not very confidence inspiring. Which leads me to this; it is amazing the number of times you have to reboot the camera on-set. Every time you have to make a battery change, you have to reboot. This is about once every 1.5 hours, and inevitably occurs right about the time that the director is asking to look at the frame. A boot takes about 90 seconds, assuming it boots the first time. This is a real time-waster, and I know I’m not the only AC complaining about it. In fact, I read just the other day that rental house Fletcher Chicago has implemented a dual-battery Anton Bauer Battery Mount that enables you to swap batts on the fly, no reboot required. Very useful, and something that every Red camera owner should have. They’ve also reconfigured their Red packages to be completely tool-less and supposedly better-balanced than the standard Red rails config. Bravo!

My bottom line is this; there is no doubt in my mind that the Red One is an incredible camera that has changed the game 100%. There is no question that it has started a revolution. The image quality is stellar, and it offers so many options to creatives. All the “big” things are there. Where it falls all over itself is in the little things, the tiny details that nag you daily and require countless workarounds. Is it worth it? Probably, particularly if you are on a tight budget. Should we have to put up with these constant workarounds and gotchas? No. This camera is supposed to be production-ready. I realize that perhaps some of these issues are probably just our bad luck with this particular camera, but there are too many things broken to just dismiss as such. Audio and playback should work out of the box. The camera should boot the first time, every time. Connectors should be hardened and able to withstand normal production use. New “stable” firmware builds should not break basic functionality, and if they do, stable build fixes should be out within weeks, not months. It’s not asking too much for the camera to work every day, as designed. Particularly when it has been marketed as a working production solution.

When you are on-set on a tight time and money budget, you need every advantage you can get to complete the shooting day on schedule. Fighting the camera all day long is not conducive to that end goal. A production camera is a system, and all it’s parts should be working together without fail on a daily basis. A production-ready camera must be so much more than just a great sensor.

I welcome your comments, workaround suggestions, and flames. Login and comment below.

The Best of Stunning Good Looks

Art Adams | 08/30

A directory of my best articles, sorted by topic.

This entry is a guide to my best articles, sorted by topic. Enjoy!

Revisiting the RED workflow, Smoke 2012 style

Marc-Andre Ferguson | 02/03

My love affair with RED Digital Cinema began in 2007, when my brief stint as demo artist in the NAB RED booth turned into a regular gig at events and trade shows.…

4K > 1080P?

Matt Jeppsen | 01/30

Why you probably don’t need a 4K TV in your living room

There’s a nice, ranty article over at CNET entitled “Why…

You must be registered to comment. This is an effort to reduce spam. Please REGISTER HERE.

I get the feeling that despite 10 years of major features produced with digital stock, we’re still very much at the dawn of the digital cinema era.

I would compare digital cinema today to the point in digital photography where we had the first dSLRs. The quality is there, we have the means, but the problems are too abundant to make that 100% commitment.

Would you agree?

Posted by Michael Critz on 03/16 at 11:07 AM

I guess it depends on what features you require on a specific shoot, and what you are willing to put up with as far as workarounds. Some people may never run into the problems I’ve been faced with, and I sincerely hope that’s the case for most! But dealing with these issues has given me a little bit more perspective when I hear of those that are unwilling to dive into Digital. I get that. If they have something that works, why risk dealing with these potential issues?

-MJ

Posted by Matt Jeppsen on 03/16 at 11:55 AM

First of all, I think EPIC and Scarlet will be what Red promised in NAB07. That’s the way RED can meet their goals.

The issue is, Red One is an FPGA (Field Programmable Gate array) and EPIC and Scarlet will be ASIC (Application Specific Integrated Circuit) systems. In practice, in electronics design one first makes an FPGA and once tested properly you ‘freeze’ the design and make an ASIC system. FPGA’s, which can be programmed, are more flexible but also more vulnerable to temporary malfunctions. Rushing directly to an ASIC is an easy way to make expensive mistakes. In this light Jim Jannard proceeds in a very consistent way and the upgrade possibility to EPIC-X is a good offer.

Second, it’s not a good idea to start to shoot with Red One without having experience with the camera. Perhaps should add, without following regularly the RedUserNet.

This situation is comparable to what happened in the past when secretaries were given computers and word processing systems. Many of the problems reported could have been easily avoided if the group had elementary knowledge of Red One and its system.

Still, it is true that operation of the camera is more challenging that operating the previous generation cameras. The rewarding part of the story is, if one takes the time to learn how to use the camera, then with the money invested in the system one gets an amazing image quality. In the next wave of Epic and Scarlet operation is bound to become easier and less challenging.

From this viewangle, jumping to the second generation is the right move from RED. At least in my view.

Posted by .(JavaScript must be enabled to view this email address) on 03/17 at 04:18 AM

Interesting points on the electronics design of the Red One vs Scarlet and Epic. I’d still argue that at NAB07 Red promised that they would deliver a working digital cinema camera in Red One, and they have yet to deliver on that promise. Also, they talked a lot about rendering obsolescence obsolete, and it looks very clearly like the Red One will be swept under the carpet when Epic arrives.

At any rate, I’m not tracking broken promises, all I really care about is a camera that does everything it is supposed to. Perhaps it wasn’t aimed at me, but suggesting that the operation of Red One is more challenging is not the answer to the issues I experienced on various Red shoots. I was not “challenged” by the Red One, I experienced numerous hardware and software bugs, failures, and breakage. My issues were not a lack of preparation or education, they were due 100% to a camera that does not deliver what it is supposed to deliver.

That was my point in this article…that Red owes users the camera they promised at NAB 2007, and users should expect and demand that. Not wait for Epic or Scarlet for a fix, Red One should be working now.

-MJ

Posted by Matt Jeppsen on 03/17 at 06:32 AM

My view is that I think that Red were desperate to be the first to market with such a camera at all costs. Admittedly Jannard was open that the first users would in effect be beta testers, but I think that after all this time it has become a bit of a joke, especially in light of the announcements of other cameras.

Red One was never the camera that was originally promised. even if it had perfectly operational firmware from the beginning. The original concept was of a camera that could be easily used by one operator, without the need for the camera to look like “The Hydra” as Jim Jannard put it. It was supposed to be relatively light and power friendly.

It is none of those things. That said I know that a lot of action filmmakers love it. Some of the guys I know in the Hong Kong and Chinese film industries are beside themselves with the ability to perfect the precise undercranking effect that they require without wasting film stock or the need to wait until development to see the effect that they ended up with.

Though I feel that the bugs really should have been worked out by now. I think that with the new camera developments that they have going they are probably spreading themselves far too thinly.

Posted by Simon Wyndham on 03/17 at 01:45 PM

I totally agree with the points you’ve made here. It doesn’t matter that this camera costs less than any other on the market that is supposed to do even half of what it does, either way you justify it, it is a lot of money and when you pay that kind of money or even if you rent, you expect to get something that does what it is supposed to and last. If you were shooting a decent budgeted feature with an HV20/30 you’d still budget in to have 10 or more of them depending on the style of shooting you were doing if for no other reason than they’re cheap and if one breaks, which would happen on a real shoot, then you have an immediate replacement. One of the reasons for spending that much more money is to remove the worry of things breaking at all. “You Get What You Pay For” is an expression that is used so often for a good reason, because it’s true. The more I work with different gear, cameras, computers, software, doesn’t matter, the more I see the benefits of paying more for certain things and $17000 for the body alone which without everything else is just a big black paper weight with a lens mount is a huge sum of money when it isn’t tried and tested to the point where these annoyances don’t happen. There are those who are fan boys to the extent where they will forgive all of these faults by saying that RED is a different kind of camera, a different kind of company with a different outlook and philosophy behind it and while all are true it doesn’t change the fact of what you as a consumer have the right to expect for the money you pay for a product. While the camera and film industry was (and still is) in need of a big kick up the back side when it comes to the technology, 4K and the infinity beyond that RED is now aiming to deliver is too much when there isn’t the workflow to deal with it in existence yet without having to compromise or have the steep learning curve that has come with this camera. If you look at the other cameras out there that are 2K and under in resolution that output DPX files you find that even they come with some hard compromises when it comes to shooting it to even get that stream out the back door of it so 4K isn’t achieved with the technology in existence without some compromises but the learning curve of a company trying to achieve this resolution ceiling is not something that should be passed on to the customer on a permanent basis. The majority of the people who have bought a RED ONE are fans of the philosophy of RED and have accepted that by buying one they are joining a “revolution” that is a constantly changing beast and that this comes with little faults that they trust will be fixed in the next firmware or the next one because they feel a part of something and that’s cool. I don’t own one but I love the vibe of the company and the community of users but at the end of the day, when it comes to a paying job, to being professional, to being on a shoot with other people in the same craft in all the departments, working in sync towards getting the shot done, that piece of the whole jigsaw which everyone wants to see in its completed form, to have things delayed and cost you more over something like having to reset the operating system of the camera several times is stupid and it’s annoying for the camera department because now everyone is looking at them like it’s their fault, that they’re letting the side down and that’s wrong.
My apologies for the length of this but like yourself, this has been building up for a while and you’ve provided the perfect chiming in point. One more thing though.
Reading this reminded me yet again of what I’ve always dreamed of in a digital cinema camera and recently found out had been delivered by Arri in the D-21, a real shutter so you can achieve all the wonderful in camera effects that you can with a film camera without the compromise you get from doing it with digital cameras, not to mention a real optical viewfinder! It’s such a simple thing and one that we all take for granted, especially in our stills cameras, without ever thinking about how useful it actually is to have it. Once you introduce a physical shutter into the camera you have what should be in all digital ones which is that the shutter speed is locked and slaved to the frame rate you’re shooting in and any variance/effect you want to achieve beyond is done through the angle of the shutter. With an optical viewfinder you don’t need high res view screens of any variety to judge fine focus, you look through the lens as it is supposed to be. This is where RED have failed from day one in my book. The camera is amazing and I have nothing but the utmost admiration and respect for Jim and what he’s started and carrying out but to omit this simple thing from all of their designs is beyond silly in my opinion. The whole selling point of the design of these new ones they have coming (EPIC/SCARLET) is that you have two cameras in one, a stills camera and a cinema camera but if they expect people, pros since no one else can afford them, to use a camera for stills work that has no optical viewfinder then they’re kidding themselves. No one wants to stare into a tiny tv screen for long periods and I can’t imagine it’s very good for the eyes either. I hated doing it with my Canon XLH1 years ago (probably got the name wrong but was a while ago) and have avoided buying anything since that has the same. If it means a slightly bigger camera then so be it, they need to be bigger so you can shoot handheld with them anyway, I would rather have a physical shutter and an optical viewfinder over a tiny form factor any day.
Anyway, there’s my two cents and the rest of the change and lint I had in the bottom of my pockets too.
Mike

Posted by .(JavaScript must be enabled to view this email address) on 03/17 at 05:35 PM

It’s interesting to read what you say, for I’ve had a Red One for a year now and although I’m making documents mostly alone it haven’t let me down in any occasion. It has been realiable in warm and in very cold.

Yes, with build 17 I ran on the field into this issue that could not toggle between Raw view and Redspace, but I figured out what was going on pretty soon. The key point is that I follow the RedUserNet regularly on a daily basis.

I also had at some point the XL H1 and without any hesitation can say the Canon EVF is not comparable to Red EVF or LCD! Neither I liked the Canon EVF, but the other way around prefer the Red LCD rather than an optical VF. Simply because there has been situation that I’ve had a camera with an optical VF side by side with Red One, and whenever I’ve looked at the optical VF thought that the Red LCD gives me much better idea of what I’m going to get.

Feel this ‘fan boy’ discussion is like under estimation of reasonable people in my ears. For me Red One is a tool just like a wrench. But it’s the biggest leap I’ve ever managed to take imagewise and believe there are many others at RedUserNet who feel the same way. Red service is the very first class and beyond what one could reasonably expect from other manufactures. All this is to say, who cares of philosophy. It’s the image quality and customer help from Red which makes the day.

Posted by .(JavaScript must be enabled to view this email address) on 03/17 at 11:42 PM

BTW, a quick note: I don’t think you’re going to find any more rugged general purpose connectors than LEMO or Fischer connectors . . . if they are not working correctly, that typically points to very hard wear and abuse that would make any other connector unusable in a fraction of the time, especially connections that rely on tight tolerances for shielding, EMI, and electrical-contact robustness. Their unique mechanical design also provides a very high-level of strain-relief, both for the solder joints on the circuit board and cable side of the connector.

BNC’s are only good for high-speed serial connections such as HD-SDI . . . they won’t work for multi-pin-parallel signals often used for analog or digital viewfinders, SATA drives, etc., which is a necessary reality for many electrical devices currently on the market. If you wanted to use a coax-signal for interfaces like that, you would end up needing a BNC for every pin on the connector . . . creating a much bigger mess of cables.

Posted by .(JavaScript must be enabled to view this email address) on 03/19 at 02:22 AM

Name:

Email:

Location:

URL:

Smileys

Remember my personal information

Notify me of follow-up comments?

Submit the word you see below:

On Set with Red: A Weary AC’s Rant