Working With the RED

[Steven Caesare]

Jim did mention the numbers - can't remember them myself. I'm much more on the image processing codec geekery

Graeme

Jim has stated that the original Mysterium sensor (ie in a non-upgraded Red One) had a sensor scan period of ~9ms.

The new Mysterium-X (ie- upgraded RedOne/Epic/Scarlet) drops thet to ~5ms.

The rotating shutter in a 35mm film camera scrolls accross the film in ~4ms.

Se we should be in the ballpark of film-cameral perfomance when it comes to skew and strobe performance.

-sc

 

[Graeme Nattress]

Steven, thanks for being my memory for me!

Graeme

 

[Steven Caesare]

And in return, thanks for the great technical explanations of the innards of digital video acquistion details in general, and the Red Way in particular.

I know you've referred to yourself as being able to "drone on" about such things, but I'm one in the audience who is thrilled to listen as long as you are willing to explain.

Thanks so much.

-sc

 

[joshua csehak]

I didn't see the article as unnecessarily complaining at all. I mean, the guy's shot 50 projects with it. Clearly, it's his go-to camera. And the frame grabs they printed were gorgeous. That's all I need to know about the positives. Why would I want to know more about what can go right? Let it go right. Tell me what can go wrong, so I know what to look out for. In that respect it was a perfect article. It confirmed some things that I suspected but thought maybe I was just doing things wrong, and warned me about a couple pitfalls I was previously unaware of (like dropped frames from loud environments using the red drive -- never happened to me, but my shoots have never been that loud).

(The tips and tricks thread has started to become a great resource in that respect. I for one would love to see it eventually contain a giant list of potential problems and the recommended solutions to each one.)

I dunno -- "There's a lot wrong with this camera, but I use it all the time in spite of that" sure feels like a positive review to me.

I will say, though, it's pretty silly to point out that the Red is not as good as cameras that cost an order of magnitude more. But when people do, doesn't that by definition imply a damn fine review of the Red, no matter what they say?

 

[Alexander Ibrahim]

I didn't see the article as unnecessarily complaining at all. I mean, the guy's shot 50 projects with it. Clearly, it's his go-to camera. And the frame grabs they printed were gorgeous. That's all I need to know about the positives. Why would I want to know more about what can go right? Let it go right. Tell me what can go wrong, so I know what to look out for. In that respect it was a perfect article. It confirmed some things that I suspected but thought maybe I was just doing things wrong, and warned me about a couple pitfalls I was previously unaware of (like dropped frames from loud environments using the red drive -- never happened to me, but my shoots have never been that loud).

(The tips and tricks thread has started to become a great resource in that respect. I for one would love to see it eventually contain a giant list of potential problems and the recommended solutions to each one.)

I dunno -- "There's a lot wrong with this camera, but I use it all the time in spite of that" sure feels like a positive review to me.

I will say, though, it's pretty silly to point out that the Red is not as good as cameras that cost an order of magnitude more. But when people do, doesn't that by definition imply a damn fine review of the Red, no matter what they say?

That's bad writing and shoddy journalism.

If you mean its a good camera say "its a good camera." Don't imply it somehow through listing its flaws with a one line throwaway about how much you use it.

I don't know Mr. Probst, but he may simply feel forced into it by the budgets on his projects. Also, he could be hugely prolific, with 300 shoots under his belt since RED was available. This may definitely not be his go to camera.

So, the positives have to explicitly stated.

Also, I think its entirely fair to compare RED to cameras costing ten times as much. Its results are in that league- and in many ways it delivers better results.

Of course, nothing is free. So, if you buy a $17,500 RED One instead of a Sony F35 ($135,000... I think?) You are buying F35 level results, with $100,000 of problems you have to solve to get it done. They are solved through additional competence/training and through specialized accessory gear.

So, in summary, I agree on most of the article's points, but I wish there were more positive points.

Heck just mentioning his opinion of the results in the article's closing would have been a huge improvement.

 

[Christopher Probst]

Hey Christopher,

I was pleasantly surprised to see that "Love Game" and "Waking up in Vegas" were shot on the RED, as I always liked the look of them, "Love Game" especially. What other music videos (with big enough artists that it's likely I've already seen them) have you shot that you are particularly proud of? (RED or otherwise)

Heres a partial list of some of music videos that I shot with the Red... I've done about an equal number of commercials... Of course, I still shoot lots of film jobs, as well as other digital cameras as well...

Chris Brown, "Forever"
Janelle Monae, "Many Moons"
Jesse McCartney, "It's Over"
Lady Gaga, "Love Game"
Lady Gaga, "Eh Eh"
Katy Perry, "Waking Up in Vegas"
Eminem, "We Made You"
Kelly Clarkson, "Already Gone"
Boa, "I Did It For Love"
Pussycat Dolls, "Hush Hush"
Black Eyed Peas, "I Gotta Feeling"
Kate Voegele, "99 Times"
Esther Dean, "Drop it Low"
Sick Puppies, "You're Going Down"
Michael Buble, "Just Haven't Met You Yet"
Chris Brown, "I Can Transform Ya"
Timbaland, "Morning After Dark"
Chris Brown, "Crawl"
Black Eyed Peas, "Imma Be"
Black Eyed Peas, "Rock That Body"
Yo Gotti, "Women Lie, Men Lie"

 

[Christopher Probst]

Alexander,

I believe I stated pretty simply that I have shot some of my favorite work on the camera. But that the Red is, like any camera system, NOT PERFECT... I then show examples of my work that I feel looks good. What more do you need?

I'm not gushing over the camera in the article, because I don't feel that way about it. Half the time on set I want to throw it through a wall. Can I make good images with it? Yes. Does it preform flawlessly and is robust? NO. Is it a pain in the ass in certain situations listed? (Like low light, warm scenes... slow motion, rolling shutter...) Yes. Do I wish the sensor was faster and that the camera offered less compression? Yes. Are there work arounds for many of the issuses? Yes. Are there NO workarounds for SOME of the issues, YES. Is there hope for further improvements. Yes. Will I use it again? Yes. Am I excited about the future for the camera system and the Epic? Yes. Can we move on already? I hope so.

This article first and foremost was my opinions. It is told in the first person from my experiences. It is not a formal review piece where I interviewed the manufacturer, cinematographers, ect... I HAVE written those sorts of pieces too, but that was not the format for this story. Should I have MADE UP A FALSE OPINION for me to write about how I feel about the camera?

 

[joshua csehak]

Hey Christopher, thanks for talking with us about your article. That's really cool of you. If you don't mind my asking, some of those videos are for top-notch artists -- I mean, you don't get much bigger than Lady Gaga or the Black-Eyed Peas -- I imagine that budgets aren't much of an issue for them. Would you mind talking about why you decided to use the Red One to shoot their videos (and so many more), as opposed to film or a more expensive digital like an F35?

Of course, nothing is free. So, if you buy a $17,500 RED One instead of a Sony F35 ($135,000... I think?) You are buying F35 level results,

So you're implying right there that the F35 is on a different level than Red. I can't speak to that, since I've never used it (it's way beyond every budget I've worked with, and frankly, I saw The Lovely Bones and couldn't tell what parts were film and what parts were Red*, so I find it hard to see the point in spending any more on a more expensive camera), but I imagine it must offer some serious benefits to someone, otherwise, why would anyone use it or anything like it?

*to be more precise, some parts definitely looked like film (and looked nice), and other parts, I couldn't tell, but there weren't any parts where I thought, "Oh, this is definitely Red footage."

 

[CJ Roy]

Chris-

I haven't yet upgraded my camera, but I'm sure I will soon. I'm going to be playing with one of the upgraded cameras in the next couple of days, and I imagine using them exclusively for a lot of upcoming shoots.

What I've seen so far, I've been impressed with. I'll be looking at some tests either today or tomorrow to figure out how to rate the camera. Rating the camera off the raw isn't necessarily fair, and should yield a low(er) ISO number. So we're going to do both, RAW & the new RedColor, since there is no more Rec709 or RedSpace. Also want to test how much of a difference there is between what we see on the monitor outputs, vs. what we get back in post.

With the old sensor, we've been pretty successful in being able to getting rid of a lot of noise in the dailies by processing the full debayer on set. I know that's not part of your or JK's normal workflow, but it's been really impressive what we've been able to get out of it. It also protects the DP's aesthetic choices if production doesn't decide to conform again from the raw... which is happening more and more. If the new sensor's noise floor is really as low as it seems, then we'll be able to push it that much further as the "usable" dynamic range has increased. As expected, the sensor does react better in daylight than tungsten. It's certainly better, but I don't know how big the difference is yet.

What I've been apprehensive about, in regards to the new sensor, is using the new RedCine-X software which is cumbersome, glitchy and difficult to use as a batch processor. Anything unstable can really f*ck us on set. The latest RocketCine software can now do the MX sensor, but RED is killing that app soon unfortunately. Obviously it also saves an enormous amount of time in getting the footage into an Avid.

I think more and more we're going to have to be doing more color on set, at least for educational purposes, to see what we can get out of it. This doesn't replace a final conform & color at all, but helps us get in the ballpark better for the final. There's no reason that the dailies have to look flat and unattractive.

That said, it's a giant step in the right direction. And the epic's ergonomics look to be much better as well. If the mock up cameras on Red Day were accurate in terms of weight and size, then the camera body will be significantly less cumbersome, although still front heavy. Art Adams and I think Brook Willard had done some basic drawing showing the module pieces being separated for better weight distribution. I really hope RED gives those ideas some attention. It'll save a lot of our backs. Also, the new EVF looks to be a huge improvement as well, which was working on the R1's thus appearing to be backwards compatible.

Let me know when you're prepping/shooting on the MX, may be I can drop by and check it out with you.

-CJ

 

[ericyoung]

Normally avoid saying this, but maybe time to move on everyone? Think we're starting to repeat ourselves...or at least in regards to the pros and cons of the article!

Thanks Christopher for taking the time to reply. And hope you've not been put off Reduser by the passionate defences of the RED!

It's certainly been an interesting discussion! :smile:

 

[Sanjin Jukic]

I believe I stated pretty simply that I have shot some of my favorite work on the camera. But that the Red is, like any camera system, NOT PERFECT... I then show examples of my work that I feel looks good. What more do you need?

Chris,

in your article everything is fine with your camera critical observations except that missing something as a sort of

"glorification" and "excitement" that you just wrote above:"...I have shot some of my favorite work on the camera".

Somehow that could make a certain balance, more sense and would make a positive tone of the whole article at the end.

A good for you and also good for all of us.

 

[CJ Roy]

Jim has stated that the original Mysterium sensor (ie in a non-upgraded Red One) had a sensor scan period of ~9ms.

The new Mysterium-X (ie- upgraded RedOne/Epic/Scarlet) drops thet to ~5ms.

The rotating shutter in a 35mm film camera scrolls accross the film in ~4ms.

Se we should be in the ballpark of film-cameral perfomance when it comes to skew and strobe performance.

-sc

I definitely was able to induce and see it one the monitors at Red Day. Like the R1 sensor, the sensor scan was dramatically improved over the course of several builds. Maybe they'll get even better with firmware updates.

 

[Christopher Probst]

Chris,

in your article everything is fine with your camera critical observations except that missing something as a sort of

"glorification" and "excitement" that you just wrote above:"...I have shot some of my favorite work on the camera".

Not to sound too harsh here, but I stated that exact sentence in the article in the 2nd paragraph...

This sort of nit-picking, without fact checking is what internet forums are rife with. Believe me, I researched and backed up everything I talk about in my article, and yet, I'm constantly getting misquoted in this forum attributing statements or lack thereof to me. People are generalizing, summarizing, or theorizing what I have written, but not going back to confirm what exactly WAS written.

And I believe a few posts above should clear any doubts about my stance with the R1...

 

[CJ Roy]

Chris,

in your article everything is fine with your camera critical observations except that missing something as a sort of

"glorification" and "excitement" that you just wrote above:"...I have shot some of my favorite work on the camera".

Somehow that could make a certain balance, more sense and would make a positive tone of the whole article at the end.

A good for you and also good for all of us.

I don't get this. The article is one professional's perspective that was developed of the course of more projects than the majority out there have served on. Why are we debating the tone of an article and not its technical merits? It sounds like the criticism of the article is to protect our investments as opposed to keeping our objectivity and move forward with a better camera.

As with any product, especially those that are being depended on by projects that cost an enormous amount of money, a fluff piece isn't going to help anyone succeed in using the camera.

Testing cameras is a routine practice in our profession and we test so that we can find out its limitations. Finding out where a system breaks, where the stress points fall and what its limitations are tell us what conditions are ideal to shoot with. Simply put, we have to know what we're dealing with. With shoots and last minute decisions made by everyone, we need to know what the camera is capable of.

Tone of an article is irrelevant, IMO.

-CJ

 

[Michael Hastings]

...

Yes, the fans are on the bottom... again, not a problem. The vast majority of computers are cooled by fans at the front and back of the case, not the top. The RED ONE has too many electronics jammed in there to make a sealed cooling system, so do you want vent holes in the top of the camera that things can fall into? And as for those bottom vents, check the airflow with the fans on high, and then notice that you can't see straight through from one side to another... because the airflow path is moving air up, over, and around all the electronics.

Nitpicking here, but I believe Epic has vent slots in the top.

I don't think there is an engineering problem with R1 cooling per se, because it does seem the flow is properly designed, it is just a limitation of an FPGA based camera. Just as the electronics package in the R1 limits the frame rates even with the new sensor, it also means you are stuck with the heat issues. And it seems to me that is the reason for the upgrade deal - we simply needed a new ASIC based electronics package to get the better mix of processing speed, lower power consumption, and lower heat generation, and it also enables even better ergonomics and the other improvements that would make "obsolescence obsolete".

I'm not gushing over the camera in the article, because I don't feel that way about it. Half the time on set I want to throw it through a wall. Can I make good images with it? Yes. Does it preform flawlessly and is robust? NO. Is it a pain in the ass in certain situations listed? (Like low light, warm scenes... slow motion, rolling shutter...) Yes. Do I wish the sensor was faster and that the camera offered less compression? Yes. Are there work arounds for many of the issuses? Yes. Are there NO workarounds for SOME of the issues, YES. Is there hope for further improvements. Yes. Will I use it again? Yes. Am I excited about the future for the camera system and the Epic? Yes. Can we move on already? I hope so.

This article first and foremost was my opinions. It is told in the first person from my experiences. It is not a formal review piece where I interviewed the manufacturer, cinematographers, ect... I HAVE written those sorts of pieces too, but that was not the format for this story. Should I have MADE UP A FALSE OPINION for me to write about how I feel about the camera?

This seems to be a pretty complete and fair summation.

I think that many of us Redusers need to remember that there are productions where things are much more critical because they may be literally spending a $1000 per minute. Low budget Indie, documentary, nature, etc. producers don't typically have that kind of pressure so some of the issues and workarounds aren't as critical. Also Remember that AC is the ASC journal so sort of by definition its primary audience (ASC members) will tend to be working on these higher pressure productions and the "gotchas" really are the critical information for them. And the article didn't need to be an "intro" because there probably aren't any working ASC members that aren't aware that the RED exists.

 

[Marc Wielage]

Of course, even with such filtering, we still have more measured resolution than the other digital cinema cameras available today that I mention above. Indeed one of the whole reasons of going to 4k and beyond is to allow you to use proper filtering to ensure the measured resolution you do get is utterly clean and un-corrupted by nasty aliasing.

I think this is open to interpretation. John Galt's "The Future of Pixels" article from a year or two ago (which I think was widely debated here) made some good pros and cons. No question, he's got his own agenda, being part of Panavision, but I think his propaganda is no less valid than opinions on Red. I think you can juggle and interpret numbers and come up with different results for every camera.

Speaking from a very unbiased post standpoint: I can tell you I did the final color correction on a major $3,000,000 network TV pilot for Disney back in March of last year. They did extensive tests between the Red and the Arri D21. At HD resolution, on the 40" Sony BVM-32E's that are standard throughout LA for TV color-correction, the Arri looked better, hands-down. (A lot of this was due to DP Theo van de Sande, who does great work.) The Arri looked consistently sharper and had less noise than similar Red shots, and in a few cases, we were forced to cut back and forth between the two cameras in certain scenes, and I had to work very hard to match them -- in some cases, slightly degrading the Arri footage.

Again, that was a year ago, and this is now. But I don't think the AC article is completely disingenuous. I think a lot of what the guy says are facts, but I think the article's approach takes a little bit of a negative slant. But I see no outright misconceptions or misstatements there.

I have every hope that the Epic will make some drastic changes for the better, particularly to improve exposure range, improve S/N, and reduce the rolling-shutter issues. As far as I'm concerned, this will benefit everyone, particularly filmmakers and post people. Anytime pictures get better on the set, life becomes a little easier for colorists.

 

[Roger Crouse]

Not to sound too harsh here, but I stated that exact sentence in the article in the 2nd paragraph...

This sort of nit-picking, without fact checking is what internet forums are rife with. Believe me, I researched and backed up everything I talk about in my article, and yet, I'm constantly getting misquoted in this forum attributing statements or lack thereof to me. People are generalizing, summarizing, or theorizing what I have written, but not going back to confirm what exactly WAS written.

And I believe a few posts above should clear any doubts about my stance with the R1...

I respect the fact that you've logged in here and had a dialogue with some of us Red users about the article you wrote. I hope you stick around and continue to contribute and challenge. And I hope you keep pushing Red's products to help move them forward.

 

[Graeme Nattress]

Marc, when I tested Arri footage, I found the in-camera downsample less than ideal. Incidentally I saw some F35 tests the other day and wasn't impressed with their aliasing response either.

Aliasing can make for a sharper image, but a less accurate one, and one which can suffer other issues through a distribution pipe. Only way RED 4k footage can look un-sharp is if it's not sharpened (because every HD camera is, by default) or in-correctly downsampled with too soft a downsampling filter. Or there's not enough noise in the image as grain or noise can create the appearance of sharpness or detail, see: http://www.cambridgeincolour.com/tutorials/sharpness.htm

Graeme

 

[Stephen Strangways]

Nitpicking here, but I believe Epic has vent slots in the top.

Yes it does, but the vent system on that is completely sealed and self contained, whereas the RED ONE is not. If something falls into your EPIC vents, the only thing it can hit or affect is the fan.

A sealed design is definitely better, but they might not have been able to do that based on the sheer volume of circuit boards jammed inside the ONE.

 

[Steven Caesare]

I definitely was able to induce and see it one the monitors at Red Day. Like the R1 sensor, the sensor scan was dramatically improved over the course of several builds. Maybe they'll get even better with firmware updates.

If you don't mind my asking: how did you induce it?

Somebody pointed out that a film gate is necessarily in front of the film plane, hence the edge is out of focus and this softer. Whereas the Red is scanning the actual pictures, hence a hard horizontal line.

That having been said, it wouldn't make a difference for strobe split screens, and the text MX footage that the Off Hollywood boys shot had some good test material (bridge supports & lamp posts out a car window) that looked to me about he same as I'd expect with film.

So, while it might be able to be artificially induced under specific torture tests, it doesn't look like a huge problem in actuality thus far.. hence my interest in your test.

-sc