Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

How is Alexa made.

I couldn't figure out how to drill down the menus in the simulator until I noticed the rotatable knob. :P
 
Great article. You can bet the inner sanctum at Red looks very much like the facility that Arri is assembling Alexa in. So similar in fact, that this is a great time to discuss the difference between strategic and tactical thinking.

Clearly Alexa is not lacking in technology or skilled and talented engineers, but I think Arri missed the boat on making Alexa a truly transformative product by opting to spend their R&D Euros tactically, not strategically. For whatever reason, Arri decision makers opted to build an essentially 2K camera with a lossy codec. Alexa is a great camera, but if you were to plot a course from the last innovation in digital cinema I doubt you could argue that 2K would be the plateau upon which to rest our collective laurels for the next 10 years. Alexa, for all her advantages, simply doesn't innovate far enough to become strategically relevant. In short, IMHO, Arri gave too much weight to how entrenched the current Post and Distribution framework is and that limited their thinking and by extension the amount of innovation they would allow... and inadvertently opened up a strategic opportunity for a competitor to push forward in resolution and image quality.
 
It was already discussed here that they choosed easy route with smaller resolution with less data to compute, less heat etc. They know its not future proof but they needed somethig now to catch the digital wave which they lost with their bulky machines. So they talk about it like camera mainly for TV features etc.
 
OK... so I'm meh about the Alexa. I think its a great camera to put into most existing production workflows... but I really wish it was able to deliver 4k finished results. There are a lot of people excited about this camera, because it gets a lot right from the camera crew perspective as well as from an engineering perspective.

I'm very interested in owning an Alexa as a compulsive camera geek... but the poor person I am has to choose and I'll choose the one I want to use most... Epic.

But I absolutely love this peek into how the camera is made and the thinking and engineering behind it.

Its pretty amazing. Its also pretty courageous of Arri. I know every other camera manufacturer out there is poring over this document with attention to detail normally reserved for military satellite photos and Internet Star Trek vs. Star Wars arguments.
 
yah alexa is EASYYYY to use..

it would be nice if epic had some similar funtions.. such as holding in the buttons for 3 seconds before certain things happen.. such as the power button..

Epic has had the 3 second hold down for power down in its spec since the Epic spec was written :)
 
Its pretty amazing. Its also pretty courageous of Arri. I know every other camera manufacturer out there is poring over this document with attention to detail normally reserved for military satellite photos and Internet Star Trek vs. Star Wars arguments.

The inside of most camera manufacturing facilities look exactly like it. While you're right about the photo-intel curiosity, I'm not sure its courageous. I think this article and the photos are meant to bolster confidence in Arri's ability to deliver some high-tech wonder and play in the ballpark of Moore's law.

But for those of us who have longed for a peek inside Red, the Arri article is probably the closest we'll get to that. Oh, the irony...

You can win every battle and still lose the war.
 
I am on the third week of a feature with Alexa right now and it is a nice camera. Its an Arri, what do you expect.

With the experience I now have with the camera, I can't think of a single reason I would choose it over Red. Maybe for commercials or episodic TV with am FCP post workflow, but even then it is the same or more work than Red footage.

If you shoot SxS cards you will want to record Log-C. This is good, but looks terrible until you apply a LUT. Nick Shaw has made a great Arri Log-C FCP plug-in, but it requires a render for playback. I guess you could get apply the LUT to the editors monitors and just cut the Log-C. The LUT must be baked into the Prores file and I much refer the meta data approach of Red.

Currently there is no sound, so everything must be synced before you get dailies or the editors can start working or anything. This will be fixed in future software and sound needs to be synced eventually anyway, so not a big deal.

Its only a 1080p camera. People don't seem to understand what that means. I was at a meeting with the Producer telling someone about how awesome the Alexa is and going through the process. When they got to the part about doing the film out and having to scale the Alexa footage up to 2k for that I thought they were going to get it, but nope! I finally got the sound guy to understand buy explaining it as him recording 16bit sound when he know he will be mastering at 24bit. Then he got it and said that is stupid, why would anyone do that.

I am meeting a lot of producers and camera people who know nothing about the Alexa and think it is the greatest camera ever made. I can't knock it, it does what it was designed to do very well. The Red workflow just makes more sense to me and having higher resolution footage seems like common sense. The transcoding time is about the same for both with the Rocket, maybe faster for Red, and Alexa needs to be transcoded to get a LUT on the footage.


Dusty
Local 600 DIT
 
There are some snags to that approach, as the simulator does things right when the Alexa not neccesarily will... But I like the idea a lot.

Who else has even approached it that way? It is remarkably like the camera menu and it at least aids in familiarization instead of having to be around the camera and "play" with it.

David
 
With the experience I now have with the camera, I can't think of a single reason I would choose it over Red.

I can't help but notice that nowhere in your post do you mention the images themselves, which are probably the primary motivator for a cameraman, especially on a television series. All of the tech might be interesting, especially here, but for all of the technical justifications or lack of them that you mention, in the end all of those things are quite secondary, because it's clear that either product can suit the technical and practical requirements. It's the things you didn't mention that are the real reasons you seek. I'm not making any judgment calls on that - that's for individual cameramen to decide for themselves on each project they do - but this is really no different than, say, choosing Fuji over Kodak. Or Cooke over Zeiss. Cinematography is not a business of absolutes, nor is it always logical. It's a series of decisions based on personal taste and artistic merit. All of the other stuff, assuming it all works, is secondary.
 
Arri digital cameras have always made pretty images. Alexa does too.

The raw image on the Alexa is an ugly Log-C image until a colorist makes it look good. Raw image on the Red is ugly until the colorist makes it look good. I think Red has more room for the colorist to work, but you can do a lot with Alexa's Prores 444 Log-C.

Since so many TV shows are using or considering 5D, I don't know how important the base look of the camera is. A good colorist can make almost anything look good.


Dusty
 
The raw image on the Alexa is an ugly Log-C image until a colorist makes it look good. Raw image on the Red is ugly until the colorist makes it look good. I think Red has more room for the colorist to work, but you can do a lot with Alexa's Prores 444 Log-C

As a colorist, I have to say that's completely untrue. All images have built in characteristics that are determined by the shooting conditions and the camera's own image signature. Log-C images are not meant to be judged directly, just like Cineon film scans are not meant to be looked at directly. They are intended to be put through a standard color matrix and conversion LUT in order to yield an image that is representative of the photography. The Alexa has a conversion LUT built in to the camera that can be baked into the video output or not, and Arri also supplies the LUT for use in post to yield the same starting point, but with the advantage of being able to apply correction to the LogC image prior to the LUT. This is very similar to what Red does with Redcolor (basically a color and saturation matrix) and Redgamma (a gamma curve that is basically doing a linear light to Rec709 gamma correction), which when used together yield an image that is very representative of the original photography. The original information is there to be manipulated if desired, but to say that the images from both of these cameras are "ugly" is to not understand exactly what is being presented to you at a particular stage of the process, and what its true purpose is.
 
This is one reason I sort of hate those demos on DVD's where the colorist shows you how he "fixed" or "improved" the color from the original film scans... it's misleading since a Log scan of a negative was never meant to be viewed directly, it's not indicative of the artistic intent of the photography nor as to what the film stock looks like when printed and viewed "normally". So in these demos, you see this washed-out and pastel image being made to "look good" -- simply by applying the gamma corrections it was always designed to use! The implication in these demos is that for some reason, the DP turned in this lackluster image that now needs to be "saved" by the colorist.

It's just like with RAWVIEW in the Red camera, some people for some reason think that it's the "true" image and try to shoot so that it looks correct as possible in RAWVIEW, when they are missing the point, just as with LOG images, the gamma was not designed to look correct for display devices.
 
What bugs me even more is when colorists say that grading log footage (vs say 16bit linear) produces a better result. As if the ingestion file format's gamma has some sort of magical properties that can't be attained with a 2D LUT to convert from one to the other.

Next thing you know people are going to start comparing zip vs LZW compression qualities. "There's just a more organic texture to footage that uses ZIP over LZW. I would say that jokingly but I know it's only a matter of time.
 
Since so many TV shows are using or considering 5D, I don't know how important the base look of the camera is. A good colorist can make almost anything look good.


Dusty

Does in fact the Red or Arri have a `base look`?..apart from colour corections and Lut`s etc..

I`m sure i read a post where someone reckoned they could always spot Red footage? in the way i understood the poster he seemed to be infering that the Red sensor somehow imparted an essential Red dna that could be detected regardless of any post processing Luts, colouring etc..

I may well be talking nonsense!
 
Last edited:
What bugs me even more is when colorists say that grading log footage (vs say 16bit linear) produces a better result. As if the ingestion file format's gamma has some sort of magical properties that can't be attained with a 2D LUT to convert from one to the other.

Next thing you know people are going to start comparing zip vs LZW compression qualities. "There's just a more organic texture to footage that uses ZIP over LZW. I would say that jokingly but I know it's only a matter of time.

It's already happened...

In the audio word there is a segment of folks who inisist that "uncompressed" _MUST_ sound better than "mathematically lossless" compression... even though the resultant bits are the same.

I've seen at least one reference to the Epic's planned lossless 2.5-1 compression "likely producing almost no artifacting".

:(

-sc
 
In my opinion, the default Red look (Redcolor & Redgamma) looks better than the default look of Alexa rec709, but I don't know anyone who is going to live with those default looks. Someone (a colorist maybe) is going to tweek them to look better.

Isn't a LUT just a color correction that someone has created based on a standard? Red has had several different LUTs. Different apps use different LUTs, they are all doing the same thing, but differently, and they all look slightly different in my experience.


Dusty
 
EPIC shouldn't be made like an Alexa otherwise in that case EPIC would cost even 3x times more than Alexa base package

just because EPIC has already packed 3x times more features than Alexa and also EPIC base body (brain) is 3x times smaller than Alexa, etc,...
 
Arri's business is somewhat diversified and divisional. Each division is probably maintained and managed as a stand-alone business. When you see the Alexa Team pictures, I assume that that's the digital cinema division. Organizationally, they are likely an adjunct to the the primary film based cinema team and thus represent a larger labor ratio to camera production than they would likely have if they made either all film or all digital cameras. This, plus the need to maintain two separate kinds of industrial infrastructure accounts for the higher cost basis of the Alexa and her siblings. These two divisions are also cannibalizing their respective markets to some extent and I don't think this posture is sustainable. At some point very soon, Arri has to go all-in on digital. The decision to make Alexa resolution limited isn't helping them, it puts them behind the curve. Its just interesting to watch.
 
Jeff, these are incorrect assumptions. Something closer to the truth can be gleaned from a careful reading of the FD Times article.
 
Back
Top