Arri Alexa and Mysterium-X...

[Sanjin Jukic]

LINK>>>

LINK>>>

 

[M Most]

I think the magic gain tricks Arri do...the use of the magic gain tricks...

What they're doing -whatever that is - is just as valid a technique as a magic math trick that is designed to give the impression that the output of a camera never clips. As Jim said, I think it's much more useful to discuss these things without punctuating the discussion with unnecessary snark. Their approach is their approach. Your approach is your approach. Both yield results that a lot of people will like and some people won't. That doesn't make one right and the other wrong, and it doesn't mean that one is some kind of parlor trick and the other isn't. I very much like what I see from Red - especially the M/X - in real life photography that I deal with as a colorist every day. I've also very much liked ArriRaw material that I've dealt with from the D21. When I see Alexa material under similar conditions, I'll form an opinion on that too, but it will be based on the reality of what I see and not on what people attempt to claim before the product even ships.

 

[Andrew M.]

1) as analogue gain pre-a-to-d.
2) as digital gain post a-to-d before recording.
3) as digital gain in raw development.

Graeme

If I really crave for WB correction to gain some DR, on RED, I just use color filter. Works better and simpler than baked one. It keeps noise level in all channels the same.

Baking it in before A-to-D will always change noise characteristics of each channel so it is big learning curve for post guy each time there is a significant change in WB

 

[Jeff Coatney]

They did say that, but because we've always stated it was compressed it's quite reasonable for us to describe REDCODE RAW as a compressed raw format, the word RAW applying to the lack of image processing and lack of demosaicing rather than the use of compression.

Graeme

This is exactly what I understand RAW to be. Compressed, yes. Baking in image processing, no. Every manufacturer should have a proprietary compression scheme to manage the RAW bandwith. But people buying and depending on a camera have the right to know what is happening to the data from the sensor and if it is being processed in any other way that affects the image.

The reason is not to make one camera system "look bad" compared to another, but to make the end-user aware that the term "RAW" does not mean the same thing from one camera to the next-- so you should adjust the way you shoot accordingly.

Anyone shooting Red and shooting Alexa should know what's happening to the image when they hit record. Obviously there is a difference in how the sensor data is being recorded.

I understand that Arri and Red are competitors, but by not clarifying this difference, we run the risk of creating a lot of confusion about what RAW means from one camera to the next. The pressure on Michael Bravin to elaborate is so that we can clear this issue up sooner rather than later. Its not like the information is going to change-- it is what it is. Let's just be clear about what we're talking about first, then we can debate the relative merits of each approach.

If WB is baked in-- that's a significant piece of information and there should be no reason why Arri wouldn't want that to be made clear. It will affect how you shoot the Alexa camera vs. how you shoot Red and what kind of flexibility you have downstream in post.

 

[David Mullen ASC]

Not sure why Michael Bravin would continue to respond to this thread, with so many of you ready to nail ARRI to a cross for daring to do something differently than Red does to the RAW signal. It's become a no-win scenario for ARRI to post information on RedUser; half of you have already made up your minds what to think, regardless of the results.

Already on the online digital HD Magazine, you can see Marc Weigert's results of ARRI baking in WB in terms of noise in the individual RGB channels, if that's what happens -- from these samples, it doesn't seem to be harmful:
http://www.definitionmagazine.com/

 

[Lee Saxon]

Already on the online digital HD Magazine, you can see Marc Weigert's results of ARRI baking in WB in terms of noise in the individual RGB channels, if that's what happens -- from these samples, it doesn't seem to be harmful:
http://www.definitionmagazine.com/

The issue isn't loss of image quality, it's loss of options.

 

[David Mullen ASC]

Practically speaking, it's noise that limits your options. Even with Red, it has been stated that its best to begin grading by white balancing the converted RAW signal. So as long as the noise is minimal, your grading options are not limited by working with a balanced image as a starting point.

All that matters are results and practical workflow issues, not sticking to some ideal spec on paper. If specs on paper were all that mattered, then most people would stay away from Redcode's 10:1 compression -- but it WORKS, that's what matters. So I don't see why everyone wants to take the high road in regards to one aspect and be so accepting of some other aspect. If Red only allowed uncompressed RAW recording, then at least they could claim to be taking the highest road possible, but instead, they've chosen to be PRACTICAL and use compression wisely -- and we judge how successful they are by the results. Which is how we should judge everything.

 

[Jeff Coatney]

Not sure why Michael Bravin would continue to respond to this thread, with so many of you ready to nail ARRI to a cross for daring to do something differently than Red does to the RAW signal.

I think what we're hearing is not people already making up their minds that its bad, but rather a desire to have some straight answers about what Arri is doing to give the end user an image.

I doubt Arri will allow "Their Graeme Nattress" to explain how they get from A to B, but this is exactly the type of information people want.

And let's not forget that Arri and Michael Bravin have a long way to go before they are subjected to the same level of hostility and resistance that Red has had to endure since they launched their program. This is by now par for the course.

Nobody is going to let any company put forth a candidate technology to replace film without that technology undergoing a tremendous amount of scrutiny from the people who will use it.

The idea that Arri should somehow be exempted from this trial by fire is what I find surprising and telling.

 

[Kemalettin Sert]

Not sure why Michael Bravin would continue to respond to this thread, with so many of you ready to nail ARRI to a cross for daring to do something differently than Red does to the RAW signal. It's become a no-win scenario for ARRI to post information on RedUser; half of you have already made up your minds what to think, regardless of the results.

well said sir..

 

[David Mullen ASC]

What I find surprising is the lack of maturity on this site -- now it's a "Mommy, he started it first!" sort of justification? Like I said, I don't see any reason now for Michael to respond here to such a hostile crowd. I can practically SEE the torches burning...

It's really adolescent to take the view that any hostility towards Red now has to be paid back to its competitors, tit for tat. It wasn't right the first time when it was done to Red so it's not right now to do it to anyone else.

When the hysteria comes down a bit, perhaps THEN Michael will feel inclined to supply more information, but there's honestly no point now, not when so many of you are ready to pounce on him if he confirms that WB is baked into RAW.

 

[Jeff Coatney]

If this were a sporting event, David, I would agree with you. The fact is, Red got tons of push-back and Arri is getting a pass and its all from the same crowd.

Both Red and Arri are asking producers and filmmakers to literally bet millions of dollars on their respective technologies. You're willing to pick one over the other without a thorough vetting?

Red made it through okay, I'm sure Arri will be fine.

 

[Mike Gross]

The idea that Arri should somehow be exempted from this trial by fire is what I find surprising and telling.

They will be given the benefit of the doubt being an established player at first. Once people start using the cameras and encounter skew, possible side effects from the DGA, and workflows hurdles, then the criticism will be amplified.

The expectation is set extremely high for ARRI. If they release too early and there are problems, they will attacked for not delivering the legendary ARRI quality and design.

 

[J. Eric Camp]

"Some one was rude to me so that excuses my being rude to someone else" does not hold water. The worst of which has been people grabbing mis-information and trying to nail Arri with it, (saying they were stripe pattern being an example) RELAX. Why do many have to hate the other to love the one they are with? It will do nothing but drive people away. Let the man speak and judge him, or more correctly the product, on those words. Or better yet, go pick the damn thing up, when ever we are allowed, and give it a whirl and then tell me how it's crap or perfect. Which ever.

A few questions were raised on an internet forum. I am pretty sure that Mr. Bravin along with many of his peers have a great deal to do with their days, and making sure to troll the "competitors" forums for question directed at him isn't top priority. Just because he hasn't responded in the last 24 hours does not equate to him ducking the question.


Relax.

 

[Jeff Coatney]

If you care about the legacy of Arri, if you care about there being a real competitor to Red in the market, if you like Michael Bravin and want him to succeed...

You must be willing to provide constructive criticism of their products and their technology. If the criticism is unfair, it will be called out for what it is.

Nobody benefits from playing softball with these companies-- especially the companies themselves. Now is a crucial time for Arri and Michael Bravin has to trust the feedback he's getting in order to provide a viable solution with Alexa.

When studios spend millions to put subjects in front of these cameras, they have to know that the system will deliver as promised not just on the set-- but in post, in exhibition and in the markets that have yet to be created.

If you give anybody a pass at this stage-- you are harming the development cycle and the people whose jobs depend on these products being viable will lose.

No matter which camera system you "want to win", you do that company no favors if you hold back your criticism. Flaws and shortcomings in any product eventually come to light and the resulting market correction can destroy the companies behind them, regardless of their history. Its better for those shortcomings and flaws to come out sooner rather than later. Its the difference between launching a Kindle and launching an iPad.

 

[Graeme Nattress]

Not sure why Michael Bravin would continue to respond to this thread, with so many of you ready to nail ARRI to a cross for daring to do something differently than Red does to the RAW signal. It's become a no-win scenario for ARRI to post information on RedUser; half of you have already made up your minds what to think, regardless of the results.

Already on the online digital HD Magazine, you can see Marc Weigert's results of ARRI baking in WB in terms of noise in the individual RGB channels, if that's what happens -- from these samples, it doesn't seem to be harmful:
http://www.definitionmagazine.com/

That wasn't recorded RAW though, according to the article, because RAW wasn't working at the time of those tests. Burning in WB into RAW would be a non-issue for when the camera is used as a traditional video camera recording video on the HD-SDI outputs.

We've had some robust discussions here on REDUser on all aspects of video technology, from compression, to sensor colour filter array patterns in general and Bayer patterns in specific, resolution (and measured resolution). I don't see any reason why we can't have a useful and interesting discussion on the merits of where white balance should be performed in different places in the pipeline. I've had those discussions in person with people at RED Day where Arri's D21 method was vigorously defended against our preferred method of leaving it to post where it can be concatenated with other operations to avoid the clipping of analogue or digital gain and multiplying the rounding errors of non-integer digital gain.

Graeme

 

[J. Eric Camp]

You must be willing to provide constructive criticism of their products and their technology. If the criticism is unfair, it will be called out for what it is.

Correct. And criticism before verification of fact = flawed. My only point.

Once a spec is stated, or better yet delivered, we can all go to town on it. I was one of the first to mention the old ArriRaw Pre -WB and thought it was crap then and is crap now. My point is, lets find out if the Alexa actually has it in it.

 

[Jannard]

What I find surprising is the lack of maturity on this site -- now it's a "Mommy, he started it first!" sort of justification? Like I said, I don't see any reason now for Michael to respond here to such a hostile crowd. I can practically SEE the torches burning...

It's really adolescent to take the view that any hostility towards Red now has to be paid back to its competitors, tit for tat. It wasn't right the first time when it was done to Red so it's not right now to do it to anyone else.

When the hysteria comes down a bit, perhaps THEN Michael will feel inclined to supply more information, but there's honestly no point now, not when so many of you are ready to pounce on him if he confirms that WB is baked into RAW.

David... I completely agree, and have just posted, that posts and questions need to be packaged respectfully. Read my 1st post of this thread. I'll ask all redusers once more to please be respectful or risk getting posts deleted.

Having said that, it would be helpful for Michael to respond and clear up any misinformation. I asked the same question 3 times in hopes to understand what Arri is doing. I didn't have a preconceived notion of the answer until there was no response. My expectation remains that WB is not baked in Arri Raw until told otherwise.

Let's do our best to ask questions respectfully, post accurate info and keep this place safe for all posters.

Jim

 

[David M]

I've had those discussions in person with people at RED Day where Arri's D21 method was vigorously defended against our preferred method of leaving it to post where it can be concatenated with other operations to avoid the clipping of analogue or digital gain and multiplying the rounding errors of non-integer digital gain.

Graeme

I think the only words in there that 95% of the people here are going to understand are "leaving it to post":emote_rainshower:

How much advantage is there to trying to get the white balance at least approximately right before the light even enters the lens? With 3-CCD cameras you can individually adjust the gains of the red, green and blue channels, which you obviously can't do with a single-chip sensor.

 

[Graeme Nattress]

I think the only words in there that 95% of the people here are going to understand are "leaving it to post":emote_rainshower:

How much advantage is there to trying to get the white balance at least approximately right before the light even enters the lens? With 3-CCD cameras you can individually adjust the gains of the red, green and blue channels, which you obviously can't do with a single-chip sensor.

Every method of wb has pros and cons. No single one is best, but we think our method fits in best for us (obviously).

With a Bayer sensor, you can do individual gain in the raw, via either analogue or digital means. Nothing different there just because it's a Bayer pattern.

Doing it optically lowers over-all sensitivity

Doing it analogue increases analogue noise and either looses code values and precision on gain <1, or clips highlight data on gain >1. There may be a shadow advantage if the analogue gain looks better than digital gain in the shadows.

Doing it digitally (pre recording) either looses code values and precision on gain <1 or clips highlights and looses precision on gain >1.

Doing it digitally in post allows you to concatenate the white balance with matrixing, so avoiding pre-clipping data before the matrix and increased precision can be gained through doing these operations together as a single high-precision calculation.

Graeme

 

[Scott Crawley]

If you care about the legacy of Arri, if you care about there being a real competitor to Red in the market, if you like Michael Bravin and want him to succeed...

You must be willing to provide constructive criticism of their products and their technology. If the criticism is unfair, it will be called out for what it is.

Nobody benefits from playing softball with these companies-- especially the companies themselves. Now is a crucial time for Arri and Michael Bravin has to trust the feedback he's getting in order to provide a viable solution with Alexa.

When studios spend millions to put subjects in front of these cameras, they have to know that the system will deliver as promised not just on the set-- but in post, in exhibition and in the markets that have yet to be created.

If you give anybody a pass at this stage-- you are harming the development cycle and the people whose jobs depend on these products being viable will lose.

No matter which camera system you "want to win", you do that company no favors if you hold back your criticism. Flaws and shortcomings in any product eventually come to light and the resulting market correction can destroy the companies behind them, regardless of their history. Its better for those shortcomings and flaws to come out sooner rather than later. Its the difference between launching a Kindle and launching an iPad.

Are we forgetting that this open development model that we are participaing in at RedUser is an annomally in this market? I wouldn't presume that Mr. Bravin has come to RedUser to poll the public and get design advice or feature requests this late in the game. Alexa is about to ship, so that particular ship has sailled. Mr. Bravin is largely here to go head to head, make comparisons, and promote their product... I presume.