Hannover Hands On comparisons of RED F35 & D21

[geoffboyle]

I have done a series of OBJECTIVE measurements of DR using the well known method of the Stouffer transimitive chart.

This is of course where we are always going to differ.

My clients really don't give a flying F' what it measures like, they want to know what it looks like.

That's why I test the way I do.

 

[Evangelos Achillopoulos]

Geoff, we can agree that you don't give a "FF" for what I do...

But our difference is that the subjective methods will always let everyone to have there opinions...

Two cameras, twenty DoP's, one scene to shoot... we have 2x20=40 different opinions... of the same reality...

So everyone can say whatever he likes...

On the other hand an objective method based in simple measurement methods will always going to result to very straight forward numbers that not always be in favor of the one or the other, since they don't allow diplomacy and fuzzy assumptions because they speak the truth and only that...

You see there only one reality that exists in our physical world... in the virtual/marketing/fantasy/personal-reality world we can create different realities... but in order to do that we have an enemy... is these stupid guys that want to do thinks objectively...

No offense, I have also done tests with real scenes... ALWAYS they just verify the measurements I do objectively... when I try to grade images for filmout only the ones with the best numbers are giving the maximum... On the other hand when I tell to a DoP as a DIT that with this camera (any camera) you have i.e. +3,5/ -4 and light accordingly I always get the best results... my numbers are always sourced from these simple measurements...

 

[Patrick Tresch]

That's why I test the way I do.

I undestand now why you put a window with changing objects (cars) in the back ground and probably changing light (clouds or ...) as your way is not """objective""" but made for your clients on a "real world" experience.

I miss interpreted your test as a keylight test as there was a chart showned with some T stops marked on it...

Pat:mad2:

 

[Blair S. Paulsen]

Apples to Oranges

Apples to Oranges

Testing disparate camera systems referenced to a particular target will inevitably yield results that are "colored" by the target choice. The most cogent issue in a lot of recent tests is the choice of "industry standard" DPX file format which uses Cineon log. In theory, a perfect transform from a linear light RAW image into the Cineon log space would preserve full granularity with intact relationships between all values but that, AFAIK, does not exist.

What real world alternative is there to DPX that would allow you to get images from all these different cameras side by side for evaluation? Valid question. Until there is a good answer, the notable tests of 2009 (ASC/PGA, BSC, Geoff's) will always have an asterisk in my book due to lin to log issues if nothing else.

Perhaps of more impact in the case of Red material, it is less "processed" at the point it is first accessible than other cameras. The matrix of choices that one has in the next steps of processing are different with Red that the Sonys or the D21.

How much of the color science is built in?

For example, the D21 has a rep for having terrific color before DI.

How much image processing on board?

Correct me if I have this wrong but it is my understanding that the F23, F35 & Genesis all process the image off the sensor(s) in 19 bit space prior to hitting any of the spigots or writing to tape at 10 bits. By "painting" the camera on set the DIT can leverage the robust on camera DSPs to get a signal off the camera that is much closer to finished product than a camera like the RedOne that leaves that to the user.

IMHO unless these fundamental differences in topology are accounted for in testing design the impact on the results will be greater than most people imagine. YMMV.

Cheers - #19

 

[geoffboyle]

I undestand now why you put a window with changing objects (cars) in the back ground and probably changing light (clouds or ...) as your way is not """objective""" but made for your clients on a "real world" experience.

I miss interpreted your test as a keylight test as there was a chart showned with some T stops marked on it...

Pat:mad2:

We checked that the levels matched on every take of the shot including exterior.

And I didn't say I didn't give a FF about the poster earlier, just the objective method.

We work in a very subjective business.

I'm done here.

I'll not post any further links to tests.

You all obviously know so much more than I do and have done tests like this so many times that it's a complete waste of time me doing this.

I'm sorry for wasting all of your time.

 

[Tim Sutherland]

We checked that the levels matched on every take of the shot including exterior.

And I didn't say I didn't give a FF about the poster earlier, just the objective method.

We work in a very subjective business.

I'm done here.

I'll not post any further links to tests.

You all obviously know so much more than I do and have done tests like this so many times that it's a complete waste of time me doing this.

I'm sorry for wasting all of your time.

Wow, I'm really sorry that it had to come to this. If people don't like the way a test was done, the best response would be to test themselves.

Again, this is a shame.

 

[Robert Frank]

Yeah, it wasn't the noise that I was looking at when I made that decision more the detail in the grill on the front of the amp

Well that is because Peavy amp distortion sucks balls... wait... wrong forum..
:emote_head_explode:

I'm a Mesa Boogie guy myself.

 

[Robert Frank]

I'm sorry for wasting all of your time.

Geoff... dude.. stick around. Really.

Don't let the (forum) noise get to you.

I, for one, truly appreciate your efforts.

 

[jimhare]

I agree, owner egos on this forum are easily bruised and we can be a defensive mob when it comes to our baby.

And when anything goes awry in testing, it quickly becomes martial law and the arrows fly.

But those sentiments calm quickly, we all regain composure and then get back on track.

That's where we are now.

I for one would like to see the ultimate results, when all testing concerns have been addressed.

In the end, I do sincerely want to know the differences of each camera on a level playing field.

Won't change how I feel about my Red, still the best thing that ever happened to me!

 

[Dylan Macleod CSC]

Wow, I'm really sorry that it had to come to this. If people don't like the way a test was done, the best response would be to test themselves.

Again, this is a shame.

Do you really blame him?

Unbelievable attacks from a few ingrate individuals who are too lazy to do their own tests and quick to criticize a veteran who made an honest mistake and corrected it (breath...)

 

[Blair S. Paulsen]

I sincerely hope that at some point in the future Geoff will re-engage with the RedUser community. As far as I can tell he conducted tests in good faith and was simply sharing what he discovered. He copped to his error in choosing the wrong piece of software to create the dpx files, something any of us could have done.

If my post about inherent methodological flaws in several well known tests was taken personally I apologize. Geoff can be curmudgeonly, but from my POV he is as Mike Most and David Mullen have described him - someone who embraces new technology when it is proven to suit his needs on a particular project.

Cheers - #19

 

[sergio arguello]

Do you really blame him?

Unbelievable attacks from a few ingrate individuals who are too lazy to do their own tests and quick to criticize a veteran who made an honest mistake and corrected it (breath...)

dylan,

i ask respectfully, have you never studied your craft by reading? Or have you taught yourself? Just asking because i believe that these forums are, for alot of people, a sorta Digital Film school, and when they see that someone has taken the time to do a TEST they kindly take it serious. Dont you? Im not being sarcastic nor critical of anyone at this point, but i will say that there are alot of eyes and ears watching and listening to what is RED and therefor I believe the testers need be aware of this....agree?

IMHO

 

[Patrick Tresch]

You all obviously know so much more than I do and have done tests like this so many times that it's a complete waste of time me doing this.

I'm sorry for wasting all of your time.

I over reacted on your mail about Evangelos "objective" test.

Sorry if I had a bad attitude.

Pat

 

[Dylan Macleod CSC]

dylan,

i ask respectfully, have you never studied your craft by reading? Or have you taught yourself? Just asking because i believe that these forums are, for alot of people, a sorta Digital Film school, and when they see that someone has taken the time to do a TEST they kindly take it serious. Dont you? Im not being sarcastic nor critical of anyone at this point, but i will say that there are alot of eyes and ears watching and listening to what is RED and therefor I believe the testers need be aware of this....agree?

IMHO

I think the people viewing the tests need to be more aware. Yes - take it seriously. Look at the person who did the test and judge how you look at it accordingly. In this case a mistake was made (rare for Mr.Boyle) and was corrected. Thanks to reasoned input from Jim et al. Otherwise Geoff's tests are quite reputable.

There is theory (reading, forums,viewing other people's tests) and there is practice. Doing your own tests. I have done (and continue to do) my share of both. The people that so boldly criticized Geoff's test and went a step further - to criticize the ASC and the "old guard" WTF?

While at the same time not showing any evidence of doing this leg work themselves. And to not offer any comparison's of tests they have done? That's not constructive diaologue.

I just did a whole series of exposure tests with my RED. Did them with Macgregor's SC_41 and normal as well. Thought about posting some results here, but forget that.

For me, this thread has dented the spirit of sharing that exists in many cinematography circles.

 

[D Fuller]

This thread is the most disgusting example of ignorant mob dynamics I've seen on-line in a long time. Get a f***ing grip people, this is a camera, not a religion!

When you find yourself attacking Geoff Boyle, David Mullen, and Mike Most, it's time to ask yourself what you're doing. And to the moderators: when you see that kind of thing going down, it's time to step in and calm the waters. And that doesn't mean make snide comments about walking into the Apple store and saying "Apple sucks". Nobody did that in this thread.

I own a Red One, and I love it. But I deal often with people who look at it like farmers in 1900 looked at the "stinking, belching horseless carriage" that scared their horses. What's going to convert them is real information, not the evangelical ranting of the faithful. Geoff was trying to provide that, and despite technical problems with the files, I for one thank him for his effort. (I have made enough mistakes in my life to cut him some slack on the bad DPX's.)

A lot of you guys need to get out in the real world more, where you actually have to defend the choices you make to clients and producers.

 

[George A.]

I am a proud owner of a RED camera - with all its strengths and weaknesses (like any other system out there).

I am proud to be a small part of the RED community, in general.

I am ashamed to see how certain members of our group behave toward other people (and fine people at that).

Simply incredible...

George

 

[Sanjin Jukic]

"Redacted" or everything can be adjusted before DPX output

"Redacted" or everything can be adjusted before DPX output

Geoff Boyle's CML Hannover test methodology should be futher "investigated" even if he doesn't want

to participate any more here because of the forum's "high noise" about his "sort of DC comparison test".


Boyle wrote at the CML page HD Hannover workshops the following:

"ARRI D-21 was recorded log C on an S.two DFR

Sony F35 was recoreded S log and S Gamut on an S.Two DFR

RED One was recoreded to 8Gb CF cards all camera setting "normalised" 5000K/320 ISO"


Let's continue to talk about main part of the test methodology at a "level" that is about:

How cameras or data have been "adjusted" before the pictures were sent to DPX output?


About the cameras: Have a look more about sort of Cine Log curve from D21 and Sony F35 at the following three documents (the links are below).

Also find more about Sony F35 with a focus on S-Log (Sony Log) and S Gamut...



Arriflex D21/D21 HD Instruction manual download here>>More about Log C at the page 57>>>


Sony F35 Operations manual download here>>>Especially at the pages 57 and 123


Sony S Log Whitepaper download here>>>



If this case means that Sony F35 and Arri D21 make cine like images that were already a bit "set", "adjusted" or "redacted"

BY A FACTORY CUSTOMIZED CINE LOG CURVE in the camera and that is not the same case as with RED1.

RED1 image was recorded as a "default" or "normalized" by user/tester custom at the values of 5000K/320 ISO.

As we know well that R3D RAW image data can be further more adjusted in REDCODE RAW converters like a RED ALERT, REDCINE, etc,...

then can look much better and be more competitive for any test.


Also could be that I'm (again) totally wrong here but that is on my side anyway and I'll take a full responsibility...

 

[Frank Cueto]

And here I always thought it was a place to share opinions and idea/experiences. Not a place to suppress opinions or points of view, SPECIALLY when backed up with home work that FEW here have done (including myself). I for one hope that Geoff will see thru the few rotten apples and grace us with his company a lil longer.

sad...

 

[Evangelos Achillopoulos]

And I didn't say I didn't give a FF about the poster earlier, just the objective method.

Geoff, I understand your frustration about the level of attack against you...

But this "FF" was totally unappropriated for me, since you know who am I and that I have never attack to anyone, and especially to you...

We disagree I know that... its old news... for me at least...

I was just trying to say something about methods and F23 performance against RED... I'm a professional in this market and I do work with everything from DV to anamorphic Film scans. So let us keep a respectful attitude between us.

 

[Dylan Macleod CSC]

Makes you wonder what else was done incorrectly.
How can we trust these tests with something as simple as outputting DPXes is done wrong. I put alot of faith in Goeff having sound testing methodology but I'm about to throw that trust out the window. What there any documentation on the test methodology was this just a quick hack?

Hacks like this shouldn't be publicized. It does damage to those of us needing real solid information. Arrrggh!

And what happened to the "real names" policy? Ariana who?