Zacuto Camera Test

[Jagger Christian]

Zacuto posted a filmed testing of the following cameras on their website

Arriflex 435ES
Red One
Panasonic AJ-HPX3000
Sony HDW900R CineAlta
Panasonic AJ-HDC27H Varicam
Panasonic HVX200
Sony PMW-EX1 XDCAM EX

Here is the link...
http://www.zacuto.com/ZacutoCameraComparison.mov

Your thoughts on the results...

[Gunleik Groven]

Just a couple...

How was the raw files processed before transfering to D5?
There's quite a lot of we compression going on in there.

Not judging anything here, just curious

[Essor]

They made the RED look really bad there.

Made it have more noise than the EX1, it would seem after one view.

I'd like to know more about how this test was done.

[Jagger Christian]

Essor,

I agree with you. I've seen so much beautiful footage posted from the Red One Camera. I was surprised at the quality of the Red One clip. I'm not that well versed with all those cameras first hand. I was wondering how good of a test it really was. Is it really accurate or did they just not know how to get the full potential out of the Red One Camera?

I don't have any info on the test. Just found the clip on the Zacuto website. They say they showed the footage at NAB 08.

[Essor]

First thing I noticed is that the Red footage suffered tremendously in the compression stages. There is obvious posterizing on the girl's neck.

Also some stunning amount of noise in the neck area. A lot of the cameras got noise there. What were the light levels on this test?

Who does all these tests anyway?

[Zk2007]

Yeah RED didn't look very interesting in that test, maybe the way it was set up. But the most interesting is how well the low-end cameras, especially the EX1 compared to 35mm and the high-end HD cameras. The EX1+Letus Extreme and 35mm film comparison looked pretty impressive, at least at this resolution. I can see this set up looking pretty good on a big screen.

[Tico Llaurador]

Interesting. And a good effort, but it has obviously lost a lot after being prepared for download or streaming over the web. Everything looks over-processed, IMHO.

It's like having a competition where the world's best chefs prepare a fabulous meal with the finest ingredients available, only to have it be frozen and then microwaved before being served to the panel of judges.

It'd be nice to see a test like this be made specifically for theatrical projection. Maybe they should have hooked up with the USC Entertainment Technology Center (http://www.etcenter.org/digital-cinema-lab) folks for assistance. They'd probably be all over a test like this.

Thanks for the test!

[Gunleik Groven]

There were all kinds of strangeness in there, and I'm unsure if that really told me anything about any of the cameras...

[Essor]

Several TIFF stills out of each camera's most uncompressed output available would have been a lot better.

Preferably with film scanned at 4k.

[Pawel Achtel]

Over-saturated and clipped colours; substantial compression noise on all clips - all much more compromising image than ANY difference among the cameras that I could see.

As a comparison - nice and completely useless! :bleh:

[Stephen Pruitt]

Yeah. And last time I checked, Soderbergh wasn't planning on using an EX1 with a Letus Extreme 35mm adapter.

Stephen

[Essor]

Yeah. And last time I checked, Soderbergh wasn't planning on using an EX1 with a Letus Extreme 35mm adapter.

Stephen

This kind of defensive attitude ruins the mood around here.

The RED footage in that test, and the comparison itself - or any comparison for that matter - can be analyzed without resorting to name-dropping and other manner of childish behavior.

[Gunleik Groven]

Thing is, that if you've worked some (and I have still just worked some...) with RED footage, the result of the RED footage is strange, and then I guess that goes for all the cams. Particularily I find the noise in the window and in the "red water" challenging to comprehend in the Red Shot, and I suspect (but of course I don't know) that it can be a result of a workflow choice...

The HPX stands out as the great cam I think it is, though :)

Gunleik

[Álex Montoya]

Pfff, I can't make much of it, really. Every camera seems to suffer in their final output.

[Matt Gottshalk]

I'd like to see the unprocessed, un color corrected still frames from each camera.

[Gunleik Groven]

That'd be more helpfull...

[Paris Remillard]

One thing that I noticed is that it doesn't mention a Redcine or Red Alert step. It mentions the 24fps telecine on Spirit2K for the film footage, but no processing step for Red footage. I guess it doesn't say anything about the actual processing of the film, but... As others have said, I'd be interested to know how the Redcode RAW was processed.

[Fredrik Callinggard]

I'm sorry to say guys but noise is an issue with RED. This seem to be a very straight forward test were they done similar channels in workflow. My guess is film to DPX and RED to DPX etc. They've graded in a very professional suite "Da Vinci" with excellent noise reduction possibilities (I assume).

The compression artifacts such as posterizing exists also in the 35 mm shot, just not as much because the resolution is higher and better in 35 mm.

I love RED but it's far from perfect and it is very noisy. Hopefully this is something that will be fixed in near future. I for one certainly can't wait for Epic and REDCODE 100.

Fredrik Callinggard

[Essor]

That much noise?

Haven't seen that until now.

I also haven't played with RED footage myself.

So let's hear more opinions about the AMOUNT of NOISE and POSTERIZING in that shot.

Because frankly, it makes the Red look worse than the HVX.

[Jason Ing]

Frankly, the test doesn't reveal anything about the cameras.

However, it does reveal that they didn't know what they were doing.

The list could go on, but to use just one point as an example... What's the point of color correcting with a da Vinci if you don't know how to color correct? Look at the detail of the yellow flower in the HVX200 shot versus Red AND 35mm. You're telling me that Red and 35mm can't do what the HVX200 can in not clipping out all that detail? And watch the "whites" in the highlights, bathtub and wine label... blue, green, yellow, orange... come on... please... some consistency in the neutrals please. Even if the label is a "cream" color... it should be cream in all the shots. And the SKIN COLOR... ???? Was that for the SCI-FI channel? lol! :)

[Graeme Nattress]

Maybe they're only comparing DOF and bokeh? And that's why they graded them all to look similarly bad?

Graeme

[Fredrik Callinggard]

Frankly, the test doesn't reveal anything about the cameras.

However, it does reveal that they didn't know what they were doing.

The list could go on, but to use just one point as an example... What's the point of color correcting with a da Vinci if you don't know how to color correct? Look at the detail of the yellow flower in the HVX200 shot versus Red AND 35mm. You're telling me that Red and 35mm can't do what the HVX200 can in not clipping out all that detail? And watch the "whites" in the highlights, bathtub and wine label... blue, green, yellow, orange... come on... please... some consistency in the neutrals please. Even if the label is a "cream" color... it should be cream in all the shots. And the SKIN COLOR... ???? Was that for the SCI-FI channel? lol! :)


I think they've kept it to a basic correction for a reason anything else would defy it's purpose.

I think you should take the differences as in that's what the cameras produce - everything else need to be corrected :sarcasm:

As for the noise... the problem exists in that they lit the background with blue gels etc. Now if they're clever they did that deliberately, since all cameras has noise in the blue channels, some more, some less.

If they shot this in tungsten light with blue gels on RED it is trouble - it's RED's Achilles heel.

Fredrik

[Mike Prevette]

The test doesn't outline if any noise reduction was done during the film transfer. My hunch is the colorist had some turned on by default without even thinking of it. The Red footage on the other hand probably got "defaulted" out of whatever processing program they were using.

[Jason Ing]

Good point. The title cards in the second, third, etc. series of tests seem to imply that.

But the title card of the first series of tests states they used the da Vinci, which implies to me that color grading is something to judge as well... or what would be the point of using a da Vinci and advertising its use? Not to mention that there are better setups to judge DOF. The setup here seems to be about color and dynamic range; there's flowers (vibrant colors), a model (skin tones), cool light on the left, warm light on the right, shadows on the girl and background and highlights off the glass.

[Fredrik Callinggard]

Good point. The title cards in the second, third, etc. series of tests seem to imply that.

But the title card of the first series of tests states they used the da Vinci, which implies to me that color grading is something to judge as well... or what would be the point of using a da Vinci and advertising its use? Not to mention that there are better setups to judge DOF. The setup here seems to be about color and dynamic range; there's flowers (vibrant colors), a model (skin tones), cool light on the left, warm light on the right, shadows on the girl and background and highlights off the glass.

Speed, flexibility in 2K grading???

[Jason Ing]

I think they've kept it to a basic correction for a reason anything else would defy it's purpose.

I think you should take the differences as in that's what the cameras produce - everything else need to be corrected :sarcasm:



Okay, so why color correct at all?

Or... if they did do just a basic correction, then do a basic white balance... make the neutrals consistent in all the shots. That would make sense and be "fair". But the neutrals are all over the place.

[Jason Ing]

And there are better DOF and bokeh setups, right? Put some candles or pinpoint lights in the background to see the bokeh. No?

I hit their site and all they have is the download. Did I miss a comments, opinions, summary somewhere?

EDIT: Oh, okay. I took a look at their site a little more. They're just a rental company. It seems to me that they just threw together some footage and called it a "test". There's no validity. It's really just advertisement for them... they showed it at NAB. I guess they're still trying to say their HD equipment can look just as good as 35 or Red... so please, please, still rent their equipment.

But they did a disservice to themselves, in my opinion, because the footage is just okay. They should have at least cc'd their footage to make it look better. I wouldn't want to rent their equipment if the end product looked like that.

Sort of reminds me of walking into a circuit city or best buy and the salespeople fooling around with the tv settings to make the one they want to sell look better. (Except all of the tvs are getting bad video feeds anyway! lol!)

Just go Red. :)

[Stephen Pruitt]

Geez, Essor, lighten up here.

My post was exactly appropriate for the reason expressed: If the Hollywood biggies thought that the RED was incapable of producing very high-quality footage, they wouldn't be using it for Hollywood films.

I, personally, have NEVER seen RED footage as noisy as that test. It was a mess!

This "test," or whatever it was, does not reflect the footage I've seen shot on the RED by anyone else, here or elsewhere.

Stephen

[Stephen Pruitt]

Geez, Essor, lighten up here.

My post was exactly appropriate for the reason expressed: If the Hollywood biggies thought that the RED was incapable of producing very high-quality footage, they wouldn't be using it for Hollywood films.

I, personally, have NEVER seen RED footage as noisy as that test. It was a mess!

This "test," or whatever it was, does not reflect the footage I've seen shot on the RED by anyone else, here or elsewhere.

Stephen

[Jagger Christian]

I, personally, have NEVER seen RED footage as noisy as that test. It was a mess!

This "test," or whatever it was, does not reflect the footage I've seen shot on the RED by anyone else, here or elsewhere.

Stephen

That's good to know! Thanks!

[Clayton Harper]

The grade, lighting and production design are all easily explained in this clip. This footage is a teaser trailer for a new adult film called: The Babysitter Gets Drunk. :pinch: