Full Frame Cinema Shootout 2020...

[luigivaltulini]

Thanks Evin.
Always nice to see how much redusers give to the community for test.
Ok I'll try :) eheheh

Although I have not yet tried the Canon 500ii and sony fx9 (I don't know the difference between cinetone and log3) and sigma.

1 test:

1. Alexa mini lf
2. Canon 500ii
3. Red monstro
4. Sigma
5. Fx9 cinetone?
6. Fx9 log3?
7. Venice

Study test:

1. Alexa mini lf
2. Canon 500ii
3. Red monstro
4. Sigma
5. Fx9 cinetone?
6. FX9 log3?
7. Venice
8. Gemini
9. Helium

[Noah Yuan-Vogel]

Study test:

1. Alexa mini lf
2. Canon 500ii
3. Red monstro
4. Sigma
5. Fx9 cinetone?
6. FX9 log3?
7. Venice
8. Gemini
9. Helium

[paulcurtis]

Helium, methinks. :)

Nah, mine doesn't crush like that and i play in the dark a lot.

I think it looks like something through a baked codec, the decision to crush there to not show noise maybe. It's difficult to do an empirical test when everything is running through a creative LUT at the end but it's great for a side by side general feeling.

For a mini B cam the sigma is fun - not an A cam. It was my stop gap until Komodo arrives!

cheers
Paul

[Björn Benckert]

Yes that is normal.
Something has to be done to the negative.

Tests like these are always camera + development tests.




Then you would compare cameras + ACES implementation. And upscaling algorythms.

All of these cameras are capable of superior imagery and a lot gets lost in translation.

Upscaling algorithms, is as I see it more fair then scaling down and then comparing.
But sure, you can downscale all to standard def and compare as well but that to me that tells less. More fair to do the comparison the other way around with upscale as I see it. Kind of same reasoning as having them all in somewhat the same / float colourspace and see how they compare in there. Sure, ACES implement different import algoritms for each camera.

But what I mean is I can only really judge them when pushing / pulling the image around in similar to such environment, to see what holds, breaks etc. Watching images perfectly balanced towards each other on vimeo is telling less, the camera differences then hides / drowns in all kind other differences. but yes you need to develop the footage some how to compare it visually and that process of course has an impact as well.

[Michael Tiemann]

ISO with a little bit of fine tuning with exposure.

Hmmm.

I'm a little surprised at how much difference there is in white balance between the cameras in the studio setup (normal, over, under). I would have expected that in that environment the WB and ISO would have been normalized to the neutral background between the faces.

Then again, perhaps the idea is to show off how differently the different cameras behave when set to a fixed WB that is supposed to be some well-known number (according to a calibrated color meter).

I will also say that looking at CIE chromaticity readings, I strongly suspect that #5 is a RED camera. I don't have enough experience with other cameras to pick them out of a lineup.

[Adam Marsh]

Magnificent and thank you for sharing.

For the next shoot out, it would be more helpful to keep the name / number of the camera on the entirety of the shot. It makes it far easier to scroll through & compare between them when it doesn't flash for a few seconds!

[Ben Roper]

Thanks for this! Very interesting! When will you reveal the results?

[Hrvoje Simic]

Upscaling algorithms, is as I see it more fair then scaling down and then comparing.

For texture and definition, no scaling.
Unless context is delivery standard.

But sure, you can downscale all to standard def and compare as well but that to me that tells less.

Depends on what you test, for this overall performance it is irrelevant.

More fair to do the comparison the other way around with upscale as I see it.

If you want to present two cameras as superior and make of others more computated fake image.

But what I mean is I can only really judge them when pushing / pulling the image around in similar to such environment, to see what holds, breaks etc.

Sure, but that is not the purpose of this test.

Watching images perfectly balanced towards each other on vimeo is telling less,

That is not "perfectly balanced".
If it was you wouldn't be able to tell the difference. Which would defeat the purpose of the test. Basic camera maker's transform and white balance is least biased. While we are at it, two routes used here are not standard and introduce more variables and a price on the material quality.

the camera differences then hides / drowns in all kind other differences. but yes you need to develop the footage some how to compare it visually and that process of course has an impact as well.

Yes, digital cameras are typically superior than one can see in graded tests.

[David J. Buchanan]

Thanks Evin.
Always nice to see how much redusers give to the community for test.
Ok I'll try :) eheheh

Although I have not yet tried the Canon 500ii and sony fx9 (I don't know the difference between cinetone and log3) and sigma.

1 test:

1. Alexa mini lf
2. Canon 500ii
3. Red monstro
4. Sigma
5. Fx9 cinetone?
6. Fx9 log3?
7. Venice

Study test:

1. Alexa mini lf
2. Canon 500ii
3. Red monstro
4. Sigma
5. Fx9 cinetone?
6. FX9 log3?
7. Venice
8. Gemini
9. Helium

If this is the case then my favs would be the Alexa and Sony

[Patrick Tresch]

As always Monstro has been over exposed.
Next time expos it at 800 iso and put an ND3