Welcome to our community

Be a part of something great, join today!

Log in
  • 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 :)

RED ONE MX and SONY F35

Here I made a more accurate PSD. It's a little less favorable to the F35 but I think it more accurately represents the two images. I might have been clipping the F35's blacks and just lifting the clipped noise free blacks slightly. Although I can't guarantee that's not actually what RED did as well. Again it's all 8bit and RED has mild noise reduction. Definitely something a little off on the F35 going a little yellow in the green corner, not sure what that's about.

The curve is just in an adjustment layer (in the layers menu). It lets you look at the exact curve I used and you can try to play with your own LUT to try and improve upon it if you like.

PSD: http://www.sendspace.com/file/bq1apg
RED: http://www.sendspace.com/file/cmh0r9
F35: http://www.sendspace.com/file/w9ac7t

Also one other unknown in this test is how each sensor rates ISO. The images were so very different in their RAW states that I don't think their gray points are defined the same. F35 might be slightly underexposing. Noise without context isn't the entire story. After all if you pushed 50 ISO 4 stops to compare to a 200 ISO the 50 would perform badly in the "Grain test". Similarly if the F35 is actually a tungsten camera you would want to see how the two perform at 1250 ISO is tungsten. Personally I think that all high ISO tests should be performed in Tungsten and all low ISO tests should be performed in daylight since that's where the two would most likely be used in a real shoot.
 
So it seems you applied (great) adjustments to the F35. It looks very nice, minus that yellow left in the grays. However you didn't try to do anything to the RED image. Have you tried getting the same look you chose for the F35 out of the RED 1080p? Or the 4k viewing at 50%? And it seems the saturation on the F35 could be a little less saturated to match perfectly. Which makes the F35 feel more attractive at initial impression. If matched perfectly it might actually help reduce chroma noise a tinie bit though. Right?

Also, the noise difference is more than just significant starting all the way from the fourth bar from the top in your comparison. This is certainly not just "a little less favorable" when you consider projection. I am sure you would agree and hope you don't feel I am nitpicking your words.
 
As close as I need it to be to show what I want to see. It even rhymes.


 
Roberto Lequeux said:
So it seems you applied (great) adjustments to the F35. It looks very nice, minus that yellow left in the grays. However you didn't try to do anything to the RED image. Have you tried getting the same look you chose for the F35 out of the RED 1080p? Or the 4k viewing at 50%? And it seems the saturation on the F35 could be a little less saturated to match perfectly. Which makes the F35 feel more attractive at initial impression. If matched perfectly it might actually help reduce chroma noise a tinie bit though. Right?

Also, the noise difference is more than just significant starting all the way from the fourth bar from the top in your comparison. This is certainly not just "a little less favorable" when you consider projection. I am sure you would agree and hope you don't feel I am nitpicking your words.
Click to expand...

True, while my previous bar pretty much put them at a dead tie I would say in my last wedge the MX has a distinct advantage, but even then it's far less significant than say the MX to the original Mysterium. I also think it's close enough that all of the other caveats could be significant enough to change the outcome of the comparison. Which is why I say "A little".

Similarly like you said if you graded the RED to the optimal F35 footage then you would just crush that last stop which is where most of the noise is coming in. There needed to be a an even blacker black to tell if the RED was starting to clip or not.

One thing the Mysterium X does really really well is that last stop. However if the F35 has more dynamic range then in a fair comparison you could just clamp it. If the Mysterium X has that great last stop and more dynamic range then it would be cheating to crush the blacks and would be giving the F35 an unfair advantage.

One thing is clear from the test, both have great sensors to be that clean at 1250.
 
If the future is 4K (and it seems to be), the comparisons need to be done there, not at 1080P. The differences certainly get magnified at 4K. So why 4K? Do you want future-proof footage? 4K delivery is coming... trust me. And 1080P does not qualify to scale to 4K and pretend it is 4K. 3K measured (film) or 3.2K (4K Bayer) does, as does 4K measured from 5K Bayer.

When 1080P came to be... no one scaled it down to Standard Def to perform comparisons with "older technology". No reason to do that now with 4K material.

It is pretty obvious that studios will be re-releasing their libraries in 4K. 4K TV is in the works. 4K and 3D are the new standards to shoot for. IMHO.

Jim
 
Jannard said:
It is pretty obvious that studios will be re-releasing their libraries in 4K. 4K TV is in the works. 4K and 3D are the new standards to shoot for. IMHO.

Jim
Click to expand...

I was at Best Buy the other day. A big truck driver type guy was buying a new flat screen TV with his burly buddy. No doubt for the Superbowl.

He exclaimed to his friend: "This one is 1080P and that is the best resolution there is ever going to be"

I looked at the salesperson, thinking he might correct him. However, from his blank look, he seemed as clueless as his customer.

Ah ... Ignorance is bliss!
 
Gavin Greenwalt said:
Yeah don't know what's going on in the greens. Not sure if it's in the footage or something I did in the curves/levels.
Click to expand...

Gavin

Might be from shooting a Tungsten balanced camera (F35) at 5000K, this could wonk out the VA's. Since the chosen scene was lit for daylight, A clip shot with the F35 set to Digital 5600 and then shooting them both at 3200 might show a different result in green and blue channels. Proper camera set-up/use of the technology is still pretty important with all the systems out there presently.

Or not, depending on the intended result.
 
Gavin Greenwalt said:
One thing the Mysterium X does really really well is that last stop. However if the F35 has more dynamic range then in a fair comparison you could just clamp it. If the Mysterium X has that great last stop and more dynamic range then it would be cheating to crush the blacks and would be giving the F35 an unfair advantage.

One thing is clear from the test, both have great sensors to be that clean at 1250.
Click to expand...

When going to iso 1250 on the F35, it's doing so with Analog gain so it will be losing some latitude on the top end. I don't know how much but it would be good to know.
 
Michael Bravin said:
Gavin

Might be from shooting a Tungsten balanced camera (F35) at 5000K, this could wonk out the VA's. Since the chosen scene was lit for daylight, A clip shot with the F35 set to Digital 5600 and then shooting them both at 3200 might show a different result in green and blue channels. Proper camera set-up/use of the technology is still pretty important with all the systems out there presently.

Or not, depending on the intended result.
Click to expand...

Your new girlfriend is going to start getting jealous if you keep talking about your old one, eh?
 
Deanan said:
When going to iso 1250 on the F35, it's doing so with Analog gain so it will be losing some latitude on the top end. I don't know how much but it would be good to know.
Click to expand...

Did they actually set ISO to 1250 on the F35 or just expose for 1250 ISO and apply a curve in post? I would hope they shot the entire spread at the same analog gain. HDCAM-SR + SLog should be more than adequate to contain all the dynamic range without needing analog gain right?
 
Michael Bravin said:
Gavin

Might be from shooting a Tungsten balanced camera (F35) at 5000K, this could wonk out the VA's.
Click to expand...

Most likely the case. I was just re-white balancing using curves (2D LUT) in PS. You would need to use a 3D LUT to correctly white balance without throwing off all the colorimetry right?
 
It would be an interesting test of the F35, whether it's better to use analog gain or not for high ASA ratings when recording in S-Log. Gain would come before compression and recording, though in the case of an SR recording, that's not much of a factor.

Of course, one advantage of using analog gain is if there won't be anyone color-correcting F35 dailies, so that the footage will come into the post house already with the correct brightness.

But none of this has much to do with the Red One.
 
Deanan said:
Your new girlfriend is going to start getting jealous if you keep talking about your old one, eh?
Click to expand...

They've met, at the Alliance Arena in Munich, one night early last fall.

Alexa is very sensitive and F35 is the grand damme so we had to be respectful.

She was introduced to Alexa as Competitor A, it was a bit noisy when the lights went off that night :-)

There are pictures on the internet if you care to see :-0

http://www.arridigital.com/teaser/sensitivity
 
Michael Bravin said:
Gavin

Might be from shooting a Tungsten balanced camera (F35) at 5000K, this could wonk out the VA's. Since the chosen scene was lit for daylight, A clip shot with the F35 set to Digital 5600 and then shooting them both at 3200 might show a different result in green and blue channels. Proper camera set-up/use of the technology is still pretty important with all the systems out there presently.

Or not, depending on the intended result.
Click to expand...

I think 5000º K. is in the real world too.
 
Michael Bravin said:
They've met, at the Alliance Arena in Munich, one night early last fall.

Alexa is very sensitive and F35 is the grand damme so we had to be respectful.

She was introduced to Alexa as Competitor A, it was a bit noisy when the lights went off that night :-)

There are pictures on the internet if you care to see :-0

http://www.arridigital.com/teaser/sensitivity
Click to expand...

Looks pretty similar to the difference on Post #83. I expect the new Arris to be very good for the industry.

Jim
 
Michael Bravin said:
They've met, at the Alliance Arena in Munich, one night early last fall.

Alexa is very sensitive and F35 is the grand damme so we had to be respectful.

She was introduced to Alexa as Competitor A, it was a bit noisy when the lights went off that night :-)

There are pictures on the internet if you care to see :-0

http://www.arridigital.com/teaser/sensitivity
Click to expand...

One day when "very sensitive" Frau Alexa(ndra) meet Herr Roth M-X this gonna be a fatal night for her.
 
Sanjin Jukic said:
One day when "very sensitive" Frau Alexa(ndra) meet Herr Roth M-X this gonna be a fatal night for her.
Click to expand...

Why would that be? Both sensors seem to be around 800 ASA as a base. I don't see any evidence yet for such a claim. For now, I would guess they are similar but we'll see whether Alexa has more dynamic range or not, or whether uncompressed RAW will lead to any advantage or not. We don't know yet.
 
David Mullen ASC said:
Why would that be? Both sensors seem to be around 800 ASA as a base. I don't see any evidence yet for such a claim. For now, I would guess they are similar but we'll see whether Alexa has more dynamic range or not, or whether uncompressed RAW will lead to any advantage or not. We don't know yet.
Click to expand...

David,

they could be similar and that's exactly why, for RED1 owners, to have not so much reason for switch on Alexa especially

when you count on R1 M-X upgrade that costs $5750 with other great deals (stages) to buy Epic X.

As a RED1 owner I ask myself what should be a reason to get or switch on Alexa that

pricing begins at "only" €50K Euros/68,380.00 US Dollar (!!???).
 
Jannard said:
3K measured (film) or 3.2K (4K Bayer) does, as does 4K measured from 5K Bayer.
Click to expand...

After many complains that RED R1 is not 4K camera, finally I see Jim got a good way of expressing the precise resolution.

In old days we used to refer to Bayer only when expressing the resolution, now days we have to be more precise.

R1 is 4K Bayer, 3.2K measured, 3.2K delivered

Also I will add “delivered” parameter to be more precise when comparing to projected film resolution.

For film I would say 3K measured, 2K delivered.

Film has 6K particles in horizontal line of good 35mm film, but since it is not a Bayer pattern but rather random one and it refers to all the colors I skipped the first parameter for the film since we will be comparing apples with oranges here.

You can read more about it here:
http://pic.templetons.com/brad/photo/pixels.html

Andrew
 
You must log in or register to reply here.
Back
Top