F35/F65 Sony CineAlta F65 Dynamic Range bench

[AlejandroMartin]

Hi, using my own setup, I got some nice material.

Date of recording: 15 June 2026, 22:00 hr (10 p.p Spnish Time)
Location: Lloret de Mar
Camera: Sony CineAlta F65 R (Global Rotary Shutter)
Lense: Arri Master Prime 50 mm
Display to Charts: Samsung QN700C at 4K (same res as RAW from F65), as TV supports 8K, but was at 4K
Monitor Brightness: 1400 NITS
Lights on recording:; none


I use this kind of setup, see bloew pics, but by the night, at about 10 p.m, so with no sun



this on field pics are using cheap mobile



all scenes recorded with no light, only the light from the Samsung monitor (at max 1400 NITS brightness)



I used as well following charts on the TV, so total


























I uploaded and share a folder in mega dot nz to share , but importatn images results on next

 

[AlejandroMartin]

Distance from Sensor to TV about 3 to 3.33 meters, in front, as horizontal and perperdiculare better aligned as posible.

I see, correct me if I am wrong, 16 stops over the noise floor.

I Did many test, to this chart, 25 shots, from T22 up to T1.3, any step was a line (small or big) on the Arri Master prime


To me, on below grab, I see 16 STOPS over noise floor, 17 stops counting the just within noise floor, and 18 stops, but hey be aware that Im not sure at all as the chart was not coloured, just black to white. Fotget the "AFter" tag, its just same as "Before".

But what I know is that the colours aro so goos, se more below





This to me, bewlo, is about 15 Stops




Below at T9 as T22 and T1.3 were not good (remember filmed on night with no more ligth that the TV in front of sensor)

Green channel at T9
16 stops over noise




Green at T.9: down to visible 14-15 stops, this TV altougtjh supports 444, with the NVIDIA Quadro RTX 6000 I was using a display link (usb C to HDMI) so not sure if the picture shown was 422 (as I suspect as camera records in 444, so red green and blue should be equal)




Check the blue



Whats your opinion ?

 

[AlejandroMartin]

For those who want to look at the RAW files, or a selection I shared all the footage and in scene videos here

47.15 GB folder on MEGA

184 files and 87 subfolders

mega.nz

Also I made a post in compuwood website, high classic tech section

Sony CineAlta F65 Dynamic Range Bench

Hi, as I posted recently (yesterday) on DVXUser.com, the F65 from Sony has a Dynamic Range of between 15 and 16 stops generally, above noise. Thats superior from the claimed 14 stops usually its…

www.compuwood.com

Videoclip in REC 709

A true pure Gibson Les Paul Goldtop 1971, ACEScct

And finally a clip, showing myself with a lover in a PEPSI spot (PEPOUI), REC 2020 PQ
Using Arri hardmatte 50mm to apply for a 4:3 aspect ratio, like the original cine was. I even think in film in 1.19:1 like Alfred Hitchcock and some Sherlock Holmes shorts / movies, instead of the 4:3 (1.33:1)

More (longer than 3 minutes videos) at my channel on youcrude https://www.youtube.com/@donaitech

 

[DustinSchmidt]

Very interesting approach!

I'm curious, how does shooting charts on the tv compare to something like using the Xyla 21? Wouldn't you run into issues with isolating the patches or interference from the brighter stops when shown on a TV like that?

I may have to give some of the test footage a look, I've never seen RAW files from the F65. Thanks for sharing!

 

[AlejandroMartin]

My resources doe not allow me to purchase Xyla's, and or use IMATEST, which are also a bit complicated. The easyiest way, give ever, a better result.
As if you use complex system , the results may get many escenartios, so more comples¡xity.

I did as more natural and pure, as I was able.

If you see the videos on the folder /Phone (TEL) from the shared mega . nz url link, you'll see that I was using only the light coming from the TV, so exactly what a pro will do using the X's 21 Chart

I've considered to use a Matter box (Arri MB-20 who can be seen on a clip on the phone folder) to get out other light, but on the end, as no light (no Sun) were there and no artificial light was ON) then no issues on interference. The AMOLED Samsung does not emit interferences (It0s not an old school CRT) and even if did/do, at the safety ued distance of 3 or 3.3 meters, ideal for the best suited lense, a 50mm (what you SE is what you GET) I think, that I did a one good aproarch.

A few minutes ago, I got an email from Sony EU, they agree with me on the 17 stops, but they are saying something .. that I prefer to not disclosure, they're so good.

I shown also Green charts, it 16.1 stops

Thats the Green Mona Lisa (Judas Priest)

 

[cyvideo]

All very interesting but in reality I've found Geoff Boyle's Tests on the F65 to be more in keeping with real world use. As he states, quote:

"The only objective here is to establish the true ISO of any camera and to find a precise usable dynamic range. I emphasise usable as although more may be measurable it may be too noisy to use or the colour may change too much."

Note! He states "... a precise usable dynamic range..." Which on his pretty comprehensive tests shows about + 4 and -4 stops giving a 'usable' range at 800 ISO of no more than 8 stops. Look at the + 4 stops over, and the facial highlights have been blown. Look at the - 4 stops under and the image is well into the mud. So a decent working range of around 8 stops of latitude. Which was about par for the course for high-end cameras back in 2012 when the F65 was released. 14 years is an eternity in sensor latitude developments

Don't take my attached JPG assessment as gospel, see for yourself at the test site. If you really want to look close-up into the results download the ZIP file with 17 high quality 0.5 stop .EXR images and judge for yourself. All can be found at the link below.

Chris Young

https://www.cinematography.net/Sony-F65-UHD-2017.html

 

[AlejandroMartin]

Hi, look at my videos , for example the Gibson one, thats a lot of DR,. also see the pics on the waveform.
Thats above noise floor, takle note that many people did not realize that, as many tools were not available on those back days.

This grapth shows 16.1 to my eye , or close to, above noise floor (in the Green) and look on the RED, men that close to 17 stops.

You can download the RAW files I uploades on my pay mega .nz account (which I pay for other to be able to download free), this is history !

 

[AlejandroMartin]

also as you may know, green has more info as the original sensor 45 degree, has 25% red, 50% green and 25% blue, A-OUI.SOME (pronounced awesome for many many people, a lot)

 

[AlejandroMartin]

Geoff Boyle's Tests on the F65 date back from almost 10 years ago, 2017.
He not did it good at all, e.g. his words "Lighting was a pair of BB&S Area 48's courtesy of Lux&co" and Why the hell you put light on a scene when you only need to capture the light coming from the colors chart ????

Menn, you cannot use light to do a bench, the scene to bench need to have light only pointing to, or better, to have light just only from the source (the color chart).
Also his chart, does not seems to be back - wel -iluminated at all, or if ilumintad, not as the 1400 NITs the QN700C monitor I used has-.

This is , in resume, why Sony only state, if state officialy back in those days, which I have doubt, for a 14+ stops (hey 14, PLUS which means they were aware that with his tools back in the day they only might got 14 stops), but ttheir superb engineerrs knew that mny many more stops were being produced by this history of digitasl film.

To me , the number one camera digital in the earth, even as of now, even better than Alexa 35 who has no Global shutter, and no Rotary mechanical ND filters

These cameras are, a piece of museum, avery hard to find one for sale,and , still usable, and a lot as on todays days we have SSD, and Tbytes of capacity, back in that 2012 days, were not SSD massively, and to store TBytes of info.. what a piss back

Today is the best camera one can have to film in digital, at least under my opinion

 

[AlejandroMartin]

Aslo this 2017, with Light (nooo lights please , no external light or as less as possible just fpcusing opnly on the chart, not on entire room), he states
"Grade corrected for best exposure at .." WHAT !!!! no men, a professional bench must have not color corrected grade, just source RAW in source recording mode (either slog, rec709 or whatever), but never, never, color corrected

Also take note that his target was to find about ISO, and secondary to get DR "The only objective here is to establish the true ISO of any camera and to find a precise usable dynamic range" ,look at.. AND which must be written as "The only objective here is to establish the true ISO of any camera COMMA, and to find a precise usable dynamic range"

 

[cyvideo]

I see what you are saying, but I have no real idea how you are going about your testing as I have never seen a camera tested in such an unusual way. And I've seen a few cameras tested in my fifty plus years in the business. By the way Sony only ever claimed 14 stops for the F65. Had they been able to claim more I'm sure they would have done so.

I use CineD for a quick reference, because I'm lazy, but you will find many other sources stating and furnishing supporting evidence to that fact.

Sony don't even claim anywhere near 17 stops for their current Venice 2. They are claiming 16 stops. But in reality a usable 11.7 stops. You can check CineD's results here:

https://www.cined.com/sony-venice-2-lab-test-rolling-shutter-dynamic-range-and-latitude/

Also confirmed by Lighthouse Rentals own Venice 2 tests where they are saying:

"In conclusion, there is a total dynamic range of 14.3 stops, with only 12.5 stops being practically usable."
https://en.lighthousecamerarental.com/articles/sony-venice-sensor-test/

The widest dynamic range CineD have ever recorded on the bench has come from the Arri ALEXA 35. A camera that has been hailed across the industry as delivering the widest range to date. CineD state:
"That’s astonishing – a new benchmark for our lab test! Comparing that to the ALEXA Mini LF that we tested here, that’s about 1.7 stops better at SNR = 2 and 1.4 stops better at SNR = 1 (the ALEXA Mini LF had 13.4 stops at SNR = 2 and 14.7 stops at SNR = 1 for ARRIRAW).
ARRI ALEXA 35 lab test – Summary. As expected, the ARRI ALEXA 35 sets a new benchmark in our lab test for dynamic range and latitude. ARRI claims 17 stops – this stop sits already inside the noise floor, as can be seen in the waveforms above. However, with 13bits ARRIRAW using some noise reduction this stop is usable!

It will be interesting to see if other camera manufacturers are even remotely able to follow our new (and old) leader of the pack here."

So color me a little sceptical when you are claiming 17 stops of exposure latitude from the F65. I simply don't buy it. And I don't think many others will.

I've said my piece.

Chris Young

 

[AlejandroMartin]

Hi Cris, you said. "By the way Sony only ever claimed 14 stops for the F65" but that's not true, the commas and symbols must be quoted,
Sony said "14+ stops": means 14 and more, which equals to 14 or 15 or 16 or 17.. as they did not have the tech on those days to match what i did a few days ago.





"I use CineD for a quick reference, because I'm lazy, but you will find many other sources stating and furnishing supporting evidence to that fact." as you may know, as you're a pro, CineD is not a true cinema people reviewers, they have tons of adverts, bad cookie, and reviews od things such as DJI, Blackmagic plastic cameras, etc.. And with respect to the people that says that are supporting CineD reviews, the same apply.

With that said, Im not saying that Nino and his friend (who both received my email and opened a few times) are not good at all, the're doing well on his segment.
He have good material to bench, but sometimes , when you think you're perfect, you're still not.

Again , my apologies, I don't want to offend anyone, just to share what I did, being the RAW files only and clips showing how I did, with no such big resources.
To have the more expensive tools, thing, card or camera, does not mean the result will be the best of both worlds. In fact, the essence, of all, is easy, with a bit of love (passion) and warfare.


you said "Sony don't even claim anywhere near 17 stops for their current Venice 2. They are claiming 16 stops. But in reality a usable 11.7 stops. You can check CineD's results here"

They say "con 16 o más de 15 pasos de latitud", which is "16, or more than 15, which equals to what I said in the first lines of this message"
Any way take care that as bigger sensor is, less DR in theory. I don't have any Venice, buy I hope to soon get a test on new F65's, sorry on R65 (rialto)

you said "Also confirmed by Lighthouse Rentals own Venice 2 tests where they are saying:" but again, I am not showingf any Venice bench, just a F65 bench, which to my knowledgement, no one has made public as I did , showing the how to, sharing files, video scenes, etc..

Next camera I will test in a few months is RED Ranger Helium, nd might be , by the time of or after the R65 (Rialto 65) the Alexa 35

 

[hamburger]

It's a good camera, sir.