F5: Sony Announcements--NAB Press Conference

Mark, I think Sony is wise choice. You went with a camera from a company who has been building cameras for decades.

Regarding ProRes, I too am crossing my fingers for 12-bit 4444 even if only 2k as it would again increase the longevity of my F55 and it would evade having to spend the extra money on RAW recorder, RAW cards, which also increases the overall size and footprint of the camera.

If ProRes does become limited to 2k then it also perfectly justifies moreso the purchase of an F5 because I'd rather have the sensor qualities and bit-depth over resolution if I had to chose between it all.

Anyway, congrats on your F5.
 
Cheers mate. I'm actually a little bitter about the whole thing - Firstly, I'm already in the middle of spending a small fortune on lights and grip at the moment, and secondly, I never had any intention of moving such a short way up the camera chain (F3 to F5). And wouldn't have done it if I'd been able to pick up a 4:4:4 upgrade for the F3 at a reasonable price (because then I would have only had to add a O7Q for 4:4:4 1080p50 - which was all I really have a need for).

But when I added up the full price of the 4:4:4 upgrade with a complete O7Q package with media - the price difference between maxing out the F3 and selling it to move up to the F5 (whilst it was on a big promo deal) was too small to ignore. I'm having a lot of HFR requests from clients at the moment, so the F5 answers all those needs.

Most of all I'm looking forward to 4:4:4 internal recording. The stress it will spare me, is probably worth the added thousands I didn't really want to spend - particularly on a camera I've never actually seen or shot with before! (though I've used the F55, and I assume that's practically identical).

I'm very interested to see how a 2000(!) ISO camera goes under the Australian sun though. I suspect I'm going to be using ludicrous amounts of ND.
 
Price difference aside, a smaller footprint is always better to have than not having it. Less external peripherals = less to break, less to wire, less to maintain, etc.. then like you say HFR, 444 internal, prores down the road... what more does one need.
 
You never know, I expected CD to do it too. I've really fallen for the Alexa footage though, I presume it's the colour science,

Do you think this is Alexa footage? I would love to know what camera this is?


In my experience, yes it is Alexa footage. The biggest giveaway for me is how the brightest parts of the flames are rendered...Alexa has a particular look to how it renders fire. The brightest parts of a flame will never go white with an Alexa...it always keeps a little bit of it's color even if it's clipping. But often it's never even clipping.

The best way to I have to describe Alexa material is "silky".

And yes, Arriflex had some very good eyes on while they were developing their color processing/LUTs. It's the best in the business.
 
...but often it's never even clipping.

That's the magic of highlight wrangling built right into the camera. Much like an audio limiter/compressor in which signals are never allowed to reach a specific level. Alexa is the only camera I know of in the world that does this.

There was an awesome highlight camera test done with Alexa, RED EPIC, and the Sony F55 not too long ago that illustrates how the Alexa just hardly ever clips no matter what and strives to keep highlights smooth and soft.

Found it...

http://www.youtube.com/watch?v=mSbsI7Tn9Cg
 
Price difference aside, a smaller footprint is always better to have than not having it. Less external peripherals = less to break, less to wire, less to maintain, etc.. then like you say HFR, 444 internal, prores down the road... what more does one need.

Very true. And the ability to upgrade to a global shutter when I can't stand jello a minute longer - well that's genuinely reassuring.

In my experience, yes it is Alexa footage. The biggest giveaway for me is how the brightest parts of the flames are rendered...Alexa has a particular look to how it renders fire. The brightest parts of a flame will never go white with an Alexa...it always keeps a little bit of it's color even if it's clipping. But often it's never even clipping.

Yep. That was my experience too. I think the test Dennis linked to shows it clearly. I did my 'Alexa' film almost entirely in reverse key lighting, putting big bright windows behind my foreground subject - and the damn thing wouldn't clip. It should have been close, it should have been on the very edge of clipping, but I had my false colour monitoring on permanently - and even when I opened up the iris one, two-stops when those windows were right on the edge - they still wouldn't go above 100IRE and into clipping!

That's the Alexa's really secret sauce I think - on top of beautiful images, it's almost idiot proof - which is why I'd still prefer an Amira! :p
 
Some more pictures of the new build-up rig at NAB. The rep was saying it should be around $10K. Certainly makes it a PDW800 type of camera, but you would need the Cabrio or Canon 17-120 to make it really ENG-like.

plate1.jpgplate2.jpgplate3.jpgplate4.jpgplate5.jpgplate6.jpgplate7.jpgplate8.jpg
 
10K. Hahaha! Classic old school broadcast prices. I guess they're targeting networks with this and their deep pockets.

$600 card readers, 5K evf's, $800 filters etc etc.
 
The F65 is a 4K camera, the Epic is a 4K camera (they both give you enough colour information for real 4K), the F5(5)/BM4K/AJA... not so much.
In the CML tests the 1:1 crop shows F55 has more resolution than Epic dragon. It's not only the sensor pixel count- it's OLFP filter tech, DSP, debayer and , in case of Red , compression/decompression before the debayer.
Sony processing experience and prowess are unparalleled and it shows...
 
In the CML tests the 1:1 crop shows F55 has more resolution than Epic dragon. It's not only the sensor pixel count- it's OLFP filter tech, DSP, debayer and , in case of Red , compression/decompression before the debayer.
Sony processing experience and prowess are unparalleled and it shows...

Certainly, but 1:1 of 4k versus 6k doesn't really tell the full story - 6k is a LOT more resolution than 4K, so your gains come when you compare the 6K frame at the same size as the 4K one (not on a pixel-for-pixel basis).

I'm at a total loss as to why people keep upscaling 4k/2k images to compare them to 4k/5k/6k/8k ones - we should be downscaling the higher resolutions to match the lower ones - that's where the benefits of oversampling kick in, the grain structure of your noise shrinks down and you fight back against the havoc wrought on colour reproduction by bayer-pattern filters.

People were complaining about Dragon having the same noise levels as MX on a 1-to-1 pixel comparison(?!) But when you downscale the Dragon footage, that noise shrinks substantially... and voila! Lower noise levels.
 
The way I see it a 4K bayer sensor - gives you a 2K camera (with full colour information). The F65 is a 4K camera, the Epic is a 4K camera (they both give you enough colour information for real 4K), the F5(5)/BM4K/AJA... not so much.


The F65 is an 8K camera. It downscales to 4K. It is far and away beyond the Epic or anything else out there.
 
The F65 is an 8K camera. It downscales to 4K. It is far and away beyond the Epic or anything else out there.

Well I'd say the F65's '8k' is a bit of a grey-area given how they get that number, though I'd agree it goes beyond anything else currently available - I've seen fine details from the F65 at 4k that I don't think anything else out there at the moment can match.
 
Certainly, but 1:1 of 4k versus 6k doesn't really tell the full story - 6k is a LOT more resolution than 4K, so your gains come when you compare the 6K frame at the same size as the 4K one (not on a pixel-for-pixel basis).

I'm at a total loss as to why people keep upscaling 4k/2k images to compare them to 4k/5k/6k/8k ones - we should be downscaling the higher resolutions to match the lower ones - that's where the benefits of oversampling kick in, the grain structure of your noise shrinks down and you fight back against the havoc wrought on colour reproduction by bayer-pattern filters.

People were complaining about Dragon having the same noise levels as MX on a 1-to-1 pixel comparison(?!) But when you downscale the Dragon footage, that noise shrinks substantially... and voila! Lower noise levels.


GRUG -

What are you saying? Are you saying that the best way to compare different sensors is to use the same size screen, or are you saying that you should enlarge the screen in order to compare the sensors? Personally I think that since peoples TV's and theaters screens don't physically change shape and size based on how many "K's" of resolution the footage was shot in - we would compare the footage based on using the same size screen for all tests.

I think we both agree on this, but wanted to clarify

Jason
 
Sex sells, that's all it is. I was more amazed by the look of the footage, than what goes on.

Only thing you can do is teach your children right from wrong, then I'm sure you have nothing to worry about. Other than that, it's all a matter of chance.

CaptJman, what do you think to the AJA CION? http://www.aja.com/en/products/cion#options



JACK-

I don't see the AJA CION selling many units.... well, at least not to people who shop around before buying at least. The real camera to keep an eye on is the Blackmagic URSA because it has the exact same specs as the AJA, costs 40 percent less

AND

Everything is upgradeable on this thing including the sensor. (exactly what I said all of us Sony users should demand) - Why throw away all the good parts of the camera? Why throw away the basic structure of the camera. This is absolutely brilliant and watch- 14+ stops of dynamic range will cost just an upgrade fee, same with 6k and 8k - whats that, you want to upgrade from EF to PL, great. Whats that, you don't want to spend thousands of dollars on an optitek adaptor and put that money into glass? Great Idea :) Whats that, you think it's a rip off to charge 1800 dollars for a shoulder mount on the F55? (only recently dropped to 800 bucks) - apparently so did BM because they threw it in for free. There is a lot of bang for your buck here- still not as good as the F55 - but with everything being upgradable it may be a platform we all have in the not too distant future. AJA?? I don't think so.

Comparison

Feature AJA CION Blackmagic URSA

Dynamic Range 12 12
Global Shutter YES YES
Price $9,000 $6000-6400
Mount PL (only) PL, EF, B4
SDI 3G-SDI 12G-SDI
Monitor Extra 10" Monitor Included
Upgradable Sensor NO YES
Changeable lens mount NO YES
Media AJA SSD (cheaper) CFast (more expensive)
Frame Rate 120 60
 
GRUG -

What are you saying? Are you saying that the best way to compare different sensors is to use the same size screen, or are you saying that you should enlarge the screen in order to compare the sensors? Personally I think that since peoples TV's and theaters screens don't physically change shape and size based on how many "K's" of resolution the footage was shot in - we would compare the footage based on using the same size screen for all tests.

I think we both agree on this, but wanted to clarify

Jason

Yep, gotta compare them on the same size screen/medium. Any other approach is illogical (if the purpose of your test is comparison).
 
The F65 is an 8K camera.
No - it's fiddling the definitions to call it that, though it becomes complicated since 4K, 8K etc only really has meaning for Bayer sensors - which the F65 isn't.

For Bayer, 4K is generally taken to mean 4096x2160 photosites - roughly 9 Mpixels. Similarly, 8K may be taken to mean 8192x4320 - or around 35 Mpixels.

Key is to look at the TOTAL count for the F65 - which is around 19 Mpixel, so only half what would be expected of an 8K Bayer. It's comparable to the number you'd expect for a 6k Bayer.

None of this is to denigrate the F65, it will give "true" 4K resolution for luminance (which all the 4096x2160 Bayer cameras won't), and it gets there by arranging the photosite rows and columns diagonally in a very clever manner, Q67 ( http://provideocoalition.com/sony/story/pixels_vs._pixels/P2 ) - but any system which defines sensor resolution in terms of horizontal/vertical counts (such as 4k, 8k) is inappropriate here. It's wrong to call it 8k.
 
Back
Top