androidlad

48 minutes ago, Anaconda_ said:

But then why can't you take a camera that could record external ProRes raw, plug it into the VA and record Braw? If you can, then why is their EVA1 support even a thing? If that's correct, by default BMD also support any other camera that can output raw over HDMI.

I'm under the impression that the VA can't wrap any raw signal into Braw. BMD need to know the sensor data for their partial de-mosaic stuff. With that said, I still feel that Braw is sensor specific.

Of course, please correct anything that's wrong here. Would love to understand it better.

Most cameras that output ProRes RAW at the moment are mirrorless cameras with HDMI output, and Atomos developed the RAW over HDMI protocol, they only license to camera manufacturers for free.

For those that output RAW over SDI, BMD need to develop support for the their RAW spec (EVA1 outputs 10bit Log-encoded RAW, Sony CineAlta outputs 16bit linear RAW). And the same applies to Atomos, but Atomos has its RAW over HDMI protocol and it's being widely adopted, so they pretty much have full control over the RAW spec.

So instead of saying BRAW is sensor specific, you can say it's brand specific.

12 hours ago, Anaconda_ said:

The codec is built individually for each sensor, so Braw for the Pocket 4k will be different to Braw for the Pocket 6k and every other camera that currently records it.

Nope.

BRAW is just a codec, it has nothing to do with sensors or camera models. It requires BMD's FPGA for the encoding.

Same for ProRes RAW, Apple has licensed the encoder to Atomos and DJI, it can encode any incoming RAW signal.

A9/A9 II uses 12-parallel ADC and DRAM to achieve 162fps full sensor readout at 14bit (internal speed, subject to I/O limitation). 

Obviously due to power and thermal requirements, DRAM is disabled and only 2-parallel ADC are used in video mode.

30 minutes ago, Alt Shoo said:

They addressed the issue with the lower DR and global shutter. 

They did their best to optimise the DR. For a charge domain based global shutter, it's doing ok, but it's poor compared to conventional rolling shutter sensors.

It's positioned primarily as a high-end crash cam, only global shutter can guarantee zero skew and zero flash banding.

According to a source, X-H2 is likely to have a conventional Bayer CFA.

1 hour ago, Anaconda_ said:

It could be that the shots with these custom lenses didn’t make it  to the trailer.

She speaks about wanting Milan to have a magical quality, but the trailer is action packed, so I think the more ethereal styles scenes haven’t been shown yet.

34 minutes ago, KnightsFan said:

There's definitely some hefty CA in the trailer I just watched, example:

Speaking of which, that's a pretty awesome shot. I haven't paid any attention to Mulan based on the quality of the other live action remakes, but that is a great shot, both for content and the optical quality. It reminds me of showdown with the Wild Bunch at the end of My Name is Nobody. It also shows the impact that vintage styling can make.

It was a custom made Petzval 85mm lens, and used sparingly in the film only for some portrait shots of Mulan. For all other shots, they were Panavision Sphero 65 lenses, very vintage but with some modern touches, CA is noticeable.

Mandy Walker simplified it a bit too much in the interview, it's not really about the CA, it's the distinctive field curvature and sharpness roll-off from centre to edge, it's another way of isolating the subject instead of just using very shallow depth of field.

This is one of the shots with the Petzval 85mm lens:

YouTube re-compresses all uploads into H.264.

For popular channels, YouTube uses better quality compression.

2 minutes ago, James_From_Raglan said:

So the XT4 doesn't have an crop mode in the sense of using a smaller portion of the sensor (equivalent to the a7sii's "APS-C mode") in order to punch in and get more range from lenses?

And does anyone know if the S1 has this function? I assume it does since it's full frame but I can't seem to find a clear answer anywhere.

XT4 has a Sports Finder mode with 1.25x crop for stills, and a 1.29x crop option for 4K video.

Export H.264 directly from the timeline, with a x264 based plug-in (for example, Aftercodecs), and use the highest quality encoding parameters, and bitrate 12-15Mbps for 1080p depending on the length of your deliverables.

YouTube is resolution agnostic now, so do not add any bars, export the picture in its native aspect ratio/resolution.

If you want to disincentivize people to watch it on phones and smaller screens, then maybe don't upload it on YouTube.

I like her cinematography in the 2008 film "Australia".

Personally, despite her attempts for a vintage feel, I think "Mulan" still looks too modern. They shot it on Alexa LF and 65 with spherical lenses and cropped it for 2.39:1 cinemascope, with many shots that just scream "we have a large sensor!!"

Would be nice if it was shot on super 35mm film with true anamorphic lenses, like what Disney did on "Cinderella".

There are speed booster adapters from Viltrox and Kipon, but Metabones ULTRA adapters are supposed to have the best optical quality.

https://www.metabones.com/article/of/Speed_Booster_for_Fujifilm_X

FINALLY!

Strange that the comparison was shot with "DCI-P3" color and gamma 2.6, both the DCI white point and the P3 gamut are not going to look correct on a BT.709/sRGB monitor people watch YouTube on.

23 hours ago, Andrew Reid said:

In the future digital will be so far gone... so far manipulated with computational photography... that documentary filmmakers will use film, to prove what they shot was real. Digital cameras are going to be reality distortion fields.

The archival of so much digital data is still an ever present challenge for such large file sizes. If you shot everything at 2400Mbit/s you should add a zero to the cost of your camera.

Digital archiving relies on LTO tapes, which are very very cheap.

10 minutes ago, deezid said:

If they're coming with a new sensor maybe, they still may add huge amounts of NR which have been quite destructive on past Sony cameras and cannot be bypassed by using an external recorder.

A7S II has the unique advantage of going way beyond ISO25600 and still provide a usable image. Using an external recorder bypasses most of the NR and compression artifacts.

BBC used A7S II with external recorder for some ultra low light scenes in Planet Earth II.

8 hours ago, sanveer said:

This dynamic range chart seems curious.

Usually the industry standard Xyla charts are available in 15, 21 and 26 steps. But this one seems to be somewhere in between 15 and 21 steps. Maybe RED invented their own chart? 

RED tests their cameras using Xyla 21.

2 hours ago, Patrick B. said:

Good question, and I would think it’s higher than 12 considering how high Helium scored on DxO’s test and that Red’s are very close to Alexa:

https://www.dxomark.com/red-helium-8k-dxomark-sensor-score-108-a-new-all-time-high-score2/

But if Alexa is rated at 14 then maybe it is 13... hmmm

https://nofilmschool.com/2017/02/shootout-arri-alexa-mini-vs-red-epic-w-helium

 

RED Helium 8K S35 scored 14.12EV in DxO measurements, most likely 14bit ADC and some NR, so total DR may be a match to Alexa's 14, but Alexa distributes way more DR in the highlights, while RED (and Sony) tend to have cleaner shadows with more monochromatic noise pattern.

8 hours ago, Andrew Reid said:

The actual 4K recording on my cards from the GFX 100 doesn't really look like 11K oversampled.

It looks more like a modern pixel binning technique. It's a bit above the A7R IV but not as detailed as an X-T4 and there's some aliasing.

So I doubt it is doing an 11K sensor output.

Fuji once did a presentation on "their best solution for 4K" from the GFX 100's sensor. And it had a quad bayer pixel mix going on.

If it does have 11K RAW video in the buffer before it gets downsampled on the image processor... I am sure it could be downsampled while remaining RAW data. Just as there are medium and small RAW stills options.

11K horizontally yes, vertically 2/3 lineskipped.

That presentation was about an X-Trans version of the GFX100 sensor, which was designed with 3 x 3 colour aware binning. But that project has since been shelved.

Yes the sensor will have to be heavily subsampled for ProRes RAW. It can’t send 11K RAW.

2 minutes ago, Andrew Reid said:

Is it Quad Bayer?

Nope.

It'll mostly likely be heavily subsampled to maintain Fuji's no crop philosophy.

Internal recording uses 11604 x 4352 oversampling which exceeds HDMI 2.0 bandwidth and obviously Atomos Ninja V's processing power.

Hands on reviews are being released now.

Regarding Cinema Pro app on Xperia 1 II, only the main 24mm lens/sensor supports 4K 60P, the 16mm wide and 70mm tele are limited to 4K 30P.

1 hour ago, UncleBobsPhotography said:

The S20 Ultra actually has a bigger sensor (1/1.33 instead of 1/1.7), but seems like a failure.  The 960 fps on the S20 ultra is even upscaled 480 fps...

The S20/S20+ has the same sensor size as the new Sony.

S20/S20+ main 12MP camera is actually the same sensor as Xperia 1 II, IMX555

If you wanna do it properly for professional work, there's a dedicated plug-in for adding motion blur called ReelSmart Motion Blur:

https://revisionfx.com/products/rsmb/premiere-pro/

It has a number of parameters for fine tuning and allows you to add precise amount of blur based on captured shutter angle and desired shutter angle.

1 hour ago, sanveer said:

Considering the breadth of the features on this one, the Least they could have done, is given multiple bitrate options.

Also, I wonder if this is also 10-bit like some of the other flagship Xperias. Interestingly does seem to have HDR (though the A Series has it in 8-bit 100mbps). 

It's 10bit with HLG HDR option available (on sensor DOL-HDR)

And all turned to mush with its 55Mbps bitrate

9 hours ago, Andrew Reid said:

I would like to see the image in HDR 24p mode with that dual conversion gain enabled. Probably quite nice. Is low light still challenging with global shutter? The Blackmagic Production Camera 4K had a lot of ciruits on the front side taking up room. These are not BSI Sony chips are they?

No, that sensor is from CMOSIS. Sony only introduced BSI to its global shutter sensors in the very recent 4th gen Pregius S.

Even with BSI, the memory node still takes up some space on the photosite, plus the reason Sony went BSI is to further shrink pixel pitch, so it cancels out each other.

With dual conversion gain, the readout noise doubles and it's already quite high at 2.1e-, so I don't think low light sensitivity will be any good.

The ultimate global shutter sensor design would be pixel-parallel ADC (one ADC per pixel), but it's still in very early stages of R&D.