What I'm saying is that when you scale the 4K to 100% you are only seeing 1/4 of the image. This means your monitor is therefore the equivalent of 1/4 of a screen 4 times the size of your monitor. So when trying to evaluate the noise levels you should consider how big the image would be and adjust your viewing distance to match.

Consider the reverse, if you scale your HD down to just 1/4 of your screen it will appear to have a lot less noise, unless you view it very close up.

Piotr, you possibly can't export to 4K as you may not have a 4K codec installed. You cannot at the moment export XAVC from Premiere, this is supposed to come very soon.

You cannot at the moment export XAVC from Premiere, this is supposed to come very soon.

I think it came in the October release...
Adobe Premiere Pro CC October 2013 Release (http://blogs.adobe.com/premierepro/2013/09/adobe-premiere-pro-cc-october-2013-release.html)

Piotr, Vegas 12 will render to XAVC and XAVC-S. I mainly use Edius which will edit native XAVC just fine at full frame rate but not full rate for XAVC-S. Edius will mutlicam several tracks of XAVC on my PC so a lot easier on the PC than XAVC-S. I have used Vegas 12 to convert my XAVC-S files from the AX1 to XAVC and also to Canopus HQX they take a little time to convert but both files edit easily after that.

Ron Evans

Well, even though the link provided by James indeed mentions exporting XAVC, and I'm working on a most recent version of PP CC - there is no such output format. The only 4K-capable output format is DPX sequence.

As to VP 12's capabilities of outputting both 4K and UHD (in XAVC or S-XAVC), Ron is correct.

Yes, Adobe talked about XAVC encoding in the press release, but it is still yet to be included in the software :-(

You can export 4K ProRes on a Mac plus DPX or H264 on both PC and Mac.

Well, actually - after much digging - I found it in PP CC, too. You must select MXF, and then in the codec tab, select one of the XAVC codec... But for S-XAVC, we still need to wait, I guess.

Well done Piotr!!!

I have been looking for this for ages. I did look under MXF, but only at the presets, never looked under the codec settings. Thanks!

Alister , do you know if the HDMI output from the FDR-AX1 is 10 bit 4:2:2 like the PXW-Z100 ?

Ron Evans

I mainly use Edius which will edit native XAVC just fine at full frame rate but not full rate for XAVC-S.
Ron Evans

Can this be edited in proxy mode, then swap?

I have not tried proxy mode as I never use so not sure of the time Edius would take to create the proxy. Decoding XAVC-S is the issue. I have proxy turned off on my user presets. May as well convert to HQ and have a usable file. Prefer to edit native or convert to HQ. I may see the use of proxy on a laptop with limited disc space compared to a HQ file but other than that I do not think it is of much use.

Ron Evans

Hi,
I would need rather urgent for an upcoming shoot your recommendation on which XQD Card to use with the PXW-Z100 recording in 4K at 25p.

On Sony's website XQD? Memory Card | Sony (http://www.sony.co.uk/hub/memory-storage-batteries/memory-cards/xqd) they present two alternatives: The pricy S-Series or the low-cost N-Series for 4K shooting.

Of course you can always go for the expensive alternative and be on the safe side. But maybe someone here has really made the test and found out, that also the N-Series cards are reliable and working in 4K.

Thank you so much.
Best, Jürgen

For 60p at some 600 Mbps, Sony says the S-Series is "recommended" (not even "a must"). But for 25p at some 220 Mbps, I'd say the regular XQD card should suffice...

Well done Piotr!!!

I have been looking for this for ages. I did look under MXF, but only at the presets, never looked under the codec settings. Thanks!

Alister,

I cannot take part in the event you will be holding at Malvern, but please come back to us to sincerely and frankly report how the Z100 is doing in the noise department, basing on true 4k watching experience. So far, I learnt that the I-frame XAVC clips are easy enough to edit (in both PP CC and VP12 which I'm using) on my current PC, while the Long-GOP clip from the AX1 (S-XAVC) is a hog to decode, so - especially in VP12 - it crawls at a poor 3-4 fps:(. But unfortunately I don't have the means to actually connect a Z100 to one of those UHD Bravias that support it, and see the PQ with my own eyes...

Cant wait for your opinion!

Piotr

For 60p at some 600 Mbps, Sony says the S-Series is "recommended" (not even "a must"). But for 25p at some 220 Mbps, I'd say the regular XQD card should suffice...
Hi Piotr,
thank you very much for your reply. The "regular XQD-card" would mean, the N-Series?
There is also the H-Series (also with 125 MB/s)...
It is confusing.

Who has tested the different cards for 4K? I would also assume, that the "N-Series" (125 MB/s) will work in 4K 25p. But a confirmation by someone, who acutally used it, would be very helpful.

Piotr, where can I find Sony's recommendation which you quoted above?

Thanks a lot,
Jürgen

Piotr, where can I find Sony's recommendation which you quoted above?

Thanks a lot,
Jürgen

Juergen,

At this site: PXW-Z100 (PXWZ100) : Features : United Kingdom : Sony Professional (http://www.sony.co.uk/pro/product/broadcast-products-camcorders-xdcam/pxw-z100/features#features)

- they say (in the "Features" tab):

"* XQD cards have a write speed of up to 800Mbps. The use of S Series cards with 180MB/s writing speed is recommended"

Unfortunately, my knowledge ends here (and is purely theoretical :))...

Piotr

From our debate (and my personal tests) on the Z100 vs. AX1 cameras (or rather the XAVC vs. S-AVC codex), my conclusion so far is that - with limited budget for upgrading all the other components of the 4k edit/delivery chain - the Intra Frame codec of the Z100 is not only better in terms of color resolution (4:2:2 10 bit vs. 4:2:0 8 bit), but surprisingly it's also easier to edit with PP CCC or even VP12 on my current editing PC (just a modest 2600K@4.2GHz with the GeForce 580 GPU). However, I still have one important question to those more knowledgeable here:

- assuming 10 bit color support by the graphic card and monitor (for full WYSIWYG experience), is Adobe PP CC fully 10-bit aware, in both editing and rendering/encoding? I'm asking, because Vegas Pro is not, afaik. It''s important for those trying to choose between the 8-bit AX1 and 10 -bit Z100 whether they'd be able to push the material harder while grading and color-correcting...

Anyone, please?

Piotr

Piotr, no. In the Vegas project settings, for the editing side of things you could choose to work in the 8-bit 16-235 video space, but by switching to full range 32 bit floating point at the time of rendering you'll get whatever the format is. If it was AVCHD that would be 8 bit 4:2:0. In the case of XAVC, it would be 10 bit 4:2:2.

You might be thinking that 10 bit is truncated because the 0-255 scaling for the histogram/waveform monitors don't change, but in full range 32 bit mode, it's working with full precision.

Am I the only one who thinks this footage from Austin looks noisy? At 50% it looks pretty good but at 100% that fine grain moving around gets really annoying.

I have to agree based on my limited exposure with the footage as well. Edited the clips in V12, exported to 4K MP4, high profile@ 250Mbs. Viewed on a Panasonic 4K set in 4K and the noise levels are very high, like some weird Instagram filter was applied. Even viewing on smaller screens had the same results: 27" iMac (2560x1440), 17" laptop screen (1600x900), 55" LG HDTV (1920x1080) and on my Galaxy Note 3 screen (1920x1080).

I'd like to see raw XAVC on a 4K monitor/display for final judgement. I was really looking forward to obtaining this camera, but if it's that noisy in good daylight footage, I'll wait out.

The first camera I got was crazy noisy. The replacement is much better. At 50% it looks great, at 100% it looks like HD at 200%. I haven't been able to look t it on a 4K display yet. Alister says that it looks fine. The noise may only mean that 4K on a 2K display is just not gonna work at 100%.

The HD that it shoots is superb.

I shot with the camera all day today and going through the shots there are a couple that knocked me out. The camera can make a beautiful image. Some shots were noisy, others much cleaner. It could be a keeper. Still trying to sort it out.

The CA really sucks.

Piotr, no. In the Vegas project settings, for the editing side of things you could choose to work in the 8-bit 16-235 video space, but by switching to full range 32 bit floating point at the time of rendering you'll get whatever the format is. If it was AVCHD that would be 8 bit 4:2:0. In the case of XAVC, it would be 10 bit 4:2:2.

You might be thinking that 10 bit is truncated because the 0-255 scaling for the histogram/waveform monitors don't change, but in full range 32 bit mode, it's working with full precision.

Hi Tom, it's been a long time since we last exchanged posts together:)

Thank you very much for this explanation; I assume (knowing you) that it's 100% accurate. Funny thing is that as a long time VP user, I mainly was drawing knowledge from the SCS Vegas forum, and the established opinion there has always been VP truncates to 8 bits no matter what...

Piotr

EDIT: Actually, it's even better than that; I just checked the piece I rendered out using the XAVC codec in the newest build of VP12, and it is 10 bit 4:2:2 even though I didn't switch to 32-bit precision before rendering... Would that be possible?

Hi Piotr,

It is always nice to talk to you friend. :-)

You expressed the question of whether V12 is "10 bit aware" to which the simple answer is that it has to be, inherently, to read and write 10 bit codecs like Sony XAVC, Avid DNxHD, Main Concept AVC high profile and others.

Then the question went to whether it truncates to 8-bit. Obviously it can't if it is to work as described above, but what if the 2 least significant bits were always zeros? In that case we would still be working in 10-bit but only using 8-bits. That's a little different then not being 10-bit aware. So let's describe the case that would apply and use XAVC 10-bit 4:2:2 in the example.

We know that for HD, rec. 709 coding uses SMPTE 8-bit reference levels 16-235. In Sony cams using hypergammas and log gammas, the color depth is much wider than that, up to 16 bit acquisition in some models.

With Vegas 12, in order to render greater than default 8-bit rec. 709 output, you select customize template from the rendering template, click on the tab for project, click the drop down for color space, and there you are presented with a choice of many, ACES, S-Log2, ADX 16 bit, DCI, CIE XYZ and of course...rec. 709. In order to have anything other than zeros written into the last two bits, you will need to render using the 32 bit full range option and choose one of the full range color spaces, not 709.

So Vegas 12 is certainly 10 bit aware, but whether you can view it depends on your monitor (8 bit rec. 709 most of them), and on your choice of acquisition and rendered codecs.

Tom - thank you for the perfect (as always) explanation; I've just become a better Vegas Pro user :)

Cheers, my Friend

Piotr

So let's describe the case that would apply and use XAVC 10-bit 4:2:2 in the example.

We know that for HD, rec. 709 coding uses SMPTE 8-bit reference levels 16-235. In Sony cams using hypergammas and log gammas, the color depth is much wider than that, up to 16 bit acquisition in some models.

With Vegas 12, in order to render greater than default 8-bit rec. 709 output, you select customize template from the rendering template, click on the tab for project, click the drop down for color space, and there you are presented with a choice of many, ACES, S-Log2, ADX 16 bit, DCI, CIE XYZ and of course...rec. 709. In order to have anything other than zeros written into the last two bits, you will need to render using the 32 bit full range option and choose one of the full range color spaces, not 709.

So Vegas 12 is certainly 10 bit aware, but whether you can view it depends on your monitor (8 bit rec. 709 most of them), and on your choice of acquisition and rendered codecs.


Tom,

Just played with the available color space options, and with an XAVC clip on the timeline, there are 3 different places the choice you describe is given:

1. Project Settings: when pixel format is set to 32 bit fp (full range), the "View Transform" drop-down list becomes active and anything (including "off") can be selected. I guess one of its uses is when an S-log video is being edited, but our example XAVC is not this case

2. When Properties of a clip inside Project Media (or already on the TL) are accessed, one of the color spaces (including ADX-10 bit and various Sony S-logs) can be chosen; I *imagine* this is used to match color space to the project properties set as per Point 1 above, right?

3. In some rendering templates (like MC/Sony AVC, but not Sony XAVC), indeed a 10-bit color space (or anything different than the REC 709, for that matter) can be selected.

Now, I understand you're talking about the last of the above 3; but with color space selection lacking in the Sony XAVC render template, would it mean it's always encoded in a 10-bit color space by default, and regardless of the project pixel format being 8 or 32 bit?

Please explain :)

Piotr

Almost had this answered before you asked the question. See if this helps, and be sure to read the link. :)

Piotr, refer to the link below. Sony Vegas Pro 12, supports the ACES workflow, and you should too. With Aces workflow, you are not working with RAW, but are working in the ACES color space which is in a virtual way, infinite and camera independent.

The 3 steps to the workflow are IDT, ACES and RRT.

For IDT (input display transformation), you go to the media properties for each clip, and set the color space, ex. Sony S-Log2 (5500k). You do this for each clip, thus allowing you to mix and match different cameras. This is the feature that makes possible the unwinding of the S-Log2 curve so that the image appears normal.

For ACES, and to enable viewing the unwound s-log, you go to the project properties, select full range, 32 bit, 1.000 (linear) gamma and ACES RRT (sRGB). This is what you will be grading on. Go to your first S-Log2 clip and add the Sony Levels plugin. Punch in gamma equal to 2.2, drag the slider for output end to 0.45, (the inverse of 2.2 gamma), and start your grade. For a rec 709 display, you'll be 95% there already. For your waveform display settings, make sure the box for 16-235 is unchecked.

For RRT (Reference Rendering Transform)
In this stage you select what the target color space is for rendering. Depending on the target color space you will lose information (ACES RRT 709). You choose this from the rendering template, customize template, project tab, color space.

So 1,2,3, done. That's a pretty good summary on how to work with ACES in Vegas. That should get rid of the noise because the color space is being remapped, rather than having the noise amplified by mathematical computations, truncations and low numerical precision issues. It's also going to show a full color gamut.

ACES – Academy Color Encoding Specification | Niwa (http://www.niwa.nu/2013/05/aces-academy-color-encoding-specification/)

I haven't found any function like Interval Record in the menus. Is there a way to get this to shoot time lapse in 4K?

Piotr, when you select an 8 bit color space (ACES RRT Rec. 709) from the rendering template, you are not telling XAVC to not be a 10 bit codec, but the render will only have 8-bit precision. The rec. 709 output will be mapped into the 10 bit space, so 16-235 becomes 64-940. And if you render a 10 bit number with 8 bit precision, it can only be as good as 8 bit, not better, thus the reason for using full range 32 bit floating point in your project properties. You only need to do that just before rendering, at other times during preview and edit you could use 8 bit, but would not be able to use the ACES RRT (sRGB) input transformation for viewing.

Good review of the Z100 by Alister...
A Tale of Two Cameras – The PMW-300 and PXW-Z100 reviewed. | XDCAM-USER.COM (http://www.xdcam-user.com/2013/12/a-tale-of-two-cameras-the-pmw-300-and-pxw-z100-reviewed/)

I knew this camera wasn't the best in low light, but my personal excitement over this camera took a big hit after seeing the pics of him in front of the Christmas tree and it's subsequent noise performance hit just to get it up to the PMW-300's 0db sensitivity. Bah humbug! Not many other options with 1080p60 and 3GHDSDI out which is what we need, and while we don't need "the best" low light cam on the block, this is somewhat brutal. Bummer.

Yes, it IS a bummer.

I had high hopes for the Z100, especially after I found out the XAVC can actually be edited on my current hardware (+ 10 bit IPS monitor), and with my current software (thanks again Tom for realizing me the potential of Sony Vegas Pro 12 I never used before).

But after Alister's hands-on review, it seems the sensitivity and S/N of this camera is more or less equal to that of my old good EX1 without any gain capabilities - if the Z100 picture is only tolerable up to 12 dB, and at that being only as bright as the EX1 (or 300) at 0 dB :(

Alister - thanks for the report nevertheless; after all you are just a messenger...But the message to me is that I'm not ready for 4k just yet; I cannot tolerate low S/N and above all the noise of the Z100, and the FS700/Odyssey combo I just cannot afford.

One should not be surprised with the performance. The sensor is cropped about 1/3 " from the full sensor to match the lens which came from the AX2000/ NX5U, it has more than 4 times the sensors of these AVCHD cameras as they have less than 1920x1080 pixel sources and interpolate to 1920x1080. A better comparison is to the small single chip Sony's like my CX700 or NX30U and it still has 4 times as many pixel sources as these and they too are back illuminated sensors. So the FDR-AX1 and PXW-Z100 are not likely to perform as well as these single chip Sony's and they do not!! In testing my FDR-AX1 ( which I think has the same sensor as in my Cybershot HX30V ) noise performance is just like this HX30V in comparison to the CX700 and NX30U 's is about a stop slower than these small Sony's. The FDR-AX1 is at best a stop slower than the NX5U at around 0 db but gets to be over 2 stops slower as it gets dark. The NX5U is about a stop slower than the EX1/3 so there is no way that the FDR-AX1 or PXW-Z100 is close to the performance of an EX range camera and will be at least 2 stops slower. If you want a camera with a 20x zoom that doesn't cost a fortune then about 1/3" is the sweet spot and there is likely no way around the issues with present technology. The RX10 gives me a view of where things may go though.

What I have found though is that shooting in HD mode is not as good as shooting in 4K and downconverting to HD with software. Noticable difference in my tests so I will not be using the HD setting. I used Edius Pro7 and used Lancsoz 3 for downconversion. Even at 21db it will provide a nice image better than the CX700 or the NX30 for the same scene which they would be at between 15db and 18db. ( dark stage set with few actors ) I had hoped to use the FDR-AX1 in 4K mode so that I could later crop in to HD but for dark scenes the noise then is too much. In good light this works and for minor cropping works great. For my uses as an unattended full stage camera it works but will look forward to a larger sensor version in the future. Even though it does not have the fancy iAuto modes of the consumer cameras just using AE shift at -1EV it has worked well on the 3 shows I have tested it with. All dark or flashing lights with most of the time it hovering at 21db gain but with the camera fixed and not moving of course !!! Maybe a camera with the sensor/lens from the RX10 with 10 or 12x zoom !!!

Ron Evans

If you want a camera with a 20x zoom that doesn't cost a fortune then about 1/3" is the sweet spot and there is likely no way around the issues with present technology.

There's the new Sony HXR-NX3 with three larger 1/2.8" sensors, 20x zoom, and IMO, an excellent price at $3,500.

Yes but its not 4K its really the replacement for the AX2000/NX5U with new sensors and different recording options like AVCHD 2.0 and just memory cards no FMU128 option. etc Nice camera but not enough for me to change my NX5U.

Ron Evans

Just a general question ... when is this camera going to get its own sub forum, now that units are shipping ?

But a more specific question to current owners ....

[1] Is this camera worth buying as a replacement and upgrade to the NX5U? i.e Are you super happy with your purchase ? ** The samples on the net have underwhelmed me but again, 4K is not meant for a HD screen or YT.

[2] Is this camera just a waypoint to a better option 4K soon (maybe another brand), hence making this a poor investment. Any guesses ?

I am looking to possibly pick one up in the USA when I pass through in January, because what I want is better resolution, better data rate, better colour sampling (without a Ninja) in the same R&G form factor and 20X lens as my NX5U. Just looking for some comfort before dropping 6K.

[3] Which retailer currently has best price ? USA or Canada

Thanks in advance

Justin.

The Z100 will not be as good in low light. The pixels are much smaller compared to the NX5U. We are a DVINFO sponsor, have the camera in stock and have some footage from the media card that you can download and import into your edit system:

Omega - Sales - PXW-Z100 (http://omegabroadcast.com/fmp/omega-detail.php?-manuRecID=firstpage&-recid=4326)

I have the FDR-AX1 and an NX5U and I do not think either of these 4K cameras are a replacement for the NX5U for a number of reasons. They are maybe 2 stops slower so if you work in a low light environment they will not be good for you as your main camera. Although switch layout is much same as the NX5U they do not have smooth gain switching. Expected on the FDR-AX1 like the AX2000 and NX5U differences I expected this feature to be on the PXW-Z100 but I cannot find in the manual so a major omission if it really isn't there. In good light in a semi auto mode the FDR-AX1 produces a beautiful image much better than the NX5U. PXW-Z100 with 10bit XAVC codec should have an even better image but will use memory at a fast rate !!! Just not viable for long form recording which is why I got the FDR-AX1.

Edited in Edius and downconverted to 1920x1080 with Lanczos3 the images are lovely even in low light and are better than the CX700/NX30U for example in similar light conditions with just a little more noise since it is about a stop slower than these cameras. Downconverted this way there is definitely more resolution in the image compared to the NX5U or the CX700/NX30U I use though .

I will wait to replace my NX5U until there is a larger sensor model to give better low light. Dream is a camcorder with the 1" sensor in the RX10. Not too big so that I can get good larger depth of field but big enough to have good low light performance.

Looking at examples on the WEB may not do justice to these camera as one has no idea how the video was prepared for the site. I find also that in camera downconvert is not as good as software downconvert using Lanczos3, another parameter that may degrade the images one sees on the internet.

Ron Evans

I almost see this camera as a direct descendent of the Sony Z1 in a lot of ways.

When the Z1 dropped, we were used to the low light performance of the PD150/170 and the Z1 certainly was a light pig compared to those. The HD footage was decent - not great, but decent - but where it shone was downcoverted footage to SD.

I shot two international docs for broadcast on the Z1 - one in Germany and one on the Tamil Nadu coast of India. I LOVE the way the footage looks in SD. The HD? Well...

Then not all THAT long after, the EX1 rears it's head. We initially didn't embrace it, despite the leap forward in image quality and low light performance, due to the media cost.

I think we are again at THAT time in camera development where everyone is scurrying around to produce the best 4k camera they can at a price point that will actually get folks to spring for them, knowing full well that whatever decision they make will be obsolete in a year. Because frankly if 4k doesn't catch on and sell some units, there is no impetus to make cameras.

The answer to the age old question of "Do I buy?" is "Can I make money on it RIGHT NOW?"

I'm in the market for a 4k solution right now. I prefer the look of the footage coming out of the JVC but don't want to be tied to 4 SD cards mainly due to the fact I'm a run-and-gun videographer, not a filmmaker so media management scares me a little around multiple cards as a solo shooter. The Z100 footage seems underwhelming. The F55 would be an option if I had that sort of cash. I don't want a RAW workflow so RED is out...

It is a complicated market for sure. Some WANT RAW, some want an easily editable but decent codec. At the same time, everyone wants long run times and cheap media.

I'm just thankful I remember the days of BetaSP 30 minute loads at $30 per tape. Keeps things in perspective for me.

Manufacturers skipped 1080p60. Why ? The jump from 480i60 to 720p60 was enough to sell HD over NTSC/DVD since just about every cable TV channel is 720p60, including the best-looking channels such as NatGeo. Why wouldn't it be the same with 1080p60 on larger 75+ " displays ?

1080p60 wouldn't be any more of a light hog than 720p60 since most full-raster 1920x1080 sensors are down sampled to 720p60.

It's more like broadcast standards skipped 1080p60. They're going straight to 4K60 or 4K120.

If they do choose 4K120 for broadcast, the F5 and F55 are already obsolete.

4K120 WILL NOT happen as a broadcast standard.

No chance.

8k will happen first.

RE: At the same time, everyone wants long run times and cheap media.

Agree Shaun. The most concerning issue with the Z100 right now is the record time of 4K onto the QXD cards. For quite expensive media, the record time is very short; too short to be practical.

I was not aware of the JVC at the time of posting my original questions, but having looked up the specifications I can see their 4 SDXC cards is an interesting approach and offers better record times, but buying 4 of those cards at a decent size starts to add up too, especially if you factor in the breakage rate of those cards is high. (the read/write switch has always been a weak point).
** The 10x optical zoom is also a deal breaker, with the JVC as I manly shoot outdoors.

The Z100 will get a long GOP option if Sony keep their promise to release a firmware update (the Nx5U did not get much post release love...); but by Q2 there is likely to be other 4K options around.

Agreed. Some people proposed 120 Hz as some new, unifying frame rate. I don't know what they were thinking. 120 Hz isn't a multiple of 50 and it's too fast for shooting under fluorescent lights in 50 Hz countries. A 1/100 shutter and 100 Hz frame rate is as fast as you can go under those conditions. 300 Hz could be a unifying standard for 50 and 60 Hz video systems but that's too many bits and too much of a lighting demand for too little gain.

Certainly, people are looking at fast frame rates for sports, but there are practical issues to be considered. Higher resolution is much easier to solve.

The more immediate question is whether we should do away with the 1/1.001 frame rates (23.976, 29.97...) and go to integer rates (24, 30...) Sounds good on paper but how do you intercut legacy content while avoiding all potential glitches and errors?

I'm thinking it would be digital cinema that would use 120 first to divide to future HFR frame rates. Broadcast would still be 4K60p.

The math would be in the neighbourhood of 8x the potential bandwidth (before considering increased efficiencies due to duplicate pixels or suppressible detail and whatever is gained by h.265) in a jump from a maximum of 1080P30 to UHD/4kP60...

Not for a while...

I'm really not convinced that we will see 4k 60p any time soon. We could right now convert much more easily from current 25/30fps HD to 50/60fps HD, but it's not happening and I don't think it ever will.

In my opinion and from what I'm hearing from broadcasters etc there will be an increase in the use and distribution of 4K at 24/25/30fps first. Later there may be some higher frame rate usage, but broadcasters and production companies etc are pretty scared of the massive files that this would involve for production and distribution. Heck most of them are scared of 24p 4K.

HEVC looks like it will offer similar bandwidth for 4K as for current HD. Higher frame rates means more bandwidth and where that would come from for traditional broadcasting no-one knows. Web distribution is a possibility.

4K at 24, 25 or 30 will start on the web first, as Netflix is experimenting with that at the moment. Web is not the domain for 50p or 60p, so that's where everything's gonna start.

I will say that the F5, F55 and even this camera with 50p and 60p lays the foundation for that for broadcast. I do think H.265/HEVC will be in the next broadcast standard to deliver 50p and 60p.

I really do not like the slow frame rates so we will miss the nice smooth motion of interlace if we cannot get 60P. Personally I would prefer 60 interlace to 30P. I shot a few clips at 30P on my FDR-AX1 for a piece for final output to Youtube but nothing was really moving so it was fine !!!

Ron Evans

I too hope 50p and 60p come, and I bet when proper HDMI 2.0 comes that it will be possible, just not at 4:4:4.

For broadcast, I don't think you are going to see 4:2:2 4k right out of the gate either... any bits they can throw away without the end user complaining will be discarded... 4:2:0 at 24/25/30 is my early prediction with 60 as a possibility...

I understand Ron's want for 60 frames as an option for motion smoothness but I think we have seen the last of interlacing... interlacing was designed for the hardware of the time. Current panels are exclusively progressive scan.

Yes... I PERSONALLY want UHD at 60 fps at 4:2:2 but if that is IP delivery and clogs the entire internet... well...

I'm quite prepared to be realistic and pragmatic... roll out what you can as you can and build infrastructure for scalable roll-out as technology and infrastructure allows.

Heck, even downsampled 4k/6k to 2k/HD looks better than most HD sourced material...

My hopes for 4:4:4 4K 60p are more for computer monitors than broadcast, but for sure we are using 4:2:0 for any type of delivery in the near future.

However, the expectations of Rec. 2020 and Deep Color can't be met with 8bit 4:2:0. Closest thing is the F55 or F65 to capture color close to the Rec. 2020 space.

People are increasingly watching content on mobile... the set-top box in the living room is on life support.

The iPod generation wants to CARRY their content, not watch it in stunning quality.

DVD Audio never caught on because MP3s meant you could carry your entire catalog of music on something SMALLER than a deck of playing cards...

Video streaming is the next logical step.

Not saying I LIKE that... just saying it is a reality that is working AGAINST us ever seeing 8kp240 4:4:4 at 12 bits on-demand in stunning quality...