I fully agree that for shooting, 4K makes the most sense, as you say simple crop to UHD. I was just looking at distribution and how 4K/UHD will creep in to homes in the coming years.

All the same arguments came up in the mid 1990's when HD was on the horizon, yet in Europe it took less than 2 years between the launch of the Z1 and the arrival of full time HD broadcasting (via satellite TV) and a further 2 years before the majority of main stream programme commissions were

I haven't made arguments against UHD, I'm all in 100% for 4k acquisition.

But I am expressing the reasons why the uptake on the consumer end will be protracted and incomplete just as HD is, only this time the consumer doesn't have a federally mandated analog to digital switchover, isn't switching from heavy curved glass crt and rear projection sets, and will need larger panels or closer viewing distances to see benefits in the home.

Even Netflix have stated they expect the uptake from the subscriber base to be small in the near term, thus only a negligible affect on their bandwidth.

As these panels are shipping now with less desirable LCD display quality compared to OLED and plasma, are shipping without HDMI 2.0 which is necessary for 2160/p60, and are marketed that all your upscaled content will look better in 2160p, we really have seen that before.

To argue against the inevitability of the future is likened to "whistling past the graveyard." Definitely I'm not doing that here, but some skepticism is justified.

Let's see what happens, in the meantime, have a beer!

I decided on the AX1 and have had it for a few days now from Crutchfield. I posted in the AX1 thread that I am having trouble rendering good quality blurays. Could someone please visit that thread and post some suggestions?

Thank you in advance,

James

Now in stock at B&H.

I would suggest that the NX5 and Z100 sensors have similar sensitivity. The NX5 has bigger pixels and uses a prism while the Z100 although it has much smaller pixels see's an approx 8db sensitivity boost per pixel because of the back illumination. Overall they should come out very similar.

However the Z100 has more sophisticated image processing. When I tested the Z100 alongside a PMW-200 I found the Z100 to be 1.5 stops less sensitive than the 200 at 0db, but... and this was the big surprise the Z100 was showing less visible noise (or at least the noise was less apparent). Raising the gain on the Z100 quickly results in more noise and some image softening, especially above +9db. So I think you'll see similar sensitivity to the Z5 at 0db with possibly a cleaner looking image from the Z100. But if you start adding in lots of gain the Z100 performance will start to fall away quite quickly.

Any ideas of how the Z100 compares to a non-sony camera like the HMC150?

Basically, I'm curious to see any updated info on this camera and how it compares to a couple other cams with regards to sensitivity (HMC150 & AX1 for reference). It's now officially out and it seemed like most of these statements I'm reading were done on pre-production models. I'm currently rocking a couple HMC150's for live event shooting and it's time to move on. I am wondering how the Z100 will compare with regards to sensitivity to the HMC150. I never really push the HMC's beyond +9db of gain (maybe 12db if absolutely necessary). With the BSI sensor on this camera, and Alister Chapman's note above that it gives a camera about an 8db boost, I'm curious how it stacks up against the older HMCs in this regard. For example, at 0db, which is more sensitive? And when an HMC150 is at +9db/+12db...what does one have to set the Z100 at to match that? And how is the noise at that level of gain?

I know the initial reports are that the AX1 and Z100 aren't the best in low light compared to the PMW-200, etc. but I need a camera with 60fps and a 3G-HDSDI out which leaves me with only the Z100 and the FS700 at this point. Does the Z100 fare any better at all than the AX1? Or is it the same? I'm hoping to get some updated info now that it appears both of these cams are available.

In terms of sensitivity AX1 and Z100 are the same. I can't comment on comparison to cameras from other brands. I have used a production Z100 and it was no different to the pre-production units.

Alister did you play around with S&Q recording when you had the camera? The settings only seem to be available in HD mode, at higher resolution no S&Q. Is that correct? Thanks-

S&Q only works in HD, but you can shoot at 50 or 60p in 4K.

In terms of sensitivity AX1 and Z100 are the same. I can't comment on comparison to cameras from other brands. I have used a production Z100 and it was no different to the pre-production units.

Thanks Alister. Does the Z100 have the focus distance readout in ft/m that I've found infinitely handy on other cameras?

"Thanks Alister. Does the Z100 have the focus distance readout in ft/m that I've found infinitely handy on other cameras?"

It does.

"Thanks Alister. Does the Z100 have the focus distance readout in ft/m that I've found infinitely handy on other cameras?"

It does.

Sweet!

Thanks Mark!

Hopefully now we'll start to see some test/comparison footage from this camera. Wish it had XAVC-S out of the box...oh well.

I had to send mine back to the dealer because there was so much noise in the image. I requested an exchange rather then a straight return because I figure it's got to be defective. If the next one is noisy as well I'll do a return.

Does the Z100 fare any better at all than the AX1? Or is it the same?

Based on my playtime with the Z100 and a colleague's experience with the AX1, note that the Z100 is recording 10 bit 4:2:2 Intra, the AX1 is recording 8 bit 4:2:0 Long-GOP. I'm becoming a bit of a 10 bit convert as a result.

One would expect a difference between 150Mbps LongGOP and 600Mbps 10 bit 4:2:2 !!! As far as noise from the sensor is concerned is there really a difference though. Because of the ramping on the zoom even in good light my AX1 will have a reasonable amount of gain compared to the NX5. Doesn't take much to make it go to 12db and at 15db one starts to see some obvious grain though not too unpleasant. I would really like to know if the HDMI on the AX1 is 10bit so that an external recorder could match or exceed the Z100.

Ron Evans

One would expect a difference between 150Mbps LongGOP and 600Mbps 10 bit 4:2:2 !!!
Not necessarily. There is no reason longGOP can't be 4:2:2 and 10 bit, and in terms of general compression quality, then you can expect a longGOP system to be more efficient in terms of quality v bitrate than an I-frame only one.

Downside is increased processing power needed to compress and process - upside is the same quality at much lower bitrates (all else equal).

And 600Mbs is what XAVC uses in I-frame mode for 4K 60p. This codec allocates a given amount of data per frame, so increase the frame rate and the bitrate scales up proportionally. This is not true for longGOP codecs whose relative efficiency increases substantially with higher frame rates and progressive mode. That's because there is less change between each frame at 60p than 30p - less time has elapsed between them. It's reasonable to keep the same TIME interval (typically 1/2 sec) between I-frames, so going from 30p to 60p means the same no of I-frames per second, but many more difference frames. Since the I-frames are far larger in terms of data than difference frames, that's why a doubling of frame rate means a lot less than doubling data rate.

There's talk of moving to even higher frame rates, 120fps or even higher. If that comes to pass, expect to hear much, much more about long-GOP systems! :-)

I understand David, but the FDR-AX1 is 8 bit 4:2:0 LongGOP, 60P at 150Mbps and the PXW-Z100 at the same 60P is 600Mbps 10 bit 4:2:2. I would expect a difference between these two. I believe next years update to the firmware will bring LongGOP to the PXW-Z100 and your comments will be absolutely correct.

Ron Evans

I understand David, but the FDR-AX1 is 8 bit 4:2:0 LongGOP, 60P at 150Mbps and the PXW-Z100 at the same 60P is 600Mbps 10 bit 4:2:2. I would expect a difference between these two. I believe next years update to the firmware will bring LongGOP to the PXW-Z100 and your comments will be absolutely correct.
OK. The only thing I'd say is that remember for 4K, chroma resolution at 4:2:0 will be 1920x1080 - full HD luminance, in other words. Since there is debate about "will we actually notice the difference between 4K and 1080?", then will 4:2:2 at 4K really make much difference compared to 4:2:0.....?

Colour subsampling may have been pretty important in the days of SD and interlace, I'd argue much less so with progressive and HD, and less still in 4K days.

Of course, you may still argue that if you're getting a 4K camera primarily for HD - but with a cropping ability - you need the higher chroma/luminance ratios in the crop......?

Alister Chapman took the PXW-Z100 for a spin while in town for our workshop. The card had 25GB of footage but I deleted some of the files and left eight clips. I zipped it up and put it on media fire for you to download and try out in your own system. The video is of downtown Austin. Last weekend was winter. Yes. Winter only lasts one weekend here. This is it. Go to this page and scroll down. Download time depends on how fast your connection is. This file is 2.5GB.

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

I posted this in a different thread but saw there was a lot of conversation here too.

Holy moly, Thank youuuuuu!

Not necessarily. There is no reason longGOP can't be 4:2:2 and 10 bit, and in terms of general compression quality, then you can expect a longGOP system to be more efficient in terms of quality v bitrate than an I-frame only one.

Only thing is the spec for XAVC-S is currently limited to 4:2:0 and 8 bit. If there is 4:2:2 10bit Long-GOP, it almost certainly cannot be in the XAVC-S spec. It would have to be a new XAVC-LongG spec.

I must have missed the info but is there a plugin I would use to view the sample PXW-Z100 clips using Premiere Pro CS6. I'm only use to editing FS700 (AVCHD) files and the DVCAM types files. So I'm not sure where to find the 4K XAVC codec for the PXW-Z100 in PP CS6.

Rovi produce a rather expensive plugin for XAVC in Premiere CS6 (PC Only). There is a watermarked demo version.
TotalCode for Adobe Premiere (http://www.mainconcept.com/fileadmin/newsletter/individual_letter/Sony-TCAP_v5.html)
Otherwise you will need Premiere CC which has XAVC support built in.

I really quite like the Z100. The colorimetry is very good and noise levels remarkable low when the light is reasonable. After shooting with S35mm cameras almost exclusively for the past couple of years it was really nice to get back to a camera with a 20x zoom. It isn't as sensitive as the new PMW-300 or an EX1 etc, but that is to be expected, it's about a stop less sensitive. In addition it doesn't have the dynamic range of an EX1, I would estimate 10 stops. For this reason the Z100 does not have Cinegammas.Like the PMW-300 it's nicely built and feels very solid in the hand. My biggest gripe is that the focus and iris response in manual is a little sluggish which is annoying and takes a little getting used to. I found it harder to precisely focus the Z100 than my F5 with a DSLR zoom as the deep DoF means that it doesn't snap in and out of focus as a camera with a shallower DoF does. The focus tools are OK with coloured peaking and image magnification which is just as well as you really do need both to ensure your 4K is sharp. I did end up with a couple of shots very slightly out of focus. With a bit more use I'm sure I could improve my focus technique with this camera.

It was quite shocking to fill a 32GB XQD card in just 15 mins when shooting at 4K 24p, although I'm kind of used to that with the F55 in 4K.

I have both a PMW-300 and a Z100 for a week and I'll be reviewing both in some depth as well as creating some picture profiles over Christmas.

Tested the latest version of VLC and it can play XAVC from the MXF file.

At 100% crop, I'm still not convinced about the noise levels, and the codec brings out the noise that is hidden by more compressed codecs. Specifically I looked at Clip0031.MXF and I'm very impressed by how the codec held up, but am very unimpressed by the grain.

But VLC's XAVC playback is not optimised. It does not include the DPPS processing. Dynamic Picture Parameter Set, which alters the interframe encoding profile and decoding algorithm based on the scene and meta data stored within the stream. VLC uses a generic AVC decoder without interfame variable decode parameters, which is unique to XAVC and makes quite a difference to the image quality.

Rovi produce a rather expensive plugin for XAVC in Premiere CS6 (PC Only). There is a watermarked demo version.
TotalCode for Adobe Premiere (http://www.mainconcept.com/fileadmin/newsletter/individual_letter/Sony-TCAP_v5.html)
Otherwise you will need Premiere CC which has XAVC support built in.

I have both a PMW-300 and a Z100 for a week and I'll be reviewing both in some depth as well as creating some picture profiles over Christmas.
First of all thanks for the pointer to the TotalCode plugin for PP CS6. I did in fact visit the TotalCode website and went directly to their purchase plugin page and noticed their blurb about supported codecs didn't mention XAVC so I downloaded and read their TotalCode fact sheet but it also didn't mention support for the XAVC.

However that could have been an oversight so to make sure I wrote them directly asking for an updated spec sheet for the PP CS6 TotalCode plugin that does state it has support for the 4K XAVC.

After reading your very informative article on the merits of using FS700 with the Sony AXS-R5 recorder I'm thinking that might be a better route for me to go for right now. I would just purchase the HXR-IFR5/AXS-R5 RAW combo and use that till the new Sony 4K style run and gun sized 4K cameras mature a little bit more.

However the PMW-300 looks like an awesome camera so I can't wait to see what you say about that unit. Perfect 1080p run and gun. I still want to be able to drop 4K XAVC files from the new Sony 4K handycams into PP CS6 so I will continue to pursue getting a solution there.

PS Alister if your ever in South Florida and need access to a Sony F65 for one of your various tests and comparisons I have a fully outfitted F65 available.

Douglas: Given the choice between an FS700 + R5 or Z100 I would go for the FS700 every time. It's a much more capable system, even though the ergonomics are less than desirable. Where the Z100 fits for me is as a companion camera to a large sensor camera for pick-up shots where you really need that 20x zoom and a deeper DoF.
For example I'm looking at the Z100 for my storm chasing shoots to act as a grab camera for the times where I only have seconds to get the shot. It would not replace my F5, just work along side it.

Douglas: Given the choice between an FS700 + R5 or Z100 I would go for the FS700 every time. It's a much more capable system, even though the ergonomics are less than desirable. Where the Z100 fits for me is as a companion camera to a large sensor camera for pick-up shots where you really need that 20x zoom and a deeper DoF.
For example I'm looking at the Z100 for my storm chasing shoots to act as a grab camera for the times where I only have seconds to get the shot. It would not replace my F5, just work along side it.
That sounds like a great plan!

Total Code is somewhat flakey with XAVC, so I wouldn't get too excited by it.

Total Code is somewhat flakey with XAVC, so I wouldn't get too excited by it.

It looks like it costs about 10 months of Adobe CC access. Doesn't seem worth it if CC supports XAVC.

Any place to download an Z100 I-frame 50p (or at least 24/25p) 4k clip, to give it a run in Vegas Pro 12? Or PP CC?

Any place to download an FS100 I-frame 50p (or at least 24/25p) 4k clip, to give it a run in Vegas Pro 12? Or PP CC?

Assuming you mean the Z100, Phil Goetz posted some clips shot by Alister Chapman earlier in this thread:

http://www.dvinfo.net/forum/digital-video-industry-news/518710-sony-launches-pxw-z100-4k-handheld-xdcam-18.html#post1824317

Yep, sorry for the typo - corrected.

Unfortunately the file is not available under that link any more :(

Yes it is. I just checked. :)

Scroll to the bottom of the page and click on the block of text under the words "Sample Footage".

Thanks Michael. Just downloaded using IE; my Chrome somehow couldn't see the archive.

The good news for me as Vegas Pro user is that those mxf play well enough on my PC to be edited! I tried the same with the AX1 sample (the goose), and it wasn't possible (that could only be played back using the Mercury engine in PP CS, which I'm only trying out now).

PS. What may be the reason those mxfs crash my VLC? I updated it to the most recent version recently, and it plays fine with my 220 Mbps I-frame nandFlash files... Any special settings required?

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.

No. Scaled to 1080 it looks really nice, and very sharp, but at 4K it bothers me too.

The good news for me as Vegas Pro user is that those mxf play well enough on my PC to be edited! I tried the same with the AX1 sample (the goose), and it wasn't possible (that could only be played back using the Mercury engine in PP CS, which I'm only trying out now).


Why Is the GPU acceleration implementation so poor in VP12?!!

While the I-frame is playable (and just barely editable) in Vegas, the Long GOP version is crawling at 1-5 fps with GPU load at 19% max. The same clip plays full speed in PP CC, with the same Quadro 4000M loaded at up to 100% most of the time!

Is the PP's Mercury engine patented, or what?!!

Piotr

No. Scaled to 1080 it looks really nice, and very sharp, but at 4K it bothers me too.

This is a disaster for both the Z100 and AX1, because to see all the details in 4k one must watch from a very short distance - and this is when the noise roars its ugly head...

That said, it's not as bad as in the EX1 at HD....

This is a disaster for both the Z100 and AX1, because to see all the details in 4k one must watch from a very short distance - and this is when the noise roars its ugly head...

That said, it's not as bad as in the EX1 at HD....

I always thought the EX1's noise was at least acceptable, and if you really didn't like it, shoot -3dB in bright conditions. I will admit though that the footage scaled down from these cameras actually does look perceptually cleaner than the EX1 at 0db.

However, using a consumer sensor that is basically a cell phone sensor (The Xperia Z1 is the exact same sensor) will have major drawbacks such as this. What we need next are native raster sensors (4096x2160 preferably, and not limited to 16x9) and super high resolution prisms. IMHO, bayer and supersampling only works on S35 and above.

I think the sensor in my AX1 is the same as the one in my HX30V. Characteristics are similar for low light. I just did a shoot with my NX5U, the AX1 , CX700 and my wife used the NX30U. It was a school dance show so the lights were flashing all the time, dark, red etc etc a real challenge and I usually leave even my NX5U in auto exposure with spotlight setting. The CX700 was set AE shift -1.0ev as was the NX30U and also the AX1. There were no max gain settings for the CX700 or NX30 but I set the AX1 to max 21db which it stayed at more most of the show !!!! To my surprise looking at the data code for the NX30 and CX700 afterwards they too were at 21db most of the time with occasional 24db !! Scaled to 1920x1080 the AX1 had a better image than the CX700 with no more noise that I could see but the interesting effect of the increased resolution was the faces were never blown out by the lights. None of my other shows will be this dark so I am happy !! Not sure if it will be as good as I want for cropping as this brings out the grain too much at these levels. In good light it works fine.

As to my shots of the Canada geese I can get it to play close to real time in Edius Pro 7 and will play in Vegas Pro 12 if preview is set to draft good. Good enough to cut by I think. My PC is an i7 3770 running WIN 7, 64bit with 16G RAM, lots of hard drives.

The AX1 is not a replacement for my NX5U.

Ron Evans

I think your viewing methods are suspect to say the least. Of course the noise will be more noticeable when scaled 100%. Just remember when making a comparison to an HD camera that assuming you have an HD monitor when you scale the 4K to 100% you are looking at an image 4x larger than the equivalent HD, so I would expect to see noise similar to the equivalent HD camera.

When viewed on a large 4K TV, the footage looks very good and noise levels are perfectly acceptable at 0db and +6db. As well as Austin, we had one in Toronto for the Vistex 4K event. This was feeding a 65" 4K Bravia and looked very nice, even close up and most people felt the images looked very clean. It's only when you go above +6db that it really starts to suffer as the noise reduction introduces a lot of extra smear and smoothing that tends to soften and blur the image.

Ron: Resolution won't make the difference between faces blowing out or not. That's dynamic range.

Zack: Who says this is the same sensor as the Experia Z1? I doubt that very much, if you've seen the pictures from the Z1 you'll know the Z100 images are far superior, much better dynamic range. 4096x2160 is not 16:9 it's 17:9. But of course any sensor with a similar pixel size will have largely similar performance, that's the laws of optics and physics at work. I wouldn't expect miracles from a micro sized prism at 4K, the added diffraction and dispersion effects will restrict the useable aperture range and negate any benefits over a bayer sensor. In addition we are already pushing the limits of current lens technology with these high resolution small sensors. Larger sensors and bigger pixels are still the way to go.

...we are already pushing the limits of current lens technology with these high resolution small sensors. Larger sensors and bigger pixels are still the way to go.

Which is why I think a 4K EA50 or FS700R replacement is coming down the line. APS-C may be a good compromise for people unable to access Super35.

[QUOTE=Alister Chapman;1824559
Ron: Resolution won't make the difference between faces blowing out or not. That's dynamic range.

[/QUOTE]

What ever the reason it works for me as an unattended camera running in semi auto even at high gain settings.

I also agree with Jack that there will be a larger sensor camcorder some time next year. Maybe using the RX10 sensor which would be great and maybe then a replacement for my NX5U. The NX3 is nice but not a big enough difference for me to change my NX5U. I want the longer zoom with good low light so an APC sensor may then require a very large lens with too shallow depth of field for my uses.
Ron Evans

Could someone help me with what is the right way for editing the XAVC clips in PP CC? As per the screen grab below (coming from one of the on-line tutorials), it should be there to chose but - unlike the other listed - I cannot find it anywhere. So I'm using the RED 4k format for my sequences instead, but is there a better way?

Disclaimer: I am a complete newbie to PP; always used Vegas Pro :)

PS. Perhaps the 30-day CC trial has some limitations?

I think your viewing methods are suspect to say the least. Of course the noise will be more noticeable when scaled 100%. Just remember when making a comparison to an HD camera that assuming you have an HD monitor when you scale the 4K to 100% you are looking at an image 4x larger than the equivalent HD, so I would expect to see noise similar to the equivalent HD camera.



Alister,

If you are talking to me, please let me explain: using Vegas with those mxf XAVC clips, I can use the option "Scale output to fit the monitor" and - on my 50" plasma hanging some 1m in front and above me - I'm then watching gorgeous HD material. When I uncheck this option, the plasma shows an almost pixel-to-pixel image and there is a lot of black splotches in the picture, which I call noise.

Similarly, using Premiere Pro CC, if I set the scale of Export format to 100 (with Export Preview set to HD, as I can't set it to anything higher), I'm getting quite nice HD. Blowing it up 4x (400%) I can see the same noise as in VP, and this is the real clip resolution!

What am I missing?

There's no preset for XAVC, but drop an XAVC clip into an empty timeline and it will adjust the parameters to match.

If your using a 50" monitor and you scale 1 to 1 pixels then you are viewing the equivalent of a quarter of the equivalent to a 100" monitor which really is a very big screen. You wouldn't normally watch a 100" screen from just 1m away. It's like watching your 50" screen from just 25cm and I'm quite sure that that close you will clearly see any grain or other artefacts. It also assumes that your software/computer can playback full frame 4K and take a crop and process this in to an HD stream in real time without compromising the image quality.

You're spot on, Alister - I tried to imagine 4x the area of my 50" plasma and came to the same conclusion; were it 100" and UHD, I would back away by another meter or two, which I did. Unfortunately, the black noise was still visible (in both PP and VP)...

Oh, I forgot about the XAVC preset: you're right the PP sequence will adapt, but somehow I cannot find a way to export 4k from PP (neither to a preview monitor, nor to rendered-out media... Please give me a hint as PP newbie :)

There's no preset for XAVC, but drop an XAVC clip into an empty timeline and it will adjust the parameters to match.

Yeah, that's what I did in PPCC and it worked fine.

Alister, excited for your review of this camera. I have to ask, working with Sony, did you know the NX3 was coming out long before it arrived? If so, do you have any word on any other sony large(r) sensor cameras (eg. 1 inch or above) under $10k coming in early-mid 2014? Perhaps you are not allowed to say but I had to ask.

No, I did not know about the NX3, but NXCAM is not a line I tend to be all that involved in. I have not heard any rumours of any more large sensor cameras, but that does not mean anything other than I don't know and if I did know I wouldn't be allowed to say.

Thanks for your response Alister. Are you saying that when I scale a 4K clip to 100% on my HD monitor that it's equivalent to scaling an HD clip to 400%? Trying to get clear on this… Remember that 4K is brand new territory for a lot of us.