Sony X70 4K - Lowest bit rate in the industry!

[Cliff Totten]

Look, If Jody's position is that Sony's XAVC offers no quality advantages above 60Mbp/s than so be it. What can I say to that? It's his opinion. He has his own reasons.

Sony designed many XAVC modes that go WAY up the ladder in bit rate. (With 60mbp/s at the very bottom) Sony markets them as higher and higher in quality as they go up the list. They also charge more for each mode as it goes higher into the model lineup. It appears that 100Mbp/s is the new "basement" or "low standard" for 4k/UHD cameras as low as $499 and higher. I'll bet allot of money that Sony will never make another 4k camera at ANY price that has a maximum bit rate below 100Mbp/s.(Especially when their direct competition offers 100-150Mbp/s) Yeah,..I know...absolute crazy talk,..right?

I dont think that Sony actually realizes how good their 60Mbp/s is?? They certainly don't know that it's just as good as their 100Mbp/s. Stupid Sony, I guess.

Me? I agree with Sony engineers and XAVC designers. The science behind higher and higher bit rates in MY opinion is on Sony's side. Sony built it, they know XAVC best...in my opinion. (But maybe Jody and his people know things that Sony doesnt?..possibly)

I'm also certain the Motion Picture Experts Group will also agree that higher h.264 bit rates exhibit less and less artifacts and capture motion more accurately. This is why they increased the level restrictions over all the years for h.264. I do not believe they did this because it does nothing for quality. But hey,...again. More radical, crazy talk on my part I guess.

For me,..id rather shoot at the same bit rate that their "cheap and REALLY cheap" 4k models now shoot in today. (No I don't want the AX300's sensor,..I just want it's h.264 bit rate) Is that a terrible thing to want? Ridicules and horrible of me? What an absurd proposal on may part! How dare I ask this of an XDCAM camcorder?

I don't doubt Jody or his associates but If you polled colorists around the world, the general return would be: "The more information the better". (Matching codec for codec of course)

The guy doesn't agree with me or Sony. What do you want me to say to that?

I suppose we are now just beating a dead horse at this point. There is no way that anybody is going to convince me that much lower bit rates are just as good as higher ones within the same compression algorithm. Yes, if we compare H.264 vs H.265 HEVC, than yes, I will agree there. But this is all within the SAME MPEG H.264 LIBRARY.

One thing is certain: Neither one of us needs each other's "approval" of our opinions. I think we will both sleep perfectly well tonight.

Sony has stated they are working on this and I have a "feeling" (or hope) they they will succeed. Competition and market forces are pressuring Sony on this.

I just hope Jody doesn't persuade them that it's not necessary after all.

lol ;-)

[Jody Eldred]

I own an F55 and can shoot 4K Raw all day long.

My X70 is merely another tool I can use in many situations, and I know exactly what to expect from it because I've done the tests. I've posted videos of the results here and on the X70 Shooters FB page. Do your own or ignore mine or whatever. Makes no difference to me. Sony knows they have a terrific product and my tests have further validated that to them. (Yes, to SONY. I have done real-world testing and evaluating of a dozen camera systems for them over the past decade +, some of which can be found on their websites. I have met with the Japanese designers on many occasions and am proud to have had input as to some of their recent high-end camera systems.)

Just sharing my experiences. And in my 31 years in Hollywood, this adage has served me well:

"A man with an argument is no match for a man with an experience."

;-)

[Cliff Totten]

For the record, I love my X70. That's not the problem.

And all this experience you have collected in your career gives you the ability to declare that significantly higher h.264 bit rates do nothing for image quality or color grading?

Than why in the world would you ever want to shoot with your "F55 in 4K raw all day long"?

You "experience" goes against all the "experience" that Sony and MPEG have.

Here is some more crazy talk from me:

"Higher h.264 bit rates = more image data = more information to work with in post"

or...

"Lower h.264 compression ratios = more image data captured"

or...

"Higher h.264 bit rates allow for more bits to be allocated into the shadows of an image and less macro blocking as well as banding"

or

"Higher h.264 bit rates allows smaller block encoding of moving pixels and allows finer and more accurate motion detail and entropy estimation"



Yeah,...radical, weird crazy talk that is all completely FALSE!! Your "experience" says that I am completely dead wrong on this. In fact, I'm clearly displaying my lack of experience here. MPEG and Sony would CLEARLY not agree with these statements I have listed here!

Where do I get this stuff from? lol

;-)

[Cliff Totten]

Woah,..woah,...let me hit the breaks here.

If is the real "Jody Eldred" and not some "impostor"....

I just Googled you....I KNOW YOU! We have met before! I have talked to you several years in a row at the NAB Sony booth!! Many, MANY times actually!

Hey, for the record, I think you ARE are pretty cool guy. Sorry I did not recognize your name until I Googled you. Dang! Nah man, I dig you and respect what you have done with Sony. You are actually one of the cooler guys to talk to at the Sony booth. No B.S, I mean this sincerely, I really do.

I don't take back my stand on bit rates and maybe I have been a bit too sarcastic at times but I do certainly respect your work and experience.

When I see you next year at NAB, you might want to kill me but I'll still want to shake your hand. I'm totally cool with you. Sorry, I didn't know this was you all this time. If I had Googled you earlier, I would not have changed my position but I would have changed my tone.

I hope we can just laugh at this next year. ;-)

[Christopher Young]

Originally Posted by Cliff Totten

So with that being said,...more image data is being dis-guarded in 60Mbp/s than it is in 100Mbp/s. The compression ratio is significantly higher in 60Mbp/s.

This compression vs bit rate argument has been around for a long time and was discussed to death many times in the early days of XDCam 35-mbit vs 50-mbit yet their corresponding compression ratios were very much the same. As Jody says "How does it look." If it holds up in post and especially if transferred to a higher level codec and works in the real world and everybody is happy with the result that should be the final arbiter.

I've shot countless hours of broadcast on both the 50-mbit and 35-mbit XDCam flavors and never had an issue with networks here and overseas and subjectively it's it very hard to tell the difference in the picture quality given that both come of correctly set up cameras. I've lost count of the number of hours we have shot in 35-mbit that has been rendered to 50-mbit for delivery to TV and never once had a query on quality.

As Alistair quoted way back in 2009:

"4:2:2 = (1920×1080 + 960×1080 + 960×1080) x30(fps) x8 (bits) = 995Mb/s. Divide by 19.9 and we get 50Mb/s

4:2:0 = (1920×1080 + 960×1080) x30(fps) x8(bit) = 746Mb/s. Divide by 21.3 and we get 35Mb/s

So from this we see that the compression ratio for EX is 21:1 and for XDCAM HD 422 20:1. This is extremely close and in terms of compression artifacts means there will be little, if any, difference between the two."

HD Warrior » Blog Archiv » XDCAM 422 v XDCAM EX by Alister Chapman

In the real world I have found this to be pretty true so we can't judge everything on bit rate and compression ratios alone.

It is also well know that Sony have their own proprietary compression algorithms that comply with the H.624 standards. Therefore we are only surmising when we say there must be a difference between the 60 and 100-mbit bit rates re artifacts when we don't know whether Sony is using a different proprietary algorithm for the the bit rate compression between the AX100 vs the X70. On that basis alone it makes it almost impossible to come up with an objective assessment of the differences between the two bit rates.

I've long ago given up nit-picking with the numbers. If it was all based on numbers then the Go-Pro would never have had a look in in the broadcast world yet they are everywhere like a plague. As Jody says if it looks good, feels good and works okay for the project you are working on run with it.

Chris Young
CYV Productions
Sydney

[Cliff Totten]

I want to be really clear on something here...and maybe this has gotten lost in my CODEC complaints.

I strongly believe the Sony PXW-X70 is EASILY the best "palm-corder" on the market today. Canon's offering is not even close. JVC's model is nice but I'll take Sony's larger 1 inch-type sensor and front end over that JVC model ANY day. (even though the JVC sports 150Mbps h.264)

I am VERY happy I bought it and I'm very happy I bought the 4k upgrade too.

My writing is NOT a "bitch session" about a camera I hate...far from it. It will be a great tool of mine for many years to come. (or until Sony replaces the model, and I buy that one. hehe)

I am a die-hard Sony freak and over all, they have gotten more money form me that I want to think about. Sony is ONLY company I buy camcorders or photo cameras from. I work for a global media company and I buy TONS of Sony products too at work.

With all this said clearly. I do get bothered at times when I see cheaper Handycam's and Alphas receive features that more expensive models don't get. The new RX10-II has 100Mbp/s AND SLOG-2...this bothers me. A while back ago the NEX5 got peaking and the far more expensive VG-10 was not allowed to have it. I called Sony and protested and complained...do you know what happened? At the very end of the VG10's sales life, as it was being discontinued and the new VG20 was being released, Sony released a tiny VG10 firmware update. Guess what the firmware did?....Added peaking! (no, I cant take credit for that. Allot of customers demanded it on the internet and blogs and forums)

That's the only reason why I do this. If Chevy allows the Tahoe to have leather seats but then limits it's Escalade sister to be locked down to cloth seats only,...there is something wrong with that.

I just hate seeing features in cheap cameras that are locked out of much more expensive ones. (I know consumer Alpha, Handycam and Pro departments debate this stuff internally within Sony too)

I'm sure that you know the FS7 vs. F5 scandal...That wasn't fair and Sony fixed that with an upgrade option for F5 customers. Good job on that.

That's all I'm trying to say here. Sony, please try to be "feature consistent" as cameras go up the ladder.

I can already see this coming. Sometime down the road, Sony will have two 4k brother and sister models. An XDCAM "pro" version and a Handycam "consumer" version. What is going to happen? Sony consumer division will somehow give the Handycam version 4K 60P and the more expensive XDCAM will be stuck with only 30p for a long while. It sounds funny,...but it's certainly quite possible! haha

[Cliff Totten]

Originally Posted by Christopher Young

This compression vs bit rate argument has been around for a long time and was discussed to death many times in the early days of XDCam 35-mbit vs 50-mbit yet their corresponding compression ratios were very much the same. As Jody says "How does it look." If it holds up in post and especially if transferred to a higher level codec and works in the real world and everybody is happy with the result that should be the final arbiter.

I've shot countless hours of broadcast on both the 50-mbit and 35-mbit XDCam flavors and never had an issue with networks here and overseas and subjectively it's it very hard to tell the difference in the picture quality given that both come of correctly set up cameras. I've lost count of the number of hours we have shot in 35-mbit that has been rendered to 50-mbit for delivery to TV and never once had a query on quality.

As Alistair quoted way back in 2009:

"4:2:2 = (1920×1080 + 960×1080 + 960×1080) x30(fps) x8 (bits) = 995Mb/s. Divide by 19.9 and we get 50Mb/s

4:2:0 = (1920×1080 + 960×1080) x30(fps) x8(bit) = 746Mb/s. Divide by 21.3 and we get 35Mb/s

So from this we see that the compression ratio for EX is 21:1 and for XDCAM HD 422 20:1. This is extremely close and in terms of compression artifacts means there will be little, if any, difference between the two."

HD Warrior » Blog Archiv » XDCAM 422 v XDCAM EX by Alister Chapman

In the real world I have found this to be pretty true so we can't judge everything on bit rate and compression ratios alone.

It is also well know that Sony have their own proprietary compression algorithms that comply with the H.624 standards. Therefore we are only surmising when we say there must be a difference between the 60 and 100-mbit bit rates re artifacts when we don't know whether Sony is using a different proprietary algorithm for the the bit rate compression between the AX100 vs the X70. On that basis alone it makes it almost impossible to come up with an objective assessment of the differences between the two bit rates.

I've long ago given up nit-picking with the numbers. If it was all based on numbers then the Go-Pro would never have had a look in in the broadcast world yet they are everywhere like a plague. As Jody says if it looks good, feels good and works okay for the project you are working on run with it.

Chris Young
CYV Productions
Sydney

I agree with everything you say here Chris. I'm not going to name names here but...the media company that I work for setup an HD deliverables standard for "Gold", "Silver" and "Bronz". Sony was very unhappy with where we placed 35Mbp/s XDCAM EX and really wanted us to change our standard. We didn't. At a SMPTE meeting in our office, Sony sent their CTO (I wont type his name but he was EXCELLENT btw) and even he tried to pitch XDCAM 4:2:2 and EX to the group. He even brought monitors and gear with him on his demo. He showed us phase-inverted-canceled A&B shots to inspect artifacts....really cool stuff! In the end, we still didn't elevate XDCAM EX on our standard (because of the "numbers") and I think Sony was really upset by that.

Meanwhile, I know for a fact that we were airing a major show with boats and drama on the water that was shot mostly on HDV (25Mbp/s 1440x1080, non-square pixel) tape at that time. (go figure?) Ehh,...maybe I have said too much already ;-)

Anyhoo....I agree with you on your post Chris.

[Ron Evans]

I think if Sony made the Prosumer/Professional difference as just SDI, Timecode in/out, Genlock , World camera specs, at each price point ( which would be features different ) it would make everyone's life a lot easier, including Sony.

Ron Evans

[Cliff Totten]

I'm perfectly OK with a paid upgrade sales model. I would certainly pay additional money for a second 100mbp/s liscense. I would certainly pay for an SLOG-2 X70 upgrade if the price was reasonable.

Speaking of SLOG-2....we are now seeing SLOG-2 make its way deep into the lower consumer markets. Multiple Sony Alphas and Cybershots now have it.

Who would have thought we'd all see the day when my Grandmother could go to a Best Buy and get her camera with an SLOG option it?

It's funny how cheap consumer cameras get things that many pro market cameras are not allowed to have.

SLOG-2...wonderful dynamic range recording! Get in on our high end super35 models or get it in the cheapest low end market! (Not available on any mid level pro market camcorder) $8k and higher or in several consumer $2.5k and below.

Strange?...

[David Heath]

Originally Posted by Cliff Totten

Sony designed many XAVC modes that go WAY up the ladder in bit rate. (With 60mbp/s at the very bottom) Sony markets them as higher and higher in quality as they go up the list.

Well, not quite. Yes there are many XAVC modes, and yes they go up in bitrate..... but a large reason for that is to accommodate I-frame only versions of the codec. Such don't NECESSARILY give better quality, but there are reasons why they may be desired, right at the acquisition stage.

And unlike long-GOP, their bitrates are directly proportional to framerate, so up the framerate and ........

Originally Posted by Christopher Young

As Alistair quoted way back in 2009:

"4:2:2 = (1920×1080 + 960×1080 + 960×1080) x30(fps) x8 (bits) = 995Mb/s. Divide by 19.9 and we get 50Mb/s

4:2:0 = (1920×1080 + 960×1080) x30(fps) x8(bit) = 746Mb/s. Divide by 21.3 and we get 35Mb/s

So from this we see that the compression ratio for EX is 21:1 and for XDCAM HD 422 20:1. This is extremely close and in terms of compression artifacts means there will be little, if any, difference between the two."

The maths is correct as far as it goes, but I'm afraid the reasoning is wrong on several levels.

Most importantly, you don't need twice the bitrate to code 4:2:2 chroma as you do 4:2:0. The reasoning is similar to why when you compress still images, for a given quality level the file sizes don't scale linearly with image size. If you've got a 4:2:0 chroma image, you can effectively "guess" at in between values. ("Guessed" 4:2:2.) So to code the true 4:2:2, you only need the values of how different the true values are from the guesses. Which can be represented by a lot less data than the original 4:2:0 chroma signal.

Secondly, Alister assumes the same compression ratio for luminance and chrominance, and for MPEG2 this is not necessarily true. It's frequently the case that the chroma signals get far harder compression than luminance - this is why when banding may sometimes be seen (esp on digital broadcast TV), it's frequently most noticeable on areas of plain saturated colour, more rarely on more monochrome scenes.

So you can't do a simple sum of total numbers of chroma and luma pixels, compared with total bitrate and reach any meaningful simple conclusion. It's far more complex than that.

[Cliff Totten]

Off topic question but I cant resist.

"The reasoning is similar to why when you compress still images, for a given quality level the file sizes don't scale linearly with image size."

I dont know the answer to this at all, but maybe you do...

Two files:

1.) If you encode a 500x500 frame size image with specific h.264 setting and achieve "X" quality.
2.) If you encode a 1000x1000 frame with the same exact h.264 setting to achieve the same "X" quality.



A.) Will file #2 require 2 times the bit rate to achieve the same "X" quality? (because its frame is 2x larger?)

B.) Does H.264 "gain" higher and higher efficiency levels as frame size increases? (So maybe it only needs 1.5 times the bitrate instead of 2 times to achieve "X" quality)

C.) I believe that H.264 operates in fixed block sizes. Or is this wrong and are these block sizes (4x4 or 8x8 or 16x16...etc) are infinitely scalable? I know that h.265 operates in different block sizes and supposedly is a major advantage over h.264 because of that. (one simple reason on many) If block sizes are fixed, than that would suggest a more linear or proportional scaling bit rate???

I'm not sure about any of these questions but maybe you or Jody or anybody else knows?

[Craig Seeman]

Might I suggest someone do a short recording using the AX100 and x70 in 4k and open and compare the files in MediaInfo. While it won't tell the whole story it will tell you what OTHER differences there are besides bit rate. Things like Entropy and GOP Structure can result in a lower bit rate file looking better than a higher bitrate file.

[David Heath]

Originally Posted by Cliff Totten

Two files:

1.) If you encode a 500x500 frame size image with specific h.264 setting and achieve "X" quality.
2.) If you encode a 1000x1000 frame with the same exact h.264 setting to achieve the same "X" quality.

A.) Will file #2 require 2 times the bit rate to achieve the same "X" quality? (because its frame is 2x larger?)

And - like with a lot of things(!) - I don't think the answer is simple.....

Firstly, I think you mean "Will file #2 require 4 times the bit rate to ......", don't you? You're quadrupling the number of pixels, so "common sense" would seem to indicate you'd need 4 times the bitrate?

The answer is generally taken to be "no", but it depends a lot on what variables change. You're changing the image resolution - but are you changing the size of the viewed image as well? If your original image is viewed in a frame 8"x8", the answer will vary depending on whether file#2 is also viewed as 8"x8" - or scales up proportionally to 16"x16".

That's important when you think of the difference between 4:2:0 and 4:2:2 chroma sampling. In such a case you are far more analogous to "fixing the size of frame" - as the 1920x1080 luminance aspect is unchanging. And it then follows that although there is a doubling of chroma samples, it shouldn't necessarily be thought of as needing a doubling of bitrate for equivalence - it should be less.

[Jody Eldred]

Originally Posted by Craig Seeman

Might I suggest someone do a short recording using the AX100 and x70 in 4k and open and compare the files in MediaInfo. While it won't tell the whole story it will tell you what OTHER differences there are besides bit rate. Things like Entropy and GOP Structure can result in a lower bit rate file looking better than a higher bitrate file.

As reported earlier (in this unnecessarily very long thread) I've done that with a colleague's AX100 and my 4K X70 in a 4K color grading suite at Roush Media in Burbank, CA. Projected in 4K on a 15-foot screen. They both looked very good (for what they are... I'm used to seeing F55, RED, or Alexa 4K up there.) The X70 graded better, chroma noise was a bit less, the compression appeared to be superior as there was slightly more digital artifacting in the AX100.

I have no way of sharing this as we did not have opportunity for outputting, much less doing so in a comparative side-by-side edit. (You'd not be able to evaluate it online anyway do to compression and the wide variance in monitoring.) I have posted some 4K X70 that Roush Media graded on this website. It'll give you some idea of what you can do. Here again is the link:

First 4K Sony PXW-X70 Color Grade, by Jody Eldred - YouTube

[Jody Eldred]

Originally Posted by Cliff Totten

Woah,..woah,...let me hit the breaks here.

If is the real "Jody Eldred" and not some "impostor"....

I just Googled you....I KNOW YOU! We have met before! I have talked to you several years in a row at the NAB Sony booth!! Many, MANY times actually!

Hey, for the record, I think you ARE are pretty cool guy. Sorry I did not recognize your name until I Googled you. Dang! Nah man, I dig you and respect what you have done with Sony. You are actually one of the cooler guys to talk to at the Sony booth. No B.S, I mean this sincerely, I really do.

I don't take back my stand on bit rates and maybe I have been a bit too sarcastic at times but I do certainly respect your work and experience.

When I see you next year at NAB, you might want to kill me but I'll still want to shake your hand. I'm totally cool with you. Sorry, I didn't know this was you all this time. If I had Googled you earlier, I would not have changed my position but I would have changed my tone.

I hope we can just laugh at this next year. ;-)

No prob. Hope to see you there.