Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Galaxy Touch 3 4k 60p Korean video, is it true?

W

Wayne Morellini

Well-known member
Joined
Jan 3, 2008
Messages
6,157
Reaction score
0
Points
0
Hi there are Korean you tube videos titled 3840x2160 60p, but I can't load up the pages to translate the Korean. Can anybody here read it to see if it really is 60p not the advertised 30p? If so, then this phone just became interesting.

http://www.youtube.com/watch?v=FcOMokZwqt8&feature=c4-overview&list=UUsaH_yBrLlBhy0tirv6f59A
http://www.youtube.com/watch?v=J74rxu-I4Gg&feature=c4-overview&list=UUsaH_yBrLlBhy0tirv6f59A
http://www.youtube.com/watch?v=7ez2j4Uzh2Q&feature=c4-overview&list=UUsaH_yBrLlBhy0tirv6f59A

It makes sense that it could do it, the new Acer Liquid S2 does 24p and uses the same snap dragon chip as some editions of the Note 3. The chip supports the function and is used on a number of phones already, so it should be possible to for some of them to support 4k if other parts can handle it.

http://nofilmschool.com/2013/09/acer-liquid-s2-4k-video-recording-smartphone/

The frustrating thing is, that there is no information I can find as to the quality or data rate of compression. I don't hold out much good of anything good, but if it can go to 80-100mb/s of quality h264 compression (like ambarella circuitry on chip could do), it at least will be watchable at 4k.

One step closer to a personal pocketable on demand shooter.
 
OK, the LTE versions if the Note use a snap dragon with 4k 60p support, and the 3g uses the octo core. Maybe the LTE version can do it and the 3g can't , but it could be they both are locked down to 30p.

Also note, that in different marketers they will use different cores and screens apparently. The most obvious reason to checkout the Samsung versus the Acer is for the amoled screen, which might be replaced with an lcd in some countries. It certainly is nice to use in filming (I've tried my tablets 7.7 inch amoled).

I still wonder what performance is possible. I've found mention of maxinum of 20mb/s 24fpsfor the Acer, and slower on two articles. But I think that 20mb/s is just stupid, unless it was h265, then it should be bad, a typo of 20MBytes/s would probably be too much, as that would be 160mb/s, which is basic TV professional. And advertising as 60p and delivering 24p seems unlikely too, but it is only prototype demo models at the moment. When we look at it, the ambarella like coprocessing circuits available for licensing should only take up a small corner of such a chip by now. So, somewhere between 80-160mb/s maybe possible on a tablet sized phone sometime.
 
The Youtube sample 4k video from the note3 is full of rolling shutter jello, and has no stabilization. I think it was 120 Mbytes for 20 seconds.
I think it's a nice addition, I plan on moving from my note2 to this phone. The LG phone has optical stabilization, but no 4k.

If you are looking for good 4K at a cheap price, the new Sony Z100 camcorder looks potentially good. 10 bits deep at 500 megabits. 20X lens. About $5500.
 
I'm looking at getting a cheap camera for carry around and an professional. The 100 is nice, but sort of in between what I want. 43rumors.com is going to have information on the direction of the Panasonic GH series in around a week and a half (so much for a rumor sites that they have to wait, talk of a upcoming camera has been around for months).

That jello effect, you could simply add a electronic shutter in front to reduce that, like I proposed around 9 years ago. Oh, that is right, Red has just announced that.

Pity about Black magic, hopefully they can have cameras ready for NAB next time, the pocket does not do 50p or 4k, there is an upcoming anti aliasing filter for it. Now they are going to run into the next GH release, maybe finally the still/video camera, already after the gh3, and probably $500 ambarella based 4k cameras and these programmable android based devices. It is time to pick it up.


The 40-48mb/s thing is barely tolerable. If it means 96mb/s for for 60 or 50 fps, then that is OK. The jello effect probably indicates it is reading the sensor at 30p. It is a matter of what professional camera app can be run on it, and I call on Red to sponsor one, I can scrape up people that can program it, on app store $10-$100. The Red Real Camera app for the professionals off time. (not to mention learning young camera people). No hardware manufacture lines needed.
 
Les, was it 60p ultrahd?
 
OK, here it is, according to this review, the 4k mode of the phone does 47mb/s, which is pretty bad unless it is footage from a static scene going through a variable codec. Hopefully camera software can boost this to 100mb/s.

http://www.gsmarena.com/samsung_galaxy_note_3_preview-review-986p5.php

It appears that only the 4g LTE N9005 version with snapdragon 800 will gave ultra HD recording.

http://www.gsmarena.com/samsung_galaxy_note_3-5665.php

The Acer turned out 4k p24 with trouble maintaining that, so fairly limited:
http://www.gsmarena.com/acer_liquid_s2_ifa_2013-review-976p3.php
 
What do you mean Nicole?

The "4k more" above was meant to be "4k mode" sorry, so it is reportedly 47mb/s, OK for 1080p60, but not so good for 4k.
 
OK, tested out the Note 3 tonight. At first it was in full HD and latitude looked fairly limited, the white paint of the shop across the corridor burned out unless the scene was unnaturally dark in the corridor. I checked the resolution and changed to ultra HD, and to my surprise the latitude became much better. I don't know why I did not have time.

I tried some testing to see if this has variable compression as lg camera app seems to use. Under a steady test and complex movement test the data rate remained at about 48mb/s.

The rolling shutter skew was bad, maybe low light mode low frame rate was making it worse.

Unfortunately I could not try lg camera app on it to see if it could increase data rate variably or increase the frame rate to 50p reducing the rolling shutter skew. A warning, the app does not report the true rate used, it might be trying to do it but you have to run an app like MX player and go into tools to look up derails to confirm data rate, frame rate and resolution.

Handling, well the android front end has some nice much needed improvements. which includes the camera app, but not much. Holding it in my hands, despite the new security cradle and stiff short elastic cord on the demo unit felt good. It feels good, small compared to filming with the 7.7 inch tablet, but as expected, screen size was good (5-8 inches is a good handheld screen size range for filming, but weight makes the difference, but a 10 inch that is very thin and lightish would be interesting). But this camera needs a good app and better external controls. I have such designs that would fit from my handheld camera design work. The thing is that people shy away from small cameras for a professionally (in videography terms) usable image but it is possible, I have being studying the design requirements over the last several years.

Maybe they will make another hybrid phone camera with 4k based on this.

Altogether, if frame rate and data rate could be increased the camera could be promising. There are silly people out there on the internet, I've come across some, that say such a thing, or using a machine vision camera, or whatever is not workable. But in the pre-red times we used machine vision cameras and conditioned normal cameras to produce "remarkable" images, sure it might not be Scarlet quality, but all we need is something that looks good enough for the price. I think if we can boost this camera it can be conditioned to a potentially remarkable image.
 
I forgot, the color balance, as on the amoled screen, was clearly off. There was a rainbow sign nearby that had color shift on the phone. I don't know if this is the footage or the display though.
 
OK, this is what I was talking about. I was planning this for my own cinema camera design years ago, plus a fair bit more. It may have been able to achieve 80% the quality of a scarlet, but at less than 20% the price (next years price). The rest of the world is catching up as usual, but it gives an idea what can be dine even with phone footage:

http://www.redsharknews.com/technol...tphone-the-galaxy-note-3-is-surprisingly-good

http://www.redsharknews.com/technol...lenses-are-coming-soon-but-would-we-want-them
http://www.slashcam.com/news/single/Software-instead-of-cine-primes--10937.html

You will see a bit of cross over with my theoretical work in deriving 3d from 2d footage that I was heavily figuring many years ago, again they now talk about this:

http://www.redsharknews.com/technology/item/1105-this-is-a-big-step-towards-sentient-robots
 
I quickly skimmed over the last lens correction article I linked to, and saw that they were going my direction. I found another article and realised just how close they are coming:
http://nofilmschool.com/2013/10/simple-lens-imaging-algorithm-technique/

Yes, one of my plans was to use a single lens to capture footage once cheap cinema camera through the same techniques, which if using my software technology can be done a lot faster. If you look back to my alternative solution to the 3d lytro lens at dvinfo 6 to 7 years ago, you will see where I was going and have been since the 90's. I have asked Red carefully about their 3d firmware update that is supposed to supply 3D image on a 2D system without modification. I asked about all the 3d methods I came up with except the ones I was actually going to use. I leave out those things until somebody else comes up with it. However, if Red plans to do 3D through a standard lens they need help of experts. Funny how much Red does is similar.

However, the single lens is only the start and I moved onto something totally alien. There has been another advancement since then that touches on my more advanced idea.

For a number of years I worked out cheaper and cheaper ways to design a cinema camera and got it down to under $100 + storage 4-6.5cm square and as skinny as possible, maybe 5 years ago. It could possibly do 4k 3d now. This drives some people in the industry nuts, as they are not intelligent enough to figure it out or see how it is possible. Pity I'm older now and near Parkinson problems I'm still dealing with, so I'm not usually that brilliant now.

Now the future is coming on the industry, the high end of the market is facing oblivion. 80%+ of the real work (most is not cinema) does not require a camera more than $500 using these techniques, anybody aiming at over $500 camera market is starting to have problems with volume. But indeed there is going to be a market for high end cameras, but not much volume except as you go beyond 8k into holographic video, for the decade as it will take advances in processing and storage technology to cheaply handle these (next decade processing of holo resolutions 'maybe' cheap enough).

I'm known to be very prophetic with these things and inventions. It is a matter of directions.

So work on android iOS Ubuntu Windows RE, cinema software for mobiles and cameras might be a promising avenue, when they get rid of rolling shutter. The trick will be to combine these techniques with consumer and professional oriented control modes to drive volumes.
 
If it was a religious subject I would say Amen.

I have seen and lived the future, but yours is about the most intelligent conversation I've had around here. Let the struggle start.
 
No, but I did try and reinvent it. ;). I just don't know if I can practically get it to take the forces between 500-1000kmh. The design was pretty unique. Reinvented many things.

You'll have to try a bit harder to get to the top level of intelligent conversation Les.

I'm just happy at the moment.

Les Dittert said:
Didn't Wayne invent the wheel ? ;)
Click to expand...

Actually, I should go and review my notes on time travel. If I sent myself back in time and bumped myself on the head to forget everything I'm pretty sure I could invent the wheel soon after seeing the problem. Though it would also help if one took a yo yo with them, or some round rocks. ;)

(Stated in the past tense, as if I have actually invented the wheel I already have gone back from the future)
 
Has Apple lost.

Has Apple lost.

I speculated that compressed raw could possibly work at 4k on the Samsung Note 3, but now a much more powerful 4k phone seems to be coming and it might be very likely it could handle compressed raw 4k, maybe even at p60.

There is already talk of the Samsung Galaxy 5 with 4k video, 16 64-bit cores at up to 3ghz+ and up to a 16mp camera, for release January or February.. The shocker is that it uses IGZO+ LCD instead of amoled. I know that IGZO is good but to replace amoled on a flagship product seems strange as it offers a number of advantages in conventional implementation. Samsung used a 100% efficient form of oled in the S4 but only in two of the three color channels, the remaining camera had 25% efficiency and had to be boosted, as that color has a short life in the new oled technology. There is another issue of color purity, as efficiencies can hit off the primary target and be used instead. IGZO has 90% light pass through, so could beat amoled given higher efficiency light source, but probably not?

The rumored processor use a 20nm chip making process, so these naive comparisons between Arm and Intel Atom that stupidly comes up, might have some hope of being based on something closer to equitable (though Intel 3D transistor is superior and they are moving down below 20nm towards 6nm. However we are closer to the point where normal smaller processes will stop yielding much power reduction. Somewhere below 20nm speed increases are going to come at extra power consumption for a number of years.

http://galaxysv.com/samsung-galaxy-s5-release-date-price-specs-features-and-images/
http://galaxysv.com/s5-processor-specs-features-release/
http://galaxysv.com/samsung-s5-features/
http://galaxysv.com/samsung-galaxy-s5-specifications/


Talk about the new Note 4 in may.

http://galaxysv.com/samsung-galaxy-note-4-s5-specs-features-release-date/


Now let's see what Nokia does.
 
There are various schemes to increase processing to energy potential to overcome the topping out of conventional chip making technology. Even before these alternatives are ready, they could be included in hybrid forms with present processor technologies, as a lot of heavy work could be handle by array processing technologies instead of the rest of the chip. Inserting small processing elements, or arrays, of the new technology allows them to handle the bulk of the heavy work in processing images at reduced consumption, before a complete solution is ready. I have a few optical computing proposals myself that have decades of clear growth path, but as I can't afford to research and develop that, I am interested in looking into Quantum dot cellular automata, and Quantum cellular automata. I am interested in putting a very lean very efficient microprocessor design on it, based of my virtual os designs. These are the technologies I was referring to in the previous post that may come in and replace current chip processes and reduce the power consumption again. Of course everybody at Red should know about these already and be on top to adopt them.

http://en.wikipedia.org/wiki/Quantum_dot_cellular_automata

The QDCA reminds me a bit of my magnetic transistor proposal that whole chips could be formed from (of course having a workable idea and getting it to work competitively is another thing). However, I have a few short cut new design proposals that could drastically improve existing chip making processes and apply to QDCA.

In the mean time industry has to research into 3D transistors and barrier technology to improve the power consumption to performance ratios. One design proposed I think 50Ghz (I think) low power 3d transistors in 2009. In reality, in a complex chip design the chip ends up working at a fraction of this speed, so 5-10Ghz might be more likely, or faster again in the hands of people I know. The Intel technology tries this. But it is not just the chip process technology that offers vast improvements, there are still wants to design a chip that can also help. In design of virtual systems I have come up with some innovative new ways to do so. We should be able to get ten to a hundred times more processing out of these technologies before hitting new chip making technologies.
 
@Les, I'm happy to announce I have also found another cinema related technology I had plans for has being released for mobile production. However, this one I am keeping to myself as I have my own methodology. We are getting to the point that a low/no budget indie film could be done on mobile (plus rig of course). By the time you spent money outfitting the mobile though, you could have bought a proper camera for $1000-$2000 in the future and got maybe better results.
 
You must log in or register to reply here.
Back
Top