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 :)

NAB 2014 - REDRAY IN STOCK

I'm guessing my RedRay will end up being sold off at some point to one of my clients as an on site player for a UHD display installation. I'd love to see RED throw serious resources behind developing a robust .RED distribution ecosystem - if they see a future for it. I see RedRay in it's current iteration as more proof of concept, than a dialed in product ready for wide launch. At the right price, with a nice UI and some hardware DRM chops a second generation RayRay-ish device could go somewhere. That said, only RED management can decide whether to allocate the substantial resources I'd imagine would be necessary to compete in the evolving IP media player market.

The Nuvola NP-1 is another option for a compact UHD server, it's no bigger than a paperback book so it could be hung on the back of most panels, or rigged to the mounting hardware. I think retail is $300. I've been waiting for them to get HEVC implemented before jumping in, it launched using H.264 and the footage I saw in H.264 looked soft to me. I was hoping to see where they were at in Vegas but as usual NAB swallowed me like the whale that it is and I never got there. Does Neil or anyone have the latest from NanoTech Entertainment on these units? I'm really hopeful they will offer an affordable device with decent quality and look forward to giving it a proper test.

Cheers - #19
 
Last edited:
Blair I like your thinking :)

Part of day job is dealing with Digital Signage and I've always though that 4K and up is a perfect pairing.

I've had to pleasure of testing Christies MicroTiles and would love to go higher than my last test run using a beefed up Mac Pro.

With the cost of RR it rivals (actually even cheaper) most of the vendors suggested players GPU cant hit that sweet spot (res).

All we need is better control in the UI such as playlist and scheduling and I would say the RR would be perfect for large res installs.
 
Blair S. Paulsen said:
I'm guessing my RedRay will end up being sold off at some point to one of my clients as an on site player for a UHD display installation. I'd love to see RED throw serious resources behind developing a robust .RED distribution ecosystem - if they see a future for it. I see RedRay in it's current iteration as more proof of concept, than a dialed in product ready for wide launch. At the right price, with a nice UI and some hardware DRM chops a second generation RayRay-ish device could go somewhere. That said, only RED management can decide whether to allocate the substantial resources I'd imagine would be necessary to compete in the evolving IP media player market.

The Nuvola NP-1 is another option for a compact UHD server, it's no bigger than a paperback book so it could be hung on the back of most panels, or rigged to the mounting hardware. I think retail is $300. I've been waiting for them to get HEVC implemented before jumping in, it launched using H.264 and the footage I saw in H.264 looked soft to me. I was hoping to see where they were at in Vegas but as usual NAB swallowed my like the whale that it is and I never got there. Does Neil or anyone have the latest from NanoTech Entertainment on these units? I'm really hopeful they will offer an affordable device with decent quality and look forward to giving it a proper test.

Cheers - #19
Click to expand...

It will be at least another two years before these next-gen codecs land in hardware. Right now, Nvidia/Intel/AMD are working on bringing VP9/H.265 decoding to their GPU products and they will likely try to launch products with both implemented. It took about ten years just to get H.264 decoding properly implemented in GPUs so let's hope things go quicker and smoother with the new codecs.
 
I would like to Congratulate the red team.
.RED and Redray are just mind blowing.
Wow! This is some serious Ground breaking progress in Codec development and hardware playback.
I m sure the way RED cameras revolutionized the camera industry, .RED will revolutionize the codecs, and REDRAY the delivery.


Today morning, I finished encoding my feature film "The Acceptance" using RREncode,
And then ingested it, and watched it on Redray. I watched it in mere HD since I don't own a 4K display/projector yet.
.RED looks much better than DCP, It is better than Prores HQ, and it beats the hell out of Blu Ray.
Who is second best behind .RED, At this point I don't see anyone by miles.

Hats off to the RED team.
 
Yogesh Khubchandani said:
(snip) .RED looks much better than DCP, It is better than Prores HQ, and it beats the hell out of Blu Ray.
Who is second best behind .RED, At this point I don't see anyone by miles.
Hats off to the RED team.
Click to expand...

Why do you think .RED "looks much better" than DCP? What comparison did you do? IMO they are close, which is impressive considering the DCI spec is 250mb/s 12bit xyz color vs initial .RED implementation of 9, 18 or 36mb/s at 8bits. In terms of perceived IQ .RED holds it's own against ProResHQ surprising well on most material considering its bit rate is roughly 50 times less, though it's kind of an odd comparison since ProRes is aimed at post processing environments rather than distribution/exhibition.

What would be fun is to test all three flavors of .RED against various HEVC encoding rates. Early white papers on UHD distro based on HEVC (H.265) aimed at 20mb/s for 2160/24P as a viable balance of quality and size. So far, typical demo material deliverables specs for UHD/4K have been 80mb/s for H.264/VC-1. This is essentially taking the nominal Blu-Ray 20mb/s 1080 encodes and allocating enough bits to fill out the 4 times larger raster. I am under NDA on a couple of things, but as one might guess most folks are more focused on HEVC/H.265 encoding for UHD/4K and are not putting any serious engineering into optimizing H.264 in 2014.

One would expect the 18mb/s .RED IQ to crush 20mb/s HEVC IQ, would love to do some empirical testing to see how obvious, or subtle, the differences are. For high motion 2160/60P material like sports I am dubious that anything less then 40mb/s in HEVC will carry enough IQ to showcase the delta over 1080 distro, and (after a couple more years of optimization) 60mb/s is probably where high motion sports start to really shine. I understand that many codecs can operate a higher ratios on larger frames. That said, 8x more area than 720/60P and 4x more area with 2x fps means roughly 8 times more data for 2160/60P to encode than current HD formats.

Cheers - #19
 
Blair S. Paulsen said:
What would be fun is to test all three flavors of .RED against various HEVC encoding rates.
Click to expand...

I've been "having fun" testing and comparing .RED against HEVC H.265 and Google's VP9 for a while now. All encoding to these formats are slow today as none are truly "optimized" yet. Which means very CPU intensive encodes.

We'll get there, but there's a lot in the air in regards to getting that all accelerated and implemented everywhere. What I do like right now is the use of all of these formats as finishing formats. Though I early on was exploring the concept of using .RED for "dailies", which is certainly doable though time consuming to export large quantities of material for.

Much of my tests have been at projected bitrates or published specs. My tests thus far with HEVC and VP9 have been at 3, 6, 9, 12, 13, 15.6, 18, 30, 32, 36, 60, 72, and 108Mbps. I think a couple other weird ones in there too. I have some other thoughts once some other features are "dropped" into those formats. Right now we're seeing the HEVC 15.6Mbps delivered via NETFLIX.

The .RED comes in 3 flavors, all good. 9, 18, and 36Mbs. I deliver most of everything at 36Mbps because "why not". I'll have a couple .RED/REDRAY files later this week for folks to upload to their REDRAYs to play with. All 36Mbps. One thing I like about the .RED codec/format is how it handles grain at higher bitrates. Grain is the one thing where modern H.264/X264 encodes get a little wonko on. All of these new formats really handle high frequency detail better, which is nice for all of us.

Interestingly enough .RED is the highest quality format I have at my disposal at my house for realtime "quality" 4K playback. That I dig.
 
roryhinds said:
interesting to hear .RED for RedRay is 8bit.
DCI DCP's are 10bit, no?
Click to expand...

.RED is not 8bit.

Unfortunately, the HDMI 1.4 outputs on the REDRAY player are restricted to 8bit at 4K. It's an HDMI thing, not a .RED thing… This is also why the REDRAY projector had a built-in REDRAY player. That way it could bypass common interfaces and their restrictions to delivery the necessary data right from the image processor to the laser/display engine. For lower resolutions, the REDRAY supports 10 and 12 bit output, also within the confines of HDMI spec.
 
The .RED spec might support greater bit depth, but based on what the RedRay supports today (and, based on RED's disclaimer about no plans to update firmware that might be the same as ever) and what single link HMDI supports; I was confident the poster had only seen .RED content in 8 bits at a maximum of 36mb/s. I have been impressed by the dozen or so .RED encoded pieces I've viewed from the RedRay. I'll reserve judgement on codec metrics until I can view some proper material from DPX/TIFF/OpenEXR media side by side with .RED transcodes of the same stuff. Without a valid comparator its virtually impossible to know which metrics in the signal chain are most influencing perception.

Mike makes a valid point about how the actual presentation hardware at the exhibitor's site (server, PJ, I/O) often dumbs down the signal to less than 12 bits. I'd still rather have a 12 bit xyz master for the hardware that can use it.

I'm also wondering why the poster considers the .RED imagery superior to the ProResHQ imagery? Perhaps the .RED codec imparts a smoother feel to the picture that some viewers might prefer, even if most pixel peepers wouldn't. As they say about opinions…


Cheers - #19
 
I have very little time right now. But I would still like to share how I arrived at my opinions

I have NOT done a formal scientific A/B testing
Whenever I reviewed the final master in each format or have a screening in cinema theatre I kept notes on what I observed and sensed


My MASTER is 4K DPX files.
Same 4K master was used to generate all the other formats - DCP (2K 12 bit) , Prores HQ (1080 p 422 10 bit) , Blu Ray, .RED 36 Mbps

At my studio, I use a 1080 p 10 bit 4:2:2 projector calibrated to 709 space.
Prores HQ, Blu Ray master, and .RED were all viewed on this same projector.

My notes on DCP 2k 12 bit are based on screening at Landmark sunshine theaters NYC in 2010.
I did not note down the exact models/versions of the equipment at the time in 2010, so can't tell you for sure that those servers and projectors were 12 bit.
But it was a huge digital 2K DLP projector and the dcp was ingested on a proper dcp server.
And we had a professional projectionist who I believed maintained that projector in very good shape.


This is the summary of my notes

DCP 2K (12 bit) - FEELS DARK. Too much contrast in 7 shots subtle shadow details were lost. Aliasing/ moire in 2 shots.

Prores HQ (10 bit 4:2:2) - MISSING OOMPH. Aliasing/Moire in 2 shots. Contrast is lost in 6 shots, feels mushy/murky. Feels soft in 5 shots.

Blu-Ray - EH. IT'S OK. Sense Banding in 4 shots, Grain is weird in 2 shots. Aliasing/moire pattern in 3 shots. Subtle Shadow detail lost in 6 shots. Details feel soft in 9 shots

.RED 36Mbps - WOW! HAS SAME OOMPH as 4K DPX MASTER. Colors and contrast are exactly as I wanted them. Grain is very pleasing and natural. Amazed at no aliasing/moire even in the 2 shots where every other format showed problems.


I kept producing different formats, but was never really happy with the quality of any format, until .RED


We have not even talked about the a) file size, b) the price points to master these different formats, and c) the costs, logistics, and environment friendliness of distribution using each format. :-)


Best regards
Yogesh

ps: at the heart of DCP is Jpeg 2000, isn't that archaic and inefficient.
I m sure you guys are more knowledgeable about codecs and can certainly discuss that aspect based on scientific facts
 
Appreciate the effort you made to test options Yogesh. From what I've seen, .RED looks like a terrific distribution codec that puts a lot of goodness into a tiny package.

Not sure how you managed color space and gamma in authoring the DCI version. There are LUTs for rec709@2.2 to P3@2.6 conversions of course, but perhaps some of your scenes needed a trim pass to hold shadow detail, etc.

Yes, JPEG2000 is the codec used for DCPs and it's long in the tooth. That said, assuming proper exhibition standards are met, it's only real "sin" in my book is that it can be a touch soft due to the entropy encoding scheme.

Would be curious to look at settings and the signal chain used to create the ProRes version. FWIW I have been underwhelmed with ProRes as a delivery codec a couple of times myself considering its bit rate. I have a couple of pieces ready for encoding so I'll attempt to compare .RED36 to PR444 as apples to apples as I can manage using the Sony X850A 65" 4K panel and the Sony VW600ES (500ES in Europe) 4K projector for viewing. Sadly I don't have a RAID around at the moment that can play 4K DPX or TIFF sequences in real time (roughly 1GB/s).

CHeers - #19
 
You must log in or register to reply here.
Back
Top