REDLINE [3xR][Blu-ray.1080p.H264.FLAC]  

Just transferred the 720p offer by Silmeria to you. Reason being he doesn't actually have the 8-bit release to seed lol.

Leeching outside the baka and preparing for the next wave... I'll try and pitch in my 4mbps up soon as it gets approved.

Took me some time but I got all three offers and took screenshots.

BSS : 1864*1048 (16:9) at 23.976 fps AVC (High@L4.1) (CABAC / 4 Ref Frames)
REVO : 1864*1048 (16:9) at 23.976 fps AVC (Header stripping) (High 10@L4.0) (CABAC / 4 Ref Frames)
3xR : 1860*1048 (16:9) at 23.976 fps AVC (High@L5.1) (CABAC / 8 Ref Frames)
3xR H10p : 1860*1048 (16:9) at 23.976 fps AVC (High10@L5.1) (CABAC /8 Ref Frames)
ALL : Japanese 48KHz 16 bits 6 channels FLAC
Note : you'll notice that only REVO is in High 10p, 3xR has released it in High 10p but I'm still grabbing it.

3xR Vs BSS
3xR Vs REVO
REVO Vs BSS
3xR Vs 3xR H10p
3xR H10p Vs REVO
3xR H10p Vs BSS

All screenshots, ScreenshotComparison has bandwidth restrictions.

Edit : added comparison with 3xR's H10p release. Hope you enjoy the time I've put into this.

XxX Betra

hmm overall i think the 3xR releases are of a better quality and the h10p version loses to the normal release. mostly becuase of the difference in bit-rate. the 3xR seems more detailed altough the transition of the colours is more harsh than h10p

I'd recommend ppl with a normal TN monitor to download the normal version.and ppl with other panels the h10P but its a shame its only half the bit-rate

Honestly, between H10p and normal when you watch the movie and are in the action, you just cant get the defference. Well I can't.

XxX Betra

-----
Of them all, the 10p 3xR was the blurriest, 2nd most blurry was the regular 3xR.
The sharpest was the BSS, and 2nd sharpest was the REVO,...

(reviewed on a 30" 2560x1920 monitor...   I could review it on my HD monitor which is larger but lower res....)....


So why does anyone think that 10p can get by w/a lower bit rate and look 'better' when has the lowest quality
out of these...

I'd love to see more comparisons, as 1 sample is not conclusive, but the claims that one can display 30bits of color vs. 24 and end up with small file sizes and NOT have quality suffer sound ridiculous.  we are talking 25% more data / **dimension** so that's
56.25% more data / frame.  So at the same frame rate, and same compression we'd see 56% larger file sizes.

There haven't been any major increases in compression efficiency, so the only way to get small file sizes is to make it more lossy.   If it was the same file size, one would have to scale down detail by a factor of 0.64 (or 64% of the detail of the non-10p version).  If the new file is 20% smaller, then the detail  reduction factor would be .8*.64=.51 ...

It's certainly possible for a 10p release to provide better quality at larger file sizes, but usually when you think you are getting 'free benefits for nothing', you find it was just a pipe dream.

It may be that on some sources a 10p will look better than an 8p at the same bit rate due to unique attributes of the color of some video source, but in general -- not.

Comparison: in mp3 encoding, if you reduce bit rate, many encoders will try to shift from separate stereo (
2 x 16 bit x 44.1k-frames /second), down to joint (1.2-1.5 x16bit x44.1k fps), down to mono, (1x16bx44k fps)... then they'll start lowering the fps, 32k, 22k, 16k, etc.....  ).

If you want smaller file sizes, you will lose information.   Why do people believe video is exempt from the same factors that are well understood for audio?   Maybe there are _likely_,  0 devices out in the real world that can actually display 10-bits/channel, so no one will notice?

I dunno...yeah, I'm a bit on the cynical side, but on things like this, I like to see alot more technical evidence and something that shows how they compress the 1 liter of water down to 750 cc's (or the equiv...)...

It is 25% more data per pixel, not per dimension.  With the same number of pixels, that comes to 25% more data in total.

But yes, I've been wondering the same thing.

Edit: That is to say, I understand that, at the same bit rate, a 10-bit source could potentially lead to a higher quality encode than an 8-bit source, assuming that the 10-bit to 8-bit decimation introduced at least comparable levels of lossyness as the reduction in bit rate from the uncompressed version.  So in that case, I expect you could potentially get similar quality for lower bit rate.  But I would expect the encoding quality for the 8-bit version to be primarily limited by the bit rate, not the colour depth.  In which case I'd expect roughly the same quality for the same bit rate, regardless of 8-bit vs 10-bit.

to me it seems the 3xr vs Revo, 3xR wins it just looks a little bit more detailed, and it looks like Revo is the one slightly blurrier

(reviewed on a 27" 2560x1440 H-IPS)

and it is indeed a shame of the compressed HI10p release....

@ those above :
I agree with a lot of things and there is a lot to discuss and I'll quote

I experienced this a fgew weeks back with image compression and it's hard to disagree with that sentence.
The fact or the matter is, watch the many different versions like I did many times (for comparison sake) and it's hard to notice wich release is wich. Starring at screenshots will point out diffferences wich is the point of release comparison, but during playback ?
Try it yourself, grab REVO or BSS and try to guess by messing up the file names and guess wich is wich.

XxX Betra

----
100% agree... which is why I looked at it on my hi rez monitor -- when a video is in motion -- hell, I can easily watch a 480p w/o it being too noticeable.  The only time it really makes a difference if the the source material is really high quality, -- and very few anime are source-created in sufficiently high resolution for me to really notice a difference when it is   in motion...  but then that's the mark of a detailed video -- when you stop-frame it and zoom in and see detail...I *usually* don't do that.  But since the screen shots were there and someone took the trouble to put them up...well, I had to look!
[quote=eba]
It is 25% more data per pixel, not per dimension.  With the same number of pixels, that comes to 25% more data in total.
----

Ok, say you have a picture that 2 bit/pix and is 2pixels high x 2pixels high. That would give us a size of 8 bits total.

If we increase the data by 50%/pixel (I''m using 'easy' numbers')... we would now have 3bits/pixel or ...um
12...um..oh...nevermind.   ...     ...well still..  25% more info is still 25% more...to expect the files is going AGAINST the grain -- just not at the squared velocity I thought it was..  

\o/
In the end as far as I'm concerned I'm gonna keep 3xR in H10p since my hard drives are starting to get full ><' Now I need to find a french sub ...

XxX Betra

This movie was absolutely AMAZING!