That’s a lot of bits

Posted on April 7, 2009 by Stephen Dawson

I’ve just been analysing the new Blu-ray release of the 1958 musical Gigi. This gets Dolby TrueHD 5.1 channel sound for the English track, backed up with Dolby Digital 5.1 at 640kbps. Foreign language audio is provided in French, German, Italian and two varieties of Spanish (Castellano and Español). All of those are in mono, Dolby Digital 1.0, no less, with a bitrate of 192kbps. Except for Español, anyway. It, also, is in Dolby Digital 1.0, but its bitrate is a massive 640kbps! I suspect someone may have pressed the wrong button in the mastering suite.

Posted in Audio, Blu-ray, Disc details | Leave a comment

Finalising The Fifth Element

Posted on April 5, 2009 by Stephen Dawson

Well, I have finally completed my almost-proof that the Australian Blu-ray version of The Fifth Element has identical video to the US remastered version (see two posts down). To do this I extracted the first 999 ‘I’ frames from each version, saving them in *.png format with identical parameters. I then compared the resulting files, one from each of 999 matching pairs against the other one, using the DOS ‘fc /b’ command. In every case the result was ‘FC: no differences encountered’.

These 999 frames covered the first 738 seconds, or 12 minutes and 18 seconds of the movie, which took it up to the start of Chapter 3. It is possible, I suppose, that the 17 or so non-captured frames for every captured one may exhibit some differences, but it seems exceedingly unlikely.

Posted in Blu-ray, Disc details | Leave a comment

7669

Posted on April 4, 2009 by Stephen Dawson

While examining The Fifth Element I noticed that there was a one minute long video file on the disc, with no apparent linkage from the menu. So I dragged it into a video player and discovered that it was a set of four test patterns, shown below (although they are of course delivered with a resolution of 1,920 by 1,080 pixels). They were being delivered at a health 15,013kbps in MPEG2 format. Each appears for fifteen seconds.

Test Pattern 1

Test Pattern 2

Test Pattern 3

Test Pattern 4

That triggered a memory, and a quick bit of googling determined the code: 7669. That is, when the main menu is showing, you key in that string of digits on the remote and this starts the test patterns running. Something I read suggested that this applied to all Sony Blu-ray discs. So here I am, as I am typing this entry, I’m checking them all one by one. So far I’ve done 24 and every single one of them has this set of test patterns, accessible the same way, except for a couple of US imports (The Terminator, and the original version of The Fifth Element), which have two minute test pattern tracks with four additional patterns.

UPDATE: (Sunday, 5 April 2009, 10:32 pm): I’ve checked all 58 of the Sony Blu-ray discs I have here. Every single one of them has this … except for three: Across the Universe, Almost Famous, and Good Luck Chuck. Every Australian disc has the one minute version. All three of the US discs (mentioned above, plus the Sony version of The House of Flying Daggers, have the two minute version. What a strange difference!

UPDATE 2: (Monday, 6 April 2009, 11:12 pm): Craig from Channel 9 reminds me that, of course, 7669 isn’t just some arbitrary number. It is the telephone keypad sequence for typing ‘SONY’.

Posted in Blu-ray, Testing, Video | Leave a comment

Slander Sony through jumping to conclusions

Posted on April 3, 2009 by Stephen Dawson

Thanks to reader James from Sydney for lending me The Fifth Element on Blu-ray. It arrived in the mail this morning. Thanks also to Theng from Perth, who also offered to make it available.

The reason I wanted to examine the disc is because it has been alleged that the version made available in Australia by Sony Pictures Home Entertainment is a substandard version, earlier withdrawn from the US market in the face of complaints. The allegation appears here, for example, and here. The source for these claims appeared to be here.

This posts were brought to my attention by a person who had purchased this movie on Blu-ray in Australia, and was consequently concerned that he had a substandard version. I weighed into the thread, which starts here, in case you want to follow the whole thing through. My review of the Australian release of this movie on Blu-ray suggested that our version was very different to the original substandard US release.

In that review, I made a mistake. I assumed that the Australian version was the same as the remastered US version. So towards the end of the first page of the thread, I brilliantly muddied the waters by claiming that the US remaster does not have Dolby TrueHD sound, since the Australian version didn’t. I should have spent a couple of minutes checking trusted sites, because that would have shown that the US version does indeed have Dolby TrueHD (plus LPCM 5.1) for English, whereas the Australian version has Dolby Digital and LPCM sound for English.

Nonetheless, it was obvious that the Australian version was not the original shoddy US version for several reasons:

  • the Australian version is a dual layer disc, the old US one was single layer;
  • the Australian version is 35.70GB in size, the old US one was 22.81GB;
  • the Australian version has an MPEG4 AVC video transfer, the old US one had MPEG2;
  • the Australian version has four audio tracks: English and Spanish, each in both LPCM 5.1 and Dolby Digital 5.1; the old US one had three tracks: English in LPCM 5.1 and Dolby Digital 5.1, and French in Dolby Digital 5.1;
  • the Australian version has its movie in file 00011.m2ts (35.3GB), the old US one had its movie in file 00001.m2ts (21.7GB)
  • the Australian version has a ‘Fact Track’ as its only special feature, the old US one also had previews for Resident Evil: Apocalypse, and Underworld: Evolution.

Oh, and of course, the picture quality was obviously better in the Australian version!

But there was one problem with my analysis. The Australian version I was using was a ‘Testmold’ disc from the Sony DADC mastering centre. It was not an actual off-the-shelf Australian disc, but what Sony anticipated it would be. The theoretical possibility remained that Sony had ‘dumped’ the poor discs in the market, while advising me otherwise. I had already made one assumption too many. Thus my plea for the real thing.

Thanks to James, it is here. I can say already that it measures virtually the same as my Testmold. All the above points are identical. The only difference is that BDInfo 0.5.2 tells me that James’ purchased disc is 38,369,228,794 bytes in size, whereas my Testmold is 38,368,180,218 bytes in size. I’m not quite sure why yet. (Yet the plot thickens. Having just rescanned it, BDInfo now tells me that my Testmold is identical in size to the purchased disc! Same software, same drive, same computer. Just a couple of weeks apart.)

So identical are they that when I extracted the first 250 ‘I’ frames from each and compared them bit-by-bit, they were also identical.

You can identify the Australian release by the yellow ‘PG’ classification on the cover, by the barcode number 9 317731 044236 on the rear, and the disc code E-25275-BD printed on the disc’s label (it also appears on the box spine as BD25275).

So, in summary, there are at least three versions of The Fifth Element out on Blu-ray:

  • the original single layer US version, released in 2006, with poor quality MPEG2 video and LPCM/Dolby Digital sound;
  • the remastered dual layer US version, released in 2007, with high quality MPEG4 AVC video and LPCM/Dolby TrueHD sound; and
  • the dual layer Australian version, released in 2008, with high quality MPEG4 AVC video and LPCM/Dolby Digital sound.

All three of these are Region Free.

I think it is highly likely that the video of the Australian version is identical to that of the US version. I hope to confirm this in the near future, since my brother has the US remaster and I can use this for a comparison. In the meantime, I note that the file sizes of the main movie in the US remaster (info taken from here) and the Australian version are very similar, with the differences being easily accounted for by the the different audio standards (and subtitles). I also note that this review of the US remaster contains a screen shot comparison just a few frames away from one of my own, and the differences are very similar to the differences shown by mine.

The only technical reason one might prefer the remastered US version to Australian version is that the Dolby TrueHD sound is claimed to 20 bit rather than the 16 bits of the LPCM (the former is apparently reported as being 24 bit by BDInfo). However, it is suggested here (first post, under the heading ‘Dolby TrueHD (lossless compression) 20-bit/48kHz’) — on what basis I do not know — that the 20 bits is a mere upconversion from 16 bits, which would accord with standard recording practices back in the late 90s when this movie was made.

So feel free to buy whichever one you want, confident in the knowledge that you will have an excellent rendition of the movie.

Be aware, though, that Sony Pictures Home Entertainment has recently sold the rights to this movie (to who I don’t know). When the retailer’s stocks are gone, you’ll have to hold out for the new company to release its version (of course, it might be the same as the Sony one, depending on what they’ve purchased from Sony).

And one final quirk in all this. According to this review of the US remaster, the only special extra on that disc is the fact track. Noticably not mentioned are the two previews that appeared on the original low quality Blu-ray disc. Now according to this BDInfo scan of the remastered US version, there are two other significant files on this disc. One is 2:29 in length, the other 2:21. As it happens, those are the lengths of the two previews on the original lousy disc. Furthermore, the file size of the longer one, reported as 348,463,104 bytes, is exactly the same as the file size of the Resident Evil: Apocalypse preview on the original lousy disc, while the file size of the other is very slightly (0.17%) larger than that of the other preview on the lousy disc.

Did that review overlook the previews? Or did Sony accidentally leave those files on the remaster, without any link to them? Those files are not on the Australian version.

UPDATE (Friday, 3 April 2009, 5:44 pm): For the past 4.5 hours BDInfo has been scanning the Australian Blu-ray release of The Fifth Element on my computer. This is the result:

Fifth Element bitrate chart

The average bitrate for the movie, as reported by BDInfo, is 27,880kbps. That’s a rather higher than average bitrate for Blu-ray.

It is also the exact same bitrate as reported for the US version here (first post, under the heading ‘Dolby TrueHD (lossless compression) 20-bit/48kHz’, scroll to the right). It seems to me that this is getting very close to conclusively confirming that the video on the Australian version is the same encode of the same copy of the film, as used in the US remaster.

UPDATE 2 (Sunday, 5 April 2009, 10:10 pm): Exceedingly strong evidence that the video is identical between the Australian version and the US remaster is laid out above.

Posted in Blu-ray, Disc details, Rant | Leave a comment

Is the Australian Blu-ray of The Fifth Element good quality?

Posted on April 1, 2009 by Stephen Dawson

It has been alleged on a forum that Sony Pictures Home Entertainment has dumped its substandard initial Blu-ray encode of The Fifth Element on the Australian market. I find this highly unlikely, but have been unable to get ahold of a copy of the disc to check for myself. I have a pressing, but it’s a pre-release version sent by Sony. I cannot guarantee that this is the same as that sold in the shops.

Are there any Australian readers who purchased this movie on Blu-ray from a regular Australian retailer who would be prepared to lend me a copy for a couple of days? I would like to extract technical information from it to see whether it is as claimed on that forum or, as seems to me more likely, the Australian disc offers the same glorious picture quality as the remastered US version. I don’t want to buy one myself because I already have four copies of the movie in different formats!

Posted in Blu-ray, Disc details | Leave a comment

Serenity re-encode improves quality

Posted on March 30, 2009 by Stephen Dawson

A couple of posts ago I mentioned that the Universal Pictures (Australasia) Pty Ltd Blu-ray release of First Blood had an identical video encode to the previous HD DVD release. I have now checked both Total Recall and Serenity, which Universal had also sent me. Total Recall, as can be seen here, is also identical to the earlier HD DVD. I confirmed that by comparing 737 ‘I’ frames from each and finding them bit-perfect matches.

Serenity, though, is very different. I shall write it up in due course in the appropriate place. Here, though, I will mention that besides the obvious differences (the Blu-ray has two PIP functions, for example), when I extracted frames and compared them, they were quite different. The ‘I’ frames were even in different places for the most part.

The extraction of frames from the HD DVD version fell over after only the first one hundred ‘I’ frames. I don’t know why. So I had a limited number of chances to find matching frames between the HD DVD and Blu-ray versions. But I did manage to find one. Here it is, shrunk down:

Serenity frame

Now here is a 250 by 300 pixel detail from that frame, unscaled, and compressed only in the lossless PNG format, with the HD DVD capture on the left and the Blu-ray capture on the right:

Serenity HD DVD vs Blu-ray comparison

The differences are there, but by no means obvious. In fact, I chose this section because it was with the forested hills that the differences were most marked. If you examine them closely, you will see that the dark parts are a little darker on the Blu-ray, and the detail a bit clearer. It’s as though a fine veil or haze overlays the HD DVD version. But an almost transparent one, just enough to knock the edge of clarity off.

The full PNG frames from which I extracted these details were the following sizes: 2,201,983 bytes for the HD DVD extract, and 2,421,047 bytes for the Blu-ray. In other words, the HD DVD frame was 91% of the size of the Blu-ray frame. That suggests that it was inherently more compressible than the Blu-ray version, given that the application I used employed the same compression parameters (for example, it produces identical HD DVD and Blu-ray PNG shots for Total Recall, Corpse Bride and First Blood). That in turn suggests that there truly is more detail in the Blu-ray version.

I expect to have more up under the review in a few days.

Posted in Blu-ray, Disc details, HD DVD | Leave a comment

Wasting Space?

Posted on March 29, 2009 by Stephen Dawson

Last night I watched Pale Rider, a Clint Eastwood Western released on Blu-ray by Warner Bros. I strongly suspect it is the same as Warner Bros releases of the movie elsewhere in the world.

I find this disc interesting for a couple of reasons. One is that blu-ray.highdefdigest.com gives it a good write up on picture quality, whereas I thought it was one of the weakest titles I’ve seen. Mind you, I’d just finished watching the deep, deep blacks of Sweeney Todd. Still, the blacks of this movie were a deep muddy brown, and there was a persistent paler band down the right hand side of the frame. Some brief sections of the film were clearly out of focus, for which I suppose I should blame cinematography, rather than transfer. The 20.88Mbps average video bitrate for the VC1 encode is probably higher than average for Warner Bros.

The other reason is the size of the disc. Both the highdefdigest review and Nero DiscSpeed 4 report that this is a dual layer disc. The maximum capacity of a single layer disc is, according to Wikipedia, 23.28GB (using 1,024 as the divisor for the ‘thousands’). This disc comes to 23.15GB, measured using BDInfo, AnyDVD HD and dir /s. So this might be the first disc that actually wastes more that half its capacity.

Posted in Blu-ray, Disc details | Leave a comment

Know It All

Posted on March 24, 2009 by Stephen Dawson

Since 2004 I have been writing a column for Geare magazine called ‘Know It All’. In each column I endeavour to explain how some bit or other of technology, or other working thing, actually does work.

I’ve uploaded the first 26 columns here. They include guns, power distribution, fuel cells, SCUBA, even us!

From the column on digital signals:

To understand, and accept, digital signals requires us to accept that analogue signal transfer is not perfect. What digital allows is for us to embrace those imperfections, draw a line in the sand on them and say, ‘This far, and no further.’You see, digital signals are inherently inaccurate. Instead of trying to transfer an analogue signal from form to form, device to device, perfectly (but always doomed to fall short), it imposes a defined level of imperfection on the signal, and then absolutely refuses to let it get worse.

Another suggests that switching off your TV at the power point is rather a waste of time.

Check them out. There’s about twenty thousand words of reading for those who are interested!

Posted in General Tech, How Things Work | Leave a comment

Windows Vulnerabilities

Posted on March 24, 2009 by Stephen Dawson

A number of years ago I wrote a review of Microsoft Word for Windows 6.0 in which I concluded it has a significant list of problems, making it a lousy word processor. Except for all the other word processors on the market, which were even worse.

These days, Microsoft receives a hammering for poor security in Windows, in Windows Explorer, and in any other of its products where a security flaw is found. This has seeped into cultural discourse and belief as evidence that Microsoft has poor programmers, and just doesn’t care anyway. That, notwithstanding the fortnightly updates and patches issued by the company.

My view has been that Microsoft is pretty good at what it does. Very good, in fact. Indeed, I suggest that it is its very success that has made it seem so vulnerable to viruses.

You see, Microsoft dominates the operating system market with around a ninety per cent market share.

So let’s say that you are one of the twisted individuals who wish to wreak havoc throughout the Internet, or upon strangers’ computers, whether for your own warped gratification or for money. Would you write your virus or worm or trojan horse or whatever for Macs? Linux boxes? Or course not. You’d concentrate on where the volume is. That is: Windows.

My brother draws my attention to this story about a hacking convention:

Security researcher Charlie Miller hacked Safari in just 10 seconds, then used a remote-execution exploit to take over the up-to-date MacBook and make it do his dirty bidding. Firefox and Internet Explorer 8 (which you can download at noon today) fell within a few hours to Nils, a master’s student who busted all three browsers wide open.

So the main Mac browser was broken in ten seconds, while the main Windows browsers took ‘a few hours’. All those attacks against Windows and Internet Explorer seem to have paid off with a more secure product.

Posted in Computer | Leave a comment

DTS-HD High Resolution

Posted on March 23, 2009 by Stephen Dawson

I’ve been after a disc — any disc — carrying DTS-HD High Resolution audio. This is, of course, the improved version of DTS. It remains lossy, but can carry more bits of data, so it is less lossy than regular DTS. While regular DTS maxes out at 1.5Mbps (1,536 or 1,509kbps, depending on the tool you use to report it), on HD DVD, DTS-HD HR could use up to 3Mbps and on Blu-ray up to 6Mbps (see here).

Universal has now sent me three of them: First Blood, Total Recall and House of Flying Daggers. In each case, DTS-HD HR is used for some of the non-English audio tracks.

I also have the first two in their HD DVD incarnations, and in both cases these claimed also to carry their non-English sound as DTS-HD HR. And maybe they did, but if so I couldn’t see or hear it. According to the info display on the Toshiba XE-1 HD DVD player (with the latest 4.0 firmware), this audio was regular DTS. Setting the player to bitstream the audio signal in original format, my Yamaha RX-V3900 informs me that the signal is indeed regular DTS. Except that the Yamaha doesn’t report its bitrate. Yet the Yamaha happily reports the 1,536kbps of DTS from a DVD played on the Toshiba. On the other hand, it won’t report the bitrate from the regular DTS track on the HD DVD of The Italian Job.

So as far as I can work out, the claimed DTS-HD HR on those HD DVDs isn’t there.

But it is definitely there on First Blood and House of Flying Daggers, and I have no doubt will be on Total Recall as well when I get around to examining it. Thanks to BDInfo, I can say that the DTS-HD HR on First Blood has a bitrate of 2,046kbps.

Meanwhile, I compared the video of First Blood between the Blu-ray and HD DVD versions. I am pretty confident that they are identical VC1 encodes. To confirm, I extracted the first 257 ‘I’ frames from each and compared the resulting graphics files. They were bit perfect matches, every single one of them.

Posted in Audio, Blu-ray, Compression, Disc details, Testing | Leave a comment