HD For Indies

Tom's Hardware Guide: Tom's Hard News has an excellent article about why Microsoft is backing HD DVD rather than Blu Ray. Initially neutral, Microsoft yesterday (and I didn't report it, to busy sorry) announced they were backing HD DVD.

I'm always quick to suspect Microsoft of skullduggery, and some of the interactivity standards stuff still might be questionable, but ASSUMING the other factors are correct, HD DVD is making a lot more sense. Here's the rundown of why MS decided to go HD DVD - they had 6 key areas of interest/functionality:

1.) HD DVD has a plan to let users make authorized copies to hard drives for home theater/media PC usage. (I'm guessing it'll require sticking the disc in every so often so that the "rent/rip/return" model used for DVDs won't be valid). Blu Ray is noncommital on this issue (perhaps Sony, the big backer behind Blu Ray, as a movie producer as well as manufacturer doesn't like that idea?)

2.) Support for hybrid discs - both have talked about it, but Blu Ray's combo disc with standard def DVD on one side, high def on the other doesn't exist outside the lab, and is nowhere near manufacturable.

3.) Keeping production costs low - HD DVD has always been rumored to be lower production costs, since it is closer in spec to regular DVDs. Blu Ray looks to be more expensive.

4.) Similarly - low disc replication costs - same kind of thing. Keep costs low to make it market viable.

5.) Disc storage capacity - while Blu Ray looks like an obvious winner with single layer capacities of 25GB vs HD DVD's 15GB, Blu Ray's promise of 50GB with dual layer looks like vaporware, whereas 30GB HD DVDs look very doable and manufacturable. Of course, that's just for read only media. I'd imagine that the evolution of single vs. dual layer burnables would be similar to that of DVDs - it's been years that we've had single layer burners, but dual layer burners are new in the last year or so.

6.) Interactivity standards - Blu Ray was going with a Java based solution, which major studios purportedly don't like as complex and unwieldy (I can't comment on that, dunno).

Mike's (further) analysis: All of this is pretty suprising. When I read that MS was backing HD DVD, I figured it was because they had a hand in developing the interactivity layer of HD DVD and didn't want Blu Ray's competing Java based standard. MS wants the world dependent on an MS based standard, period.

But the other reasons are surprising - now that the technologies are getting close, and actual, deliverable product is getting close as opposed to plans and intentions, HD DVD is looking better if all the above comments are true about what is and isn't ready for mass production.

One thing that has always bugged me about Blu Ray was the fact that the media gets written 1/10th of a millimeter away from the surface of the disk - what about scratches? In theory, I'd love to see Blu Ray win, but it sounds like they are way behind in their implementation.

I'd figured that Sony's inclusion of a Blu Ray player in the PS3 would be a massive stealth push of Blu Ray discs into the population, but with both Intel and Microsoft backing HD DVD, the computer world is now wrapped around HD DVD pretty tightly. Where will Blu Ray support come from for the Windows users of the world? That gets pretty confusing, and if there isn't native Windows driver level support built in, that's a huge blow to Sony and their camp.

So it might be HD DVD that wins this one by default from the computer side of things - remember, DVD-ROMs were the stealth door in for consumers to get used to DVD Video.

OK, so why no posts for a couple of days? Let's see:

-record heat in Austin - old records beat by 7 or more degrees, new records every day for the last WEEK
-108 on Sunday
-no air conditioning since Sunday
-IRS has mistakenly decided I owe them money, have to jump through hoops to show that it doesn't owe taxes for a period in which that business had no income
-17 year old cat goes missing, in 100+ degree heat, moving back from place-other-than-home

Oh, and then on top of all that, the part you guys care about - we have been working on a project to color correct a 15 minute short shot in HDV at 1080i60 using Final Touch HD. How's it going? Colorist's done coloring and I'm dickering with some XML issues to get all the rendered stuff back into Final Cut exactly right. The results look great, I'm feeling better about HDV's ability to accept moderately aggressive color correction.

My gut still says that as a format, regardless of camera of origination, HDV doesn't hold up to HDCAM (as expected), but that it is doing better than I thought it would. HDV is definitely the everyman's HD in much the same way that DV is the everyman's SD, but that you need bigger/better/faster everything to handle it in an equivalent fashion.

OK, gotta blaze....

-mike

UPDATE - Cat found, AC fixed, IRS hordes pushed back across the Rubicon, client snargle managed.

It's all about problem solving.

: )

Total Training: Coming Soon

Well durnit, I wanted to be first (and I still could be if I hurried), but Total Training has announced they'll ship in a month or two HD versions of their training materials for Final Cut Pro and other applications.

-mike

CinemaTech: SPMTE Digital Cinema forum in November

They'll be talking about the DCI specification, JPEG2000 compression, and NATO (National Association of Theater Owners, not the North Atlantic Treaty Organization) will discuss their present thinking on a rollout model for digital cinema.

That last part I think will be the most newsworthy. I've blogged extensively about the DCI spec here, here, here, here, and find more in the April 2004, and April 2005 blog Archives.

It's in NYC in November, I can't attend (I'm in Austin, TX), but if any reader can attend I'd love to see a report on it. Please email if you'll be able to attend and would like some free publicity.

-mike

CinemaTech: Links from vfxblog:

More fun stuff via CinemaTech:

"Don't the guys at vfxblog ever sleep?They've got great posts on: A visual effects featurette from 'The Lion, The Witch, and the Wardrobe' An interview with Mike Johnson, co-director of 'The Corpse Bride' A Hollywood Reporter piece about the Jim Henson Company's virtual puppetteering tool, called the Henson Digital Performance System"

New Moves To Eliminate Online Piracy - Yahoo! News

I have to quote the article:

The music and film industries are taking a new approach in their efforts to thwart illicit file-sharing on the Internet. They are releasing free software that can detect and remove programs used on PCs to swap copyrighted tunes and movies.


Mike's Comments: HAHAHAHAHHAhahhaaaaaa......LAAAAAAAAAAMMMMMMMMEEEE.

Ahem. Excuse me. Had to get that out of my throat.

OK, so here's the scoop - by making this software available for free (can you imagine anyone buying it? I mean, really, who'd want to?), this sets up a circumstance whereby at some point the record or movie industry could claim that a business, university, or other institution could be claimed to have been negligent by NOT running this software.

The same game that the SPA (Software Publishers Association) plays with businesses when a source (read disgruntled employee) reports that there are copyright violations (bootleg software) on site, then the SPA comes in with all 3 left feet, finding all installed software and installers and demanding proof of ownership of everything, whether in use, installed or not. Ugh. Ugly. An acquaintance of mine had a 99 user key that was not his in a text file, there was concern that his employer would get hit for a 99 seat license that they never used.

By shipping this software, the industry is trying to establish a precedent for them to be able to do the same thing.

Would this eventually establish a precedent whereby by joining a network you give permission for your hard drive contents to be searched? I could see the content industry wanting to do that, but there is no such precedent in any other physical medium. You bought a CD from Best Buy? Best Buy's private cops can search your home. Right. I don't think so.

Anyway, no sane private citizen derives any benefit from running this software, and businesses only benefit from preventing possible lawsuits - avoiding a negative outcome rather than seeking a positive one for their efforts.

Blah. This kind of thing ticks me off.

If I try to put on my producer hat, however, I try to look at it this way - if I were making indie movie content, would something like this help me?

Probably not. This software and future similar software will probably only be searching for content from major publishers. As a small indie, the ability to get your content on their "what the software looks for list" will probably be a challenge, only benefitting the big players.

So feh.

-mike

Gadgeteer has an article about a "plague" that spread in a game. In short, a Big Bad Guy could cast a spell that others carried with them like an infection, and like any virus, it spread in a hurry. There's a teleport feature that allowed players to spread it to other distant lands in minutes.

This models/mimics real life epidemics, but this is the first virtual infection I've heard of.

What does this have to do with HD? Absolutely nothing. But as a geek, I find this very interesting. I'm reading Snow Crash to Melissa (girlfriend) at night before bedtime, which involves a virtual virus as well (and then some), and has long been on my short list of Favorite Books of All Time. (Neal Stephenson is definitely my favorite author.)

In any case, reality is catching up with science fiction realllllly fast.

If you haven't read Snow Crash or The Diamond Age (a sort of sequel), do yourself a favor and start.

-mike

Apple wants Intel chips early, but that may not matter.

The chipset that Apple is believed to be basing their first MacTels on are the Woodcrest and Merom chipsets, which won't be ready until 3rd quarter 2006, well after June 2006 shipping date Apple has stated.

The article states "unless it has decided to use Yonah processors." My understanding is that Apple DID plan on using Yonah processors from the get-go, so this article and it's fearmongering headline might just be wrong from the start.

But what does sound plausible is that if Apple were trying to push Intel to ship chips faster, Apple doesn't have the leverage to do so, what with Dell shipping 20 times as much product. In that scenario, why exactly would Intel give preferential treatment to Apple instead of Dell? Exactly. Not.

-mike

Reader Mail time:

Michael from Germany wrote:


Hi,

Living in a PAL region I just learned that the new FINAL CUT STUDIO does NOT work with my new JVC HD101 in 720p/25. I can't believe this since this is the ordinary format to film with this camcorder. In iMovie all scenes run a little too fast (30fps?)?! (I ordered FINAL CUT STUDIO six weeks ago and the Apple store sent just the books and no DVDs. I'm waiting for the software since then... - in vain, because they will be of no use??)

Regards from Germany,
Michael


My response:

Natively, no. Nor is 720p24 supported, either, and this has been a known shortcoming of FCP since NAB. Purportedly, LumiereHD 2.0 will support non-30fps rates, but it was supposed to ship back in July and I haven't heard anything new in a while. Last time I inquired, I was told it was coming along but slowly.

I pinged Frederic Haubrich over at Lumiere HD and he said:

No set dealine but we are working on it and will support the 25p mode as well.


-mike

Ahmed wrote in:

Hello
Does the Kona 2 support all the Sony HVR-Z1U HDV codecs on Final Cut Pro and
allows us to do Real Time Preview?
thanx


Mike's Response

Yes, it does, and I've been using it quite a bit recently. I can personally verify that HDV 1080i60 works fine with Kona2 with realtime effects, handles the scaling and everything. Checking their website, they list HDV support. The Sony camera only "really" does two framerates, 1080i50 and 1080i60. The CineFrame mode is still written to tape at 30i. So yes, the Kona2 does support the Sony's HDV formats. (And so do the BlackMagic DeckLink HD cards).

-mike

Reader John wrote in with a smart question:

Hi Mike,

How come it is more expensive (not done) to have larger image sensors with the same resolution as these consumer HDV cameras? I thought the expense would be in cramming more pixels onto a smaller surface?

There are so many products aimed at trying to emulate the 35mm depth of field, and it is such an important issue for indy cinematographers, why has no one tried it yet?

I know the professional cameras have larger chips, but they (have more ‘lines of resolution’ and) haven’t really tried to come close to the size of a 35mm gate (except for maybe the more recent 4k uber-expensive offerings).

I was hoping Kinetta was exploring this, but now I realize that they are going for a higher rez, more expensive price point.

You’d think you build a sensor with old SD technology, only bigger, you’d end up with a pretty good sized HD sensor.

Thanks,

John.


Mike's Response

Yeah, I've wondered the same thing too, and asked around a bit. Ted Schilowitz from AJA and Graeme Nattress of Nattress plugins have been good sources of info on this stuff, but if I have it wrong it's my bad not theirs.

My limited understanding of the issue is as follows (Graeme or other more knowledgeable folks feel free to append/correct and I'll post up here as well):

-cramming more pixel sensors onto a smaller chip has an inherently bad effect on the resolving power and contrast capability of the imager.
-larger pixels on larger chips provide better contrast, true resolution, better signal to noise, and all kinds of other goodness. MTF gets involved in here too.
-manufacturing larger chips is expensive - the bigger the chip, the more costly the per unit cost. Chips are made on wafers (think CPU chips as well), and the more they can cram on each wafer, the better the profit margin, since there is somewhat of a fixed cost per wafer of a given size and quality.
-therefore, manufacturers try to make their chips as small as possible in order to get as many usable chips out of a given wafer
-From a straight business/manufacturing point of view, to try to make a 35mm chip would not only cost significantly more to make if you were making them in the quantities involved in Sony Z1U quantities, if you WEREN'T making them in the thousands, and it was for a small manufacturing run (lower quantities), costs would spike substantially even higher as the manufacturer would have to amortize their R&D, marketing, etc. costs onto a smaller number of chips. If it took the same amount of R&D to make a 35mm chip (large assumption), but you only sold fewer, you have to divide the R&D costs across that smaller number of units - higher costs.
-Kinetta is specifically and definitely using a bigger imaging chip, with the long term goal of using a 35mm chip
-the Drake camera is also trying to do larger imagers as well

-mike

Toast 7 Titanium (Review)

This is slightly off topic, but for Macheads Toast has long been used for burning discs. It now has various video encoding capabilities.

I wouldn't use it's encoding for serious projects, but for tests, comps, client review, etc., could be handy.

Macworld UK - Aspyr releases DOOM 3 demo

Doom 3 for Mac. Time to see how that dual 2.7 GHz G5 with ATI Radeon 850XT really holds up.

-mike

Ulead announces free HDV and MPEG-4 upgrades for VideoStudio 9.

The HDV Plug-in provides Sony HDR-HC1 users with a cost-effective solution to capture, edit, and output high-resolution HDV 1080i video - enabling users to incorporate HDV fully in their video projects, from capture through output.

MPEG-4 is a flexible format rapidly gaining acceptance for applications ranging from DVD-quality playback to video delivered over the Internet and mobile devices.

The Ulead MPEG-4 Plug-in for VideoStudio 9 allows users to easily create video for high-quality playback, streaming over the net or display on mobile phones, PDAs or the Sony PlayStation Portable (PSP).


(found from a link on FresHDV.com

Director (Soderbergh) Zigs From Stars to Nonactors in a New Film - New York Times

Least relevant but most interesting quote:

"'I'm just like Fassbinder,' Mr. Soderbergh says in his pleasant deadpan, 'but without the drugs and the whores.'"

Soderbergh challenges 'out of whack' studios

More info on how Soderbergh and 2929 are challenging the traditional film distribution model of the studios.

It's quite likely that this is the future model of how movies will be distributed.

So read up.

-mike

HD-DVD vs. Blu-ray: And the Winner is...

...potentially holographic storage. I'm not as optimistic about timetables, but the idea of 200GB optical storage in about a year is pretty enticing. With an eventual capacity of nearly 4 terabytes and a throughput of 1Gbps (that's 125 MB/sec), that's a pretty darned useful backup media.

Read the article for more.

The author agrees with my premise that HD DVD and Blu Ray will flounder in the market with no clear leader, much like DVD-A and Super Audio CDs.

The TelevisionBroadcast.com two minute drill is a daily audio report I listen to. Today they had some further details on devices to record via FireWire the output of the two recent interesting HDV and DVCPRO HD cameras from JVC and Panasonic.

JVC DR-HD100 recorder for GY-HD100 - audio, video, timecode over FireWire

-40 & 80 GB capacities, for >3 and >7 hrs of HDV recording
-HD or SD recording
-SD DV25 as .dv, AVI type 1 or 2, 2 type 24p, QuickTime 30i or 24p (and other AVI formats as well for Matrox etc.), in HD it records 720p24, p25, p30 .m2t files for NLE systems

Focus DTE for the HVX200

Focus DTE (Direct To Edit) FS100DTE for the Panasonic HVX200

-DVCPRO, DVCPRO 50, DVCPRO HD recording in the field, can edit directly from the FireStore expected March 2006 with a price under $2000

UPDATED: Solutions to get 5.1 mix onto an otherwise 4 track capable HDCAM tape have been added, including a specific vendor to use. Live and learn.

Coupla reader mails recently:

5.1 audio mix out of FCP to HDCAM?

Hi Mike,
I have a question… I hope someday to repay you for the wealth of info your blog has provided me with. But first, if you can answer this quickly with out it taking too much of your time I would be very grateful. We have a client that requires a 5.1 mix of an HD film which is being screened at the world’s largest Toyota Show Room in Oman. (where do these bizarre jobs come from?). Anyways, they need 5.1. The audio post house I use out here, assure me that they have loads of experience with 5.1, but not so much with HD video, so rather than have them fart ass around with a deck, I want to see if it is poss for them to export wav files which we can lay into an FCP sequence in our online and playout. Does fcp 5.0 w/ a Kona 2, support 5.1 audio output to HDCam??? Have you had much experience with it? I tried to search your blog without any result…

Much appreciated,
Paul


Paul -

There are multiple pieces to this puzzle:

1.) Does the Kona2 support that many channels of audio out
2.) Does FCP support that many channels of audio out
3.) Does HDCAM support that many channels of audio out, and can they be configured for a proper 5.1

1.) YES, the AJA Kona2 card does support up to 8 channels of audio. Done.

2.) Final Cut Pro 5.x and QuickTime 7.x finally, Finally, FINALLY properly support more than stereo audio output. Version 4.5 of Final Cut Pro does not, Not, NOT.

3.) HDCAM supports 4 channels of audio at 20 bits. Suddenly, OOPS, you're in trouble on that one. (Update: no you're not, see below). I'd take a look at either the Panasonic D-5 format or HDCAM SR. Both will do 10 bit, 4:2:2, full raster (no loss of horizontal resolution). HDCAM is 8 bit 4:2:2 and NOT full raster (only 1440 pixels of horizontal resolution, and 3:1:1 color sampling not 4:2:2 like the other two).

UPDATE: An astute reader wrote in with a fix from a prior project he worked on:

After that I did a little research and found out you can record 5.1 if you use dolby E mix.
here's a excerpt from an article I found online:
"Dolby Laboratories has addressed this problem with the format "Dolby E" Dolby E allows 5.1 surround audio to be compressed into a bit stream taking up the space of 2 channels of audio (stereo). That way, HDCam can record 6 tracks into 2 channels of Dolby E and have it decoded on the other end back into 6 tracks of audio"

Then reader Dave Gutt wrote in:

Mike-
With our 5.1 mixes, the audio post house exports each track, iso, as an aiff file. We resync each track in FCP, then export each track, yet again, as an AIFF. These are then put on a hard drive and sent to a post house for them to encode them to HDCAM tape. They add the 1.5 sync shift, etc.

The post house we use is Digital Arts in NYC. They are very very good.


Next question:

iPod as DV, HDV, or DVCPRO HD field data/video recorder?

hey mike,

i was wondering if the firestore drive does record HDV... i emailed them, but got no answer... but from the technical side there shouldn't be a problem with recording the HDV stream via the firewire out to the firestore setup??? is that solution the only one??? why is there no one out there writing s/w for using the ipod for this (capturing DV/HDV streams to harddrive)??? imagine using the 60G ipod for this, much cheaper and smaller... i thought, seeing the pix of the firestore drive, it's very huge/bulky...

cheers

f


I wrote back:

FireStore is going to have to make new models to work with HDV and DVCPRO HD.

You have to remember, a hard drive is just a dumb storage device - it needs some instruction as to what to do. There has to be some brains in the gadget that understand what the format is, how to segment it into individual take files, know when to start/stop, etc. But there already is an HDV version, and a DVCPRO HD version is in the works for the Panasonic HVX200.

Basically, I think they just need new firmware in the unit, and while the HDV version's throughput requirements are similar to DV's (about 3 1/2 MB/sec), DVCPRO HD has much higher requirements - up to about 15 MB/sec.

So the DVCPRO HD version would need a bigger drive to hold similar quantities of footage, updated firmware, and perhaps bigger batteries to have similar field operation times with that higher data rate footage. My amateur's guess, based on understanding the throughput capabilities of these types of drives, is that no special hard drive will be required, the 60-120GB laptop drives available today should provide sufficient throughput for DVCPRO HD's higher needs.

As a follow up, reader Frederic wrote:

isn't the output thru FW a simple stream? the 25mbps/3.8Mbyte/sec shouldn't be any problem for the internals of even an older ipod... i think it's just a s/w thing to get the ipod to just "record" that stream to HD... the stream doesn't have to be encoded after it leaves the camera, that's why i think there doesn't have to be much procressor power on the ipod side... the advanced features of the firestore i don't need. i only need record, not record, on the ipod, maybe a little window saying how much space is left on the ipod HD....


While in theory that might work *, as a practical matter I don't see Apple going to the trouble of altering their incredibly successful iPod line for this very niche application - they'd leave it up to third part developers to come up with their own solution. Also, keep in mind the iPod was designed to work heavily with a RAM cache - briefly spin up the drive to load a song from the drive into RAM, then spin down the drive until it's time to load another song into RAM. By only having to spin up occassionally, it allows the battery to last a long time. Field recording, on the other hand, would require constantly running the drive, which a.) the drives weren't intended for that kind of usage and would have a vastly shortened lifespan, and b.) the battery wouldn't last long at all under that usage scenario. So I don't expect an iPod mod for direct to disk (or Direct To Edit as FireStore products brand it) anytime soon. But that KIND of an idea - a drive with some brains and a FireWire connection - is entirely valid, as FireStore's DV success has shown to date.

That's a bit like asking your SUV to drive the Baja 1000 or your car to run at 140 miles an hour all day every day - yes, it is vaguely that kind of a device, but is not at all designed for or capable of that kind of heavy duty usage where performance has to be sustained and flawless.

* (and tests have shown that to work in some circumstances - Blackmagic staff uses an iPod in FireWire disk mode to run test DVCPRO HD footage through their systems with footage playing from the iPod itself)

-mike

CinemaTech: Movielabs: The message is still about piracy

Good summary of what MovieLabs appears to be about, and links to the Forbes article.

It is too late and I'm too tired to synopsize, but if you care about how you'll be watching your digital media in a few years at home, or if you care as a content producer about how your content will be protected from piracy, read all you can about this stuff.

Or if you're both a producer and a consumer and want your stuff protected but free access to everyone else's, welcome to the human race - take a number...

-mike

OK, this one's a two parter:

First off, there's the NY Times article:

Hollywood Unites in the Battle to Wipe Out Movie Pirates - New York Times

I love their optimism in that headline that implies "of course they'll succeed!"

Then there's the ripping-it-a-new-one analysis by commentary from Freedom to Tinker

I think they will probably come up with some more restrictive DRM stuff that will, to use the technical term, suck balls * for the regular consumer, but still be circumnavigable by the elite geeks who always crack this stuff anyway. And once those elite geeks crack it, it'll still find it's way onto the P2P boards in short order.

Its pretty much inevitable that if you're trying to keep secrets from someone with encryption, it gets very very difficult to do so when the people you're trying to get the information to are the same people you're trying to keep the information from. And since those same folks get to see both the encrypted input and the decrypted output, it's really tough to keep it all a secret and keep it all secure.

(found from a link on the rockin' CinemaTech.com)

TCode 2.1 - VersionTracker

From the versiontracker.com description:
Product Description:
TCode is a SMPTE compliant time code application to assist editors in logging live or prerecorded shows only requires a OS X compliant sound card and a VTR or Time Code Generator capable of outputing Linear Time Code (LTC). TCode now supports multiple client copies of TCode on the network.

What's new in this version:
Web server & client for wireless logging using wired or 802.11 networking
Time code shows in the menubar, next to the clock
Time is "typed" in any application with a simple hot key
Floating clock window
Autoselecting timecode input
Sync local computer's clock to timecode
Compare timecode to the system clock
Enhanced preference window
PAL, NTSC, & film framerates supported
Simultanious view of multiple LTC inputs


Product Requirements:
Mac OS X 10.3 or later

So I continue to work with Final Touch HD on a daily basis, and that's part of why blogging has slowed down. Also today, for the first time ever, I had to factor in the possibility of impending hurricane into our scheduling efforts - "OK, we'll color correct on Friday, but hold of on the overnight render until Monday because the hurricane might hit Friday night/Saturday morning, and we want the gear off, unplugged, and up off the floor before then." Rita is headed for the Texas coast, and the line of greatest probability last I saw was aiming straight for Austin, my hometown. Fortunately, we're a 3-4 hour drive from the coast, so it'll slow way down and "only" give risk of tornados, not generalized 150 mph winds everywhere. Texas coast - get out if you can, and hunker down wherever you land. Best of luck for this weekend.

Over last week and this weekend we got through our first project in Final Touch HD, a 90 minute DV based doc that mixed 23.98 progresive and 29.97 interlaced footage.

From an engineering/support perspective, when I get new stuff in I like to ease into it, test it a bit at a time, starting simple and slowly adding complexity. To dive into a 90 minute doc with over 500 cuts in one 30 minute section is a bit like taking your brand new car and driving it out of the dealership lot, turning left, and flooring it to begin the Baja 1000 race. (Yes, I have Dust to Glory on the brain, since the DVD has been waiting for me for days to watch.)

Not exactly easing into it.

I would have preferred that we try a music video, then a short film, then maybe a 44 minute TV show, then a feature. This was Fighter Pilot 101 of "Turn key, begin combat. Try not to die." Of course, time was short, there were delays getting started due to Final Cut Pro snafus (that Media Manager, I tell ya - you can't depend on that sucker to behave!). Lesson learned - do NOT expect Media Manager to properly reduce a file, trimming clips down and adding handles, and copying it all over, and relinking, when you've mixed frame rates. Kerblooey.

After that issue was circumvented (in a not pretty, "I say we nuke the entire site from orbit, it's the only way to be sure." kind of a way), we were finally ready to start coloring.

We had a couple of snags in Final Touch HD workflow - we couldn't assign saved looks to keyframes, which I found dumbfounding. Much, much, much to Silicon Color's credit, they said while they had not thought of doing that, it certainly made sense, and made a whole new build of the software with that feature available to us and everyone else within about 24 hours. WOW. THAT is customer service. If you ask Apple or Adobe to roll in a new feature, they'll listen, nod, say OK, and maybe bugs will be fixed in weeks or months, and new features added in the next scheduled major release, about once a year. Silicon Color not only fixed what we considered to be a bug (they called it a feature), they also added a desired feature in the next build too within 24 hours. Damn, that rocks like nobody's business. That gave me great, great confidence that problems encountered will be addressed in a timely manner by these guys.

Colorist put in some very serious time over the weekend to get it colored, and I babysat the renders Sunday afternoon and evening to get the deliverable (QuickTime movies) to the client so she could add graphics etc. and kick out a DVD for a screening to some investors/clients/distributors/unknown important people of some sort.

As for Final Touch HD, there are definitely still some known issues that I'd like to see resolved, but Andrew, Roland, Brian and the rest of the team are clearly open and receptive to figuring out ways to fix these things. There are a lot of little UI quirks that bug me, but they don't absolutely prevent the software from working, just slow you down. There are still some functionality things that I wish were there - the control surfaces aren't being used to their full potential, for instance, there are unassigned buttons & knobs that could be used for features that aren't in use yet. I'd like to see some modal usage - hit a button and the trackballs become positioners for masks rather than secondary color correction adjusters, for instance - that would increase operator efficiency. XML import and export hasn't made me entirely happy to date, and that's a very scary thing that I want to see nailed down rock solid perfect for this application to be useful as intended, and not a "yeah, it's neat, but doesn't live up to it's potential." I'd hate to see that come to pass. I've had a couple of test projects that haven't flowed in and out as smoothly as I'd like (or in one case, at all, problem as yet undetermined). It's still a young product from a small team that is rough around the edges - if Adobe or Apple shipped something like this, a lot of folks would throw it back over the fence (but the bleeding edgers wouldn't). But it is improving quickly, and that is what will make or break it long term. And that small team is also MUCH more responsive to small fixes and changes (overnight? Hullo, does ANYBODY else do that in the graphics software world?). My one concern would be whether they could fix BIG things with that small team, especially if it involved something beyond their control (such as some XML issues possibly). Time will tell, but I'm optimistic today.

Anyway, back to the present nitty gritty at hand:

So we got through that first project and the clients seem happy - they charged off with it to get ready for their screening and we haven't heard a peep back, so they are presumably happy. I feel quite quite confident I'd have heard back in the wee hours of the morning if there had been a problem.

So today, I spent a couple of hours dealing with some of the workarounds and current shortcomings of FTHD, but I'm not too worried that I'll always have to do this mundane fixit stuff.

Today we began our second project - an 15 minute HDV short film project, shot on a sony HVR-Z1U at 1080i60 by a "real" DoP, not somebody who just bought the camera and thought it'd be fun to make a movie. They are interested in a 24p version of this thing, so I'm debating whether to use this one as a test project for the big 24p test I've been meaning to do to try DVFilm, Graeme Natress' plugins, Magic Bullet, Compressor, Compression Master, Cleaner, After Effects, Algolith, my cat staring funny at it, and anything else I can throw in the mix at it to convert 60i to 24p.

But for coloring, for the moment, we're looking at keeping it 60i. The cool thing about Final Touch is that when we're done, we won't have one giant long single file, we'll have replacement files for Final Cut Pro that will swap out with the existing media so we can then 24p-ize THOSE color corrected clips. I was about to get in a tizzy about having all 25 GB of HDV footage converted to uncompressed 24p (about 25 times as much file space) when I realized I could do the 24p thing after the fact and do an even better job. Whew! What a relief.

So this HDV project will be a good test - it took something like 6+ hours to render out all the 24p SD footage from the 90 minute doc, I'll be curious to see how much slower the HDV project is. In order to get it into FTHD, I need to convert it to a supported codec. Since I'm interested in getting the best possible quality out of it, that means uncompressed 10 bit 1920x1080 60i footage which is biiiiiiiiiiiiiiiiig. It took more than 3 hours to get Media Manager's Recompress To feature to convert the more the 25GB of HDV 1440x1080 source material to 1920x1080 10 bit uncompressed footage.

Tomorrow I'm going to bang on FTHD a bit to make sure everything behaves as expected to prep for our Friday session with the client. Once colorist gets a colorin', I'll probably keep messing around with this 24p stuff.

Oh, and colorist seems very pleased with it as well - we had a long, long discussion today about "Is there anything you could do on your daVinci (hooked up to a telecine or doing tape to tape) at your old place that you can't do here with FTHD?" and we came to the conclusion that the answer was no. He likes the Tangent control surfaces but wish they could do more. I'll get some quotes from him tomorrow.

-mike

WiMax Wireless May Directly Usher Disney Into The Living Room - Yahoo! News

Disney was messing around with movies on demand to be sent over unused TV channels, now they're looking into WiMax, the long distance cousin of WiFi.

Sounds interesting - we'll see if Disney is techno-savvy but talent deprived.

-mike

Macworld UK - Apple Expo: Steve Jobs talks Macintosh

No new Macs.

He did say Apple is on track for June 2006 Intel based Macs (didn't say which one).

I had thought we might see some mini or laptop models as early as Feb/March, but nope, June it is. That leads me to think it'll be second half of 2007 before we see the Heavy Iron (Power Macs, servers) move to Intel.

There are likely to be at least two rounds of new G5 based Macs before we see a new Intel based one. One late this year/early next year, another late 2006/earliest 2007, then new PowerMacs (will they even be called that?) second half 2007.

But, as always, that's just my conjecture.

-mike

Spoke with a client today who was in a bitg of a pinch - he had a G5 with some hard drives in it and was trying to capture 1080p24 footage straight from an F900 after a shoot.

He had a whole slew of problems that an indie could face, let's march through them one at a time:

Capturing uncompressed HD footage directly from an F900, no deck involved:

Normally you'd need a deck to capture material from a tape. These high end HD cameras aren't like DV cameras where you're expected to be able to capture footage from a tape in the camera. They expect you to use an HD tape deck. As such, the cameras don't have all the things you'd want to capture footage in a controlled way - namely, a way to control the deck.

HD decks have a deck control interface port, usually a 9 pin deck control cable. The F900 doesn't have this. So you can't control the deck remotely. "That's OK" one would think, "I'll just use Capture Now in Final Cut Pro to capture footage..." That works technically, but leads to other problems...

"No problem" the client thinks, "the HD-SDI interface on the camera has embedded audio, video, and timecode - I'll get timecode from there." The problem is that HD-SDI capture cards expect timecode to be coming over that 9 pin cable. Woops. So you can't control the deck and you can't get timecode. So any footage capture is "wild" - that is, you have no reproducible means of acquiring that footage again, as an exact match, if you need to later. The captured footage to disk becomes the new master in this sense, unless you want to manually line up the captures by hand - a lame a difficult proposition. There are ways to work with the footage in this context to get a good deliverable, as well as be able to do an offline edit, but it has to be done carefully. If it's an important job, you also want backups in case a drive fails, otherwise you're hosed.

Offline workflow in this context

You can use Media Manager to make offline footage from your HD source. So if you want to use anamorphic DV for your offline edit, or DVCPRO HD 720p24 footage, or whatever suits your needs. Later, you can use Media Manager again to make an offline timeline that you can relink to the original high def source.

Capturing SD from HD source

Working with the client, it was interesting to notice some new options in the BlackMagic capture presets - there are choices for capturing 1080p23.98 as 1080i29.97 footage (presumably adding 3:2 pulldown as you go) and other related options. You can also downconvert upon capture - so does that mean to capture SD 30i from an HD 1080p24 source you set the BlackMagic card to downconvert on input and the Final Cut Pro settings to capture SD 23.98 converted to 30i?

A whole new set of interesting complicated choices. But it DOES mean that if you have access to an HD deck, you can make your own downconverts and get them into your system in realtime without needing a DV deck and two passes. Of course, DV decks are much easier to come by than HD decks, but you get the idea...

RAID difficulties

The client also had recently bought a Swift internal drive mounting kit and three IBM/Hitachi 500 GB drives (the 7K500 series presumably). He also already had two internal 250 GB drives he'd striped into an array. He had bought the Swift and 500GB drives with the express intent of using it for HD work. Now, opinions vary, but this is NOT what I would recommend. For starters, he now has 5 internal drives in a system intended for two. The power supply in the G5 is overbuilt, but I just don't like the idea of loading a ton more drives in there for a variety of reasons, including cooling, noise, and power draw inside the G5. Add to that the near impossibility of moving that array somewhere else if you need to work on it on another system.

He was having some Finder lockups and other oddnesses, too. We went through a bunch of his System Preferences and found a few reason for that and they shouldn't occur again.

The biggest inconsistency I saw was when he did test his 3 disk array with the BlackMagic Disk Speed Utility, he had made a perfectly reasonable guess as to what the proper block size should have been (you now have more choices in Tiger's Disk Utility than you did under Panther). But the performance numbers he was seeing - 135 MB/sec write speed on an empty array - seemed far too low. His read speeds were up around 170 MB/sec, and that seemed too large a discrepancy.

Troubleshooting the RAID

He was using a Sonnet 4+4 eSATA card. We checked, and he was running the latest firmware (v1.2 at the time). So the culprit wasn't there. He tried the block size I suggested and the throughput jumped to 185 MB/sec. So there's a good lesson to be learned - block size can make a SUBSTANTIAL difference in RAID speed.

I also explained that while the data rate for the format he wanted to capture - 8 bit 4:2:2 1080p footage at 23.976 frames per second - was about 96 MB/sec (plus audio), he needed even more throughput from his disk array in order to allow for the overhead of Quicktime and some safety margin. And since hard drives get slower as you write to the inner tracks (as the drive fills up), the array will get slower as it gets full of data. So that 185 MB/sec would likely drop to about 100 or 110 by the time the disk was full - and that would most likely not be enough to capture the footage he wants.

In the end, I recommended he get an external enclosure for his drives, and add a fourth drive to the array to make sure that it would be fast enough all the way through for the kind of work he wanted to be doing (he wanted to finish in 10 bits/channel), and set it up to spec to ensure adequate throughput.

Conclusions

In short, it's very easy to assume that something is going to work and not have it work. I've skipped over the conversation he and I had about trying to get the footage in as uncompressed SD in case he couldn't capture as uncompressed HD (the client wanted an SD deliverable, but if he could offer HD that'd be a nice option).

So check, double check, test, don't assume, and when in doubt....feel free to contact your friendly Internet based consultant.....ahem....

-mike, the friendly Internet based consultant, whom you can reach at mike at hdforindies dot com

(overt enough for ya?)

MacMerc.com: Sonnet Announces Internal Drive Mounting System for G5

....and then this happens. I just finished writing that I'm leery of internal drive solutions, but here comes this from Sonnet, whom I otherwise like and endorse their other SATA drive related stuff. It's an internal G5 mounting kit for G5s to put 3 additional SATA drives in there.

-mike

HPshopping.com - HP f2304 23" High-Definition Flat-Panel LCD Monitor

HP has their own 23" LCD panel, that uses the same panel as Apple's 23" model. Except Apple charges $1500 for theirs, and the HP goes for $900 at the moment. Six hundred bucks - nothing to sneeze at.

-mike

AppleInsider | Apple to introduce new hardware at Apple Expo

Rumor has it, according to AppleInsider, that either new Powerbooks, new G5s, or both will debut at the upcoming Paris Expo.

New G5s are rumored to have two dual core IBM PowerPC 970MP chips, which are dual core CPUs. That is, each chip has two processors on it, so two of them makes for four processors. There is no mention of PCI Express in the article, which Apple has been needing for some time.

So maybe we will, maybe we won't.

I had previously read the tea leaves to construe that if there was no keynote by Steve Jobs, that there would be no new significant hardware. An astute reader pointed out that maybe Steve DOESN'T want to be present if a new, faster PowerPC based Mac was introduced a few months after Steve announced Apple was moving to Intel, as this would Con-Fuze The Masses. And I think that's a very believable theory.

So I dunno. I wouldn't be surprised either way if Apple did or didn't roll out new G5s. If they do, I'm more into PCI Express than I am into twin dual core processors.

But that's just my geeky self looking for my own particular desires to be met.

UPDATE - The game continues - of course, Think Secret thinks no new hardware is coming for the same reasons I did - they were going to roll out new hardware, but didn't have it ready in time so they bailed.

Whee! Yes, no, yes, no, maybe, whatever. Wait and see.

-mike

(or at least new info to me)

Reader Stephen attended the ResFest demo of the Panasonic HVX200 camera. First public exhibit outside of IBC.

He also has some news about the Canon XL H1's native CCD resolution.

Now, keep in mind, this is what a rep told him. I have found it notoriously difficult to get good, solid information on new products from reps, because it's all new to them, too. So this is what we've been told, by company representatives, but it does NOT mean that it is definitely true. So I can verify that this is what they told Stephen, but I can't verify that it is all factual, since the reps don't always have it right at the beginning.

His report, starting with the Panasonic camera with a rep talking about it (my bold for emphasis):

After a rather interesting seminar, focused on differences between codecs, she showed us some live footage of 1080i, and it looked pretty good. Then we had a question/answer session.

The camera can do 1080 in 60i, 30p, 24p. Variable frame rates in 720p. This is interesting because when I asked the question "What is the native resolution of the chips?", she got a bit flustered, bullshat for a minute, then just said, "I'm not at liberty to say" and then went on to say how it didn't matter, because what you got was 1920x1080. This is even more interesting after I talked to the canon rep, who said that canon really wanted to go 1080p, but a chip that could actually do 1080p would have made the camera cost $100,000, and they had no choice but to make the compromise. (on that note, I have information on this 24f thing) From my conversation with him, coupled with Panasonics avoidance of the question, I'm pretty sure the chips are not natively 1080. So its not a question of 1440x1080 or 1280x1080, its a question of how much smaller than that it is. Bummer.

From what she said, you can get anything out of the camera that the camera can shoot, I believe she said through firewire, but there is a catch. when recording to P2, you simply get 24 frames. coming out of firewire, you get 60 fields, so you have to do a pull down, which is a pain in the ass, and makes file sizes larger (i didn't realize that).

Audio, on p2 or over firewire: Uncompressed 48k 4 channel w/ dvcpro100 Uncompressed 48k 2 channel w/ dvcpro50

Variable Frame rates are still not decided, however the following are "definite": 12, 18, 22, 36, 48, 60 Only in 720p

Low light rating. She wasn't' sure, the only thing she could say is that its f8 @ 2000 lux, while the dvx was f11 @ 2000 lux, and it has 7 stops of latitude.


The model camera only did 1080i, and the image looked like shit. the demo footage was done by someone obviously not skilled with a camera, and out of a hotel window, was shot on an overcast day and was grainy as hell.


p2 started sounding a little more reasonable after the seminar. I didn't realize that the idea is that you never buy more than 3 cards. Its not a storage solution, just an intermediary between the camera and the storage. After the costs of tape, and the convenience of not having to log/capture, in the end, just might be cost effective.

XL H1: Native 1440x1080 chips. It does 24i, as in field a and field b have temporal differences. He did say that it is not as noticeable at a slower frame rate since the interval between the fields is less time. (wrong - see MikeNote1) I wasn't quite convinced. The camera did look pretty cool.

I didn't take the best notes on other peoples questions, and realized so afterwards, but if theres anything else your curious about, shoot me a question and I'll try to remember.

-Stephen


Mike's Comments: This all sounds very interesting.

Storage issues: FireWire vs P2

The most interesting tidbit new to me is that the footage that goes to P2 is not the same as goes to FireWire. The P2 card is smart enough to not need to work like the Varicam, putting the flagged 24 frames in a 60 fps stream. Since there is no tape transport involved, it can just record the 24 important frames and not record the others. Going out over FireWire, 720p60 is the format. To make it do just 24p, they'll have to be flagging frames again. (If you are unsure of how 24p works with Varicam, read the archives. Google has broken archive searching at the moment, very very irritatingly). So recording to the FireStore FireWire based device will require more storage than the P2 cards will if shooting 24p. Two and a half times more, since 24 goes into 60 2 1/2 times. So the two 8GB P2 cards you get in the $10,000 bundle with the camera (total of 16GB) will hold as much 24p footage (regardless of 720 or 1080 resolution) as 40 GB of FireStore device. Eeek. Not quite as groovy as when we started. I haven't seen a quote as to the size of the FireStore device, but I'd guess it would be in the 60 to 100 GB capacity. The largest drives that will fit into the form factor he's going to use are 120GB drives, but they are quite expensive on a per GB basis as compared to the 100 or 80 GB drives. An 80 GB size would be a safe assumption of minimum size, perhaps 100GB. So an 80 GB FireStore would hold as much as 4 P2 cards in 24p mode, but at about 1/4 of the price - so still a good deal, just not as good a deal as we would have hoped, where it could have held as much as 10 P2 cards.

UPDATE - a sharp eyed reader caught a flaw in my math - the above ratios work for 720p, but not for 1080p. When in 720p mode, you're working with a 60 progressive fps stream of frames. When in 1080p mode, though, you're really only working with a 30p stream of frames. So the difference between P2 and FireStore is much less when dealing with 1080p compared to 720p. 1080p over FireWire to FireStore would only be about 20% less efficient than 1080p to P2. Because it would be a difference between 24 and 30 frames per second, not 24 and 60.


DVCPRO HD resolution issues

Based on her response, I'm guessing the HVX200 has a 960x720 imager, with the green pixels offset half way between the red and blue to give an effective 1440x1080 resolution. This would give you a perfect and clean DVCPRO HD 720p of 960x720, but could be interpolated up to 1280x1080 for DVCPRO HD 1080p and 1080i. Even if I'm wrong with my first guess, and it's a guess, and the CCD is 1280x720, 1080p or 1080i mode would be the way to get the most resolution out of this camera. And based on Stephen's report, something in what she said implies that it is less than 1280x1080, likely 1280x720 or 1280x1080 at best, and likely even less. Only if the CCDs were a non-offset 960x720 would 720p mode be the optimal way to get best quality out of the camera without stretching the image up. Because stretching the image up simply makes it bigger, not better. If the imager was 1280x720, 1080p/i DVCPRO HD would be optimal, otherwise 720p would only record 960x720 to P2/hard drive due to the DVCPRO HD codec.

Canon's CCD resolution and optimal workflow for best 24p resolution

MikeNote1: The rep purportedly said: He did say that it is not as noticeable at a slower frame rate since the interval between the fields is less time. OK, that can't be right. Maybe we're misreporting what he said, but that is factually incorrect. At a lower frame rate, each frame lasts LONGER, not shorter, so the interval between the fields would be more not less time.

At this price point, I can believe that 1440x1080 CCD resolution is real, but I would like to get confirmation. And is it a "real" 1440x1080, or the compromised 1440x1080 like the Sony HDV cameras? As for the 1080i 1440x1080 resolution of the Canon camera, from a strictly format point of view, it should be possible to get more resolution out of the Canon than out of the Panasonic, all things being equal (which they aren't). The Canon records natively to HDV and the Panasonic to DVCPRO HD, and they have different kinds of lenses (the Canon's are interchangeable, the Panasonic's is not). But, assuming you have a good lens, and go out the HD-SDI to a higher end recording solution than HDV, the Canon should be theoretically capable of sharper images. But there are waaaay to many real world variables to calll that for sure at this point. But XL H1 out HD-SDI to deck or uncompressed to disk, and 50i de-interlaced to 25 conformed to 24p, would be a HELL of a lot of work for a possibly better result than shooting 1080p24 on the Panasonic. Or possibly not. Would it be worth the added hassle, cost, time, weight, additional post work, and other factors to make it worth while? Depends on the vagaries of your expectations, shooting requirements, budget, crew, post situation, etc.

But that's the theory at this point, and theory is all we have at this point.

-mike

PS - there's all kinds of good info in the comments (link below) worth reading.

Disney Moves Away From Hand-Drawn Animation - New York Times

Very interesting article from NYTimes (aside - do I link to very boring articles?) about Disney's contentious move away from hand drawn movies into digitally animated ones.

Opening paragraph:

ON April 4, 2003, Glen Keane, one of the Walt Disney Company's most respected animators, summoned about 50 of his colleagues to a third-floor conference room on the lot here to discuss the war brewing at the studio. Disney's animators had settled into two opposing camps: those who were skilled in computer animation and those who refused to give up their pencils.


It talks about the internal struggles of the traditional pencil based animators with the computer guys, Disney's troubled relationship with Pixar, and how Disney plans to do animated movies in the future.

Other "wow" quote:

If there is any question about whether there is life for Disney after Pixar, consider the following: Last June, Disney caused a ruckus at the industry's largest computer-animation conference in Los Angeles when it set up a large poster in front of its booth - and facing Pixar's - to advertise the preproduction of "Toy Story 3." Mr. Jobs had sought to make the movie, but Mr. Eisner said no when Mr. Jobs wanted it to count toward the five that Pixar owed Disney as part of its partnership agreement.


Wow - that's about as in-your face as you can get to Pixar. And it was no mistake that they were across the aisle.

Mike's Analysis: While I applaud Disney's efforts to keep up to date, I don't think that is the heart of their problems - I think their problem has more to do with lack of heart in their recent productions.

If you look at their recent productions, like Treasure Planet, and compare those back to prior movies like The Lion King, look at the sense of heart - I didn't see Treasure Planet, it looked like a stinker from the outset to me. I still, however, distinctly recall sitting in a theater in 1994 watching The Lion King with Holly L'Italien, and she quietly nudges me points at the little boy a few seats over, who was perhaps five years old and standing up, holding onto the seat in front of him, eyes tearing up and lips aquiver wondering if the Lion King had really just been killed by his evil brother. It was powerful and moving and emotionally resonant all to hell. Can you say that about recent Disney fare? Uh, no. No way.

I don't think Disney's problems have to do with the technology of the films they produce. I think the problems have more to do with being a large corporation and group thinking and focus group testing a product to death rather than letting a clear vision come through.

Think about every single Pixar movie so far - could it have worked as a pencil and ink movie? Would the writing have been as sharp (hint hint Disney), the characters as interesting? I'm willing to bet so. Every Pixar film amazes with the look, but it's the story and characters that really suck you in.

There's something warm and homey and fulfilling about every Pixar film to date (I'm worried Cars is going to be a clunker, more on that some other time). They all connect on a much deeper level than the third (rendered) dimension.

Now think about Disney doing Toy Story 3, and not Pixar. I just have incredibly low confidence in their ability to pull it off with the usual finesse that Pixar does, who have been pretty much infallible so far. It's not that I have anything agianst Disney, I don't care that they are a huge company or whatever, it's just that they have failed to rock the house for quite some time. When/what was the last Disney animated movie that you saw that you walked out of and thought "Wow, that was just really incredibly great!" I don't know when I did that after a Disney film. But I have said that after every Pixar film to date (especially The Incredibles, it really blew me away).

Now think of recent Disney fare - Brother Bear, for instance - would that have worked better in 3D? I think not. And frankly, I think it would have been the worse for the crisper, cleaner, sparer high tech look as opposed to the comfortable level of abstraction that the hand drawn look provides.

Learn, Disney, learn.

Speak to us from your heart - warm, beating, organic and alive, if you still have one.

Not from your pixels & polygons..

-mike

PS - I found this article from CinemaTech's coverage. Scott has some nice quotes extracted as well. If you don't have time for the full four (web) page article, skim his.

The good news is that Google introduced a new service to allow searching just blogs, not the internet as a whole, and has changed their little header at the top of Blogger blogs (of which I am one).

The bad news is that this seems to work like crap to search my own blog - it either doesn't work at all, or only searches the current front page. In any case, it does NOT seem to search the archives anymore, which it used to do. Which is a shame, since the archives contain everything from April 2004 up until 30 days ago, and there are about 1200-1300 articles that can't be searched.

You can still click on the month by month archives on the list at the right to see one month at a time, but this is the hard way to find things.

I'll complain, I suggest you do the same to see if they can fix this feature.

-mike

VideoCoverage.org has some IFA 2005 Video Coverage of interest to geeks like us. It isn't text, it is links to video clips of interviews etc. interestingly, some of the coverage itself is high def. It is BitTorrent links, so you'd need the a BitTorrent client in order to download and view these.

Stuff covered of interest:

-Nero engineers talk about H.264
-Phillips has a Blu Ray prototype player, showing some Java based interactivity (another reason I like Blu Ray).
-I may be getting this wrong, but I think it's an HD DivX based player with DVI HDCP output, removable hard drive, and USB2 connection
-DivX certified DVD recorders
-LG showing 1080p screens
-Sony HDV

It's a High Def World has a geeky article up, with downloads, comparing the quality and speed of some film look plugins.

An interesting read....Magic Bullet takes about 3 times longer to render, but the authors liked the looks better of MB. Which is slightly not the real deal test I think - since both allow for controls of the effect, and what's desirable is subjective.

Read on...

-mike

(found via FresHDV.com)

German reader Jan wrote:

Hi !

I enjoy your blog very much !
But i have an question. Do you know what kind of signal lies
on the BNC output of the new Canon XL H1 ?
8 or 10 bit ?
RGB or YUV ?
I or P ?
4:4:4 or 4:2:2 or 4:2:0 ?
What kind of conversions happens to the signal between chips and BNC output ?

I`ll read your blog. As always !

Kind regards Jan


I wrote back:

Hi Jan - excellent question!

I would expect that it is either 8 or 10 bit 4:2:2 YUV coming out of there - the high end cameras that have HD-SDI typically kick out 10 bit, but at this price point, I don't know - maybe 8 bit. Wait and see. Hopefuly 10 bit (gives 1024 gradations rather than just 256).

Definitely not 4:4:4 or 4:2:0, definitely not RGB.

As for I or P, it's going to be I (interlaced rather than progressive), since it is a 1080i camera.

Stuff happens between chips and BNCs to process the image (white/black point, etc.), but it as clean a source as you're going to get off of that camera.

Everyone's excited about this feature, since an HD-SDI built into the camera is the cleanest possible digital source off of the camera (short of hacking into the CCDs directly), but others have been doing something similar by attaching a AJA HD10A converter to the HD component analog jacks of the Sony HVR-Z1U and other cameras. What's the HD10A do? Converts the analog component to HD-SDI. It isn't as clean a signal as HD-SDI source from camera, but it has similar benefits (although not as good a solution).

Everyone's excited about the possibility of hooking up a deck or recording device, until they start pricing out the recording devices - a Panasonic AJ-HD1200A DVCPRO HD deck is about $30K with HD-SDI inputs and FireWire output.

Oopsie, there went the budget.

-mike

HD Beat has a roundup of pre-release HD DVD and Blu Ray players from Toshiba, Hitachi, Sony and the like.

None of which are shipping yet, but it's a good update on what's going on with consumer playback and computer playback/recording decks and drives.

-mike

CinemaTech: Unasked questions: More on the Disney/Christie digital cinema deal

such as, will this make any financial sense?

Hands-on with the Canon EOS 5D and HDV High Definition XL H1 - Engadget - www.engadget.com

confirms 1/3" CCDs, no true 24p mode, and has pictures of the camera

UPDATE - there are some very interesting comments about why Apple would or wouldn't launch a fast PowerPC based system after the Intel announcement, and whether the OS can actually handle/use 4 processors. If you're viewing the whole blog and not the individual article, click the Comments link at the bottom of the article.end update

Think Secret - Power Mac G5 revision expected soon, signs point to dual-core

As soon as the next month - Think Secret speculates it might be at the upcoming Expo, even though Steve Jobs bowed out and there is no keynote scheduled. So I think no new hardware at Expo, unless it is another of what I'd call a "shame release" like they did with the current 2.7 GHz G5s. When they only had 2.5 GHz, and were already late for the promised 3.0 GHz, Apple sneaked the 2.7 GHz upgrade inbetween NAB in April and WWDC in June so that nobody high level would have to be on a stage to try to grin while announcing yet another underwhelming upgrade.

And if it is dual core, IBM (the chip vendor) has stated that it's likely to top out at 2.5 GHz, actually numerically lower than the current dual 2.7 GHz chips shipping in high end PowerMac G5s.

Fortunately, the chips perform 50-80% faster at some functions according to the article, so that overall the dual core system should be faster.

I had previously speculated that the next Mac update would include dual core G5s, possibly two dual core chips, for a total of four processors, and would have PCI Express (PCIe) slots.

With Steve bowing out of the keynote, I don't see how that's going to happen, unless these new systems wouldn't be ready to roll out (even in Apple's time warp of "Available now!"="next month if you're lucky"). Such that if these systems were close, they could be announced even if not ready to ship for some weeks/months (Apple does this LOTS).

So I read all this to mean that the dual core 2.5 GHz systems will be announced and ship sometime after the Paris Expo, that they will be dual core not twin dual core, and they will still be based on PCI-X not the newer PCIe.

My gut says MacWorld Expo announcement of dual core (single or twin for 2 or 4 processors) with PCIe with an acutal street availability of the fastest model sometime in March/April.

But that's just my pessimistic gut, I'd love to be wrong. We always get our hopes up, and so rarely does Apple deliver hardware that blows away our expectations (except for iPods, and dammit, I may have to get a nano Just Because.)

-mike

Apple Beefs Up Antipiracy Efforts for Intel Version of OS X - Yahoo! News

As expected - Apple is clamping down on piracy of their OS, since Apple makes it's money from hardware not the OS for computers. (Same song for iPod/iTunes as well, frankly).

They are altering their developer kits so that software won't run on anything but Apple branded/approved hardware.

After having issues with his Highpoint RocketRAID 1820A card, a reader upgraded to a 2220 card from the same vendor. Here's his latest report:


hullo mike

here's another update...

i replaced the rocketRAID 1820a with a 2220. the 2220 handles disks differently (for example single drives connected to the controller will NOT just show up in disk utility or the finder. they need to be prepared for use in the raid management console).

now the raid is back up and running again. i had to back up all the data, initialize new array (again using RAID level 5 with 8x 300GB disks), format raid volume (zero write to map out potential bad blocks), copy data back to raid.

the new management utility (called within a webbrowser) is not bad. one can change the caching strategies on the fly (write through vs. write back - the latter is about 6 times faster in my tests). however, the block size (for raid 5, also applies to raid 0) is no longer customizable or even known (i think with the old card it was 64K blocks).

the best part is that turning on the raid works without a restart! when the computer is running, i turn on the external enclosure, press the 'rescan' button in the management utility and after about 10 to 20 seconds the raid volume appears. joy!

cheers, oliver


Mike's Comments: The 1820a seemed interesting but flawed. The 2220 seems improved, with better hotswap capabilities, web based access to the RAID controller, and other improvements.

Also, BareFeats has a performance review of the 2220 as well and tested it in RAID 5 vs RAID 0 modes vs some other competing products. RAID 5 performance is improved, but not as fast as I'd like to see it for uncompressed HD work. But it's getting closer. See Rob-ART's thorough testing for details.

-mike

Canon's XL H1 has a lot of interesting features.

I've got a little time as I'm waiting for something to render right now, and I've been digging around on the specs page for the XL H1.

Items of interest:

-1080i camera at core

-"60i, 30F, 24F" and "24 Frame gives the look and motion of film" sounds like true 24 fps to me. It causes me vexation, however, that they say "24F" rather than "24P" One would think, after their experience with the XL2 and Advanced Pulldown Mode that they could do this. And the indies have obviously been clamoring for it. Only reason I can think of to not offer it? That's the XL H2.

-the Professional Jack Pack is interesting - timcode in and out, genlock, and SDI/HD-SDI output are primo features. By calling it "Professional Jack Pack," does that imply it is a separately purchasable option, or just a named feature of the camera? I don't know yet. UPDATE - I THINK I DO - It's mentioned in the interface stuff - HD-SDI, genlock, TC in/out, so I think it's a feature, not an option, since they didn't break it out as "optional" beneath the list.

"Total Cine Control"
sounds pretty thorough - you get control over the following:

Gamma
Knee
Black
Master Pedestal
Setup Level
Sharpness
Horizontal Detail Frequency
Horizontal / Vertical Detail Balance
Coring
Noise Reduction 1
Noise Reduction 2
Color Matrix
Color Gain
Color Phase
Master Red Gain
Master Blue Gain
Master Green Gain
R-G Matrix
R-B Matrix
G-R Matrix
G-B Matrix
B-R Matrix
B-G Matrix

That's pretty damn thorough.

-a 20x HD lens with optical (preferable to digital) image stabilization. Cool.

-three CCDs that are 16:9 natively, not cropped/stretched 4:3

-each CCD has 1.67 million pixels. Reverse engineering some math, that's a little over 1440x1080, the native resolution of HDV. That's encouraging - no upsampling to get the resolution. Extra pixels for stabilizing or stills or what?

-2.4" combo viewfinder/LCD panel. Hmmm...2.4 inches is kinda small, especially compared to the much much larger LCD on the Sony HVR-Z1U. "Magnifying Focus Help" sounds like it can zoom in pixel for pixel to check focus on the viewfinder, a very very necessary thing for HD. The critical question is whether it can do this after you've started rolling, or just before shots like the HVR-Z1U and HDR-FX1, both from Sony

-you can save six custom presets - look at the long list of modifiables up above, and think how handy that is. and you can save them to a SD card. Items you can save in a preset include:
"including: color gain, color phase, sharpness, setup level, V detail, color matrix, gamma, knee, black stretch and skin detail (hue, gain, area, Y level). "

-you can define two custom keys for a wide range of stuff. So maybe a zoom/focus/look position 1 and zoom/focus/look position 2? Sounds similar, but not as juicy as, the saved zoom/focus over time the Z1U can deliver.

-Time Code Issues:
"The XL H1 conforms to industry standards with SMPTE time code, as well as Drop, Non-Drop, Rec Run and Free Run modes" - this is groovy and pro level and appreciated. Can also makes bars/tone.

-the camera can, for additional costs, be modified by a dealer to handle 50i & 25F as well as the default 60i, 30F, 24F. So it isn't a one time, no-turning-back change, sounds like you gain the ability to flip back and forth (but probably rebooting the camera)

-skin detail controls (nice)

-Interface stuff:
-BNC video (HD-SDI/SDI? what do they mean here?)
-component/composite/s-video
-FireWire
-2 XLRs w/phantom power
-mike in
-headphone jack
-LANC
-genlock
-timecode in and out

THIS WOULD IMPLY THE PROFESSIONAL JACK PACK IS A FEATURE, NOT AN OPTION. EXCELLENT!

-24P? or 24F?

"Under the "HDV1080i" (HDV) specifications, 1440 x 1080 (16:9) images in 24 frames are recorded. (24F recording)"

Again, if they are using 1080i nomenclature, why not call it 24P, unless it really isn't 24p? Again, wait and see, awaiting clarification. Anybody knows, clue me in.

RECORDING MODES: The camera nomrally can record in 60i, 30F, or 24F for the NTSC model. PAL model (with the 50i upgrade/mod), 50i, 25p as well.

Enough for now.

If you know something to fill in the gaps, let me know.

-mike

The CamcorderInfo Blog: All About Shooting, Editing and Polishing Your Videos - Consumer Camcorders - Camcorderinfo.com

It's the Canon XL H1, and here's Canon's page on it

Goodies include:

-1080i resolution
-native 16:9 CCD's
-60i, 30F and 24F - (pseudo 24, not real 24p. Think CineFrame mode I think but don't know - need to read up)
CORRECTION - IT DOES APPEAR TO BE TRUE 24P, or at least is claimed to be so on their website

-straight (pre-compression) HD-SDI and SD-SDI output, for uncompressed source not molested by tape compression or subsampling
20x lens (38.9 - 778mm)

Ships November, $9000 list.

Records HDV, but had SDI/HD-SDI outputs. That last tidbit is HUGE - that means you can plug it into a real HD deck, like a Panasonic 1200A, or an HDCAM or HDCAM SR deck.

Or record to a computer if you're rigged for it...which I am.

Ladies and gentlemen, we have ourselves a very, very interesting camera here...except that it lacks native 24p.

I wonder if they'll have a PAL version to shoot 50i and convert to 24p?

-mike

Kong is King.net | King Kong movie news and rumors | News Archives

Latest in their post production diaries, talking about making digital doubles, with body scans used to let Naomi Watts (NOT Cambell as I previously mis-stated) fall several stories without damaging her perfect skin. Amazing what can be done digitally these days. WETA Digital is DA BOMB.

Not that there isn't other news to report, but I checked in on ProLost for the first time in a while, and he's got a little more up about his linear color workflow stuff, as well as some fun books he's bought, including "If it's Purple, Someone's Gonna Die" about the use of color in film. Stu's a smart interesting guy, so check it out.

-mike

LaCie - Biggest S1S: SATA - 4-Disk RAID Subsystem

LaCie has introduced a 4 bay removable SATA drive RAID. It supports RAID levels 0, 0+1, 5, and 5+hot spare.

It's $3,200. For most video folks, a pricey but secure option. Transfer speed 80 MB/sec as RAID 0, likely less as RAID 5.

I have a lengthy review of it's big brother, the Biggest S2S, the 5 drive RAID 0/1/10 solution, that I need to finish up and publish soon (working on a Final Touch HD project for client, more on that later too).

-mike

BetaNews | iTunes Image Files Suggest Video Store

One step at a time...signs that Apple is going to sell music videos. How long until short films, Hollywood films, and HD content after that?

I still think a