Les,

1.)I'm not getting YUV from the CCD's, i don't think it has been claimed otherwise.

2.)The sustained rates for drives IS what i've been looking at. Right now, i am capturing RAW data clips continously on a WD EIDE drive that i bought two years ago. I can capture as long as i want and the data is free of errors and continous. I have tested continous raw video captures up to 15 minutes in length with no problems.

Juan

Jesus........What a Thread !!!!!!! I must confess I am wet !!!!!

Juan you are incredible...

<<<-- Also: look at the sustained transfer rated for hard drives, not the interface burst rates. Big difference. -->>>

Firewire doesn't have a burst rate -- you're thinking of a half-assed format like USB 2. FW400 is sustained 400mbits/sec, FW800 is sustained 800mbits/sec.

Granted, very few harddrives can max out FW400 (50 megabytes/sec), and I don't know of any non-RAID, out-of-the-box harddrive that can max out FW800 (100 megabytes/sec).

If you go to http://www.lacie.com/ you'll see that they have a bunch of cheap FW800 drives that can do anywhere from 55 to 88 megabytes/sec (a nice cushion for Juan's ~35MB/sec). Most pros I know prefer LaCie for high-speed Firewire applications; I know one guy who does uncompressed HD on a PowerBook over two FW800 buses using LaCie drives. He was saying that the LaCie drives were the only ones that could take it...

- ben

Ben speaks the truth :)

I myself use the first LaCie FW800 drive that came out, a 200GB unit. Every benchmark i've run yields a ~70MB sustained write rate, which is more than enough what is needed for this application.

However, like I said, right now the system is not capturing to the Lacie drive but rather directly to a 120GB Western Digital WDC1200BB which i bought a long time ago. I haven't run any benchmarks, but it handles the continous writes perfectly.

I will be done with my finals on thursday afternoon, so i will finish writing the code to handle the blue frame strips and upload a clean clip in raw RGB frames....

I also can't wait to watch one of these clips uprezzed to HD on TV/projector....

Juan

Ben:

Lacie does not manufacture hard drives or hard drive chipsets. All they are doing are getting IDE drives from Maxtor, Hitachi, Seagate and/or WD and putting them into an external enclosure with a firewire (and/or USB) chipset.

Any other manufacturer or user can do just the same with identical drives and firewire chipsets. Lacie has no properiatary technology to boost the speed of their drives (unlike Medea).

Lacie is just repacking components (doing it well, especially with their big drive series), but I just put a 500 GB RAID 0 setup in my computer using Hitachi 250 SATA drives for $390. The Lacie drive goes for $580. An external box (firewire, USB and/or SATA) with RAID chipset runs around $100 if you must have it external. As long as you pick the right chipset, you save almost $100 over Lacie plus Lacie only warrants drives and box for one year whereas my Hitachi drives come with 3 year warranties.

I'm aware that LaCie doesn't manufacture the actual drive mechanism.

However, an external harddrive is more than a drive mechanism and an Oxford chip. There's a lot of supplemental electronics, and it so happens that LaCie is very, very good with that electronics.

Their latest line (the "D2 Extreme") is far faster than their first generation FW800 drives, even though I'm pretty sure they were both based on the Oxford 922 chip, and used comparable drive mechanisms. Clearly, they've got a line on how to put together a drive.

But don't take it from me, talk to any video pro who relies on raw speed from the Firewire bus. If you're pushing the envelope, you're using LaCie...

- ben

Well, I'm not a hard drive engineer, but a Lacie external drive only has three components: drive(s), a chipset (including the RAID0 bridge) and firmware. The extreme uses the Oxford 912 chipset for improved performance, not the 922.

I suggest you read this http://www.barefeats.com/fire44.html for more info. Note the Hitachi drives I just added beat the Lacie drive handily.

I'm not hard drive engineer either (and obviously get my Oxfords mixed up), but if you crack open a case, you'll see a drive and a board -- there's a lot more on the board than just a controller chip. I'm pretty sure LaCie designs their own boards.

I'm not generally given to brand loyalty/mysticism, but few friends and I have used LaCie for many years, and have never ever had a hard drive crash. Other friends have used WD, Seagate, Maxtor, Que, OWC, and they have all suffered catastrophic crashes. I know this is probably 99% what actual internal drive is used, but... there it is. Maybe it's just exceptional QA at the plant.

I must say, RAID is a whole separate issue, and you're comparing an internal RAID to an external one. To use a hated 80's phrase, "nuff said."

Anyway, this is OT...

- b

Ben:

I'm glad you've never had a hard drive crash, but Lacie used the same WD and Maxtor drives you claim are unreliable. I've had and supported thousands of drives over the years, including Lacie and they've all crashed. Brand reliability varies between model, batches, runs and a lot of luck is involved with it.

A good warranty and good back is always best.

Per electronics, the board is just what holds the oxford bridge (and in models with RAID, the RAID controlller). Note all Lacie "Big Drive" and "Big Drive Extreme" are RAID models, they do not contain single drives.

The only big difference in an internal and external drive is the bus used (firewire vs ATA or SATA) - the setup, drivers and electronics are the same.

And this is very, very on topic as Juan's biggest hurdle (IMHO) figuring out a usable and affordable recording mechanism for the drive. I think that will be key to good sales of the unit.

I'm still a fan of SATA 2.0 (specs just finalized and released) as it allows the most flexibility and performance plus already is as low in cost and cross-platform. [url]http://www.serialata.org[/ur]

Stephen/Ben:

To elaborate on this aspect of the project, i've looked at all the options, and here's what i've found.

The problem with SATA for this project is that afaik it was designed as a protocol for internal drives, so righ away i would have to solve encasing/power supply issues.

But most important of all, SATA requires special, very fast switching circuitry to implement, and I haven't found any hardware(i.e. IC's) that i can buy and use, already made for this purpose.

With firewire, buying an external FW800 drive already has an encasing and it's own method for powering it. Furthermore, it's slightly easier to obtain hardware that supports it, and for other technical reasons it is a breeze to implement in my case.

I felt that if i went with an internal drive, i would have to include it in the capture system encasing, and thus the user could not decide what size drive to use, not sure if that is a big deal...i guess i could have a mount on the bottom of the box such that you could mount it here, but then the tripod mount would be gone.

Furthermore, i don't think SATA cabling was designed to be long and as easy to deal with as a single firewire cable if the drive is to be somehow mounted on a shoulder bag.

I'm also thinking about implementing some sort of output to monitor the raw video...of course it won't be high quality, but it will allow to see the latitude and color of the raw footage for adjustments. This monitor output could be as simple as a standard analog S-video/RCA out, to an actual SDI or maybe separate YCbCr outputs...

Let me know what you guys think...i might be wrong about SATA because all i read was the low-down technical spec paper...but it seems like it's designed to be inside a computer rather than separate from the host and hanging from your shoulder :)

Juan

Juan:

SATA II parts might be a harder to come buy right now but already a number of external SATA 1 parts on the market.

http://www.extremetech.com/article2/...1546751,00.asp is one and there a plenty of DIY kits, including ones that use Firewire for the connector but SATA drive. Note that no drives are Firewire or USB native, they are IDE or SATA or SCSI (or Fibrechannel if you have the bucks) - this is the key advantage SATA has.

The specs for SATA II are here: http://www.serialata.org

I except by mid 2005, SATA will have 50% of the drive market or more.

A transition plan may be to go with SATA drives in firewire enclosures and then SATA drive in SATA enclosures as parts and costs allow. Perhaps offer bpth flavors.

SATA may get 50% of the internal drive market. But it will have under 10% of the external market.

Firewire is the monster in the external market right now, and that certainly won't change by mid 2005.

- ben

Ben:

It's only a monster in the Mac market which is well under 5% of the worldwide PC market.

USB 2.0 has already outpaced firewire for external devices, including drives. Only SATA II could change that, not Firewire 800.

Of course, probably near 100% of DV cam users have firewire 400, but only G5 users and handful of others have firewire 800.

That's why I think a SATA drive in a firewire enclosure is probably the best short-term solution followed by a SATA 2 solution. When PC's come standard with an external SATA II port.

Of course, the ultra high speed wireless USB could be interesting too when it debuts.

"When" PCs come with external SATA II ports? PCs come with USB2 and FW ports right now -- the installed base is massive. 2005 is practically tomorrow, and the spec for SATA II isn't even finalized yet. It's an RC, and after that there will be other standards applications. If SATA II ever becomes a standard port, it won't be for at least 18-24 months.

And let's not even get into the Mac issue -- is this not dvinfo.net? I wonder what the percentage of out readers are Mac users -- if it's not a majority, it's probably close. Firewire is a monster not for Mac users, but for all video/media creators. SATA is great for internal storage, and interesting as a possible external solution, but let's not get ahead of ourselves. FW800 is awesome, it's already on a huge number of machines, and it works now.

As for USB2 outpacing FW, well, history shows that there's no stopping PC users with their hearts set on an inferior technology. But I imagine that if you look at external drives that run at or above 7200rpm, it's not so clear cut.

- ben

Ben:

I see we difffer on this - however, I think your view of the storage market is incomplete and not based on actual sales numbers - a quick google does not seem to back up your points.

While you may be comfortable with Firewire 800, the actual market conditions seem to point in other directions.

Juan needs to probably rely on better experts than you or I in making a decision about market factors and interfaces.

However, SATA II is a final spec (http://www.internetnews.com/storage/print.php/3339821 & http://www.intel.com/technology/serialata/ahci.htm
and expect to see it plenty later this summer and full swing by fall.

A further problem with SATA II is that, looking at the spec, I see no mention of peer-to-peer connectivity. One of the basic principles of Firewire is that you can use it to hook two Firewire devices together without a computer. For example, you could hook a camera directly up to another camera, or a camera to a hard drive. This capability was expanded with FW800.

With SATA II, it seems that you'd either have to build some very sophisticated electronics, or rely on a computer to sit between the DVX100 and your drive. This is clearly non-optimal.

Reading the spec, it seems that the main design goal of SATA II is pretty similar to SATA: to provide an ultra-fast channel of data direct from a disk or disks (they mention RAID configurations often) to a computer.

So SATA II is never going to replace the USB or Firewire families as a convenient general-purpose interface.

And I'd think twice before I discounted FW800. Final Cut Pro already supports DVCPRO-HD over Firewire, and HDV utilizes Firewire as well. If you want to be doing HD work in the next few years, you'll still be using your old pal FW. That'll help keep the entire FW market charging forward...

- ben

"..SATA II is a final spec..."

Huh, that's interesting. Because the SerialATA website ( http://www.serialata.org/ ) calls the draft a "Release Candidate" undergoing a 30 day review as of 4/22/04.

I'm looking forward to buying some SATA II gear later this summer! I'll plug it right into my SATA II port, which is located right next to the Unicorn dock and all six FW3200 ports. We may see SATA II gear in 6 months, but its main relevance will be internal storage and high-end external RAID solutions. When they start putting SATA II ports on laptops, then we can talk.

Clearly we do disagree, and it's not just about the marketshare. But tell me exactly what you searched for in Google. How on earth did you find numbers for hard drive marketshare within the video/media industry? That's what I want to see, since we're talking about uncompressed DVX100 lest we forget, which is a niche issue.

I think Juan should rely on his own experience and intelligence, which has led him straight to FW800, an interface which is fast, readily available, and apparently not a problem to implement. I don't understand why we're still talking about SATA II and "wireless USB" and unicorns...

- ben

I did not save all my google links, but search for "sata" "serial ata" etc...

SATA is peer-to-peer based, though I don't expect it to see it in DV or HDV cams as it is way overkill. But for HD and faster cameras (like the Viper, Arri), the 3 Gb/sec of SATA II might be better than akward SCSI implementations especially now that Seagate has got Native Command Queuing supported.

I had forgotten the name but a google search for serial ata peer to peer gave this link for the chipset company that already has SATA II controllers ready:

https://www.marvell.com/products/sto...a/88SX60xx.jsp

The built in SATA power connection standard is also something I like.

I'm for the quickest, least complex and most direct solution.

A FW800 port connected to an external FW800 drive - Juan's original idea - sounds great.

It'd be nice to also be able to connect it to a FW800 equiped PC/Mac as to record to a large internal RAID array with lots of storage. Of course that means you are tethered to a computer but hey sometimes that's okay. Plus I think a firewire cable can run a long way.

I don't think it should have an internal drive. Mostly because it will probably delay getting the thing out and about. Real world production will get the proper feedback for a MkII. What, already a MkII being discussed!

A video output would be very useful for previewing and SDI output would be fantastic. Of course keeping with my no-major-added-delays theme.

If I had to add something...
When I hit record on the camera - the tape and the drive start to record. Maybe there is a way to sense that switch-over. But that might involving soldering and I am weary of that.

But really a separate record button is fine. Please add a record light though.

As far as I know, The Viper and Arri are utilizing HD-SDI. The Arri outputs just about exactly 3Gb/sec in normal mode, which is cutting it kind of close for a single SATA II connection. I think it may go over 3Gb/sec in 75fps mode. They're running the Arri over 3 HD-SDI's, each of which have 1.5Gb/sec...

Of course, maybe SATA II will come into play for these high end systems, but... so what? Weren't you the one who brought up marketshare? These systems have barely seen the light of day, and are strictly rental material. Whatever interface they're using on those giants has no bearing on anyone here...

My view of the storage market may be incomplete as a whole, but I have a pretty good sense of what professionals and prosumers are using for video. When I make guesses about marketshare, they're simply based on what I see people using, and aren't intended to be taken as empirically or statistically sound... So the next time you declare accusatorially that someone doesn't know the "actual sales numbers" and hint vaguely at Google results from god-knows-where, you might want to be acquainted with said numbers yourself.

Firewire is built into the computers and DV equipment we use now, and the HD equipment that we'll be using -- HDV and DVCPRO-HD. For that reason alone, Firewire will continue to be a massive force in the video world, and is the best forward-looking choice for a portable direct-to-disk box.

SATA II is going to be awesome for renderfarms and workstations, but until I can look at my camera and my laptop and see two matching SATA II ports, we're still living in a FW world.

- ben

The google results I was hinting at are proprietary studies from Instat and iSuppli if you must know. They are only available for purchase, not for posting and linking. But the results can be inferred form the discussions and new articles on technology news sites and abstracts that reference them.

Firewire 800 was barely mentioned in the context.

First of all, you keep using "firewire" and "SATA" as equivalents - they are not. SATA is both a drive technology and a interface. Firewire is an interface only.

Firewire is a great interface and the standard for DV interconnects. I use it every day.

But storage is another matter. Capturing tapes to NLE drives is an enormous waste of time and money. Direct aquistion to disk is the future. The best low cost disk technology is SATA - now and in the future - not firewire; as firewire is not a disk technology but an interface stacked on top of disk technology.

Some firewire drives today and all firewire drives in the future will have SATA drives inside.

I'm suggested to Juan that he start with a SATA drive and firewire interface to begin with. So we agree here since you have not indicated what type of drive Juan should put inside the firewire enclosure.

But a SATA II external port will be faster, more efficient and eventually lower cost due to volume. And they will be available soon - I'm not wrong about that.

Some numbers.

-image size: 773x495
-3 channel 16bit TIFF file = 48bits
-single image size (uncompressed) = 2.189 MB
-52.536 MB/s at 24 fps
-60 minutes at 24fps = 185.74 GB

As has been discussed 16bit tiffs are way more than needed for the 12(10 now) bit data. But it does make it much easier to work with.


Some other numbers to note.

D1 (601) 422 ntsc uncompressed video:
720x486, 0.67 MB/frame, 20 MB/s, 60min=70.4GB

HD1080 444 24fps uncompressed video:
1920x1080, 5.9 MB/frame, 142.4 MB/s, 60min=692.5GB

SATA isn't the problem that I used to think it was. There are external enclosures (although not powered which might be a problem for a hand-held camera), and the interface does support hot-swapping. Additionally you can purchase cardbus adapters for SATA, so there shouldn't be any problems on the laptop front. Frankly I think it's only a matter of time before SATA does overtake the external hard-drive market over firewire800, but that may take a year or more before stuff is readily available, right now SATA stuff is quite sparse whereas firewire800 is everywhere. So while we may argue over this fact over and over, I'd say we go with the solution that works now and gives us greater choice, because frankly if we wait around for another year just to get a hard-drive interface that we want, then they'll be something else on the horizon and nothing will ever get done.

BTW, on the other topic of 10-bit linear Cineons, all those files should be compatable with every program that supports Cineon/DPX-the reason being that the lin-log conversion takes place via a look-up table inside Shake, combustion, etc. All these programs read the log file as a linear file, that's why they look so "light" when you open them. After applying a LUT, you get the linear image the way it should look. With Juan's approach of using a linear Cineon file, you'll simply ignore the loading of the LUT.

One thing though, I thought the RGB files off the A/D converters was 12-bit. If you use a 10-bit linear Cineon, then you're loosing a lot of the highlight dynamic range. I'd say use either a 10-bit LOG Cineon to perserve that dynamic range, or else use a 16-bit Tiff file. I would find it very unfortunate that you'd have all this dynamic range and then clip off the top two bits to squeeze it into a 10-bit linear file, destroying the advantages of the extreme highlights that this camera can handle. Just my .02 cents.

I agree with Jason, and this is the same reason i came to integrate FW800 to the design. As much as SATA looks great and right now seems like it is the future, FW800 is what is available right now. Perhaps when I apply this mod to one of the new 3-CCD HD cameras, SATA will be more mainstream and we can switch over to that....besides, the added bandwidth WILL be needed then. :)

The cineon format supports 12-bit linear, so it can still accomodate all the information. I think someone said that 10-bit is the most common, but like jason says, i'd rather go 16-bit TIFF than sacrifice 2-bits/channel of color, unless the user explicitly wants to.

Juan

BTW, just curious, do the 7200RPM Hitachi 2.5" HD's sustain a high enough data rate to support the 16-bit TIFF's? I was thinking you might be able to use one of those smaller drives for a "magazine" if you will, and then offload that onto a much cheaper 3.5" HD attached to a laptop after the fact (or just keep swapping out the 2.5" drives if you have the cash :-)

I checked out all the small laptop drives (including the hitachi's) that i could find, and it didn't seem like they could handle it, at least by any significant margin.

However, if you have anything particular in mind, just see if it can do at least 42MB/sec sustained write....the numbers peter posted are correct, yet a bit over-bandwidth for the actual application, since it actually captures 12-bitsRGB(36bits) and then later converts the data to 16-bit TIFF's. So the actual sustained write rate is ~42Mb/sec.

This number is for the final application however, when dummy bits are stripped out. Right now, because of all the dummy bits present in the raw images, my test computer is recording ~63MB/sec sustained, with a hard disk 2 years old that i bought at circuit city. Stripping out the dummy data and using a modern FW800 drive is a much better situation. :)

Juan

Juan:

For some reason I'm having a really hard time making this point. There is no such thing as a "FW 800" drive. Inside, you will have to put either a EIDE or SATA drive.

You are absolutely correct Stephen. I have not seen any FW800 external cases that accept SATA drives, only EIDE. I don't think it wouldn't make a difference in speed whether a SATA or EIDE drive was used in a FW800 external case anyway.

Stephen:

I agree, and i know exactly what you mean. :)

But, what should we call my LaCie 200GB drive from now on? =)

Juan

<< For some reason I'm having a really hard time making this point. >>

It's because it's not a very good one. We're not discussing the type of actual hard drive mechanism to use, because this project does not include an internal drive. We are discussing what kind of port to put on the box. You will then connect a cable to the port to an external drive. The external drive has an internal drive in it -- and yes, if we're talking about a FW800 drive, 99.99% of the time, there's an IDE drive in there.

But you seem to be a bit confused as to what you're arguing for. Are you really saying we should put an SATA port on Juan's box? That's what you seemed to be saying a few posts ago.

<<There is no such thing as a "FW 800" drive.>>

Wow. Actually, the term "external drive" refers to the whole package, just like "computer" doesn't just refer to the motherboard. "External drive" is often shortened to "drive," if you've already made it clear that it's external. Therefore, it's perfectly reasonable to call an external drive a "FW800 drive," or a "Firewire Drive," or a "USB2 drive," etc. It doesn't matter whether there's a SCSI or EIDE or SATA II drive. If the back of the durn thing has Firewire ports, it's a Firewire drive. Unless you're on some kind of Quixotic quest to abolish abbreviations, I'll keep on saying "firewire drive" rather than "EIDE hard drive within a FW800 enclosure."

<< inside, you will have to put either a EIDE or SATA drive. >>

Or a SCSI drive. So what are you saying? That we should use SATA in FW800 enclosures?

<< The google results I was hinting at are proprietary studies from Instat and iSuppli >>

Convenient! But unless these were studies specifically of the media/video market, and unless they were conducted at least 6 months after FW800 came out, I refuse to believe that you're somehow more informed than me on this particular issue.

Bottom line: I know FW800 is the best choice, Juan knows, Peter knows, Jason knows, everyone seems to agree that FW800 is the best currently available choice... except for you, Stephen...

- ben

I think FW800 is the way to go (EIDE drive in a FW800 enclosure) using a FW800 cable from camera to drive. Yes, there are some using SATA external, including myself, but it is not proven itself yet and the cables are not as durable as FW especially on the road. My 2 cents.

Ben:

I think my point is an excellent point despite it's alleged unpopularity. And I always think having various perspectives is important.

Having a SATA drive in a FW800 enclose gives far more options than a EIDE drive in a FW800 enclosure. Especially if Juan offers a removable tray that could plug into a hot-sway SATA enclosure. That would be a nice speed boost over FW800. That would provide both FW800 and SATA options as well as being backwards compatible with EIDE systems (via SATA/EIDE adapter). Everybody wins.

I'm not confused about what I'm proposing. I was continuing previous discussions about this in this thread from several months ago that you were not a part of.

I said then, that by the time Juan had a working prototype, SATA II had a good chance of being ratified, which it is. I also felt that once he knows if he's just going to self-assemble a few unit on a low scale or look for a larger business model and outsource manufacturer/assembly, looking at SATA drives makes sense to me.

I still think Juan is some months at a bare minimum away from having a product to sell. While SATA II looks like pie in the sky for external drives to you right now, it might be different in several months.

I personally don't think calling something a "FW800 drive" in this thread is accurate. While that's fine for a label on a box, we are discussing Juan building a item from scratch and technical accuracy is very important. He could outsource and buy a direct drive recording but I got the impression he wanted to design and build his own.

Per the market share studies, I don't have the money for them either and sorry if I gave that impression. I did look at a number of the articles written about them and made my own conclusions. Which could be wrong.

Bottom line: All the bold lettering in the world is not going to make me less of SATA proponent. I understand why you favor FW800. Why not agree to disagree?

<<<-- Originally posted by Don Barzini : I think FW800 is the way to go (EIDE drive in a FW800 enclosure) using a FW800 cable from camera to drive. Yes, there are some using SATA external, including myself, but it is not proven itself yet and the cables are not as durable as FW especially on the road. My 2 cents. -->>>

Don:

There are several FW800 external enclosures on the market using SATA drives - I've seen mostly rack mount styles so far, not portables, but the chipsets must be out there. I recently was shopping for all sorts of drive options and recall seeing them.

Also, SATA II changes the external cabling and hopefully new, durable, longer cables will come to market quickly.

Anybody else desperate for more clips from Juan :)

Why not agree to disagree? Because you fundamentally misunderstand what Juan is building.

He's not building something with an internal drive, hence a "removable tray" is of no use. He's building a box that will connect to an external hard drive of your choice.

Here's the important part: your choice. So if you really love SATA II that much, you'll be able to get a SATA hard drive in a FW800 enclosure.

However, you also don't realize that the FW800 drives Juan is using are perfectly adequate for this application. There's absolutely no need for a "nice speed boost over FW800," because we're not even approaching the max transfer rate of the drive, let alone the interface.

Furthermore, although Juan may be months off from shipping a product, he's developing and testing right now. You don't develop with non-existant technologies.

I just really, really, really don't understand how we can still be having this conversation...

- ben

I think the question is why are we arguing over what's going inside the FW enclosure when the FW enclosure itself is hot swappable? You could have SATA, EIDE, SCSI, or Fibrechannel hard drives... or a RAID-style multiplex across 50,000 floppy drives for all I care. If the connection from the drive to Juan's capture device is FW800, then we, frankly, don't need to worry about what's going on inside the enclosure as long as it can handle 50MB/s sustained write speeds.

It's really not that critical. Juan said he's going with a FW800 interface, so that's what he's going with. If, at some later point he decides that SATA2 would make a better choice, then we can jump off that bridge when we come to it.

[Edit: ... yeah, and what Ben said. :P :) ]

Agree to disagree
 

Ben,

No one wants to be told that they "fundamentally misunderstand" something. By making statements like that you invite the converstation to continue.

Why not agree to disagree? Cause you can't say that in one sentence and then in the next tell someone their wrong.

Ben:

I am encouraging Juan to build a interface and drive mount box and/or belt-shoulder rig for the DVX or offer it as an option, perhaps even supplying the drives as well.

If he just does a port only, ie. just a Firewire FW800 and the stream down the port, a number of issues may come up:

e.g. controlling the start/stop record process, dealing with all the possible drive, enclosure and chipset combos, dealing with cable length and quality issues. It might be a real tech support compatability issue. Now if he just decides to hand assemble a few units on the side or build a few prototypes to market around, a simple FW800 port is fine.

But I think an elegant, secure, supportable drive solution is a vital part of this soluiton and would vastly increase it's usability. A partnership with an existing company that already does the direct drive recording is also an option.

Now we can agree to disagree.

FW800 is a robust standard that can easily support groundbreaking features such as "stop" and "start," and you need only look at the many FW DTD solutions to see that this is easily done.

I don't believe you'd have to support chipsets individually -- I think FW800 abstracts them and provides a common interface. However, you may need some way for the device to know the physical details of the drive -- can these be queried from the drive over FW? Juan may know.

Having a port is the most elegant option, since it allows the maximum flexibility. It may not be as easy as an affixed drive, but from what Juan has said, it doesn't seem to be a massive hurdle either.

Now, since Juan has all but settled on a box with a FW800 port, can we once and for all drop this aspect of the discussion?

- ben

Ben:

If you find the discussions tiresome, don't particapate. Your recent posts directed towards me appear to contain a lot of frustration, sarcasm and condensation. If that's not what you intend, perhaps you should re-word them.

I realize firewire supports start/stop commands, but external firewire drives lack controls for it. I would also like to see onboard shuttle/review controls for going back to the monitor as well a LCD display for dropped frames, time/space remaining etc.

The QuickStream DV drive from MCE has an interesting approaches to some of these issues. However, a standard firewire portable enclosure won't have these features (redundant buffers, internal battery, LCD etc.).

However, I feel this field is wide open - if I were a hard disk engineer, I think the QuickStream could be improved.

can't you guys let Juan do the work and stop bombing him with this stuff? I don't even want to read the board anymore..drop it