Log in

View Full Version : NEW DVD Burner at 16x - How fast is fast enough..?


geffroman
31st January 2004, 19:24
I am looking for input on this one... I am very interested in making my next burner a dual layer 8X machine... I see that 12x & 16x machines are being discussed. But I have to wonder how a DVD burner that is supposed to burn data faster than a computer can deliver it makes any kind of sense...

Maybe my math is off... but a disc to disc copy of 4GB of data on a typical UATA100 hard drive at 50MB/sec actual throughput is about 8 minutes... An 8x burner can burn this in about 7 mins (already faster than average computer drive can deliver it).

Are 12x & 16x burners going to be UATA100 or UATA133 or SATA150..? Will we have to use a RAID in STRIPE mode with dual SATA drives to use this burner and get the 12x or 16x performance?

Will we all need dual firewire internal drives with dedicated drive for Windows TEMP directory to speed read/write performance...? What's it actually going to take to run a 16X burner at 16X...?

Then there is the media... You have a hard time getting your DVD-Rom to READ any disc at a full 16X... How are you going to get a 16X burn on a DVD-R...

I know I have NEVER seen my CD Writer burn a 40X CD on my 48x CD Writer and I am running a RAID UATA133 that tests at 97MB/sec...

Can someone help me understand this one...? Cause right now I cannot see the advantages of anything beyond 8x until we get to MUCH faster computer bus and drive systems.

Kedirekin
31st January 2004, 19:53
I wonder ... maybe they'll be targeted at mass duplicators with robot loaders and 10 GB RAM drives.

geffroman
31st January 2004, 20:00
Originally posted by Kedirekin
I wonder ... maybe they'll be targeted at mass duplicators with robot loaders and 10 GB RAM drives.

I don't see these drives being powerful enough to replace mass duplicating efforts on lower performance DVD-R discs... I think mass producers will still contract with DVD Video disc producers...cheaper per disc... more durable product...

LIGHTNING UK!
31st January 2004, 20:42
DVD Speed is 1,350Kb/s so 16x DVD would need approx 22Mb/s read transfer rate from the hdd.
I dont think that's too much to ask.

alexnoe
31st January 2004, 20:52
Geffroman: Those 50 MB/s are a limit of the discs. An UDMA100 Interface can easily do 90 MB/s (maybe less on VIA chipsets)

However, those drives will (at least Pioneer drives) be SATA. Not because IDE is too slow, but because SATA has integrated DRM...

geffroman
1st February 2004, 00:12
Originally posted by alexnoe
Geffroman: Those 50 MB/s are a limit of the discs. An UDMA100 Interface can easily do 90 MB/s (maybe less on VIA chipsets)

However, those drives will (at least Pioneer drives) be SATA. Not because IDE is too slow, but because SATA has integrated DRM...

That depends totally on how you test... Typical read/write behavior on an IDE Bus is half that...

Let's take the math out of it and make this simple...

Do a simple copy to another hard drive on your system... Use Master to Master on Primary and Secondary IDE if you want... Copy 4GB of data on any UATA100 or UATA133 and tell me what people get...

I am guessing you'll find most people need about 5 to 6 minutes or MORE...

8x can burn that in 6 to 7 minutes
12x in 5 to 6 minutes..?
16x in 3 to 4 minutes..?

If a 16x machine gets me MORE stability on 8x dual layers discs I think it makes sense... but not for speed... especially if it requires people to upgrade their system for an SATA interface...

alexnoe
1st February 2004, 00:15
Geffroman: the hard discs are limited! Your test won't tell you anything about the speed of the interface.
Typical read/write behavior on an IDE Bus is half that... On VIA chipsets, maybe, but not on *real* IDE Busses.

But if 4 GB from one drive to another takes 6 minutes, you should really replace your VIA chipset...

Test file: 4,26 GB
Disk 1: Raid-0 2x80GB WD (WD 800 BB), both Slaves
Disk 2: WD 1200 JB (Master)
Disk 3: another WD 1200 JB. (Master)

All discs are connected to a FastTrak 100 TX2 controller, and I use a Intel chipset, so no broken VIA chipset drivers are involved.

Disk 1 -> Disk 2: 2min 30sec
Disk 2 -> Disk 3: also 2min 30sec

geffroman
1st February 2004, 01:17
Originally posted by alexnoe
Geffroman: the hard discs are limited! Your test won't tell you anything about the speed of the interface.
On VIA chipsets, maybe, but not on *real* IDE Busses.

But if 4 GB from one drive to another takes 6 minutes, you should really replace your VIA chipset...

Test file: 4,26 GB
Disk 1: Raid-0 2x80GB WD (WD 800 BB), both Slaves
Disk 2: WD 1200 JB (Master)
Disk 3: another WD 1200 JB. (Master)

All discs are connected to a FastTrak 100 TX2 controller, and I use a Intel chipset, so no broken VIA chipset drivers are involved.

Disk 1 -> Disk 2: 2min 30sec
Disk 2 -> Disk 3: also 2min 30sec

You just helped proved my point... Whether the road block to speed is the IDE controller or the IDE hard drive the result is the same... How do you put more on a DVD than the COMPUTER BUS/DRIVE can deliver... Truth is I can put the same UATA133 hard drive on a variety of interfaces and see different speeds so I guess I don't agree it's just the hard drive's limitation. You say it's the hard drive but you also blame Via chipsets which sit in a lot of these folks computers...

Your set up is somewhat like mine. I too use the Fastrak controller but it is on an Intel 845E chipset Mobo w/ P4. Also like mine yours is not typical... I get 97+MB/sec transfer speeds but MOST people don't have RAID controllers... and most people see 60% of the data transfer speed that you do... so most people should know what they are NOT going to get if they wait around for 12x & 16x burners...

In the end, I'm just looking for answers.

I'd really like to hear from others with typical IDE controllers (UATA100 & UATA133) and typical hard drives (UATA100 OR UATA133 7200RPM IBM, MAXTOR OR WESTERN DIGITAL).

If anyone would like to reply just quote this post and fill in the blanks:

CPU:
Chipset:
HD Controller: UATA100 or UATA133 or SATA

Hard Drive One
Brand:
Interface: UATA100 or UATA133 or SATA
RPM:
Size:

Hard Drive Two
Brand:
Interface: UATA100 or UATA133 or SATA
RPM:
Size:

Disc to Disc copy speed of 4GB?

DMagic1
1st February 2004, 02:21
I think it will be more like this:

8x 8-9mins - Now Availiable
12x 7-8mins - Future
16x 5-6mins - Future

quantum
2nd February 2004, 20:31
Originally posted by geffroman
Maybe my math is off... but a disc to disc copy of 4GB of data on a typical UATA100 hard drive at 50MB/sec actual throughput is about 8 minutes...
I think your math is off. I burn at 4x from the drive on another computer on my LAN. My network meter shows around 5.6mB/sec. Nero reports about 14 minutes to burn a full disk. 14 x 60 x 5.6 = about 4,700 megs (full DVD).

Extrapolate from this a 16x burner would need 22.4mB/sec. My off the shelf IDE drives can handle this, although just barely.

geffroman
3rd February 2004, 02:10
Originally posted by quantum
I think your math is off. I burn at 4x from the drive on another computer on my LAN. My network meter shows around 5.6mB/sec. Nero reports about 14 minutes to burn a full disk. 14 x 60 x 5.6 = about 4,700 megs (full DVD).

Extrapolate from this a 16x burner would need 22.4mB/sec. My off the shelf IDE drives can handle this, although just barely.

I think I agree... If the media can hand the 16x I think the average PC user has a shot at 12x to 16x burns... By average I am suggesting 1.6 to 2.0GHz with UATA100 min & comparable drive... After doing more testing on several systems here as well as data from you guys I think my math was off...

A 12x or 16x burner that can handle dual layer disc is about perfect... Not sure if I will wait or jump on the first Pioneer version of an 8x dual layer burner (guessing it will be the DVR-108/DVR-A08)...

quantum
3rd February 2004, 03:17
I doubt the CPU makes much difference. When burning at 4x, my Athlon 1.4 CPU is maybe 5%. In fact I play Halflife while burning at 4x with no problems.

alexnoe
3rd February 2004, 05:15
I'm sure you know that the first Dual Layer burner will be writing Dual Layer at 2x?

geffroman
3rd February 2004, 22:38
Originally posted by alexnoe
I'm sure you know that the first Dual Layer burner will be writing Dual Layer at 2x?

Well they will just have to speed it up... I ain't waiting no HOUR... been there... done that... LOL

alexnoe
3rd February 2004, 22:53
Philips thinks that they could have 4x dual layer somewhen 2005 ;)