internal or external dvd writer?

  • Thread starter Thread starter Dave
  • Start date Start date
D

Dave

was goin to get the internal drive as it is cheaper however had a
thought..if i get external would i be able to burn to my existing cd writer
and the dvd writer at the same time and could this also be done if was an
internal writer?

thanks
Dave
 
Dave said:
was goin to get the internal drive as it is cheaper however had a
thought..if i get external would i be able to burn to my existing cd writer
and the dvd writer at the same time and could this also be done if was an
internal writer?
Click to expand...

It's ill-advised to try simultaneous burning with any pair of
writers--internal or external. You'd be asking way to much of the
source disk's bandwidth, and the likelihood of either very slow speeds
or completely failed burns would be quite high.

If you are thinking of being able to make 2 copies of CD's at the same
time, I don't think that's possible with your typical burning
software. Mass duplicators do exist, but I believe they run rather
specialized firmware/software for doing this, and I also believe
they're tpyically SCSI based.

It's an interesting topic though, and I'd welcome those more
in-the-know to contribute more info.

Best Regards,
 
was goin to get the internal drive as it is cheaper however had a
thought..if i get external would i be able to burn to my existing cd writer
and the dvd writer at the same time and could this also be done if was an
internal writer?

thanks
Dave
Click to expand...

There is no benefit to having it external unless you need it portable,
and it will be slower on an external bus, but not much given it's an
optical drive.
 
2 HD, 2 burning programs, burning 1 Cd with data from HD 1 and another CD
from HD 2 still asking to much from the system and a high likelyhood of
errors and burn failure?
 
Shep© said:
I don't know about that as I don't have an external USB Burner but USB
2.00 can handle a lot of bandwidth?
Click to expand...

You are correct. Hi-speed USB2 mode is 480Mbps which is faster than
a UDMA/33 IDE bus 33MB/s = 264Mbps

Comparing the specs of 2 popular 8x Plextor burners shows that the USB
model can actually burst faster than the IDE

http://www.plextor.be/english/products/PX708UF.html
Data Transfer Rate
-- Burst USB 2.0: 480Mbps
USB 1.1: 12Mbps
IEEE 1394 (Firewire): 400Mbps

http://www.plextor.be/english/products/PX708A.html
Data Transfer Rate
-- Burst (UDMA/33) 33MB/s (that's 264Mbps)


All other read/write specs you can see are the same, and quite a bit
lower than the peak transfer rates of either the IDE interface or the
USB2 interface.

This leaves the question of "okay, what is the real world experience
of these specs?" The only variable we haven't accounted for is how
the drivers for UDMA differ from USB as implemented in Windows. I
don't know the answer to that question, but best i can tell, that's
the only thing that could possibly support the possibly-antiquated
notion of a USB-based writer suffering versus and internal one.

I go internal myself for the cost benefit, and relative stability of
the IDE bus. In my experience, the fewer things I hang off the USB
ports in Windows, the more stable my experience.

Best Regards,
 
Dave said:
2 HD, 2 burning programs, burning 1 Cd with data from HD 1 and another CD
from HD 2 still asking to much from the system and a high likelyhood of
errors and burn failure?
Click to expand...

Hell, I don't even wanna run an mp3 audio player when I burn CD's or
DVD's in windows.

Yeah, I think that would increase the likelihood of burn failure, or
at the very least, dumb down the burn rate because of trouble keeping
the write buffers full on teh burn.
 
You are correct. Hi-speed USB2 mode is 480Mbps which is faster than
a UDMA/33 IDE bus 33MB/s = 264Mbps
Click to expand...

No, the spec, on paper AND in real life, DEMANDS that it always be
slower. Why? Because it's STILL an IDE device. It's constrained by
the IDE bus just as if attached to the motherboard, then futher slowed
down by processing, buffering of the USB bridge chip, and then again
on the USB bus.


Comparing the specs of 2 popular 8x Plextor burners shows that the USB
model can actually burst faster than the IDE
Click to expand...

There are 3 (or 5) possible reasons for this).

1) ATA66 drive

2) SCSI drive (would be faster than USB2 using native SCSI interface,
eliminating USB as with the IDE drives).

3) Windows buffering caused misleading results.

4) They intentionally misrepresented the data rate.

5) Noting the 480Mbps they listed, they were not trying to imply that
it can actually burst at full 480, nor sustain it... there's always a
bit of overhead even if it could "try" to burst high enough to
saturate the bus. "480" is ONLY the speed of the bus, not the drive's
throughput.
http://www.plextor.be/english/products/PX708UF.html
Data Transfer Rate
-- Burst USB 2.0: 480Mbps
USB 1.1: 12Mbps
IEEE 1394 (Firewire): 400Mbps

http://www.plextor.be/english/products/PX708A.html
Data Transfer Rate
-- Burst (UDMA/33) 33MB/s (that's 264Mbps)


All other read/write specs you can see are the same, and quite a bit
lower than the peak transfer rates of either the IDE interface or the
USB2 interface.
Click to expand...

As I wrote previously, that's only the speed of the USB2 bus, not the
drive, and especially not when bridging ATAPI to USB2 then
transferring to the motherboard.
This leaves the question of "okay, what is the real world experience
of these specs?" The only variable we haven't accounted for is how
the drivers for UDMA differ from USB as implemented in Windows. I
don't know the answer to that question, but best i can tell, that's
the only thing that could possibly support the possibly-antiquated
notion of a USB-based writer suffering versus and internal one.
Click to expand...

NO, as I already mentioned, even with a perfect hardware support with
no additional drives, USB2 would be slower.

IF the drive's I/O is lower than the sustainable I/O of the USB2
controller (both motherboard and bridge chip in external encloser,
which typically peak at 21MBps but often less with cheaper chips),
then there's primarily the latency from bridging the two busses. It's
not a huge difference but only because optical drives are relatively
slow... on a HDD it's much larger difference.

I go internal myself for the cost benefit, and relative stability of
the IDE bus. In my experience, the fewer things I hang off the USB
ports in Windows, the more stable my experience.

Best Regards,
Click to expand...

I'm not knocking USB2, it's a great alternative for lowest cost
external drives, but to be clear it's not possible to be higher
performance nor even same performance, but possibly only a very tiny
bit slower than direct IDE or SCSI.
 
2 HD, 2 burning programs, burning 1 Cd with data from HD 1 and another CD
from HD 2 still asking to much from the system and a high likelyhood of
errors and burn failure?
Click to expand...


Errors and burn failure aren't near as likely as having the drives
"burn-proof" feature kick in, that the laser has to stop and start
again a few times waiting for data. I don't recall trying to burn
with two programs and burners simultaneously so I don't know for sure
that it'd work, but if it didn't I suspect it'd be windows' fault more
than anything else.
 
kony said:
5) Noting the 480Mbps they listed, they were not trying to imply that
it can actually burst at full 480, nor sustain it... there's always a
bit of overhead even if it could "try" to burst high enough to
saturate the bus. "480" is ONLY the speed of the bus, not the drive's
throughput.
Click to expand...

I do understand this. I encourage you to educate me further though on
one point.

Don't overlook this:

All other drive speed specs are identical. For example, for DVD+R it's
6X-8X: 8.100-10.800KB/s (ZCLV) -- that's just 0.064-0.086Mbps. Both
drives cite this burn rate throughput. Now, this fastest burn speed
is 0.03% of the burst capability of the IDE interface, and and 0.02%
of the burst capability of USB2.

This is why I'm having a hard time understanding why a USB2 drive is
guaranteed to be slower--because the write data rates are a very tiny
fraction of the bus throughputs.
As I wrote previously, that's only the speed of the USB2 bus, not the
drive, and especially not when bridging ATAPI to USB2 then
transferring to the motherboard.
Click to expand...

Unless bridging is going to make things over 100x slower, I still
don't see it.
NO, as I already mentioned, even with a perfect hardware support with
no additional drives, USB2 would be slower.
Click to expand...

Are we talking now about the bus speed? Or the drive's ultimate
performance?
It's not a huge difference but only because optical drives are
relatively slow...
Click to expand...

Tada! You have reinforced the point of the spec analysis. All the
optical drive throughput numbers on the specs linked above are
identical, and both are much lower than the capabilities of the bus
speed measured.

So, it's quite likely that there could be no performance differences
between two DVD-RW drives that differ only in interface.
on a HDD it's much larger difference.
Click to expand...

No doubt.
I'm not knocking USB2, it's a great alternative for lowest cost
external drives, but to be clear it's not possible to be higher
performance nor even same performance, but possibly only a very tiny
bit slower than direct IDE or SCSI.
Click to expand...

If the relative throughput of the drive itself is X, and if interface
A is capable of sustaining, say 100x throughputs, and interface B is
capable of sustaining 150x, I'm still not clear on how/why you
maintain that a USB2 version of a DVD-R drive must necessarily be
slower.

I'd be very curious to see a side by side real world burn speed
comparison to settle this analysis-based disagreement we're having.

Best Regards,
 
If the relative throughput of the drive itself is X, and if interface
A is capable of sustaining, say 100x throughputs, and interface B is
capable of sustaining 150x, I'm still not clear on how/why you
maintain that a USB2 version of a DVD-R drive must necessarily be
slower.

I'd be very curious to see a side by side real world burn speed
comparison to settle this analysis-based disagreement we're having.

Best Regards,
Click to expand...

The rate of data transfer from the drive to the USB enclosure, is the
same as it would be from the drive to the motherboard, if directly
connnected. Add to that the time it takes to bridge and transfer on
USB. Because the drive's max sustainable througput isn't exceeding
IDE or USB througput, the "overall" burn process can proceed at full
speed. However, burst rates are not nearly the same speed, likely
around 50% lower (rough guess based on benchmarks of USB2 performance
with HDDs, typically around 17-21MBps dependign on the USB
chipset/integration used). Add to that the situation with the system
taking more time per data transfer, even though the drive can get the
data fast enough, it's tying up the system longer to do it... the
system could be doing other burning-related tasks if it had finished
transferring the data over the USB bus to the drive, already.
Multitasking would become more difficult also.
 
Back
Top