IDE channel saturation

[TopHatTheater]

I have two burners and I've been wondering if I can burn two discs at once. Here's the setup:

IDE 1: 120GB Seagate Barracuda
IDE 2: DVD/CD+RW (4x DVD+R, 40x CD-RW) set as Master, CD-RW (48X40x48) set as Slave

Do you guys think I could burn files from the hard drive using both drives simultaenously or would my IDE channel be data saturated beyond that ability? Other hardware info:

XP2100 @ 2.3GHz
Abit NF7-S
2x 512MB Corsair XMS3500C2

~THT

[colordog]

You can speculate, but since every computer is different, it's probably best just to quickly test burning both at the same time - then you'll know.

[Ramirez]

Nah, don't do it, plugging such a bandwidth sucking devices on the same IDE channel it's a very bad idea , don't get me wrong though , it Will work ^ ^ but in very inefficient way, your burners will spent tons of time on preventing hundreds buffer underuns and burning process will be unbearable slow. Instead, I'd suggest you to move your HD on the SATA controller using IDE>SATA converter. (Came bundled with your motherboard)

[TopHatTheater]

I should have thought of that. I'll move the Barracuda to the free SATA channel and move one of the drives to IDE1. Thanks for the inspiration.

~THT

[ppera2]

CD burners (and readers) have low bandwith - count so: 50x150KB/sec - it's only 7.5 MB/sec.
I have CD Burner and DVD reader on same Atapi line, and I can extract DVD with max speed while burn CD - DVD read has higher dataflow than any CD writer.

[Ramirez]

This is Not about HIGH bandwidth requirements but about STEADY gapless data transfer something which is quite often very hard to achieve even on higher end hardware.

2 ATAPI devises (DVD reader\CD burner) sharing the same bus is not the same as two CD\DVD burners as they are controlled by the "Burn-proof" technology which suspends the burning process in any event when the buffer goes empty (it's happens very often even with only one burner connected and it's greatly prolonging burning times)

[ppera2]

Sorry Ramirez, but what you write is pretty obsolete. By modern MB, well set BIOS and Windows, buffer underrun almost never happens, even by running 3-10 programs at same time.
I can play games, surf etc. and burn-proof never activates. Drive's buffer never goes under 90%.
I don't know from where you get that buffer underrun so often happens...

In any case, TopHatTheater can make some tests and see...

[Ramirez]

Obsolete you said? This is all very elementary and very much known issues and if you'd ask anyone who dealing with hardware for a living you'd never hear from him "go ahead, connect 2 burners on the same IDE port, it's OK" :rolleyes:

I've seen and tested hundreds systems based on wide variety of hardware () and believe me it is still true (especially with fast 2MB Cache 48/52 burners); B.U does occurs regardless quality of hardware in use, and If you're completely free of such probs then you're indeed lucky :--/

P.S
Burn-proof is always active unless if you're explicitly turning it OFF.

[colordog]

From my own experience, I see results similar to what ppera2 states. I'm almost always encoding while I burn CDs, without my internal buffer dropping below 90%. But I also have sinked a lot of money into my system.

With a newer computer, your bus width is certainly no less than DMA Mode 2, which equates to 33MB / sec / IDE channel. Logistically, you shouldn't have trouble running two 50X CD burners on the same channel, as that's just under half the minimum available bus width. I have a CD burner and DVD-ROM on the same channel, and have used both simultaneously with no problem.

TopHat - Of course, people may say yes or no, but a 5-minute, 40 cent test will tell you for sure before you hassle with moving drives around. Just start two burns and find out. Good luck!

[ppera2]

Originally posted by Ramirez ...lucky...
Burn-proof is always active unless if you're explicitly turning it OFF.

I don't think that I'm lucky - just know limits, and how what works.

This about burn-proof is misleading statement. Burn proof is activated by start of CD burn, but it means that option is just ready to activate itself by need. When buffer underrun occurs, then burn-proof really activates - it happened by me only 2 times, when I had system block for 1-2 minutes.

[Ramirez]

Originally posted by ppera2
This about burn-proof is misleading statement..
A simple search for a "Burn-proof" would return the following description of B.P..

When a CD is burned (recorded), the data must usually be written onto the CD without any interruption. If an interruption, such as the opening of a new application, does occur, a Buffer under Run error occurs and the burn is unsuccessful. BURN-Proof technology located on firmware inside the CD-R or CD-RW drive monitors the recording process. If the drive detects a Buffer under Run error, it suspends recording. When the problem is resolved, the CD-R or CD-RW drive restarts recording data from where it stopped.
So it IS always active after all..:rolleyes:

How hard is it to keep this discussion on topic? Just to remind you this tread topic is about how feasible it is to perform simultaneous burning using 2 burners connected on the same IDE channel, and I already gave my recommendation, NO, IT'S NOT.

Just for a better illustration I've made a little test today, I've connected 2 identical () X52 burners on one IDE channel and then benchmarked the burning speed. (Both drive where burned data simultaneously)

Hardware
Intel 860 motherboard
P4 2.6 HT CPU
256x2 DDR-400
Maxtor 120GB 8MB SATA HD
2 x52 Burners (2MB Cache)

Software
WinXP-PRO
Intel MB Drivers
Nero 5.5

Burning Setup
Burning Speed: X24
Burning Mode: DiscAtOnce
Overburn: Enabled
JustSpeed: Disabled
2 RAR Archives 720MB each

2 drives plugged into one IDE ATA-100 Channel:
Nero Ultra buffer fullness: 40%-100%
Device own buffer fullness: 10%-70%
Total burning time: 18 minutes, 34 seconds

2 drives each on its own not shared ATA-100 Channel:
Nero Ultra buffer fullness: 60%- 100%
Device own buffer fullness: 40%-80%
Total burning time: 9 minutes, 39 seconds.


Done with this tread :P

Sayonara.

[colordog]

I'd suggest that your computer isn't fully optimized if your Nero file buffer isn't always at 100%. The same with your device buffer. Even on separate channels, you're indicating that you can't feed your buffer fast enough to always keep it full (especially at your low testing speed of 24x). Which isn't a bad thing, per se, as long as it doesn't bottom out. But that's why trying to answer this question is kind of silly, as the "yes" or "no" answer is going to completely depend on the individual computer setup.

My own results are different from yours, and will probably be different from someone's else. The point is that (1) it is physically possible, (2) other people have done it (3) you may or may not be able to replicate it with your setup.

Also, I must agree with ppera2 again about the Burn-Proof - just because it's a function controlled by the firmware doesn't mean that you can't disable it. I can disable burn-proof, as well as SpeedRec. Now, I'm not sure why anyone would want to disable it... but who cares anyway?

I must say, I am curious why you tested your burners at half speed - any reason?

[ppera2]

Originally posted by Ramirez

Just for a better illustration I've made a little test today, I've connected 2 identical () X52 burners on one IDE channel and then benchmarked the burning speed. (Both drive where burned data simultaneously)

Hardware
Intel 860 motherboard
P4 2.6 HT CPU
256x2 DDR-400
Maxtor 120GB 8MB SATA HD
2 x52 Burners (2MB Cache)

Software
WinXP-PRO
Intel MB Drivers
Nero 5.5

Burning Setup
Burning Speed: X24
Burning Mode: DiscAtOnce
Overburn: Enabled
JustSpeed: Disabled
2 RAR Archives 720MB each

2 drives plugged into one IDE ATA-100 Channel:
Nero Ultra buffer fullness: 40%-100%
Device own buffer fullness: 10%-70%
Total burning time: 18 minutes, 34 seconds

2 drives each on its own not shared ATA-100 Channel:
Nero Ultra buffer fullness: 60%- 100%
Device own buffer fullness: 40%-80%
Total burning time: 9 minutes, 39 seconds.
Done with this tread :P
Sayonara.

Yes, yes, I talk very topic related. I don't want more about how burn-proof works, let's focus on your results:

This is poor misery what you described. As I said, device's % never goeas under 90%. Nero's buffer is always on 100%. It stays not only for long files, but for case of many short files.

I can imagine 3 reasons for such bad results:

1: You have no DMA turned on for IDE channels
2: Motherboard has failure or is very wrong set
3: Burners are crap

And yes, I worked with 2 burners on same IDE channel, and it worked like charm.

[Zhnujm]

Im just curious why it takes almost 10 minutes for a 24x burn even with separate channels.

[TopHatTheater]

Okay, I tried it and it took seven minutes to burn the two CDs. Both buffers emptied routinely. CDs play fine though. Anyway, I'll report back when I've switched the drive config.

~THT

[lilhobo]

ok, here s a naother sotry to be told...

i have a maxtor diamondmax 9 on an external rack of a scsi system dual p3 1 ghz....with ONE ide channel

so i connect the external rack and a NEC 1300A burner to it.

the external rack is rated as ATA133.

Nero fails on 5 ocassions for 5 coasters at the same point, app. 65%, for a cd iso NOT DVD....and nero was the sole prgram running

i switched it over ot the scsi win2003 server and everything sails along....

does the nec have burn-proof or not....it seems that consistent data thru put is still important

[TopHatTheater]

Three minutes to burn both CDs after shifting the drives to the following configuration:

SATA 1: Raptor
SATA 2: Barracuda
IDE 1: DVD+RW
IDE 2: CD-RW

~THT

[mmebane]

OK, lets straighten this out, shall we.


First off: BurnProof

With the latest versions of Nero (Nero 6, anyway), it is impossible to disable. In this case, the term active if iffy at best. Yes, it is always enabled, but it isn't always in use. Basically, it kicks in whenever the drive buffer empties. It stops the burning process and starts it back up when it has more data. It is not perfect, however, as it increases the number of C1 errors on the disc.


Now for dual drives/single channel:

You should have absolutely no problems with IDE channel saturation. The max a 52x burner uses is ~8 MB/sec. Even 2 DVD burners won't saturate an ATA-133 channel. What you do have to worry about, however, is the command speed of the drive in the middle of the chain, since it has to process the data for both drives. If it can't relay data to the second drive fast enough, you will have problems. I haven't seen any objective way to measure this, however, so all you can do is try.