Burning very slow

[AlanHK]

I've just rebuilt an old computer and installed a LG IDE burner.

In my previous PC, using the same drive, the same blank (TY) discs, using ImgBurn, I could burn at 8x.
Burning now it goes at 1.4x, or less.

From the log below you can see ImgBurn recognises the disc type and sets speed at 8x.
But then it actually burns at 1.4x, despite the "Max" setting.

So, is there a Windows setting that is screwing this up, or is it a hardware issue and I'm I stuck at dead slow burns on this PC?

===
Duh. Worked it out.
DMA/PIO mode.

ImgBurn actually has a tool to fix it.

[lordsmurf]

Don't forget the channel issue. (Optical and HDD on opposite cables/channels.)

Easy to forget that in the era of SATA.

[AlanHK]

My PC is still in the IDE era.


It does what I need, so I don't have an excuse to replace it.
But it's hard to find IDE drives around here now.

[jman98]

Duh. Worked it out.
DMA/PIO mode.

ImgBurn actually has a tool to fix it.

Running XP? This actually gets into an "it's not a bug, it's a feature" argument with how XP was designed. Basically it was deliberately designed to do this under certain circumstances. In my opinion what triggers those circumstances really could have used a lot more thought by the programmers at Microsoft.

The ImgBurn tool and other simple fixes do not always take care of this issue. In stubborn cases it requires carefully changing a value in the registry and rebooting to fix it.

[AlanHK]

Duh. Worked it out.
DMA/PIO mode.

ImgBurn actually has a tool to fix it.

Running XP? This actually gets into an "it's not a bug, it's a feature" argument with how XP was designed. Basically it was deliberately designed to do this under certain circumstances. In my opinion what triggers those circumstances really could have used a lot more thought by the programmers at Microsoft.

The ImgBurn tool and other simple fixes do not always take care of this issue. In stubborn cases it requires carefully changing a value in the registry and rebooting to fix it.

Actually, Win2k.

If it makes my burner work at 1.3x instead of 8x, it's most definitely a bug.
Having good intentions doesn't make it a "feature".
Most of the time, it slows down your system dreadfully, without warning or explanation, because of a temporary glitch.
I can see over 90% of my processor occupied with interrupts when this was active.

Anyway, as I said, it's fixed now. ImgBurn's tool does it, or manually via the IDE/ATAPI controllers settings in Device Manager.
It's happened to me before, a few years ago, just took a while for the details to come back to me.

[jman98]

Without getting too verbose, Microsoft basically decided that its users were dumbasses (you can understand why they might think that) and it was just easier to have the burner work horribly slow in degraded mode (PIO) than to try to handle the conditions sensibly that made it decide to switch to PIO mode. Yes, Win2K has the same issue. Wow. Can't believe you're still using that.

There are several ways to handle this, but the best is a registry modification that sets the value of the error counter that Windows uses to FFFFFFFF (the highest possible value it can hold in hexadecimal) so that it will take many millions of "errors" (more on this in a minute) before it gets switched to PIO mode. These are not really "errors" in the way that you or I or any rational person who doesn't work for Microsoft would interpret that word to mean, but Microsoft decided that by God they were worth counting and worrying about and that were indeed just truly awful and if you encountered something like maybe 10 of them then it was full blown panic time and required an immediate switch to PIO mode. I would say that this was the right course of action less than 5% of the time this situation was encountered but they decided to make things easier on themselves by just degrading immediately rather than trying to figure out sensibly if it was a real problem or not. It's been some years since I looked into this. All I can tell you is that the vast majority of the time these "errors" could be safely ignored but they decided to play it safe and just assume the worst so that support didn't have to deal with the 5% of the time it was a serious issue. I think they honestly thought that "nobody will ever notice" when they switched into PIO mode.

[AlanHK]

MS eventually realised this was a problem.
See http://support.microsoft.com/kb/817472
It has a registry patch.

The idea seems to have been to fall back if the IDE disk (assumed to be a hard disk) was responding too slowly. But of course, if you have a damaged DVD you will quickly get a bunch of errors.

The MS devs will have very fast hard disks and burners, so never came across this problem, if they tested it at all with DVDs.

And yes, still using Win2k. Almost all "XP" apps work on it now. But within the next year I'll probably have to upgrade as some important apps like Firefox and Adobe Flash will drop support.
I use a lot of very old, even DOS, apps and compatibility with them is vital for my work, so I upgrade the OS very cautiously.

[[_chef_]]

Actually, Win2k.

Makes no diff with that issue.

For DOS you can create extra partitions and using emus and DOSbox etc. pp.