Burn Buffer Constantly Empties - Normal?

Estuardo wrote on 1/1/2003, 4:11 AM
I wonder if this is normal. I have a hunch it is not, but I'd like comments from others.

I burned my first CD-R with CDA5 using W2K, 1.4GHz, 768MB RAM and a Plextor 24x10x40 CD/RW. I selected "this burner supports burn proof" just before the burning operation.

I did a successful burn, but throughtout the burning operation, the burn buffer constantly went from about 94% to 0% at which time the CD/RW indicated it stopped burning (the LED changed from amber to green as well as the time remaining started counting up instead of down). About 1-2 seconds later the buffer would fill up and again go from ~94% to 0% and would keep repeating this buffer depletion until it finished.

The entire burn went smoothly and fast (about 4 minutes - "max" was selected) and the resulting CD sounds just fine.

My question is, isn't this burn buffer supposed to be filling up as it writes?

I wonder if I did not select that the burner supports burn proof if the buffer would simply empty and then create a coaster.

Is there any disadvantage to letting the buffer continually empty, but letting the laser stop and then start?




Estuardo wrote on 1/1/2003, 4:24 AM
Sorry, I should have searched BEFORE posting. Someone had made a post about the same probmem here http://www.sonicfoundry.com/forums/ShowMessage.asp?MessageID=138226&Page=0 .

End result - no definate answer.

One post made the most sense in that it's the way CDA handles burn proof writing and I think that's the problem.

I de-selected make a rendering file on the burn window since I had already made an image file which (I believe) is the same thing as making a rendering file except the rendering file is a temporary file. Anyway, my computer was just loafing along - wasn't doing anything during the burn. It appeared that it didn't even make an attempt to fill the burn buffer while it was writing (and being emptied).

Geoff_Wood wrote on 1/3/2003, 4:01 AM
The burn buffer fills and empties itself to the CD-R on an ongoing basis. If your burn speed is higher than the PC can feed data to the Cd writer, then the buffer will empty faster than it is filled. If you didn't have burnProof, you would have a coaster. However while BurnProof may give a usable disc, it is better to reduce your burn speed (so BurnProof is not invoked)in order to give a better quality master disc if it is going to be used as a 'production master'.

timparis wrote on 1/16/2003, 6:24 PM
Well I don't agree that it is just gear being not up to it. I have the same problem with a dual 2.2GHz with 1 GB ram, and a Yamaha F1.
Sonicade wrote on 1/16/2003, 11:14 PM
I agree that this should not be a problem and you should not HAVE to render a wav of the your project unless you have a very old cpu and have used lots of FX. The old CD Architect (4) had no problems filling the buffer on even older machines.

Geoff_Wood wrote on 1/17/2003, 4:47 AM
The old CDA4 had no effect chains or on-the-fly resample/mp3 conversion.

Given that if I load *one* effect like Isotope Ozone I can't even preview this in real time (without glitches) on a Celeron 533, it does not seem beyond the bounds of possiblity that it might still require pre-rendering on a significantly faster machine.

Sonicade wrote on 1/18/2003, 2:11 AM
Yes I can understand where pre-rendering would be very useful for complex effect chains and on-the-fly resampling, etc. However I used absolutely no frills, just plain track splicing with a few volume envelopes and the buffer chokes through the whole burn. I am using a Celeron 533mhz machine as well.

Geoff_Wood wrote on 1/22/2003, 2:49 PM
The resampling can be enough to choke it. Reducing the resampling quality (maybe from Best to Good - try it on some critical music and see if the reduction in quality is apparent)may help.

I can't do on-the-fly 48K resampling/burning on my 533, but I can on my 1G1 Celeron.

Estuardo wrote on 1/27/2003, 9:29 AM
I believe that it's a problem the way the burning program was written, possibly the way it interfaces with burn proof drives (as my Plextor 24/10/40 with burn proof).

Trying the same burn using my laptop (with a Sony CDR/RW and no burn proof), it burns a CDR just fine. But with my 1.4GHz and 768MB RAM, and essentially nothing else going on, but poor dear drains the buffer to zero at which point I can see the CDR/RW stops recording (the yellow LED extinguishes indicating it stopped recording), and then at only this point will the buffer then begin filling. As far as I can tell by observing the progress bar, no data is even being written during when the buffer is being emptied for a burn operation.

Others have well documented this problem, so at least it's conforting to know that I'm not the only one.

Maybe it's time to get another drive (LiteOn or another Plextor...).