The Burning Issue

PaulO wrote on 12/1/2002, 12:23 PM
Windows 98SE, 800MHz P3, Intel BX chipset, Plextor 2410A writer

Just upgraded to CDA5 (long-time user of CDA4) and there is clearly a problem with the burn engine compared to Nero, CDRWin, EasyCD. In fact, I wonder if the burn engine core has been changed since the days of CDA4 because this is exactly the same problem that I experience with CDA4 when I set the drive type to 'ATAPI MMC' in an attempt to get it to work with a modern burner.

If I set the burner's properties to DMA ON (in device manager), within a minute or two of starting the main body of the burn (after lead-in has completed), the burner stops dead and all hard disk transfer ceases. Pressing 'cancel' within CDA is acknowledged but the software then locks up, waiting for the burner to respond - which never happens. The only way out is via task manager.

While this appears prima facie to be a problem with the machine interface to the burner, none of the above-mentioned software (with the exception of both CDA versions) suffers from this problem or anything like it. This really does suggest that the CDA burn engine is nothing like as robust as one could reasonably expect these days.

If I set DMA OFF then the burn proceeds satisfactorily but sufficient CPU power is consumed to cause the 'burn-proof' rescue to be invoked a few times during the burn - I notice there are several earlier postings claiming that the buffer levels drop to zero a few times during a burn unless the pre-render facility is used.

To summarize - other burning software works quite happily on this machine with the burner set to DMA ON - only CDA has a problem and I think the burn engine requires some serious work to bring it up to modern standards.

Thank you.

Comments

Geoff_Wood wrote on 12/1/2002, 2:38 PM
So you are using realtime effect chains per track in EZCD CDRWin and Nero ?

geoff
PaulO wrote on 12/1/2002, 4:47 PM
>So you are using realtime effect chains per track in EZCD CDRWin and Nero ?

Geoff - that is rather missing the point! In any case, these tests were performed with NO event plug-ins engaged, and no dither in the master output section.

The problem which I am pointing out is that the burn facility within CDA5 (and CDA4) is prone to locking up when the drive is set to DMA ON, which does not happen with Nero et al. It so happens that when DMA is on, CPU utilization by the burn process is sufficiently low that CDA can keep up when processing on-the-fly. When DMA OFF is selected, there isn't sufficient remaining CPU time for on-the-fly rendering but this is not the problem.

Let's not get diverted here. My issue is that CDA's burn routine locks up in circumstances in which the other software does not. With DMA ON, CDA still locks up even when 'render to temp image' is selected.
Geoff_Wood wrote on 12/1/2002, 4:58 PM
I guess that before going off the deep end we really need to ascertain that this is actually a bug, as oppose to just a feature of your particular setup.

geoff
PaulO wrote on 12/1/2002, 5:02 PM
>we really need to ascertain that this is actually a bug, as oppose to just a
>feature of your particular setup

Well, I don't really see how a problem with CDA's burning that isn't shared by any other mainstream software can be described as a 'feature' of my setup, but I await with interest...
Cold wrote on 12/1/2002, 7:38 PM
I suffer from the same issue. 3% into a burn the system locks up and needs task manager to close it. I'm running a very different system, 1gig athlon, new yamaha crw f1 drive... This issue only happens about 50% of the time. no issue burning from sound forge. But it does make me look bad in front of clients.
Steve
Rednroll wrote on 12/1/2002, 11:48 PM
Geoff is too much into defending this worthless product that's been slapped together and rushed out the door, to understand any reasoning.

Let's keep slapping together new products and introducing a whole crap load of bugs, instead of investing time in the programs they already have and adding features and ironing out the bugs through continued support. This is definitely NOT the direction of a successful audio software company.
Geoff_Wood wrote on 12/2/2002, 1:57 AM
And red is so busy attacking everything that moves ...

Have just verified that this bug does *not* affect my system with an IDE drive running DMA, however this is XP.

My Win98SE system runs SCSI drives on a controller that is DMA, and that runs normally too, for what that's worth.

Now we need somebody with an IDE drive on 98SE running in DMA mode to verify that it can work . That must be about most users ? Maybe even somebody with the same model drive ?

geoff
PaulO wrote on 12/2/2002, 6:43 AM
>Geoff is too much into defending this worthless product that's been slapped
>together and rushed out the door, to understand any reasoning.

Guys, while I appreciate that emotions can run high about this kind of thing, I'd really appreciate it if discussion of Sonic Foundry's development philosophy could be carried out in such a way as to not hijack this thread away from its true purpose, which is to solve the obvious technical problem with the CD burning engine. Thanks.
PaulO wrote on 12/2/2002, 7:59 AM
>Now we need somebody with an IDE drive on 98SE running in DMA mode to verify
>that it can work

I'm sure that it *can* work! However, the fact that it *can* work does not prove the absence of a bug. The most pervasive fact in this whole business as stated by me and supported by others is that the burning engine on most of the standard software (Nero et al.) does not suffer from this problem, or anything like it.
Geoff_Wood wrote on 12/2/2002, 3:20 PM
Equally that there seems to be a problem on your and at least one other system (albeit not on Nero, et al) doesn't mean that there is a bug in CDA5. Just like CDA5 (and SF, Acid,and Vegas) calling up Sonic Foundry doesn't mean that they are 'spying' on us.

Hence my suggestion that somebody with a similar setup to yours without a problem report back.

You may be aware that there are *many* things that can get out of skew in a computer system, including left-over or replaced system files from other installed (or previously or subsequent installed) apps that can cause problems.

g.
Cold wrote on 12/2/2002, 4:08 PM
I agree with PaulO on this one, I'm more intrested in solutions. Perhaps there is a software conflict, my system is not a pure work station (yes, tsk tsk on me) so software comes in and out of my system all the time. I can't be the only one though,can I? Steve
PaulO wrote on 12/3/2002, 5:21 AM
I can add yet another program to the roster of software that seems to burn without problems on my system.

So far we have Nero, Easy CD Creator, CDRWin. Now we can add the almost-free Exact Audio Copy. I burned five test discs last night, DMA ON, without a glitch. CDA hardly gets through a few percent of a single disc before data transfer ceases and it locks-up.

What is going on here?
Rednroll wrote on 12/3/2002, 10:30 AM
Don't worry Paulo, we see your deductive reasoning and I would have to agree that there is some sort of bug happening in CDA. I'm sure Sonic Foundry will find it someday, but that's the whole point I've been trying to make. Stick with a program...get all the initial bugs worked out with the first release.....fix the bugs...therfore making a solid piece of software.....see what the users would like in the program....and make further feature improvements.....all the while maintaining a solid piece of core software. "Build it and THEY will come"......"But build it right and they will keep coming back". It's the later statement that SF is lacking in.

Currently I have the entire suite of SF products. ALL of them have Major bugs and simple oversights. Vegas is probably the one with the least bugs....kudos to Dave Hill. Just goto the Acid forum, and see all the users complaining about "when will the update come?". I started using the latest Sound Forge on the 6.0b release. I tried it for 2 hours and reported 8 bugs. Vegas Video has Track Insert meters that are non responsive and inaccurate...making them useless. Sound Forge has processes, that only processes data on 1 track of a stereo signal, when you have both selected. I use plugins, using the Preview feature, I hit process...and the process file sounds nothing like the preview. I open some plugins...hit process...and it totally erases ALL the data window. Sound Forge has a crash recovery option, so that it saves the data to a temporary location incase the program crashes....guess what?..it doesn't work. Sound Forge has an autoname extract from CD option....guess what again? it doesn't work properly.

So, is it better to release this new CDA5 so that now instead of having 3 bug infested programs...you have 4? I guess so....I would have just prefered they added a couple features to Vegas for CD burning and I would have been happy...Now..NOBODY is happy.
PaulO wrote on 12/3/2002, 1:07 PM
>Now..NOBODY is happy

Well, that's not strictly true. The new CDA seems to be stable on the whole, I quite like the new interface and the plug-in philosophy has been quite well thought out. However, the problems with reliable burning rather let it all down. It's rather like buying a car with lots of stunning new gadgets, but the engine isn't all that reliable. All CDA exists for, ultimately, is the burning of Red Book CDs! It could well be in SF's best interests to get CDA solid, because regardless of what you think of other products from the same factory, CDA is probably the best approach to creating Red Book CDs on the desktop, bar none. But if it acquires a reputation for unreliability...?

If CDA5 had the ability to produce ISO files or a WAV image + CUE file, I'd probably not be moaning right now. Personally, I'd have been quite happy if SF had simply provided a reliable, modern burning engine for CDA4, but life ain't like that.

Anyway, what do the SF support people who inhabit this village have to say about my problem?
SonyDennis wrote on 12/3/2002, 2:25 PM
PaulO:

I've asked someone to look into the burn issue you're having. They are working on getting our matching Plextor drive on a Win98 system. I (or they) will let you know what they find.

///d@
PaulO wrote on 12/3/2002, 5:38 PM
Thank you. Please make sure that the test is conducted with DMA on within Device Manager/settings for the drive under test.

Would you like to communicate via this forum or private email?

If the latter, what is your address?
Rednroll wrote on 12/4/2002, 9:18 AM
Paulo, One other thing to check is to go to the burners support page and download the latest firmware and install that. CD burning issues are always the hardest to solve when you're having them...too many variables. Is it the burner? Is it the software? Is it the firmware? Is it the OS? Is it that particular brand/color of CDR? Is it the burn speed? If you're in the SCSI domain like me, then is it the SCSI controler? Is it that the SCSI bios needs to be updated? Been there...done all these at one time or another.
PaulO wrote on 12/4/2002, 12:25 PM
The firmware is up-to-date; I'm fastidious about that sort of thing.

The main question would nonetheless apply: under a given set of conditions, why does all this other software (including one free program) burn successfully, but CDA fails. I'm afraid that all the empirical evidence collected so far points you-know-where.
Rednroll wrote on 12/5/2002, 9:25 AM
I'm with ya on this one, but you know not all software is written by the same programmers, so sometimes there may be hardware interface issues.

I had an issue with CDA4.0 in the past, where it would burn the disc, but it did so with a high amount of errors, and half my clients where complaining that they couldn't play back their CD's in their home players. Although when I checked it on my studio CD player, they played back fine. At first I thought they where using older players that wouldn't playback CDR's. Then I thought it might be the brand of CDR I was using. Then I thought, it might be the software, so I started to burn with Nero and CDRwin...I got less complaints, so I figured it might be CDA4. Then I updated my firmware of my Yamaha SCSI burner. Still got unsatifactory results. I actually started to burn 1 CD from my mastering PC, then transfer over to my other slower PC, which had a Ricoh 2X burner and the same SCSI card and created an image and burned it using that....This fixed the complaints I was having, I was in definite trouble when they asked me for 10 copies using this method. Finally, I went to my SCSI cards updates section, and found their was an updated Bios, where it said "this update will fix CD writing problems found with some Yamaha CDRW burners". So there just happened to be a conflict between my SCSI card, and that particular burner. Oh...the headaches I went through!!
Bill_Wood wrote on 12/5/2002, 12:28 PM
Rednroll is right about the many variables between hardware, firmware, software and so forth. I have been burning audio CD's since 1995. Ran into many hardware and firmware problems that caused problems with various authoring software packages.

Don't be so quick to blame ANY piece of software just because it coughs on your particular configuration. Have you thought about upgrading your OS? Win XP Pro is extremely stable and reliable. That might solve your problem right there! I have been using it for two years now and have very few problems. Up until yesterday I used SF 4.5h and CDA 4.0g for all my short run audio CD's, writing to a Plextor 12/10/32A burner on a Dell Precision Workstation 620. I have yet to turn out a coaster with that configuration!

Also remember that NO software package can be produced that will work perfectly on ALL combinations of OS, hardware and firmware. All we can hope for is that Sonic Foundry will address and fix the anomalies that always crop up after a product goes into wide release.

So hang in there. You did right to report the problem. Now let the good people at Sonic Foundry come up with a fix.

Bill
PaulO wrote on 12/5/2002, 6:39 PM
>Now let the good people at Sonic Foundry come up with a fix

I'm waiting to hear something...

The first step however is for SF to acknowledge a problem. I too have been burning CDs since the days when a coaster costed $25 so I'm not exactly green at this sort of thing - more gold perhaps ;-)

Meanwhile I can add yet another program to the collection of those that work just fine - Plextools, which now incorporates a burning utility. Yes, there are many variables involved but you must admit that six other programs which all burn without difficulty does rather suggest that statistically, CDA is at fault. My system shows no other signs of problematic burning. So we wait patiently...
Geoff_Wood wrote on 12/5/2002, 9:12 PM
I've also been burning since coasters cost $25, and I had an HP4060i (which possibly explains my current lack of spare cash). Actually one of my PCs has gone quite flaky in the burning dept since I installed the lastest PlexTools - dunno if that a coincidence or not...

g.
Estuardo wrote on 12/6/2002, 8:54 AM
And to think I thought I was the only one with the same problem PaulO posted - except that I am still waiting for CDA5.0 to arrive. I was using CDA4.0g (latest version) with W2K Professional SP3, 512MB RAM, 1.4GHz Thunderbird, VIA chipset and a Plextor 24/10/40A CDRW. I tested one or two wave images and they wrote fine (albeit slowly) and then, all of a sudden, it would not work - writing only the lead-in at which point it'd lock up. Can't remember if I had to use Task Manager to get out or use the old "on/off" switch. Oh yea, latest firmware update from Plextor.

I just looooove how smooth and reliable Nero is. Keep thinking how nice it'd be if CDA could/would write the same.

Wound up having to use the Colin Hill work-around to use CDA4.0g to work. And then only his Premaster Ultility v1.00 would work; the later CDPBurn v1.00 and v1.13beta both refused to work. An added plus was his work-around supported burn-proof.

Oh yea, another thing. On the couple occasions that I did get CDA4.0g to work, I got the dreaded buffer underrun error once or twice. It is hard for me to believe my PC could not keep up recording off a wave image.

Anyway, I'm anxious to test CDA5.0 and see how it works.

If I run into the same issues I'll probably just get one of the latest Lite-on CDRW's.

I'm not blaming SF, just expressing my frustrations. I'm sure it's something on my end. Did I tell you how much I like Nero for reliability writing CDR's?????

//Stu
Rednroll wrote on 12/6/2002, 2:43 PM
And don't forget to mention that Nero not only does it very reliable.....but does it on multiple burners at the same time.