Writing CDR: slower speed-> better quality???

how come a miscalculation or a skipped one results in a "skip" on an audio cd, but could result in a completely corrupt directory structure or inaccessible files on a data cd, if it's all data?

How do you know it is not a surface defect?

All CD players have some error correction involved. This is not to fix 'errors' on the CD itself, but rather 'read errors'- since a CD basically streams the data off the CD (it may also use a buffer if it is 'skip proof') but the idea is to keep reading and reading no matter what happens- so it might interpolate a bit or two here and there, since it isn't going to just stop until it figures out what that bit it sorta read was, or it isn't going to replay that part to 'fix' its read error. Some CD players are simply sloppier than others. Some use oversampling where they basically read the same data a few times and compare the results. When you consider how cheaply most CD players are made, it is a wonder they aren't even sloppier.

On the flip side, even with a fast CD drive, you might install software at only 6 or 8X (that 54X is a theoretical max)- and if it reads your data that much faster than music playback of 1X and had all sorts of errors, much of your software wouldn't even run. There would be nothing logical about an error such that a computer could ever figure out what it *should* be. Of course computers also use checksums and other error correction methods to detect problems, and it isn't always expected to perform "real time" like a stereo component.
 
Last edited:
. . .maybe the audio file gets converted to analog and back somewhere along the pathway?

that is a cool example. Keep us posted with the research! My brain is starting to hurt from all the new info

cool website, btw. . .cept you look like one of those 'tortured, intellectual, introspective artists'. You sure you're not from San Fran?

smile :D
 
Last edited:
I deleted the previous post because it turned out that the ripped file wasn't EXACTLY the same length- after editing the start and stop points, and then doing the analysis- they turn out to be bit for bit equal- all this burned at the maximum speed (8x):

screen capture

bit for bit equal-

There is STILL a ton of controversy about this at rec.alt.pro newsgroup !
 
filtersweep: I'm pretty sure you're right about EZ CD Creator not burning 100% of the Red Book Spec. I don't know if you have hardcore proof of this, but if you don't, I would be of the same opinion.

PopD: For your earlier question about skipping on an audio CD and corrupt data on a data CD. This substantiates the "Data is data is data" rule stated by filtersweep.

Maybe this analogy will help:

Think of your CD player (or CD ROM Drive) as one town, and your speakers (or computer monitor) as another town. In order to read your files, or listen to your music, the "data" has to get from the first town to the second town.

The way data is stored on an audio CD, the path between these towns is like an asphault highway. If a chunk of road is missing, whether it be a pothole or a one-mile stretch, the car (audio) might have a rough ride and bounce around a little (skip), but it still gets where it needs to go.

On a data CD: think of each file or directory crossing its own bridge. Each file has to cross its respective bridge over, lets say, molten lava. Now lets take a piece out of the "final_thesis_paper.doc" bridge. We don't know what piece - it may be a rivet in one of the support beams, causing the whole bridge to collapse and send the car plummeting into the hot lava. Or, we may only be drilling a hole in the surface, allowing the file to get by.

Bad analogy?

Anyways, data on a data CD is stored in a much more "delicate" (in a manner of speaking) way than on an audio CD. I can't think of a better way to explain it without giving you a computer science lesson (which I KNOW you don't want -- nor do I) :D

Hope that helped a little!
 
It is a pretty well known fact that Adaptec isn't redbook- and by this I mean the PQ codes, etc... in fact when I ripped my CD for the above test, the ripped file was slightly truncated or longer- I can't remember which. The data contained is identical to that of a redbook CD (of the audio files)... Adaptec doesn't allow complex indexing, etc... I'm sure if it allowed all the redbook options, they could no longer call it *EASY* CD Creator!
 
n3utrino said:
. . ..
Bad analogy?

hellll no. Y'know, I don't remember ever learning that stuff in my C++ programming sequence course:confused:

what exactly are redbook specifications? Who wrote them, and why are they not followed by any particular prog?
 
Pop D-

redbook is an ISO standard that defines audio CDs (CDDA)- as opposed to yellow (CD-ROM, CD-ROM XA, CD-ROM Bridge, Mixed Mode CD), orange (CD-R, CD-RW, Multi-session, Photo-CD), green (CD-I, CD-ROM Bridge, CD-I-Ready) , white (Modern Karaoke CD (not to be confused with earlier CD+G-like Karaoke-CDs) VCD 1.0, 1.1, 2.0), blue (CD EXTRA, Enhanced CD). Redbook goes way back to the 80s when Sony and Philips *invented* CDs-

Anyway, I'm sure I already told you more than you wanted to know ;)

If you really want your head to spin and you want to check out how the subcodes and TOC operate check technical mumbo jumbo

Basically redbook includes all the details to ensure CDs are playable on all CD players- the P Q codes, lead ins, lead outs, etc... all the stuff that is automatic with Nero or CD creator...

Red book is followed to the extent that an audio CD burned on your PC generally plays back on any player- but a glass mastering house won't want to touch it unless it was burned with a red book spec app, like CD architect, wavelab, etc.... it is not exact enough for direct mastering (the wave data is fine- but not all the subcoding).

And no, there actually is no analog conversion that takes place- that would be something handled by the audio card.

Finally- that pic was taken awhile ago... I'm much happier now, but I appreciate your concern ;) A minneapolis winter will do that to anyone!
 
oh- I should add that nero or easy CD creator are essentially crippleware- that they come bundled "free" with burners, and the general public doesn't need to edit subcodes, so companies can charge a premium for software that does allow the subcode editing- but at this point, I'm waaay off topic
 
PopD: Don't feel bad about not learning that in a programming course. I took a bunch of VERY low-level programming courses, like Computer Architecture. We learned a lot about the most basic levels of programming, data storage, HEX, binary, and machine language, stacks, etc... As for the analogy, I came up with that on my own! If it helped, then I guess I'm pretty proud of that!

Never take that course! It's hell on earth. However, it did give me a MUCH better understanding of what exactly is going on beneath all the Window Dressing.

Filtersweep: thanks for the extra information on Redbook!
 
Back
Top