Burn CD from DDP WL8 vs WL7....inconsistency

WaveLab
1

Please bear with me for a bit while I lay out some history with this issue and describe some testing that deals with a lot of variables.

As early as WL4, I believe, I had done quite a bit of testing of different CDRs for consistency and quality of REAL TIME playback (note that I am not talking about data retention, though that can sometimes be an issue). I had found that many CDRs playing back in real time resulted in jitter-like effects, which often included dramatic loss of soundstage width and apparent eq shifts. My original results concluded that only the MAM-A Gold CDRs were consistently sounding like files directly from the computer.

Before anyone jumps in and says I am crazy because it is all just 0s and 1s, realize that real time playback of a CD is dependent on all kinds of issues, not the least of which are reflectivity and error correction. When I brought this set of observations up to the head tech guy at MAM-A few years back, he told me that this was not the first time he had heard this and that several full time mastering engineers had told him more or less exactly the same thing.

If the 0s and 1s are still there, however, why would this matter? For one thing, how folks hear reference CDs matters a lot. For another, clients use CDRs as masters for duplicating operations (not replicating), and I am convinced that these issues matter there to some extent. For another, I was early on sometimes choosing mastering techniques that purposefully truncated and/or which made it hard for me to predict how the CD itself would sound…in other words, I could not really hear the end result out of the computer before burning a disc, meaning that the only real reference was for the disc itself.

As I got all my clocking issues worked out over the years and as I got better and better interfaces for my Wavelab computer, I eventually found that I could reliably tell what was going on and make direct comparisons with the output of the computer and a CDR (MAM-A Gold) playing and started at the same time, as long as the computer and reference CD player were being clocked by the same source. As I refined my dithering and mastering techniques a bit, I found more and more that this direct comparison was increasingly reliable, and that playing the CDR in the reference player was giving me very good info about the product I was making.

Fast forward to WL7 and DDP: It became clear to me quickly that burning a reference CDR from the DDP image was a better representation of what I was hearing out of the computer than burning from the Montage using a temp file. While subtle, this was consistent and reassuring.

With WL8, I have found this not to be true…BUT, there are a ton of variables involved. My version of WL7 was on my older, quad core XP machine with its built-in CD burner. WL 8 is on my Xeon 8 core with Win 7 at 64 bit, using its built-in CD burner…While both machines have Lynx AES16 interfaces, the newer machine has a PCIe card, and the older one has the PCI card.

With all that in mind, I did a ton of listening tests and determined that every disc I burned on the new machine from a DDP had an extra glassiness/high freq. component in real time playback as compared to the Montage (obviously with all the plug-ins…mostly UAD…and dither in place). This was absolutely consistent.

When I then copied that DDP back to the old XP machine and burned a CDR from it utilizing WL7, the glassiness/high frequency over emphasis was gone. This is also absolutely consistent. I have not A-Bed these new discs with project playback in real time on the new WL8 computer, but I will and I know the results will be clear.

(Although this will not affect the A-B test I am talking about above, other burns from DDP on the old computer are not now possible because WL7 won’t stay installed on it, as indicated in another thread I have started)

Because of the whole host of variables, I really can’t say why this is happening, but it does seem likely that there is either a difference in code between the two versions or some issue with the new burner.

I really don’t know, but would ask for all to put their thinking caps on. I have to find a a way to burn reference CDRs that are closer to what I am hearing out of the project.

I know that in years gone by I had done testing of burning at various speeds to find the most transparent settings, along with finding the best media. Maybe I will have to revisit all of that, but I don’t relish that prospect.

Thanks in advance for all thoughts on the matter.

2

I would contend it’s possibly (as you say) the burner. Can you tell us what your old and new burners are? I would try to get one of the old burners on Amazon and install it in an extra space in the new computer or an external enclosure. btw, how are you burning CD from DDP? (never done that). Are you creating the DDP, loading the DDP back into a montage, and burning from that montage? Or some other way?

I’ve experienced this in the past. Difference in CDR vs. commercial CD vs. montage playback w_dither or DDP playback, etc. I think CD and CDR can throw in a lot of variables.

3

In WL7, you burn a CD from the DDP from an item in, I believe, the Utilities menu. In WL8, the File/Export menu item gives you a choice to do this.

There are lots of variables, but I do find that replicated CDs (not duplicated CDRs) are pretty consistent, and that is what I am looking for…i.e., consistency in playback so that I know what I am really doing at the mastering stage. I had achieved this consistency for a few years before this experience.

With CDs going by the wayside for the big record companies, this stuff is all in flux, but artists will keep wanting CDs, so I think it is still relevant for all in the business.

Will list the two different burners in a future post…but I think there may be something else going on…

I am also considering looking for a different reference CDR blank media. I know Media Supply was advertising a new, non-gold “archival” DVDR that was supposed to have higher reflectivity than gold (silver maybe?)…Don’t know if they have a similar CDR for sale…In any case, it supposedly had the same shelf life as the MAM-A golds.

Anyone else with other thoughts, please chime in because I am dumfounded at the moment.

4

I have found out that burning at anything above 8 x gave me all kind of problems like the cd not working proprely in cheap cd readers.

5

Have you tried extracting both CDs and comparing files?
I use EAC (Exact Adio Copy) free on the internet.
1-Pop in CD
2-Save as cue/wav image (single wav file)
3-Repeat for CD 2
4-Load both wavs in Wavelab
5-Menu->Analysis->File comparator

I have edited a single sample and it detects the difference. Anything above that will definitely show up.

6

I haven’t found a CD yet that can’t be perfectly ripped to identical files, no matter how “bad” it sounds, so I’m pretty sure geezer would get perfectly identical files back from any of his CDs. But I have to agree with him that when played, those same CDs can sound different.

7

If the same bit sequence can sound different, then there is a problem in the playback chain. That it may be triggered by different CDs (by whatever arcane mechanism - influences of timing, power draw, etc) is irrelevant - it is a fault that such things can have such an effect.

8

I have not yet done all the comparisons within Wavelab (7 or 8) between the CDs burned from the same DDP in 8 on one computer and 7 on the other…I opted to use my time to actually get the Red Book master out the door first after getting my WL7 old computer back up and functional…while dealing with some video software editing issues.

I should be able to do some more of this tomorrow, but don’t necessarily expect to find a difference with the resultant, ripped files. Tests in the past found this to be true…but that is sort of the point I have been making from the beginning: Different media and different burn rates have always been known (by me) to produce different sounding CDs WHEN PLAYED IN REAL TIME, which is a major difference from reading the data from the discs and playing that data out of a computer.

The tests could help to show if there is actually something else going on, so I will try it as I get time. In the past, I had not noticed a difference across several computers and across several generations of Wavelab, but this was all pre-DDP and pre the rewrite for WL8, so I am just investigating all possibilities, and was quite surprised by this consistent difference with the discs burned in WL8.

It should also be noted that I did find a difference in the sound on the old WL7 computer between discs burned directly from the Montage with a temp file and discs burned from the DDP file. I found that to be a positive difference in the WL7 computer, but not positive in the WL8 computer…maybe Phillipe understands this, but I do not.

The new UAD card on the new computer, which was utilized heavily for this test project, muddies the waters a little further.

I will, however, do some more testing and report back in. Any other thoughts would be welcomed as I try to devine my new workflow. I really don’t want to have to go back to the old computer to burn every disc from DDP from now on.

9

I would just reiterate to the poster two above that the phenomenon of REAL TIME CD playback being grossly different using different CDR media is moderately well known by those who have done the testing, and is readily understandable when you consider that CD players were designed for certain reflectivities and certain pit sizes and configurations, and also have error correction mechanisms that respond to these variables…and that the pits must be scanned at constant speed and converted to real time audio.

This is simply not the same thing as counting ones and zeros to compile files when there is no need to maintain constant speed and constant conversion to analogue.

10

I agree. I would start from there and discard as much as possible before the subjective stuff.
Different media does sound different. CD player’s error recovery reacts different to different media.
Even within same product brand and number there are different qualities through batches.
I use Taiyo Yuden and use Plextools to test spindles. The ones below 1% C2s I use for masters… the ones above I use for refs and data.

11

I understand it fine; but if the audio bit stream can be recovered without error (if it can’t, the disk is faulty - that’s what the error correction is for, right?), then any difference in sound is a deficiency in the player. Timing and jitter issues are rendered null by proper buffering, and any interference in the D/A process from the activity of the player is just inadequate electronics design.

12

pwhodges-

I am in total agreement…with the caveat that it is important to find a way to burn useful, consistent CDR reference disks that actually give you reference playback. That is the whole goal of this exercise. Before this latest issue, I had consistency and good reference disks for a number of years.

On top of that, I do still need to run some tests to make sure that the disks being burned on my new computer are actually bit for bit (or close) to the file they are burned from. I suspect they are, but need to eliminate this from the possible sources for the alteration in the real time playback audio I am dealing with. One simple test will be to rip the disk from the new computer into the old and burn a new disk and see how it sounds on the reference player, I suppose…along, of course, with comparing ripped files on either or both of the computers.

At that point, as long as I find the ripped files to be more or less equal sounding, then I can work on finding different media for the new burner or replace the new burner and move on about my business.

That is the logical, scientific way around this problem, I think.

13

Thanks Jim. That seems to work fine with DDPs made in WL7 or WL8. For some reason it doesn’t work with DDPs that were made with the CubeTec DDP driver in WL6. If I try to burn one of the CubeTec DDPs in WL7 or WL8, I get the error message:
“An error has occurred. Information supplied by the driver: The command has failed. (tk_writecd_ddp) Error code: 0x000000FA”.
But the CubeTec DDPs will import fine to montage in WL7 and WL8. Don’t know why.

14

I skipped WL6, though have used it at a colleague’s facility. I was unaware that there was any real option for DDP inside WL6, so I guess you are referring to a parallel Audio Cube program.

Perhaps that is the issue. My friend loves the way WL6 sounds, but uses WL7 for making DDPs only. He has not used WL8 yet, but understands about the rewrite for the audio engine, and may start…Since my friend is deeply inside the software vending world, I assume he would have used WL6 with DDP if it existed within the program.

I am sure Phillipe can address the issue with trying to write a CD from a DDP exterior to the program, though I thought there was a pretty solid connection with Audio Cube and Wavelab…but WL7 and WL8 are rewrites for Mac compatability, so who knows what has changed in this regard, or what different standards may be in play.

At any rate, I just wanted to point you to the two different menu placements for this operation in WL7 and WL8. I was quite confused and disconcerted about this when I first upgraded to WL8, but happy when I found the item was still there but in a different menu.