Personal View site logo
Weird audio cd problem - does anyone have any clues what can have gone wrong?
  • I had some CDs made (replicated from a glass master, not just copied in a burner), and they have a really odd distortion which sounds a bit like an old dusty LP in the first few tracks, but the distortion / noise gets more noticeable as you go higher up the track numbers, and the final track is really, really bad. Track 1 has a bit of noise, but almost unnoticeable, track 28 is absolutely horrible. It sounds like an LP when you have grunge on the stylus.

    This is an extract from track 27 (the CD has 28 tracks in total).

    As supplied: http://snd.sc/WWAvgz

    On the CDs I got back: http://snd.sc/OWtlJF

    I've never heard anything like this and never had problems before with this company so any ideas what can have happened? Anyone know about CD replication from a glass master? Before I get back to the company!!

  • 53 Replies sorted by
  • I'm getting real hard of hearing Mark. Maybe sounded a tad flat but I didn't notice anything striking.

    I sure do love the music though!

    Correction, I put my headphones on and there is a real issue there. Sounds like electronic interference to me. I think they have some equipment problems?

  • Could be - I'm wondering if something has gone amiss when they make the glass master, but it's not a process I know that much about because it happens in the factory. Then they press the CDs from that. It is really, really weird, and nothing I've come across before.

    Thank you for the comment about the music - it's designed to be quite "close" sounding as it's for use in big halls where children have their sessions. We did 28 tracks in all, and it's been a lot of fun - this is the second run. The first run of 1500 was absolutely fine, but that was a physically different master (same company though, I think anyway - although saying that, the text font burned into the circle in the middle of the CD is a bit different, so maybe it's a different pressing factory.

    I'm just trying to make sure they don't blame me for this. Really difficult when you have 1000 shrink-wrapped CDs sitting in front of you, all of them unusable!

  • I'd like to buy a copy if they're for sale!

  • @peternap That's really nice of you, but I can't sell these CDs!! However, and even better, you can hear all the tracks (undistorted) for free on our development site. It was made a while ago to share the emerging CD contents and uses flash, but provided you have a browser that supports it (not iPod/Pad/Phone in other words) it gives you all 28 tracks of the music and some uses it can be put to.

    That music was a lot of fun to do. We have another project for children with special needs (in the pipeline) and a further one for newborns, which is tied in with music in hospitals. So quite a lot of interesting stuff going on. They specifically wanted CDs although we are going to make them more download-friendly in future. Just a shame these CDs are no good!! Anyway, enjoy!

    http://musicmushrooms.co.uk/index.html

    My favourites are Funga Alafia, Baboushka (maybe @Vitaliy_Kiselev will like that one) and the Hokey Cokey (that's me right at the end).

  • Sounds like some crazy aliasing or jitter Mark - not heard anything quite that bad before - you'd think they'd have noticed!

  • @soundgh2 Yes - it does, doesn't it? The nearest thing I've heard to that is badly tracking DAT tapes or (if you remember those) Sony PCM tapes when they used VHS to record onto. It's odd that it gets worse in the higher track numbers (which are nearest the edge of the CD, as they play from the centre outwards). But the total CD time is 50 minutes, so it's not as if I was pushing the media too much by going for 74+ minutes and straying too near the edge of the disc.

  • Hehe I still have a Sony PCM F1 propping open a door in the house lol

  • In the days before absolutely everything was digital, I always thought it was a neat trick to use them to send digital audio from one place to another via a video link. Clever technology - but yes, they are a bit of a brick to carry round. I think that's what old technology is good at - door-props and paperweights!!

  • Got a ZX81 propping the wobbly corner of me Jupiter 4 up too - must have a clean out!

  • I don't have sound on my work computer or I'd listen to it. But, I have to ask, did you have a proof CD made from the glass master before going into production? Usually if the CD blank isn't centered when pressed, you can get something similar to what you are talking about. Again, I haven't heard your tracks but as you move out in tracks on the CD, this gets worse as the diameter of the track gets larger. Could also happen if the CD is slightly out of round but that rarely happens anymore.

  • Poor tracking (of the laser beam). If your CD player has a service mode (usually accessed by holding down a few buttons while powering up the unit), you should try to enter it and see if you can get it to display tracking errors.

    Google your CD player model + “service mode.”

  • Ok, so I listened to your tracks, definitely was pressed slightly off center. You can hear it as it shows up as a slightly cyclical sound. My car CD player does this sometimes because the rubber centering thing on the spindle is worn down. I don't think you'll have any problems with having them do this again.

  • All CD problems can be analyzed quickly and easily with an old plextor CD burner with hardware coded error checking and Plextools software. You can see the error level in hardware, and you can match up the original to the copy to see if they are the same. You can also use a certified burner (there are quite a few, I have used plextor, Liteon, Yamaha and a few others) with EAC, which is freeware. The systems are pretty self explanatory, but if you need more info on what the C1, C2 level errors are and what they mean, I can do a short article on error checking in digital audio.

  • Heh. I just got rid of an old plextor which would tell me it burned CDs without errors but the CDs would have horrible errors.. Not sure it's really a good way to check!

  • Thanks all for your great suggestions and conversation!

    I have sent a couple of pressed CDs back to the company after a conversation with them, so will wait to hear. They are anxious to put it right - I've done many projects with them (runs of 1000-2000) and I would use them again in the future despite this (depending on what happens next, though). From what you've all said and from an ex-colleague I caught up with yesterday, who has a lot of experience in mastering CDs and has written books on digital audio, my learning at the moment is:

    1) the problem should have been picked up by the factory
    2) @svart very good suggestion as it does sound like cyclical errors, and I absolutely should have had a test pressing. However in dealing with this company I've never had / needed one. I did actually ask about this at the time of supplying this CD master but was told they don't normally do that - obviously that would have delayed the production run initially but also would have saved me having 1000 useless CDs shipped all the way from France only to expire in my hallway. That squealing sound is really really odd and no-one has come up with an explanation for that.
    3) My audio colleague suggested I should have told the plant that I had CD-Text as they strip that off and put their own pq codes on the CD during the data transfer to glass master. Going via a UK company to a pressing plant in France, is also added potential for things going wrong. The artwork did have a checking process - I don't know where the printing was done but that was truly excellent quality and they provided soft proofing which picked up a pdf transparency error I was able to sort out. If only I'd had the same with the audio!
    4) @DrDave Good points about checking for errors, and last time I did something like this (for a DVD project) I did a thorough check of errors on my masters before sending. I since changed machines and don't have that software and can't remember what it was, and of course I should have done this if only so I could be absolutely sure it wasn't my fault. Today I loaded KProbe (I have a liteon drive) and burned and tested afterwards 10 discs using the CD list and media I used for mastering. Results were good apart from one in the batch which had C2 errors. As I don't have the actual pair of discs I sent, it's difficult to know if there were errors on one or both of those. Then again it should have been picked up as it sounds wrong, apart from anything more esoteric! Even the disc with C2 errors plays properly and doesn't exhibit any audible issues. That's a great offer on interpreting the results - but to save you the bother I found some good info here: http://tinyurl.com/987pw3y Interesting @svart that you have experienced errors which shouldn't be there according to software checking. Just shows how mysterious and alchemical the whole process is. Next time, before submitting a master I shall sacrifice a toad at midnight to Daton, the god of CD replication, and keep a lucky heather twig on top of my computer.
    5) @MrMoore I tried several discs on all the players I have (2 car, xbox 360, 1 portable, BD player, DVD player) every combination of disc / player exactly the same issue, so it's not a replay issue. However, it was a good suggestion because it's also what got me doing an error check on the drive / media I mastered on.

  • If you have the right Plextor drive (I have several but use the Plextor Professional) and the right software you can rip the CD and compare it with the master. It should absolutely match 100 percent--except for the initial offset--and anything else is just BS. It is just a big number, and you should have the same number , or numbers on each CD, and rip it 100 times and get the same result. I keep these drives for only one reason--problems with pressed or mass duplicated CDs. And there are lot of junk CDs pressed nowadays.

  • @DrDave Would that apply in my case, where my master included CDtext but the pressed CDs didn't? It's a really good suggestion - because at least it would show whether the master is accurately reflecting what I sent. I don't have a Plextor drive unfortunately but can EAC or other software do this for a non-plextor drive? Sounds like you have a huge amount of experience with this and that would tell me if something went wrong with the pressing.

  • Yes, absolutely. You rip the CD from the factory, and after the offset if should be identical to the master. The CD text--which BTW should be in the duplicated CD, DOH!--has no effect on the bitstream. Each drive has a different offset, or starting point, which is like a lead-in basically, and after that the CD should be an exact match. You can use EAC with a non-plextor drive for bit matching, or pick up a Plextor Premium on eBay if you want to know what the typical error rates are, Basically, higher error rates mean lower quality pressing. Most CD have some C1 errors, but the CD player will read then using Reed-Solomon system. Now, you probably don't want to get into the nitty-gritty of CD encoding, but I always match the dupe to the master, and I always check the error rate. EAC will do that handily. To see what your error rates should be on your particular drive, just test them against a selection of quality CDs.

    However, I do recommend picking up a Plextor Premium (52/32/52) Premium2 712 716 755 760

    That has hardware error checking. But EAC will tell you if your CD is really copied correctly.

  • @DrDave I understand about CIRC but I don't understand how to use EAC to check one CD (ie the pressed one) against the other (ie a master I supplied) - ie that the stream matches - and the forums don't enlighten me either. I understand what you're suggesting I try, just not how to do it in EAC. What exactly do I need to do in EAC?

  • Hi Mark-- It's complicated, but very easy once you get the hang of it. In its simplest form, you can do two basic tests, compare tracks and compare the whole CD. Go to actions, copy selected tracks uncompressed in one folder Then do the same with the other CD to *a different folder

    Then go to tools, wav compare, and compare the waves. Unless you have set the offset, the beginning and ends of the wave might be slightly different, but the middle--the sound--will be absolutely identical.

    To compare a whole CD, go to actions, select copy range, uncompressed, and rip the whole CD to one big wave file. Do the same for the other CD, then compare the waves. Again, except for the offset, they should be exactly the same.

    The offset varies from CD burner to burner, but it is more or less the lead in and lead out size.

    If you can manage a bit of German language you can also use the program CD Vergleich, a free download. If you don't have Plextools, EAC is the one that is the most accurate.

    However, it does not really tell you if you have a garbage CD, that is, one that is of low burn quality. Suppose you computer has a great "reader" and can recover the data. The CD can still be really low quality. And here, you can use CD Speed quality test, which comes with Nero or there is a free trial somewhere, but plextools is the way to go. Professional CD analysis for the price of a burner--a 712 model goes for $50-$60. There is also a Lite-on version, but I don't consider it accurate. K-probe can be way, way off. I just go with plextools. Pop the CD in, and it will tell you if it is readable, or has a lot of errors (all CDs have some errors, but they are correctable)

    As for Plextools showing a large number of errors, well, you have to know how to use it. A well pressed, Swiss made CD will have almost no c1 errors, a CD from a burner farm will have tons of C1s and and a few C2s. American pressed stuff has 500-1200 errors, and some of the burner farms have really amazing numbers of errors from burner absolute junk blanks at 52x on overheated burners that should have been replaced about a year ago.

    Link to "interpreting c1 errors" http://tinyurl.com/987pw3y I consider that information useless, sorry :). I just can't see using a burner that could be off by a factor of ten. I've tested it, and it is not reliable. I tested the burners and software against my 6 Plextors.

    So here is a basic question, should you buy your 1000 CDs from a company that uses a quality blank (TY disks) with quality glass master pressing, or from a company that will will sell you el cheapo? And since 99 percent of the companies use El Cheapo, how do you know? Well, ask for a sample, pop it in the Plextor, and Bingo, EOS. You know if it is good, period. And while the CD is in there, see who made the actual disk as well.

    Time is money, save time, save money. Lots of those CDs, people buy them, they don't work, and they toss them. You never hear about it, you just lose a customer.

    As for believing anything the the company says, well, that of course is ridiculous. "Oh yeah, dude, it is 1000 pieces of junk, but don't worry, we have pro tools" I work with several companies, and I KNOW they are giving me a good disk because I test them.

    I'm assuming you burned the CD DAO, if not, simply compare each track. If your CD text did not come through, it probably means they used a burner farm like Kunaki, or they just don't know what they are doing, or there is a problem with the ripper. I mean, leaving out the CD text? There is no excuse--it pops up when you rip it, it is part of the disc. Either way, you should get a copy of your master. That's why it is called duplicating, not "sort of mostly" duplicating.

  • Thanks so much @DrDave. I had high hopes that this would all be OK as this is essentially a re-release of a CD that we did an initial run of 1500 on - at least, same pressing plant but obviously a new glass master as they don't tend to keep them.

    Running the track I linked to in the original post, through the wav compare process, and comparing my source wav with the one from the pressing, reveals 18 missing samples in the middle of the track plus three instances of "different samples" in various time-ranges throughout the rest of that track. I can't see how to save the results unfortunately. Without having reliable drives to check the CDs, I can't say for certain where the errors have occurred but clearly they are audible in any case.

    I'm waiting to hear back from the company. It's silly that they didn't spot this, given it's a major pressing plant (I'd just rather not say who at the moment but happily will once this is sorted). But next time, I'll do it via a DDP file, so that will eliminate one potential problem.

  • Well, now you know. The dropped a few samples when they ripped it, or the master had a few dropouts. But if they had run it through THEIR error-checking machine--they usually have an expensive one--they would know if there was a problem, which is what they should do. If it is one or two samples, is it the end of the world? If you can't hear it, AND the error count is low, well, you can make a call there. But chances are the error count is high, maybe not, but that could account for the differences.

    Now let's say you do everything right, and you have an exact copy. Does that mean it will play on every CD player ever made? Nope. But low errors, quality disc, exact copy, that's the best you can do. You paid for a copy.

  • What I don't know is what my master CDR was like - because obviously they have that, and if they faithfully replicated that, then I'm sunk - unless as you say, their equipment should have picked up any errors in my master and rejected it. But you listened to the before / after samples., right? Something happened somewhere along the line between my master and the final pressings, it's majorly noticeable, hence I guess they didn't even listen to the thing!! It's so extraordinarily bad that I just can't fathom why they thought that could have been intended to sound like that!

  • @Mark_the_Harp If they give you the option of providing them with the files online as opposed to via CDR (either via DDP or even just track numbers indicated in the filenames) then you can eliminate the variable of how you burned the CDR. If you are burning your own CDR, all I can offer are the usual suggestions: use the most reliable brand you have experience with, burn at 1x and test before sending.

    As far as the issue between the two recording links you mentioned: yes, night and day difference, even on cheap laptop speakers. Sometimes a scratched CD sounds like that in my car's CD player.

  • For online files you want just one file, formatted as DAO, not TAO, and there can be issues there as well, in regard to error checking. Otherwise you will have problems. Do not be shocked that there was a mess up, I have seen this happen dozens of times. Half the people working in the business do not understand digital audio. The market is geared toward thin profits on cheap materials.

    Here's my work flow: Master in Pyramix or Sequoia or Samplitude. You should have a single, long file with embedded track markers, not separate track files.

    Burn disc (Taiyo Yuden) DAO (Disc at once) with Sequoia/Samplitude and check the verify disk option. The DAW will burn the disc direct from the timeline, and then automatically check it against the original bounce or mix. You need this feature, it is a pro feature. If your DAW does not have a bit matching feature, get the $50 version of Samplitude (sometimes $15) so you can use it for mastering. Then, using the same wav file, burn and check a second master. Do NOT generate a new wave. Why is this? Because the 16th bit, when dithered, will be different, and you will get two different files that sound exactly the same but have slightly different dither. If you are not up on dither, use Powr3 smart dither. If you care about such things, you can write your own dither code or use the one from Izotope--it is very good. But Powr3 is a no brainer.

    Next, rip each disc and compare. They should match, and you have two identical discs in all respects. They are exactly the same, same bits; same offsets. Next, run Plextools error checking and save the error rate as a png if you want a baseline. If you use the Plextor drive to burn that disk, you will get super low C1 rates. In fact, I once burned one with zero C1 errors. Who cares, C1 errors are normal. Mail one, keep one.

    When you get the product, check it against the master, which you saved, which is identical to the one you sent, or just compare it to the wave you ripped from the master when making two copies. Pop it in the Plextor and make sure the error rates are reasonable. End of story.

    Burning speed: 1x does not produce the lowest error rates. Dyes are no very sensitive to light to support higher speeds. Depending on the drive, you will get the lowest rates in the 4-12x region (I use 8x), but it also depends on the media. Taiyo Yuden media and some others give you good disks that will last for fifty years or more in storage. How do you know which is best? Why, burn a disk at each speed, then check the error rates for each speed. A lot of us keep around some of the older Plextors (made in Japan...) that have slightly better lasers, but really it doesn't matter, just get the errors fairly low.