I've just this moment setup my new SoundLab Turntable for use in digitizing old records. I am currently using the pre-bundled software Audigy for output via PC.

There is no reference in the manual or on the product as to what pressure is best for the stylus, I have currently set it at 1.5g.

This is a sample of the output I am getting, the glitches are horribly clear and ruin the playback (the song is "drony"):
http://host.trivialbeing.org/up/distortion.mp3

Can anyone help me in resolving this?
What are the likely causes considering everything is brand new (including the record)?

What samplerate are you using in the software?

EDIT: If the samplerate in the software differs from what is being output by the turntable you may experience "distortion" somewhat similar to what is on your recording.

Are you talking about the clicks (three major occurrences in the sample) or something else? Do you get the same thing with other LPs or are you just convinced the problem couldn’t be the LP itself and haven’t tried others?

Those clicks could be scratches or something stuck in the grooves. If the LP is not the source, then the most likely cause is an interruption in the flow of data to the recording program. There are many possible causes of this, but the most common one is the general computer system being allowed to do too many other things while recording.

For most programs, multi-tasking works fine, the programs each take their share of resources, in sequence, and the user may not even notice that the focus is momentarily elsewhere. When recording, however, the input data never stops coming, regardless of what may be going on within the computer. If another program uses too much of something critical, such as disk access time, some input audio bits just never make it. The recording program continues on as though nothing has happened, as soon as its turn comes up again, but there is a sudden discontinuity in the recorded audio file due to some of the audio signal never being received. The symptom of this is a largish click.

Anti-virus and other security programs, automatic program updates, mail and call checks, communication paging, system logging, and various other such background tasks are prime culprits for this sort of problem. Anyone serious about recording generally has to optimize his/her system so these kinds of tasks are not allowed to run while recording in being done. Sometime soundcard buffer settings can be increased to compensate for other activity, but only within fairly narrow limits.

Sometimes there are conflicts with software such as motherboard chipset drivers or with other hardware such as ethernet cards (especially wireless) and video cards that interrupted audio data flow even without the usual software problems. IRQ conflicts are not that common on modern systems, but they can occur and result in the same symptoms.

Thanks for your comments AndyH, I'll try running it when EVERYTHING else is shut down.

It has been happening on three separate LPs, all new.

Hi! Just a little off topic, what artist / song is on that sample?

I've taken a very close look at the waveform. There are 5 obvious glitches in your sample:

1). 0.820s: This doesn't look like a glitch caused by sampling errors. Although it doesn't look very much like a typical click due to LP damage, that could be due to artefacts caused by the MP3 encoding, and my feeling is that on balance it is a vinyl defect.
2). 4.078s: I am 99% certain this is due to LP damage - the waveform has a typical "roughened" shape that is seen on some types of "scuff" damage.
3). 4.976s: On first sight, this looks quite like a sampling dropout: a sudden discontinuity. But I also see a little bit of ringing both before and after, which is not what you'd expect. But perhaps this ringing is an artefact of the MP3 encoding of a genuine discontinuity.
4). 4.986s: Very much like the one at 4.976s - a discontinuity with pre and post ringing.
5). 7.705s: Rather similar to the one at 0.820s, and like that, on balance I think it's a vinyl defect.

In conclusion, I think it's possible that you have a mixture of typical vinyl clicks and some sampling dropouts. That said, those two glitches that look like sampling dropouts certainly *could* be vinyl defects. The fact that glitch number 3 is only on the left channel reduces the chance that it's a sampling dropout.


Just because an LP is new doesn't mean it's free of defects. The LP that contains no clicks at all doesn't exist.

Thank you for your analysis cliveb, it has been very helpful. In such a case, do you think it likely that anything is being severely damaged in the process?

I have a number of expensive records I'd like to digitize but I don't want them to go anywhere near it if it could badly damage them.

TuNk77 - the track is from Ben Frost's "Theory of Machines".

just my opinion but that is sample dropout, not lp damage. there was one low-pitch one at like 6 seconds that honestly, sounds like you bumped the turntable, or maybe set a mug down on the same table the record player is on... but who knows.

i think the biggest problem is the actual recording not having enough cpu or something and so you are losing data. its also the most likely considering the setup, new equipment and everything. i've had new lps with defects before, but very rarely, and you could see them, and they sounded like white noise, not like audio dropouts.

try recording into another program, audacity maybe, whatever you have. do shut down everything else when you rip, maybe even create a seperate partition just for this.

welcome to the world of ripping vinyl, i'm here to tell you that its a hell of a labor of love.

The on-computer ‘interferences’ I wrote about are the possible causes of such “sampling errors” as clibeb described. These are on-computer events, having nothing to do with the LP. In other words, these occurrences cannot cause damage to the phonograph record.

The USB turntable and its cartridge, on the other hand, being the mechanical devices in direct contact with the disk, could cause damage. If the stylus on the cartridge is defective, or if the cartridge is not aligned correctly, or if the tonearm is not adjusted correctly, playback definitely will not be optimum, and some permanent changes may be made within the grooves. These are most typically high frequency distortion, not clicks, pops, scuffs, and such. However, one stylus I had that pointed a little off to the side, rather than straight down, caused a lot of clicks, crackle, and scuff type noises in the channel it pointed away from, but there did not seem to be any permanent damage once the stylus was replaced.

There are quite a few sites with detailed instructions on cartridge alignment and tonearm adjustment. Certain test LPs, such as the HiFi News & Record Review Test Record, can be quite helpful in making the adjustments.

It is also likely enough that such a low end TT will do a little of this in-groove damage even when it is adjusted as well as possible, but probably not a noticeable amount of damage from just a few plays of an LP.

Playing the LP with dirt in the grooves or on the stylus are also damaging events.

Have you tried the turntable on another PC or laptop? Some USB chipsets have very poor drivers.

.
Duder,

I assume you’ve cleaned the LPs meticulously? What is the PC (CPU / RAM / OS / HDD / Free Space) you are using for this work?

Does the PC have any spyware / Ad-Ware / Virus / Trojans / Worms etc. running in the background, sucking up clock cycles? Run a check on the unit with this freeware:

http://www.safer-networking.org/en/mirrors/index.html Kill everything it finds…

Click Start / Run / and type in… msconfig then hit the OK button. A dialogue box will appear, click the ‘Start-up’ tab and note what’s starting every time the PC boots up.

If you are unsure about what each statement means, use to Google to search for whatever ‘.exe.’ file is referred to in the path statement shown as you scroll to the right.

Clean up your PC then clean up the start-up list, add hardware if necessary (go from 256mb to 512mb or perhaps 512mb to 1 gigabyte of RAM for instance), and defragment the hard disk drive – more than once if you’ve never done it before.

As alluded to earlier by several fellows; ripping data to a PC requires that the PC be pared down to an extremely efficient state of operation. I would suggest that you take some time to read the excellent suggestions offered by Benchmark Media with regards to honing your computer to be somewhat of an audio workstation / platform.

Windows XP
http://extra.benchmarkmedia.com/wiki/index...k_-_Setup_Guide

Windows Vista
http://extra.benchmarkmedia.com/wiki/index...k_-_Setup_Guide

Macintosh
http://extra.benchmarkmedia.com/wiki/index..._Audio_Playback

If all of this fails, um… buy an i-Tunes account :-)

Andrew D.
www.cdnav.com


.

LP damage and digital sampling errors can sometimes sound remarkably similar, but the waveforms are usually very obviously different. The waveforms in the example look more like some sort of LP damage to me. HOWEVER... since this sound clip has been through an MP3 encode-decode cycle, it's entirely possible that the normal sharply-defined discontinuities that are typical of dropouts have been "smeared" so they look more like LP damage. So a re-examination is in order.

Let's take the glitch at 0.820s. If you ignore the pre and post "ringing", and imagine that the major steps were originally sharp, then the offending section of waveform is 44 samples long. This is suspicious, since it's exactly 1 millisecond. It also looks like it could be a repeat of an earlier fragment of waveform, although I couldn't find any previous section of waveform in the clip that matches the suspected repeat. But the MP3 encoding might have distorted the waveform enough that finding the repeated fragment isn't possible.

Now let's consider the glitches at 4.976s and 4.986s. Again, ignore the ringing and imagine sharp discontinuities. It looks like this is a single glitch, starting at 4.976 and ending at 4.986. Again, it could be a repeated fragment of previous waveform, and this time it's 441 samples long: exactly 10 milliseconds.

So on further consideration, it looks to me as if these two glitches are digital sampling errors, and the "smearing" of the waveform disguised the fact. It would help if a clip from the the original WAV file was posted, so we could examine the waveform without the MP3 artefacts.

I stand by my view that the glitch at 4.078s is LP damage.

The glitch at 7.705s is 13 samples wide, which isn't any kind of neat factor of 44100. I think that's also LP damage.

As for whether playing your records on this USB turntable will damage them: Andy has already pointed out that sampling errors won't hurt anything - it's the physical aspect that you need to worry about. My opinion is that the majority of USB turntables seem cheap and nasty, and if I had any precious and valuable LPs I wouldn't let them near such a budget device. Provided the tracking force and geometrical alignment of the cartridge is correct, the damage will only build up slowly after repeated playings - basic wear, causing a loss and roughening of the high frequencies. But the other thing cheap turntables are not good at is tracking security. If you get any mistracking (noticable as a nasty distortion during loud sections), that has the potential to seriously damage the LP grooves in just one play.

Have you tried re-ripping the same track to see if the glitches are in the same place ? That would confirm if they are due to
dirt/damage on the vinyl.