Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: "ripping" LP's (Read 37082 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

"ripping" LP's

( I think this is the right forum, I'm trying, at least, to get this right. )

Like some other old folks here, I have a lot of LP's that I would like to move over to my computer.

There are several issues.
First, it has to happen in real time, no faster. This is quite profoundly annoying.
Second, I'd like to be able to get things like album art, etc, by entering the catalog number or something like that, or searching CD databases, perhaps, for the info, although often tracks are different, moved, or present/missing.
Third, I'd like to process the stuff on the fly, using the sound card's analog level adjustment to keep me in a decent dynamic range, but providing some space for overloads.
Fourth, I'd like to be able to run things like de-clicking, de-noising, etc.
Fifth, I'd like to stuff that all into .wma Q90

Does anybody make anything like this?

You may assume I own:
A good turntable
A good record cleaner
A good cartridge and stylus
A good phono preamp/general preamp

So I can get line level signals that are properly equalized into my sound card, in fact using balanced lines so that I can even get some noise rejection.
-----
J. D. (jj) Johnston

"ripping" LP's

Reply #1
Audiograbber mentions Line-In recording as a feature, i haven't tried it but Audiograbber itself is a pretty good program.
we was young an' full of beans

"ripping" LP's

Reply #2
It is very hard to run a rip faster than realtime. First of all, no commercial LP table will spin faster than 45rpm; 78 players are generally more expensive. Second, the RIAA curve gets messed up quite a bit when you play faster, and you'll need to manually compensate.

"ripping" LP's

Reply #3
There exist a number of sites with album art that can be downloaded. I looked at some a couple years ago but the only address I have handy is
http://www.recordlabels.smugmug.com/
The password was labels some time ago, but I haven't tried any access lately.

While it isn't organized for downloading into some database, you can find out quite a bit about many albums at  http://www.allmusic.com/  CD data bases exist, of course, and information about using them is no doubt readily available. Try reading in EAC or one of the other DAE programs.

I would say it isn't worth while to try to record to computer at faster than real time. Using a 78RPM turntable would make for problems with the equalization, at the very least. If you want to do a decent job of cleaning up the audio, that is going to take quite a bit more time than recording anyway.

If you are at all serious about the quality of the finished product, you need to start right. Recording should be in PCM 32 bit floating point to avoid problems with processing. Audition is the best bet, but if you don't want to spend money, Kristal and Audacity will let you get the file created in a reasonable format.

Various less expensive programs have declicking and noise reduction abilities but I only know the quality of the ones I use. There are a number of inexpensive programs that satisfy some people but from what I've seen they really don't approach what can be done with the right software.

I'm not sure what you mean about "process the stuff on the fly" but are you sure your soundcard has an analog level adjustment. Most don't, they just reduce the bit depth. It does depend on the phono preamp, and the cartridge, but for most setups there is no need to adjust the level; just record what comes out of the phono preamp. And you certainly don't want to be adjusting levels dynamically, as thought orchestrating a live recording. The result would be strange.

I don't think there is any reasonable way, which is to say quality oriented way, to do something additional while recording. Record, then process. When everything is as you want in the .wav file, declicked and clean, then you can consider encoding into some compressed format.

"ripping" LP's

Reply #4
Quote
I would say it isn't worth while to try to record to computer at faster than real time. Using a 78RPM turntable would make for problems with the equalization, at the very least. If you want to do a decent job of cleaning up the audio, that is going to take quite a bit more time than recording anyway.

Yes, yes, I know. I'm not considering it. I did say "it has to happen in real time, no faster" after all.

The biggest problem isn't even RIAA compensation, that can be reversed with some loss of dynamic range after the ADC.

The biggest problem is the response of the pre-amp and cartridge, and how heavy you'd have to track to keep the stylus down in the groove.
Quote
If you are at all serious about the quality of the finished product, you need to start right. Recording should be in PCM 32 bit floating point to avoid problems with processing.


Well, given that I can provide a clean line-level signal in I think that storing in 32 float is a bit of an overkill.  int 20 is more than the table and preamp can muster, and they are quite good, actually. (not very new, but good)

I would think that one could quickly go, after some simple gain ranging, to int16 without any real issues.

Processing, of course, has to happen in at least 32 int (8.16.8) if not float, of course.

I guess I'll have to think about automatic recording myself, then.
-----
J. D. (jj) Johnston

"ripping" LP's

Reply #5
It is true that a good 16 bit DAC will capture all that the LP can give you, but that isn't terribly relevant to using or not using a floating point file if you intend to do more than just record. Processing in integer produces rounding errors at every operation, every time the program makes calculations. Digital audio is all math. The errors are of course cumulative.

Also, assuming you have floating point at any particular time, converting to integer means either simple truncating or dithering. Neither is something you want to do multiple times; they are cumulative too. Therefore the reasonable thing is to capture in floating point (or integer, if you are so inclined, then do the extra step and convert to floating point). Maintain floating point until you have the final polish done and accepted, then convert to integer as your last step (or at least the last step as PCM). Of course no one forces this upon you, but you cannot beat the math with alternate choices.

"ripping" LP's

Reply #6
Stylus weights:  usually magnetics tracked at 1 gram.  That is pretty light.  You can check yours to be sure by doing a Google search.  The company or two which still stock stylii will have tables showing the correct tracking weight for your cartridge.  Use it.

Your amp will make no difference: you will be using the pre-amp out as input to your computer sound card.  Use RIAA.  Everything after about '60 was on that compensation curve.  Clean the records well (http://www.loc.gov/preserv/care/record.html).

Enjoy each and every LP as you play it.  You will probably never hear it that way again.

FWIW - years ago folks would copy their LP's to cassettes on the first play to preserve the LP's surface as pristine as possible.  Every play degrades it a bit more.

Happy Trails   
Nov schmoz kapop.

"ripping" LP's

Reply #7
It's not the end of the world if all you can record at is 16/22. If you're planning on encoding to WMA lossy it definitely doesn't matter. It is probably useful for forensic or archival purposes to record at 24/96 if you can, as vinyl is quite capable of going to 30khz (admittedly with a very strong headwind). Just be sure to upsample to 32-bit float when doing processing.

There are no particularly good ways to do automatic level control of vinyl rips, unless you have an external compressor handy. Just dial the music peaks in to -10dBFS or something and forget about it. Right now I'm recording at -20dbFS peak volume for most music, due to some clipping issues with the higher gain setting on my RME, and I don't ever anticipate it ever being an issue. (Of course, if you're recording at 16 bits then it could be an issue.)

"ripping" LP's

Reply #8
Quote
as vinyl is quite capable of going to 30khz (admittedly with a very strong headwind).
[a href="index.php?act=findpost&pid=327145"][{POST_SNAPBACK}][/a]



I would sure like to see a demonstration of this capability.  Have you one??   
Nov schmoz kapop.

"ripping" LP's

Reply #9
The Cardas Sweep Record has sweeps to 30Khz. Although I have not tried it, I would suspect that such sweeps have been experimentally verified to do something useful at 30khz.

"ripping" LP's

Reply #10
Quote
It is true that a good 16 bit DAC will capture all that the LP can give you, but that isn't terribly relevant to using or not using a floating point file if you intend to do more than just record. Processing in integer produces rounding errors at every operation, every time the program makes calculations. Digital audio is all math. The errors are of course cumulative.
[a href="index.php?act=findpost&pid=327124"][{POST_SNAPBACK}][/a]


Quote
Processing, of course, has to happen in at least 32 int (8.16.8) if not float, of course


Yes, yes, of course, that's elementary.
-----
J. D. (jj) Johnston

"ripping" LP's

Reply #11
Quote
Stylus weights:  usually magnetics tracked at 1 gram. [a href="index.php?act=findpost&pid=327137"][{POST_SNAPBACK}][/a]

Well personally, my cartridge tracks at 1.25 grams, including the little damper brush that Shure conveniently built into it.  Yes, I'm aware of the setup for my turntable.

I asked about software for the COMPUTER that would take the properly played, processed, and EQ'ed signal from my preamp. I know I don't need an amp. I know what RIAA EQ is, I'm OLD, man, I've built them, for 4 ohm moving magnet cartridges, with reactive input to keep down the Johnson noise.

I guess I'll just write my own. Sheesh.
-----
J. D. (jj) Johnston

"ripping" LP's

Reply #12
Quote
Quote
as vinyl is quite capable of going to 30khz (admittedly with a very strong headwind).
[a href="index.php?act=findpost&pid=327145"][{POST_SNAPBACK}][/a]



I would sure like to see a demonstration of this capability.  Have you one??   
[a href="index.php?act=findpost&pid=327167"][{POST_SNAPBACK}][/a]


SQ Quad on an LP goes to above 40kHz.

Of course, it does so using a carrier for a second pair of channels.

Does it work well? I don't know. Never played with Quad.
-----
J. D. (jj) Johnston

"ripping" LP's

Reply #13
You want a recommendation for recording software? There are many programs that will do the job of creating a file on your hard drive. I recommend Audition. It also contains tools to clean out the noise, transient and broadband, and do just about anything else you could want with LP recordings. There are also less expensive programs for recording, including some freeware like Kristal and Audacity.

Actually, I think you are getting so many answers that you don't care for because what you want isn't at all clear. Of course that might just be vis a vis me, but I suspect everyone is making up their own mind about an appropriate answer for a fundamental reason.

Quote
I would sure like to see a demonstration of this capability. Have you one??
I didn't keep it and I'm not in mind to set up and repeat right now, so you'll have to either accept or reject as you are inclined. I recorded the Cardas Sweep tones at 96kHz. It did indeed go up to the frequency claimed. My recording level declined somewhat at higher frequencies but I have no idea how much of that was due to my equipment and how much was simply because the LP recording itself had less energy at the extreme high end.

In addition, there were four very distinct higher harmonics, which were of course also swept from low to high. They were very distinctly visible to their upper limit, which for the 3rd, 4th, and 5th meant they went to the Nyquist limit of the 96kHz space, i.e. 48kHz.

"ripping" LP's

Reply #14
To keep thing truthful, I looked at my miscellaneous recordings folder and found a Cardas LP recording. In my own defense, I point out that I made the recording two and a half years ago.

I recorded only at 88.2kHz, not 96. This is where I do all my regular LP recording so I'm sure I just followed procedure on this. This means, of course that the upper frequency limit of the recording is 44.1kHz, not 48kHz.

Looking at a frequency analysis plot of the fundamental, or 1st harmonic, I see a sharp cutoff at 31,300Hz.

Zooming in to the section of interest, in spectral view, I see that there are only three upper harmonics that I would really call "very distinct." These all definitely reach the maximum frequency limit for the 88.2kHz sampling rate. There is one additional faint trace that I can also see reaching the upper frequency limit, making the four I previously mentioned, and two more that fade into invisibility before reaching it.

I suppose I could post a screen shot (is that supported?) if anyone is truly interested, but the thing itself isn't especially fascinating.

"ripping" LP's

Reply #15
Well, there have been lots of replies in this thread with varying opinions. As someone who's been transferring LPs to digital since 1994, I have a fair amount of experience and would like to make a few observations:

1. Bit depth of recordings. In principle, using long word lengths makes sense, primarily to guard against rounding errors during processing operations. But there comes a point where increasing it beyond a certain depth is completely pointless. A *really* good LP, played on the very best turntable available, might achieve a dynamic range of 70dB with a following wind. That equates to a bit depth of less than 12 bits. 16 bit recording is more than adequate. You'd have to do enormous amounts of processing to accumulate enough rounding errors to get anywhere close to compromising the noise floor. (This assumes, of course, that the processing itself doesn't use 16 bit integer arithmetic - as far as I know there are *no* packages out there stupid enough to do that). And if you do so much processing to reach that threshold, you'll have destroyed the music signal in far more damaging ways than adding a little bit of quantisation noise.

2. Sampling frequency. While it is true that a lot of LPs have some kind of signal above 20kHz, in most cases it bears virtually no relation to the programme material - ie. it's noise, and isn't worth recording. 44.1kHz sampling is fine for recording LPs. Someone mentioned that SQ quad records go up to 40kHz. This isn't true: SQ (and QS) quad is based on phase manipulations. I think you're getting it mixed up with CD4 quad, which does indeed use a carrier signal for the rear channels. I can't recall the exact frequency, but I think the carrier is around 35kHz. But CD4 records have to be played with special cartridges using Shibata stylii, otherwise the carrier is destroyed.

3. The most important aspect of transferring LPs to digital is to get the best possible analogue replay in the first place. If you record to hard disk in a slap-dash fashion, you won't be able to recover things in software. The turntable/arm/cartridge must be set up correctly, and the record should be cleaned properly. Ideally use a vacuum cleaner such as a Nitty-Gritty, VPI or Moth. A Keith Monks machine is even better if you have access to one.

I've written up a quite lengthy page of notes that might be helpful: Here's a link.

"ripping" LP's

Reply #16
Does it make sense to do a normalise on the recording? since the volume on vinyl can vary a lot from record to record, you can't always get it perfectly set at the source..

I'm just curious.. and I'm aware normalisation isn't a popular word here

"ripping" LP's

Reply #17
Quote
Does it make sense to do a normalise on the recording? since the volume on vinyl can vary a lot from record to record, you can't always get it perfectly set at the source..

I'm just curious.. and I'm aware normalisation isn't a popular word here
[a href="index.php?act=findpost&pid=327223"][{POST_SNAPBACK}][/a]


The question is the same as always: "what would it gain?"

"ripping" LP's

Reply #18
One general reason for normalizing is to make optimum use of whatever bit depth you have. Rasing the signal up near maximum means that the lower order bits are more above the equipment noise floors, so yes, normalizing is always a good thing, never a bad thing.

I have no idea why "normalisation isn't a popular word here" but unless it has something to do with processing in some lossy format, any unpopularity is based on a misunderstanding. Normalizing is simply amplifying every bit by some calculated amount that brings the peaks to any particular desired value. Normalizing in the digital realm is equivalent to turning up the volume control in the analogue realm -- a volume control that is absolutely linear and absolutely distortion and noise free.

"ripping" LP's

Reply #19
Quote
One general reason for normalizing is to make optimum use of whatever bit depth you have. Rasing the signal up near maximum means that the lower order bits are more above the equipment noise floors, so yes, normalizing is always a good thing, never a bad thing.[a href="index.php?act=findpost&pid=327238"][{POST_SNAPBACK}][/a]

The bit depth used depends on the original recording level. Anything under the quantisation noise remains so after normalisation - you amplify the noise floor along with the signal. This is why it's fairly important to achieve decent recording levels (especially when recording at 16 bit resolution - if you record at 24 bit, *and* your soundcard has a sufficiently low noise floor and linear performance in the lower order bits, then you can get away with being a bit sloppier).

Quote
I have no idea why "normalisation isn't a popular word here" but unless it has something to do with processing in some lossy format, any unpopularity is based on a misunderstanding. Normalizing is simply amplifying every bit by some calculated amount that brings the peaks to any particular desired value. Normalizing in the digital realm is equivalent to turning up the volume control in the analogue realm -- a volume control that is absolutely linear and absolutely distortion and noise free.
[a href="index.php?act=findpost&pid=327238"][{POST_SNAPBACK}][/a]

Normalisation does introduce some rounding errors, so it's not perfectly linear. But the errors are so insignificant there is nothing to worry about. I personally don't believe these errors are audible at all.

As a matter of course I normalise my LP recordings just prior to burning them to CD (ie. as the last step after all other processing has been completed), simply to avoid having CDs that are even quieter than they need be.

"ripping" LP's

Reply #20
Of course the recording's noise is amplified by normalizing, it is all of one piece, but I did specify ‘the equipment noise floor." You move the audio further above the noise of any playback system.

It's true that rounding errors always occur. That's part of the beauty of 32 bit float (or 64 bit in some programs) -- the rounding errors are really far down where they can never effect the music. As a demo I amplified a 55 minute LP recording by 6dB, then I amplified it by -6dB. The peak differences between the result and the original came out at -150dB and the average differences at -190dB. I could have had the same results by amplifying it 100dB, then -100dB. Anyway, that's so far down that it doesn't exist in human terms or even hardware terms.

With 16 bit files the effects are going to be greater, but not relatively greater than doing decrackling or NR or other clean-up operations, I should think. With 32 bit these errors just can't ever be a consideration, they occur below the theoretical 24 bits of the hardware. I concur that the artifacts of rounding are unlikely to be heard in 16 bit files of LP recordings, but if one has a choice, why not go for the sure thing?

I also agree that normalizing is best done as the last step. A normalized file can easily go above 0dBFS through other following operations. That doesn't present any serious problem when working in 32 bit float but it means permanent clipping in 16 bit files; in floating point files you will have created the necessity of re-normalizing to get back down to the proper level.

One thing that might give normalizing a bad name is insisting on 0dB. It is best to normalized to -0.2 or -0.3dB to avoid potential clipping problems in the DAC, even though the kind of signals that can do that are not supposed to be very common in music. I've read claims that some CD players have such a problem. Also, my slow memory retrieval finally reminds me that many Creative cards have some kind of built in limiter that really squashes things unpleasantly at the upper signal level limits. I think that starts somewhere around -3dB, so a normalized file could sound significantly off if one is blessed with one of those cards.

It must be admitted that the following is not likely to be very significant in music because there isn't much music with content above the 44.1kHz sampling rate Nyquist limit, but it is real and quite obvious in the several soundcards I've been able to test. Feed a sweep signal that goes well into higher frequencies into the analogue input and record at 44.1kz sample rate. Look at the recording in spectral mode and the aliasing image will be very visible. It is quite strong for the first few thousand hertz back down from the limit and can be seen to go essentially (but very faintly) back down to the noise floor if the input continues to sweep to a high enough frequency. This is true in spite of the ADC's anti-aliasing filters and 64X over sampling.

By recording at a high sampling rate, such as the 88.2kHz I use, then software converting to 44.1, the strong part of that image is up near 44kHz (where, to start with, it is certain to be much weaker than if recording at 44.1) and quite missing from the 22kHz final result. Once again, I agree, with real music instead of test signals, it isn't very likely to really matter.

With LPs as the source there is often significant signal above 22kHz . That may not contribute anything to one's listening enjoyment if, for instance, you record and playback at 24/96 (it is most likely only harmonic distortion anyway), but it does present more HF signal, and that signal definitely will be reflected back into the 44.1kHz space. I guess one could say that the reasons for recording at a higher sampling rate are comparable to the reasons for working with 32 bit float. If one has the choice, why not go for the better option, even if you are not sure you will ever hear the difference? There isn't much down side.

This aliasing test really shows the reasons for recording at 48kHz with SoundBlaster cards, then converting to 44.1kHz in software. SoundBlasters are messy either way but the 44.1kHz recording really shows off the effects of its poor internal resampling (remember, resample in 32 bit float  )

"ripping" LP's

Reply #21
And while cliveb rightly won't mention it. as a satisfied user I'd like to mention that his Shareware program WaveRepair is an excellent software choice for all aspects of the process, having been designed from the start for this purpose.

"ripping" LP's

Reply #22
One criticism of the "LPs only have 70db of dynamic range" argument is that humans can often detect signals well under the noise floor, sometimes 20 or 30db down. 70db is a peak measurement. So it would not surprise me if one could actually hear quantization noise from a normalized recording of vinyl surface noise at 16/48. There are some other cases that could get even more sensitive, that are not especially pathological - ie, a 3khz signal, perhaps from singing, could definitely be audible 30-40db down from the noise floor.

In the most pathological of cases, dedicated instrumentation hardware can recover certain signals 160db down from noise floors, but that's not really pertinent.

EDIT: I just realized that if the quantization noise is of similar spectral content as the surface noise, it may not be audible even at -10db.

"ripping" LP's

Reply #23
Quote
Well, there have been lots of replies in this thread with varying opinions. As someone who's been transferring LPs to digital since 1994, I have a fair amount of experience and would like to make a few observations:

1.  But there comes a point where increasing it beyond a certain depth is completely pointless.

My thoughts exactly. I have, after all, played a "sine wave" record into a digital analyzer.
Quote
2. Sampling frequency. While it is true that a lot of LPs have some kind of signal above 20kHz, in most cases it bears virtually no relation to the programme material - ie. it's noise, and isn't worth recording.

Again, my thoughts exactly.
Quote
I think you're getting it mixed up with CD4 quad, which does indeed use a carrier signal for the rear channels. I can't recall the exact frequency, but I think the carrier is around 35kHz. But CD4 records have to be played with special cartridges using Shibata stylii, otherwise the carrier is destroyed.

Yep, you're right, that's what I was thinking of. It's been a while, sorry.
Quote
3. The most important aspect of transferring LPs to digital is to get the best possible analogue replay in the first place.

Of course. I use a bit of a different, home-cobbled setup, but the only meaningful difference is the lack of a label on the front.

This is what makes the whole thing such a bleeping bother, too, of course.
Quote
I've written up a quite lengthy page of notes that might be helpful: Here's a link.
[a href="index.php?act=findpost&pid=327218"][{POST_SNAPBACK}][/a]


Why, thank you. I'll go read that.

What I asked about here was software to at least somewhat automate the issue, say something that can recognize impulses every 33.3/60 th of a second.
-----
J. D. (jj) Johnston

"ripping" LP's

Reply #24
Quote
One criticism of the "LPs only have 70db of dynamic range" argument is that humans can often detect signals well under the noise floor, sometimes 20 or 30db down. 70db is a peak measurement.
[a href="index.php?act=findpost&pid=327398"][{POST_SNAPBACK}][/a]


Well, 70dB is the best you get from an LP.

Even if we ignore rumble, I suspect, eh?

Now consider, in noise one critical bandwidth wide, we can detect a tone down about 5.5 or 6 dB in energy from that noise, as long as both are above absolute threshold.

On the other hand, this is just as true of digital recordings as analog recordings, so I don't really understand why you mention this. The noise from the LP will dominate the noise from 16 bit uniform, TPD quantization at all frequencies, I suspect. I even suspect most pre-amps won't do any better than that.

Now, your statement about noise floor is true, but it applies to both analog and digital noise floors equally, unless somebody forgot to dither (oops), at least.  Say for 24kHz bandwidth (48khz sampling rate), the noise in a critical band at 1kHz will be at about -22dB re: the total noise, for anything with a white noise floor.

This puts the total "hear down into the noise" at about -28dB, give or take, relative to the wideband noise level.  This is all well and good and all that, but it really doesn't matter because it works the same for analog and digital (except that the analog noise floor won't be quite flat, even if we ignore rumble), so you might get another 10dB out of that, making the 70dB into 80dB.  that's still 16dB of headroom/tailroom, isn't it, now?
-----
J. D. (jj) Johnston