How Good Can Vinyl Sound?, Anyone got some sample clips? Options

[2Bdecided]

If you meant "during the music", then I don't think anyone would challenge you. It's easy to find music that's consistently loud enough to drown out the background noise of vinyl...

...but there's also plenty of music with sufficiently quiet moments to let the vinyl background noise shine through

Cheers,
David.

[Mescalito]

I have a Linn Sondek LP12/Ittok/Troika/Lingo/Linto set up. Now that's nearing as good as vinyl gets. I now only use it for transferring LPs which have never been issued digitally.

The net result is that I have a lot of FLACs of needle drops. They actually sound pretty good. Of course, on anything with a big dynamic range, the vinyl artifacts (surface noise & clicks intrude). My Akurate DS is better, but not much. When I get back home, I'll upload a couple.

Chris

[Canar]

I await your uploads with eager anticipation.

[Arnold B. Kruege...]

I await your uploads with eager anticipation.

Hmm, nearly a week later and nothing uploaded. Or is it someplace else on HA?

[Mescalito]

I await your uploads with eager anticipation.

Hmm, nearly a week later and nothing uploaded. Or is it someplace else on HA?

Sorry, Mate,

I didn't make it home last weekend. Ended up having to go to Tunisia on business. Hopefully, I'll be able to upload something soon.

Regards,

Chris

[Cavaille]

I wonder if I also can upload something.

To entertain myself I did a sampler for myself a month ago. I´m not sure if excerpts from that sampler would fit here since I used fairly standard vinyl equipment. As a turntable I used an old Grundig PS 3500 (itself a rebadged Technics with Direct Drive), the cartridge seems to be the product that came with turntable, my boyfriend exchanged the needle some 6 years ago. I found this turntable in the cellar of my boyfriend and I adjusted all settings like Anti-Skating to the optimum. The manufacturing quality is very good. Anyway, I recorded several vinyls with the lowest speed possible and without an RIAA-amplifier. I recorded with an E-MU 0202 USB in 192 kHz in 32 Bit (I needed that amount of data because I did some heavy processing afterwards). Most of the things were done only in software. Recording in low speed changes little crackles on the vinyl into bigger clicks which can be identified more easily by de-crackling software.

Although I´m under the impression that these vinyl rips of mine do sound very good I hated all of it. I hate Vinyl. All those pops and clicks... and the noise... bah! Still, anybody interested?

This post has been edited by Cavaille: Jan 26 2010, 12:22

[RonaldDumsfeld]

^^ Yes please. especially if you have some well known stuff.

I'd like to compare your results with what I'm getting. Which are I'm sad to report are not as good as I hoped they would be and I haven't figured out why yet.

Be useful to compare the results from a painstaking refurb job vs a £1 Mp3 download as well.

Gave up on the clicks, pops, surface noise removal stuff quite quickly except in the case of a very few much loved old favourites. If I was happy playing them on vinyl and never bothered with the CD what's the point? All credit to them what's made those wonderful tolls not withstanding.

---

TOS 8 and evidence for claims is great. HA is top.

Although I just had an amusing thought.

You cannot logically prove a negative (it even admits as much in the TOS themselves). If one is not expecting to find any difference that must be at least as suggestable as spending £100 on a new power cord and then expecting the sound quality of your PA to improve. lol.

[Cavaille]

So, here they are.

Some of this music is mainstream, some is not. I encoded them with Nero AAC for better convenience. Every track is in 24/96, coming from the 32/192 source. Every track was recorded with the lowest speed the turntable offered, which means that in case of 45 records the recording took awfully long I recorded them with my E-MU 0202 USB interface (with ASIO) that was feeded directly from the turntable. Then I did a first de-clicking and de-crackling and I also de-noised most of the tracks very, very carefully. After that I did the speedup with Sound Forge (after testing of course if Sound Forge would introduce distortions, which it didn´t). By doing the speedup after the first processing steps I was able to move possible distortions from the de-noising process out of the main frequency area. Then I did a second de-clicking & de-crackling and after that I applied the RIAA-curve with iZotope Ozone 4. I did some EQing on every track to adjust the sound to the original (coming from CD) or to my liking.

I did all processing steps very carefully. That´s why I had to remove remaining clicks manually by hand. In some cases this took an awful lot of time. I´ll now list the state of the tracks. Some records were played back wet - how could I have done this? Some records are fairly old and worn out. But listen for yourself.

HIStory, Michael Jackson (rest his soul): very seldomly played, original release (1995), 33 1/3 LP
Im Nin' Alu, Ofra Haza: worn out, severe clicks, 45 single
Theresa, Jan Hammer: original release (1987), seldomly played, 33 1/3 LP
Little Lady, Aneka: completely worn out, severe clicks, severe distortions, played wet, 45 single
Miles Away, Madonna: brand new, never played, coloured vinyl, original release (2008) 33 1/3 LP
Gremlins, Jerry Goldsmith: original release (1984), often played, 33 1/3
Smaointe..., Enya: seldomly played, played wet, severe distortions, 45 single
The Empire strikes back, John Williams: often played, several distortions, original release (1980), 33 1/3
Viva La Felicita, De-Phazz: often played, used for DJing by my boyfriend, original release (2002), 33 1/3
Woman in Love, Barbra Streisand: often played, original release (1980), 33 1/3

Before I forget: vinyls coming from before 1987 are not mine. They were bought by my parents and I saved them from the garbage can.

Equipment used: Grundig PS 3500 (DirectDrive, adjustable speed, rebadged Technics, date of manufacturing ca. 1977 or 1978). Cartridge: the one that came with the turntable (!), the needle was replaced with a new one six years ago. A/D converter: E-MU 0202 USB. Software: Sound Forge 9.0, iZotope Ozone 4, iZotope RX Advanced, Algorithmix DeClicker.

P.S.: thank God the CD was developed.

This post has been edited by Cavaille: Jan 26 2010, 14:18

[2Bdecided]

With the caveat that I have a cold, and I don't have great high frequency hearing anyway (though not too bad!)...

Basically these sound great, but I think there's something very strange in the higher frequencies on some of them. I see part of the problem is a ~15kHz tone on some of the LPs - but there's also an 18kHz resonance or harmonic or something on the first sample, and other strange (though inaudible) things on the others e.g. strong ultrasonic tones on Gremlins, weak 17kHz on Enya, etc.

It could all be on the original recordings, but on the ones I can hear, it sounds weird to me. It makes my ears ring.

I suspect that you're getting far more information from these LPs than was ever intended! e.g. on the Enya one, you can clearly see it was mastered from a 44.1kHz digital recording, because there's nothing above 22kHz.


Apart from this(!), I think your transfers are stunningly good.


EDIT: Which Algorithmix declicker did you use? Or was it mostly iZotope RX Advanced?? (Which costs more than most turntables!)

Cheers,
David.

[Cavaille]

Basically these sound great, but I think there's something very strange in the higher frequencies on some of them. I see part of the problem is a ~15kHz tone on some of the LPs - but there's also an 18kHz resonance or harmonic or something on the first sample, and other strange (though inaudible) things on the others e.g. strong ultrasonic tones on Gremlins, weak 17kHz on Enya, etc.

It could all be on the original recordings, but on the ones I can hear, it sounds weird to me. It makes my ears ring.

I suspect that you're getting far more information from these LPs than was ever intended! e.g. on the Enya one, you can clearly see it was mastered from a 44.1kHz digital recording, because there's nothing above 22kHz.


Apart from this(!), I think your transfers are stunningly good.


EDIT: Which Algorithmix declicker did you use? Or was it mostly iZotope RX Advanced?? (Which costs more than most turntables!)

Cheers,
David.

Thank you very much. I know about these strange high tones. Most of these tone can be found on the originals also (especially with the track from Enya). And the mastering for the Enya track was only partially digital: the Uilleann pipe was recorded analogue. You can´t see that anymore but I saw it before I applied the RIAA frequency curve. But I have to admit that this could also be some distortion introduced by the cartridge. In case of the Ofra Haza track the high tone was so very strong that I used a notch filter to lower it. Quite successfully, I might add.

Strange that you can hear these tones at all. I tested the tracks with several playback equipment (they had to sound good on every one of them) and I couldn´t hear them, not even with my Sennheiser HD-600 - and I can hear up to 19 kHz. Your hearing must be fantastic, I envy you. Or, it could be that the compression of the AAC codec kicks in and disturbs the tones when transients are encoded.

I also have to say that the E-MU 0202 wasn´t completely free from tones itself. I still did not find the reason. I used the Algorithmix ScratchFree for the automated declicking & iZotope RX for manual declicking (on distorted s-consonants I used spectral editing because Algorithmix introduced some noise-like artifacts during declicking)

... and yes, it is an expensive hobby.

If someone is interested... here is the mastering technique I suspect that was used which could be observed very good before applying RIAA:

HIStory, Michael Jackson (rest his soul): digital mastering with 48 kHz
Im Nin' Alu, Ofra Haza: partially digital
Theresa, Jan Hammer: digital
Little Lady, Aneka: analogue
Miles Away, Madonna: seems to be partially digital (some synths reach up to 30 kHz)
Gremlins, Jerry Goldsmith: orchestra: digital. synths: analogue
Smaointe..., Enya: partially digital, Uilleann pipe reaches up to 33 kHz
The Empire strikes back, John Williams: analogue
Viva La Felicita, De-Phazz: digital
Woman in Love, Barbra Streisand: analogue

This post has been edited by Cavaille: Jan 26 2010, 14:43

[Tony T]

I recently converted my vinyl to Flac (wav) and would appreciate any comments on the 4 samples I uploaded:

http://www.hydrogenaudio.org/forums/index....c=78189&hl=

[Canar]

I've got some samples to create and publish here too, from my ghetto-fabulous ripping set up, but first I need to purchase a breadboard to set up a proper RC loading circuit for my cartridge. The distortion is quite pronounced.

Perhaps this is what is happening with your rips Cavaille? Have you adjusted for the proper RC load for your cartridge?

This post has been edited by Canar: Jan 26 2010, 15:15

[Cavaille]

I've got some samples to create and publish here too, from my ghetto-fabulous ripping set up, but first I need to purchase a breadboard to set up a proper RC loading circuit for my cartridge. The distortion is quite pronounced.

Perhaps this is what is happening with your rips Cavaille? Have you adjusted for the proper RC load for your cartridge?

RC load? Help me with that, what does it mean? I adjusted azimuth & weight and I leveled the turntable out but apart from that I did nothing.

[Canar]

Here is the list of specifications for my cartridge: http://store.shure.com/store/shure/en_US/p...tID.104993100#t Click the "Specifications" tab. Note the "Recommended Load" part. As you're clearly doing flat transfers, you'll need to put the appropriate load on the cartridge if you want the flattest frequency spectrum on the rips you make. There's a very good chance that the strange spectrum of your rips is a consequence of not loading the cartridge.

Here is how a proper loading circuit should look:


Credit: I shamelessly ripped that picture from here: http://www.vinylengine.com/phpBB2/viewtopic.php?t=6674 That site has more discussion about RC loading. In the diagram, ignore R1 and L1, those are simply models of the circuitry of the cart. R2 and C1 are your resistive and capacitive loads, respectively. IVm1 is just a volt meter, and if I understand correctly, that's where your output lines should be wired.

This post has been edited by Canar: Jan 26 2010, 16:20

[DonP]

Here is how a proper loading circuit should look:

The 47K is the standard load, at least for MM cartridges. C1 is to absorb any radio frequency signal that gets on the line and should not affect the audio frequency response.

[Canar]

Thanks for the info, DonP. I don't know this stuff particularly well, I'm afraid. Perhaps at the end of all of this I'll put together a guide.

[Cavaille]

Thank you for the explanation, Canar. But I´m afraid that I don´t have the means of adjusting the load. Furthermore I´m not able to measure the capacitance of the cable (it´s the soldered cable from the turntable). And I might add that I´m using the turntable only once per year. Vinyl never was my cup of tea and I don´t intend of buying stuff for it... well, maybe a new cartridge in the future. Instead, I´ll continue investing in newer software since I use that much much more often.

For me these strange spectrums are all right since I adjust it via software... well, I admit, some things were left.

[pdq]

Credit: I shamelessly ripped that picture from here: http://www.vinylengine.com/phpBB2/viewtopic.php?t=6674 That site has more discussion about RC loading. In the diagram, ignore R1 and L1, those are simply models of the circuitry of the cart. R2 and C1 are your resistive and capacitive loads, respectively. IVm1 is just a volt meter, and if I understand correctly, that's where your output lines should be wired.

You should also check the input circuitry of your sound card. It will certainly have an input resistor, and 47 k is a very common value to use. It will also probably have an input capacitor, but you would need to check its value.

[Fedot L]

Just because the bandwidth extends to 30kHz doesn't mean it's musical information

…of course CD4 quad LPs have an FM carrier signal with the demodulation info up around 34 kHz (or is it 35 kHz?).

The sound reproduction bandwidth of all the four CD-4 LPs’ “discrete quadra” channels is 30-15000 Hz.

CD-4 quadra LPs carrier frequency is 30 kHz. The (left front + left rear) “sum” and (right front + right rear) “sum” 30-15000 Hz signals are cut on the left and right sides of the groove. The (left front - left rear) “differential” and (right front - right rear) “differential” 30-15000 Hz signals modulate 30 kHz carrier frequency for the left and right channels, are ANRS-compressed in dynamic range, and are cut also on the left and right sides of the groove, simultaneously with the “sum” signals.

The highest frequency “cut” in the groove is 45 kHz. Being picked-up with a “quadra”-type cartridge (having 30-45000 Hz linear FR), and the (left front - left rear) “differential” and (right front - right rear) “differential” signals being demodulated and ANRS-expanded in dynamic range, all the four signals are converted by a decoder into four discrete quadra signals and played.

On 2-channel stereo equipment, the (left front + left rear) “sum” and (right front + right rear) “sum” 30-15000 Hz signals of a CD-4 record may be picked-up with “ordinary” high-class 2-channel stereo cartridges and played as a 2-channel stereo program. Or on mono equipment, picked-up with “ordinary” high-class mono cartridges and played as a mono program.

[Fedot L]

So, here they are.

I’d say, excellent!
Only “Viva La Felicita” seemed to me a little dull. 8 kHz 6 dB boost revived it in my perception.
And your efforts to achieve the final quality of the tracks were heroic. Only really precious records are worth them, I think.

[Cavaille]

So, here they are.

I’d say, excellent!
Only “Viva La Felicita” seemed to me a little dull. 8 kHz 6 dB boost revived it in my perception.
And your efforts to achieve the final quality of the tracks were heroic. Only really precious records are worth them, I think.

Thank you very much. You´re right about "Viva La Felicita". But I wouldn´t exaggerate the sound colouring I already did with most tracks. "Viva La Felicita" were in fact of such a high quality (despite being played often during DJing) that I only needed to remove 3 clicks by hand and by using too much EQ (which I tried) I almost lost the timing. This is the one track with almost no processing done. Other tracks were treated a bit more with stuff like EQ or stereo compressing/widening. The track where you can notice that I heavily altered its sound is "Little Lady": the bouncing synth line already was a bit out of sync, after processing even more. But I´m especially proud of "Im Nin' Alu" which now boasts a very sonor bassline, something that just wasn´t there that way before. It sounded so much like 80s... I had to do something about this.

You know, despite what people are saying... you actually can produce that "warm", colourful and full sound people often (wrongly) identify with Vinyl most of the time - just by using the proper EQ settings. But why should I only use only really precious records? I have to work with the things I got and I have to get the best out of them. And some music you can only find on worn-out vinyl. I even went so far as remastering the original sound. After I was finished with my processing on for example "The Empire strikes back" I realized that it wasn´t the vinyl that sounded awful, no it was the music itself. You can observe this very easily with the remastered CD re-release from 1997 - it also sounds awful. So I ended up adjusting the sound to my taste and to newer (digital) John Williams recordings.

[jamie_P84]

So, here they are.

Despite my having registered and logged in, these downloads do not work for me.
The filenames are listed but they are not links. Am I missing something here?

However, one thing I can say without playing them is that all the years you've listed for those records are post-1979.

The problem here is that between approx 1975 and 1979, the vast majority of record cutting plants installed digital "cutting delays" in-line between the analogue open-reel tape recorders and the cutting lathe. From what I've been told by an ex-engineer, these were originally 12-bit and 32KHz, and they added an audio delay to the tape output of several seconds. Later models (in the 80s) used higher bit depths and sample rates.
These devices were installed for work saving reasons; When cutting a record, the engineer has to manually press a button which generates the "shiny space" we see between tracks on the surface of LPs. The "cutting delay" was used to give him advanced warning of the end of each track (by monitoring the live output from the tape recorder), so that he knew when to press the button.
Prior to this, the engineer had to listen to the tape once-over in full and write all the timings down, before firing up the cutting equipment.

So basically, all of the records you've listed have been mastered from a digital source, negating most of the potential advantages of vinyl.

Before someone points it out, I have found that when looking at the output of my digitally-mastered records on a spectrum analyser, there is content above 16KHz (and even above 24KHz) when in theory there shouldn't be. However, I can only assume that this is some form of distortion from my playback equipment, because it's even present on records which have been cut from digital studio masters (where the CD version is marked DDD or ADD).

This post has been edited by jamie_P84: Feb 3 2010, 21:50

[Arnold B. Kruege...]

The problem here is that between approx 1975 and 1979, the vast majority of record cutting plants installed digital "cutting delays" in-line between the analogue open-reel tape recorders and the cutting lathe. From what I've been told by an ex-engineer, these were originally 12-bit and 32KHz, and they added an audio delay to the tape output of several seconds. Later models (in the 80s) used higher bit depths and sample rates.
These devices were installed for work saving reasons; When cutting a record, the engineer has to manually press a button which generates the "shiny space" we see between tracks on the surface of LPs. The "cutting delay" was used to give him advanced warning of the end of each track (by monitoring the live output from the tape recorder), so that he knew when to press the button.
Prior to this, the engineer had to listen to the tape once-over in full and write all the timings down, before firing up the cutting equipment.

So basically, all of the records you've listed have been mastered from a digital source, negating most of the potential advantages of vinyl.

The digital cutting delays that I'm familiar with were all 16/44 or better. We did an ABX of Ampex ADD1 and found that it was undetectiable by ear.

There is an ad for the Ampex ADD-1 digital delay for cutting rooms in the Oct 6. 1979 issue of Billboard.

The purpose of the delay was to facilitate automated setting of groove pitch. The delay provided a "look ahead" signal that was used to increase groove pitch for loud passages beofre the louder, wider pitch groove was cut. About a half second delay would suffice.

[Cavaille]

Despite my having registered and logged in, these downloads do not work for me.
The filenames are listed but they are not links. Am I missing something here?

However, one thing I can say without playing them is that all the years you've listed for those records are post-1979.

The problem here is that between approx 1975 and 1979, the vast majority of record cutting plants installed digital "cutting delays" in-line between the analogue open-reel tape recorders and the cutting lathe. From what I've been told by an ex-engineer, these were originally 12-bit and 32KHz, and they added an audio delay to the tape output of several seconds. Later models (in the 80s) used higher bit depths and sample rates.
...

So basically, all of the records you've listed have been mastered from a digital source, negating most of the potential advantages of vinyl.

Yes, it´s true. I´ve deleted them since I needed the space here for other files - but that didn´t work out. Still, I find your comment interesting. During mastering (and even on the finished tracks) I´ve observed frequencies over 16, 22 or 24 kHz. Let´s take the track "Music" for example by Madonna: one (presumably analogue) synthesizer used in the production produces "ghost frequencies" and therefore a very recognizable pattern when viewed with spectrum analysis. One part of this spectrum can be found at 17 kHz, the next part at 34 kHz (still visible) and another part at ca. 51 kHz (which was lost after downsampling from 192 kHz to 96). The same case with "The Empire strikes back" where some parts of the orchestra reached up to 35 kHz.

Especially the latter part could be because of distortions. There have to be some distortions, because percussion or transients with high frequency content always reach up to the end of the available spectrum and I can´t imagine that percussion reaches up that high all the time. For sure, all of this is very interesting but I wasn´t very fascinated by that. Since vinyl is an inferior source compared to CD (at least to me) I only wanted it to sound pristine without caring so much for frequencies beyond 20-25 kHz.

[cliveb]

The purpose of the delay was to facilitate automated setting of groove pitch. The delay provided a "look ahead" signal that was used to increase groove pitch for loud passages beofre the louder, wider pitch groove was cut. About a half second delay would suffice.

Before the days of digital delays, this was achieved by having an additional playback head in advance of the main one. I realise that using a digital delay is unlikely to add audible degradation, but on the basis that you should pass the signal through as little circuitry as possible, the older method strikes me as theoretically superior. So, do you have any idea why this was replaced by digital delays? I'm just curious to know the reason.