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: Couple questions regarding the spectrals of a hi-res vinyl rip (Read 7599 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Couple questions regarding the spectrals of a hi-res vinyl rip

Hello everyone,

I have a couple of questions regarding the spectrals of two vinyl rips coming from the same source. The audio setup is: TT-integrated amp-USB soundcard (Creative SoundBlaster).

Here are 3 spectrograms (taken with Audacity). #1-2 from the first rip, #3 from the second rip.

Spectro 1: http://i.imgur.com/iGy11CS.png
Spectro 2: http://i.imgur.com/2C9oDe2.png
Spectro 3: http://i.imgur.com/As1VTEN.png

1/ My first question is regarding the external noise clearly visible at 37-38 kHz. It's not on the record because it's there on all rips. Does it look familiar to someone? I was told that infrared receivers use such a frequency, between 36-40 kHz. But I'm scratching my head at how that could enter the recording process.

2/ The cartridge/stylus is an Ortofon 2M Red. I checked the specs and the frequency range is 20-22.000 Hz. So my second question is how can there be so much audio information well above 22 kHz on the spectrograms? I would have thought that the recorded frequencies would end somewhere around 22 kHz, maybe 24 kHz but not much more. Am I wrong or missing something?

If someone can shed some light on either one of my question I'd be very grateful.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #1
Unless you actually filter it out, the noise from the stylus will extend to very high frequencies, although at very low amplitude.

 

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #2
#1 That HF noise can come from various sources as like from your display monitor or IR source sa you mentioned. I had this same type issue once and got it fixed by moving all TT related devices far enough from monitor and its cabling. Anyway, I would just use LP filter to remove those.

BTW, are you recording 24/96 format audio data using standard version of Audacity in Windows PC? I have not checked the latest updates for Audacity but it lacked the support for 24-bit recording (unless you used the 'ASIO version'). Worth to check this.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #3
I doubt infrared remotes could cause the interference, because they don't operate continuously. It is more likely a switchmode power supply somewhere nearby. Perhaps an LED lamp.

The high frequencies that seem to be audio-related are likely caused by distortion. The imperfections of vinyl playback cause harmonic distortion that manifests itself as high frequency hash at low levels.

You can't easily tell the distortion from the wanted signal in such a spectrogram when you don't have the undistorted signal as a reference.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #4
The problem of those vinyl rips is primarily that the content above 20-22 kHz is meaningless and irreproductible on common equipment. Actually it would have been better to capture at 24/44.1 or 24/48 kHz, thus filtering those frequencies out, then also the analysis of what is going on at higher frequencies becomes obsolete.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #5
Unless you actually filter it out, the noise from the stylus will extend to very high frequencies, although at very low amplitude.
I understand that you can have high frequencies on the record but what I don't understand is how you could capture frequencies that are above what your cartridge can capture. If you cartridge is limited to 22 kHz to me it seems impossible that you'd be able to capture much above that limit.

The only point of comparison that I have is my own cartridge, which has a frequency range of 10-25.000 Hz (http://audio-heritage.jp/TECHNICS/etc/epc-205c-iil.html). If I look at the spectrogram, audio is captured up to 27-28 kHz, after that nothing at all (except noise shaping at the top frequencies): http://i.imgur.com/J0Xtalf.png And as far as I know I don't apply any filter while recording.

The 3 spectrograms that I used in my original posts (and which are not from my rips) look like they've been "streched" upward if you know what I mean.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #6
BTW, are you recording 24/96 format audio data using standard version of Audacity in Windows PC? I have not checked the latest updates for Audacity but it lacked the support for 24-bit recording (unless you used the 'ASIO version'). Worth to check this.
I think that was true for the older versions of Audacity. The latests record in 32-bit float by default I think. Personnally I use Sound Forge Pro 11 to record and edit my rips.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #7
I understand that you can have high frequencies on the record but what I don't understand is how you could capture frequencies that are above what your cartridge can capture. If you cartridge is limited to 22 kHz to me it seems impossible that you'd be able to capture much above that limit.
Just because a cartridge's spec defines an upper frequency limit, that doesn't mean it won't generate anything above that frequency. An upper limit of 22kHz would probably be at -3dB. Beyond that, it will generate frequencies at a gradually falling levels (eg. 25kHz at -6dB, 30kHz at -12dB, etc).

Also, what you're recording isn't just the signal that comes off the cartridge. You're also recording the intrinsic background noise of the phono preamp - and phono preamps are pretty noisy devices (simply due to the function they have to perform - amplifying extremely low voltages).

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #8
Quote
Unless you actually filter it out, the noise from the stylus will extend to very high frequencies, although at very low amplitude.
I understand that you can have high frequencies on the record but what I don't understand is how you could capture frequencies that are above what your cartridge can capture. If you cartridge is limited to 22 kHz to me it seems impossible that you'd be able to capture much above that limit.

I think the spec sheet is pretty clear:

"Frequency range at - 3dB  -  20-22.000 Hz"

http://www.ortofon.com/ortofon-2m-red-p-317

So you can produce sounds up to 22kHz with attenuation of less than 3dB.  Above 22Khz, attenuation will be more than 3dB.  This is consistent with what you are seeing. 

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #9
I understand that you can have high frequencies on the record but what I don't understand is how you could capture frequencies that are above what your cartridge can capture. If you cartridge is limited to 22 kHz to me it seems impossible that you'd be able to capture much above that limit.
No, you seem to have the wrong idea about the technology. The frequency range of the cartridge only tells you which range of frequencies it is able to play back at roughly the correct amplitude. This doesn't mean that beyond that the levels drop off sharply. It is not even sure that they drop off linearly. If you want to know you need a measurement diagram.

The preamp will also have its own frequency response. And the cartridge may interact with the preamp in a nontrivial way, making predictions even more difficult.

In such a situation you shouldn't try to infer too much from technical data you picked up from somewhere. You are better off measuring and/or troubleshooting your setup.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #10
I understand that you can have high frequencies on the record but what I don't understand is how you could capture frequencies that are above what your cartridge can capture. If you cartridge is limited to 22 kHz to me it seems impossible that you'd be able to capture much above that limit.
No, you seem to have the wrong idea about the technology. The frequency range of the cartridge only tells you which range of frequencies it is able to play back at roughly the correct amplitude. This doesn't mean that beyond that the levels drop off sharply. It is not even sure that they drop off linearly. If you want to know you need a measurement diagram.

The preamp will also have its own frequency response. And the cartridge may interact with the preamp in a nontrivial way, making predictions even more difficult.

In such a situation you shouldn't try to infer too much from technical data you picked up from somewhere. You are better off measuring and/or troubleshooting your setup.
I obviously had the wrong idea, thanks for correcting me. To me the spectros didn't show much attenuation, the color spikes stay magenta-red up to 48 kHz and that left me puzzled. Of course I don't know how the cartridge-amp interacts or even how the soundcard behaves. There are too many unknowns.

What surprises me is how sharply my recordings drop off, like I said after roughly 28 kHz the spectrograms don't show anything (this spectro: http://i.imgur.com/J0Xtalf.png ) Could that be an influence of my amplifier? Or is it because the cartridge has a low output (2 mV according to http://audio-heritage.jp/TECHNICS/etc/epc-205c-iil.html )? Or maybe a combination of both?


Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #12
like I said after roughly 28 kHz the spectrograms don't show anything (this spectro: http://i.imgur.com/J0Xtalf.png ) Could that be an influence of my amplifier?

What's that mirror image of the signal at the top doing there?
It's not a mirror image, it's noise shaping.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #13
It's not a mirror image, it's noise shaping.
http://imgur.com/J4LMnuB

That's what I replied previously, since the spectrogram you uploaded has some noise which looks like constant. But in this picture

http://i.imgur.com/J0Xtalf.png

the noise is not constant, it begins at 0:21, now I guess it is not noise shaping.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #14
It's not a mirror image, it's noise shaping.
http://imgur.com/J4LMnuB

That's what I replied previously, since the spectrogram you uploaded has some noise which looks like constant. But in this picture

http://i.imgur.com/J0Xtalf.png

the noise is not constant, it begins at 0:21, now I guess it is not noise shaping.

The first spectro was from the middle of a song. The second spectro is the beginning of a song, the music doesn't really start before 0:21. I don't know if that has an influence or not. There's some light blue color at the top before 0:21 as well, but much less.

I will make a recording with Audacity, to see if I get a different result (the ones above were made with Sound Forge).

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #15
Indeeed, on the http://i.imgur.com/J0Xtalf.png , that is clearly imaging, which in turn might mean that you have a problem with the recording configuration. You might be recording at 96Khz, but in the middle it is working at 48Khz.

Check the configuration for the recording device on windows audio preferences. Also, ensure that your soundcard can record at 96Khz.

Edit:  For more information:  https://hydrogenaud.io/index.php/topic,104547.msg859022.html#msg859022

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #16
Indeeed, on the http://i.imgur.com/J0Xtalf.png , that is clearly imaging, which in turn might mean that you have a problem with the recording configuration. You might be recording at 96Khz, but in the middle it is working at 48Khz.

Check the configuration for the recording device on windows audio preferences. Also, ensure that your soundcard can record at 96Khz.

Edit:  For more information:  https://hydrogenaud.io/index.php/topic,104547.msg859022.html#msg859022
You're a life saviour!!! Really. Your message left me scratching my head (again), I had gone over my settings a million times and couldn't find a fault. I was ready to uninstall-reinstall my audio driver (once again, I had done that in the past without any improvements) when in my soundcard installation notice I read a small sentence that advised "to uninstall current audio drivers before installing the Creative audio driver". So I went to the Windows Device Manager to see if there were other audio drivers beside the Creative one. And indeed, there was a "High Definition Audio Driver": http://i.imgur.com/3y0LPm8.jpg I'm guessing this is the default audio driver supplied with WinVista. I deactivated it and this morning I hooked up my amplifier to the soundcard and did a test recording: http://i.imgur.com/Tz76S9J.png That does look like a 24/96 spectro now! I'm in a rush and have to leave but I will do more tests this evening when I get back home. But it looks like you helped save, maybe not my life, but most definitely my audio recordings.

PS: the hydrogenaud.io link you posted leads me to a blank page.

Re: Couple questions regarding the spectrals of a hi-res vinyl rip

Reply #17
So I did one more test just now, with the Microsoft High Definition Audio Driver still deactivated and the driver of my Creative soundcard set to 24/96 and I'm getting a spectrogram that isn't limited to 24 kHz and no mirroring like before:

- http://i.imgur.com/WuJHuWf.png
- http://i.imgur.com/jQPFHtr.png

So it looks like previously it was the Microsoft audio driver that was being used by the recording software (Sound Forge / Audacity) and not the Creative driver, despite each software showing the Creative driver as being the device in use. Creative ought to put a warning on their soundcards' installation notice in bold, red letters to deactivate/uninstall all other audio drivers. On mine the sentence is written with a picture telling you to disconnect the power cord of your computer before inserting the audio card in the PCI slot! No wonder I missed it.