IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Mastering to Tape, 96khz Useful?
will.pett
post Aug 7 2013, 19:52
Post #1





Group: Members
Posts: 4
Joined: 7-August 13
Member No.: 109519



Hello,

I have so far not seen any convincing argument for recording at ultra-high sample rates. While the advantages of 24-bit recording are clear (keeping the noise floor low during processing), recording at sample rates higher than 48khz or so seems pointless. Anti-aliasing arguments do not make sense since oversampling at extremely high sample rates is already done in hardware by all ADCs these days.

However, I am dubbing my recordings onto compact cassettes directly through my D/A converter, and I am wondering if maybe there is an advantage to using high-resolution recordings for this purpose? For example, one could imagine that the ultrasonic content of my masters might have a beneficial effect on the high-frequency response of the tape by in effect introducing a bias signal that is correlated with the high-frequency content... i.e. in addition to a single high-frequency bias signal like the one employed in most cassette decks, the audio itself would have a built in auto-correlated bias signal that would boost high-frequency response in high-energy recordings where the high-end typically suffers on tape. Or would the ultrasonics in a typical recording has such low amplitudes that they would have no effect on tape bias?

Anyone familiar with tape mastering who can comment on this?

Are there other mastering situations where recording/playback at higher rates could be useful? I am not familiar with a lot of DSP techniques out there... maybe there are some plug-ins/effects/filters that benefit from working with high sample rate audio, e.g. for anti-aliasing purposes? In this case, there would be an argument in favor of hi-res audio during processing similar to that in favor of 24-bit. If this is the case, can someone explain it to me? Would these techniques not simply benefit from up-sampling instead of recording at high rates in the first place?

This post has been edited by will.pett: Aug 7 2013, 20:02
Go to the top of the page
+Quote Post
hlloyge
post Aug 7 2013, 20:40
Post #2





Group: Members
Posts: 689
Joined: 10-January 06
From: Zagreb
Member No.: 27018



I'm sorry, you are making output to standard compact cassettes? like TDK, SONY C60 or C90? Even the metal tape had frequency response up to some 18 kHz on quality tape decks, and considering it's old analogue technology, I seriously doubt you will have any benefits from outputing anything more than 16/44.1.
Go to the top of the page
+Quote Post
Ethan Winer
post Aug 7 2013, 20:43
Post #3





Group: Members
Posts: 248
Joined: 12-May 09
From: New Milford, CT
Member No.: 69730



QUOTE (will.pett @ Aug 7 2013, 14:52) *
While the advantages of 24-bit recording are clear (keeping the noise floor low during processing), recording at sample rates higher than 48khz or so seems pointless.


I think both are pointless. A cassette has an effective bit depth of maybe 8 bits, so recording at more than 16 bits seems as pointless to me as recording at 96 or 192 KHz. Even the finest analog open reel recorders don't achieve a noise floor as quiet as 16 bits.

QUOTE
one could imagine that the ultrasonic content of my masters


By "masters" do you mean your cassettes? Is there even any content (bias leakage doesn't count) past maybe 15 KHz?

--Ethan


--------------------
I believe in Truth, Justice, and the Scientific Method
Go to the top of the page
+Quote Post
TomasPin
post Aug 7 2013, 20:47
Post #4





Group: Members
Posts: 169
Joined: 5-June 13
From: Argentina
Member No.: 108508



QUOTE (hlloyge @ Aug 7 2013, 16:40) *
I seriously doubt you will have any benefits from outputing anything more than 16/44.1.

...or 16/48 for any listening purposes.


--------------------
A man and his music: http://tubular.net/
Go to the top of the page
+Quote Post
will.pett
post Aug 7 2013, 21:33
Post #5





Group: Members
Posts: 4
Joined: 7-August 13
Member No.: 109519



Yes, most cassette tapes are only able to faithfully reproduce frequencies up to around 18khz, but the record heads in all cassette decks are capable of reproducing much higher frequencies. Common bias signals are greater than 100khz. While these are not reproduced during playback of the tape, they definitely have an effect during recording. I am curious if perhaps the inaudible ultrasonics in a hi-res recording might also have some similar effect on the tape response to audible frequencies. It seems unlikely, since they are usually so quiet, but it seems possible that there could be a measurable effect.

This post has been edited by will.pett: Aug 7 2013, 21:35
Go to the top of the page
+Quote Post
will.pett
post Aug 7 2013, 21:57
Post #6





Group: Members
Posts: 4
Joined: 7-August 13
Member No.: 109519



QUOTE (Ethan Winer @ Aug 7 2013, 14:43) *
By "masters" do you mean your cassettes? Is there even any content (bias leakage doesn't count) past maybe 15 KHz?


Sorry, neglected to mention. My masters are digital so they can have ultrasonic content, yes.
Go to the top of the page
+Quote Post
Glenn Gundlach
post Aug 7 2013, 23:29
Post #7





Group: Members
Posts: 359
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (Ethan Winer @ Aug 7 2013, 11:43) *
<snip>
I think both are pointless. A cassette has an effective bit depth of maybe 8 bits, so recording at more than 16 bits seems as pointless to me as recording at 96 or 192 KHz. Even the finest analog open reel recorders don't achieve a noise floor as quiet as 16 bits.
<snip>
--Ethan


I have a Sony test CD that a friend asked to be copied to a cassette. The disc has tones from 0 to -90dB in 10 dB steps. My Nakamichi deck can reach -65dB for the noise floor so I expected to hear nothing beyond the -60 dB tone. I was wrong as I could clearly hear the -70dB, -80dB and -90dB tones when playing back the tape. This was not a subtle thing requiring ABX tests to verify. It was obvious to everyone.

G
Go to the top of the page
+Quote Post
DVDdoug
post Aug 8 2013, 00:09
Post #8





Group: Members
Posts: 2441
Joined: 24-August 07
From: Silicon Valley
Member No.: 46454



QUOTE
I am curious if perhaps the inaudible ultrasonics in a hi-res recording might also have some similar effect on the tape response to audible frequencies.
You would have to assume the existing bias is not optimum, or assume that somehow a variable or "dynamic" bias would be more accurate. If the tape machine is set-up properly, randomly monkeying with the bias can only make things worse!

IIRC, the bias signal is very strong and any high-frequency signal is probably insignificant in comparison.

In fact if there is significant ultrasonic energy in your digital master, you should probably filter it out. You wouldn't want to "waste" any of your dynamic range on inaudible signals. And, if you listen to the digital master you wouldn't want to waste amplifier power, or maybe damage your tweeter!
Go to the top of the page
+Quote Post
splice
post Aug 8 2013, 00:23
Post #9





Group: Members
Posts: 113
Joined: 23-July 03
Member No.: 7935



There's a very good reason why you should filter out everything above 20 KHz or so if recording to cassette.
Ultrasonic frequencies will beat with the bias signal and can create difference frequencies in the audible range. The most common example is when recording FM radio, where harmonics of the 19 KHz subcarrier can beat with the bias oscillator. That's why many cheap cassette deck/radio combos had a "beat" switch, which changed the bias oscillator frequency slightly. It was cheaper to do this than adequately filter out the 19 KHz subcarrier. Some decks had filtering, but then you still had to make the filtering defeatable if the deck was capable of recording up to 19 KHz.


--------------------
Regards,
Don Hills
Go to the top of the page
+Quote Post
saratoga
post Aug 8 2013, 01:25
Post #10





Group: Members
Posts: 4718
Joined: 2-September 02
Member No.: 3264



QUOTE (Glenn Gundlach @ Aug 7 2013, 18:29) *
My Nakamichi deck can reach -65dB for the noise floor so I expected to hear nothing beyond the -60 dB tone.


-65dB is the noise floor in the time domain, while a -60dB is the intensity in the frequency domain. The actual noise power over the bandwidth of that tone is like something like -100dB or less if its really a pure tone. I expect you can easily hear the tone.
Go to the top of the page
+Quote Post
will.pett
post Aug 8 2013, 02:58
Post #11





Group: Members
Posts: 4
Joined: 7-August 13
Member No.: 109519



QUOTE (DVDdoug @ Aug 7 2013, 18:09) *
You would have to assume the existing bias is not optimum, or assume that somehow a variable or "dynamic" bias would be more accurate. If the tape machine is set-up properly, randomly monkeying with the bias can only make things worse!

In fact if there is significant ultrasonic energy in your digital master, you should probably filter it out. You wouldn't want to "waste" any of your dynamic range on inaudible signals. And, if you listen to the digital master you wouldn't want to waste amplifier power, or maybe damage your tweeter!


Hmm...
Excellent point!

Let me think this through...
Dynamic bias (like Dolby HX-Pro) can be beneficial for boosting high-frequency headroom... However, this is accomplished by decreasing the amplitude of the bias signal in the presence of strong high-frequency signals. Thus, if ultrasonics modulate/boost the bias during recording, they would actually reduce the audible high-frequency headroom when it is needed most! And if you're using HX-Pro, it seems plausible that any ultrasonic energy could potentially "soak up" the headroom boost high audible frequencies would normally get.

And yeah, I should have thought about the MPX Filter. It exists for precisely this reason.

So you are absolutely right that ultrasonics should be filtered out (or never recorded in the first place) when mastering to tape (or any other time)!

This post has been edited by will.pett: Aug 8 2013, 03:03
Go to the top of the page
+Quote Post
Ethan Winer
post Aug 8 2013, 22:39
Post #12





Group: Members
Posts: 248
Joined: 12-May 09
From: New Milford, CT
Member No.: 69730



QUOTE (will.pett @ Aug 7 2013, 16:33) *
I am curious if perhaps the inaudible ultrasonics in a hi-res recording might also have some similar effect on the tape response to audible frequencies.

People can't hear or perceive ultrasonics, no matter what you might read in some hi-fi magazines. So the only affect I can imagine is beating - which you definitely don't want! - as splice mentioned.

--Ethan


--------------------
I believe in Truth, Justice, and the Scientific Method
Go to the top of the page
+Quote Post
Ethan Winer
post Aug 8 2013, 22:45
Post #13





Group: Members
Posts: 248
Joined: 12-May 09
From: New Milford, CT
Member No.: 69730



QUOTE (Glenn Gundlach @ Aug 7 2013, 18:29) *
I have a Sony test CD that a friend asked to be copied to a cassette. The disc has tones from 0 to -90dB in 10 dB steps. My Nakamichi deck can reach -65dB for the noise floor so I expected to hear nothing beyond the -60 dB tone. I was wrong as I could clearly hear the -70dB, -80dB and -90dB tones

This brings up an interesting point. Cassette decks with Dolby or dbx type noise reduction don't really give a larger dynamic range. It only seems that way when playing back very soft content. But when playing music that's recorded closer to 0VU, the tape hiss is indeed only 50-60 dB below the music. So I'll grant you that recording cassettes using 16 bits makes more sense than 8 bits. But personally I wouldn't bother with 24 bits.

--Ethan


--------------------
I believe in Truth, Justice, and the Scientific Method
Go to the top of the page
+Quote Post
qduaty
post Aug 9 2013, 12:00
Post #14





Group: Members
Posts: 9
Joined: 10-July 13
Member No.: 109045



QUOTE (will.pett @ Aug 7 2013, 20:52) *
I have so far not seen any convincing argument for recording at ultra-high sample rates. [...]For example, one could imagine that the ultrasonic content of my masters might have a beneficial effect on the high-frequency response of the tape by in effect introducing a bias signal that is correlated with the high-frequency content...

With digital magic we have today, you can model tape and head characteristics (hysteresis loop, self-demagnetization, head slot) and record without bias. I guess you'll need ultrasonic D/A and amplifier to pass short impulses to the head. Don't forget to share your results because it's interesting.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Aug 12 2013, 20:53
Post #15





Group: Members
Posts: 3537
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Glenn Gundlach @ Aug 7 2013, 18:29) *
QUOTE (Ethan Winer @ Aug 7 2013, 11:43) *
<snip>
I think both are pointless. A cassette has an effective bit depth of maybe 8 bits, so recording at more than 16 bits seems as pointless to me as recording at 96 or 192 KHz. Even the finest analog open reel recorders don't achieve a noise floor as quiet as 16 bits.
<snip>
--Ethan


I have a Sony test CD that a friend asked to be copied to a cassette. The disc has tones from 0 to -90dB in 10 dB steps. My Nakamichi deck can reach -65dB for the noise floor so I expected to hear nothing beyond the -60 dB tone. I was wrong as I could clearly hear the -70dB, -80dB and -90dB tones when playing back the tape. This was not a subtle thing requiring ABX tests to verify. It was obvious to everyone.


The human ear can hear pure tones that are well below the noise floor depending on the listening level. The explanation in another post about bandwidth is relevant. The ear operates like a spectrum analyzer and breaks down signals based on a sort of set biological filters in a component of the inner ear called the Basilar Membrane.

The ability to hear pure tones below the noise floor is a property of both analog and digital recordings. The exact degree of this effect depends on the frequency of the tone and the spectral content of the noise floor. Under ideal conditions the tone can be dozens of dB below the RMS amplitude of the noise floor and still be audible and free of gritty sounds.

There is a belief ("Audiophile Myth") that digital recording cannot reproduce signals cleanly below the LSB, but in fact it can and do so quite nicely and totally free of distortion as long as the digital recording is dithered.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 20th April 2014 - 22:05