High Resolution Audio: Does It Matter?

[solderdude]

Not Blumlein but will take a stab at it.

The Plangent redbook releases are (so much) better because they were processed before the CD was made.
The differences between the original (non treated/processed) RBCD and treated RBCD thus have nothing to do with RBCD not being capable enough but the processing is the difference here ?

In order to do the processing they use 96/24.

 

[Blumlein 88]

Not Blumlein but will take a stab at it.

The Plangent redbook releases are (so much) better because they were processed before the CD was made.
The differences between the original (non treated/processed) RBCD and treated RBCD thus have nothing to do with RBCD not being capable enough but the processing is the difference here ?

In order to do the processing they use 96/24.

Yes that is it. The processed CD has had the scrape flutter and other issues removed. That fact it sounds better than the unprocessed version is simply an indication the Plangent process works and that CD is of fully needed audible resolution to show the difference.

Prior to Plangent a CD has scrape flutter and other issues. After Plangent those issues are fixed and downsampled to CD and it shows the difference. Had an original recording been done digitally at 44.1 sample rates, this indicates no reason to think a higher sample rate would have been better. It would still appear some aren't aware of what the Plangent process does, and why these results are exactly what one might expect. While also indicating nothing about why original recordings to 44.1 would suffer in comparison to original recordings in 96.

 

[Mark S.]

I believe some here still haven't grasped, or 'visualised' what scrape-flutter actually does to a signal, perhaps perceiving it to be little different to the plainly audible (and easily measurable) "wow and flutter" of, for e.g., consumer cassette decks

Think about a frequency of 12Khz, one of the harmonics of an instrument, well inside the audio-band, and the hearing range of most people under 50 years old with no hearing damage. If the signal is being modulated by, say, 750Hz scrape-flutter, that tone will be rapidly compressed and expanded in the temporal domain, on the rise and fall of that flutter frequency, to a new range of frequencies. The math to calculate what those would be is beyond me (or is it as straightforward as 12KHz +/- 750Hz?).

From what I gather, our hearing is acutely sensitive to such micro-variations in timing in HF. The flutter will already have had an audible effect on the sound when monitored directly off the source (the tape), and when sampled at 44.1KHz and 'reconstructed' via a steep c.21KHz low-pass digital filter ....

And to re-iterate; this can be ameliorated to a large extent simply by capturing and playing back with high sample-rate, high bandwidth. But further, if the flutter is then removed and the signal down-sampled, then the resulting Red Book file will be far better than a direct dub of the tape to 44.1, as demonstrated by Plangent.

 

[derp1n]

 

[Mark S.]

Not Blumlein but will take a stab at it.

The Plangent redbook releases are (so much) better because they were processed before the CD was made.
The differences between the original (non treated/processed) RBCD and treated RBCD thus have nothing to do with RBCD not being capable enough but the processing is the difference here ?

In order to do the processing they use 96/24.

" ... Conversion resolution is minimum 96 kHz / 24-bit, though 192/24 is preferred." And I gather 384 khz is sometimes used.

 

[Mark S.]

Witty. Positively Wildean.

 

[andreasmaaan]

@Mark S. it's me again

Our ears are most sensitive to timing variations in roughly the 300-1000Hz range. Our sensitivity to inter-channel variations is orders of magnitude greater than our sensitivity to variations that affect both (or multiple) channels uniformly. I'm not sure whether the kind of variations caused by scrape-flutter falls into the former or the latter category.

Also, still no reason or evidence has been offered as to why these releases would sound better when played back at higher than 44.1Khz.

 

[Frank Dernie]

You're not listening. Why are the Plangent Red Book / 44.1 releases so much better than any previous versions if, as you say, Red Book can "portray" the flutter/temporally modulated signal "just fine"? Be specific - hand-waving about "poor mastering", "incorrect EQ" et al isn't an explanation.

Are you referring to this company's recordings?
https://plangentrecords.bigcartel.com

 

[derp1n]

Are you referring to this companies recordings?
https://plangentrecords.bigcartel.com

https://www.plangentprocesses.com/

Read back a few pages...

 

[Frank Dernie]

https://www.plangentprocesses.com/

Read back a few pages...

I could but jumped to the end since it was a long thread and I am short of time, sorry!

Must make interesting reading for those involved in the renewal of enthusiasm for reel-to-reel tape being pushed in magazines and on the net recently!
Even as a mediocre amateur recorder of mainly choral music the shortcomings of tape recorders were obvious!

 

[Wombat]

I believe some here still haven't grasped, or 'visualised' what scrape-flutter actually does to a signal, perhaps perceiving it to be little different to the plainly audible (and easily measurable) "wow and flutter" of, for e.g., consumer cassette decks

Think about a frequency of 12Khz, one of the harmonics of an instrument, well inside the audio-band, and the hearing range of most people under 50 years old with no hearing damage. If the signal is being modulated by, say, 750Hz scrape-flutter, that tone will be rapidly compressed and expanded in the temporal domain, on the rise and fall of that flutter frequency, to a new range of frequencies. The math to calculate what those would be is beyond me (or is it as straightforward as 12KHz +/- 750Hz?).

From what I gather, our hearing is acutely sensitive to such micro-variations in timing in HF. The flutter will already have had an audible effect on the sound when monitored directly off the source (the tape), and when sampled at 44.1KHz and 'reconstructed' via a steep c.21KHz low-pass digital filter ....

And to re-iterate; this can be ameliorated to a large extent simply by capturing and playing back with high sample-rate, high bandwidth. But further, if the flutter is then removed and the signal down-sampled, then the resulting Red Book file will be far better than a direct dub of the tape to 44.1, as demonstrated by Plangent.

If you can't explain it simply you do not understand it enough or your explanatory skills let you down. More homework is needed.
On this forum there are knowledgeable and experienced members in many fields and I see you as struggling with replying to considered responses to your posits. I am probably wasting my time as you have been given many considered replies but choose to not consider them and conduct a one-way monologue on your original position.

This engagement is falling into the usual merry-go-round that indicates a lack of awareness and appreciation of existing knowledge and understanding of audio.

The earlier suggestion that you take a hiatus and educate yourself beyond your personal opinions is sound. Some understanding of rational argument would assist in your presenting a case.
I put this to you as constructive criticism. ASR is more demanding than most audio forums re opinions being credible in an evidential manner if you choose to take them public.

 

[Blumlein 88]

+10

 

[Blumlein 88]

I believe some here still haven't grasped, or 'visualised' what scrape-flutter actually does to a signal, perhaps perceiving it to be little different to the plainly audible (and easily measurable) "wow and flutter" of, for e.g., consumer cassette decks

Think about a frequency of 12Khz, one of the harmonics of an instrument, well inside the audio-band, and the hearing range of most people under 50 years old with no hearing damage. If the signal is being modulated by, say, 750Hz scrape-flutter, that tone will be rapidly compressed and expanded in the temporal domain, on the rise and fall of that flutter frequency, to a new range of frequencies. The math to calculate what those would be is beyond me (or is it as straightforward as 12KHz +/- 750Hz?).

From what I gather, our hearing is acutely sensitive to such micro-variations in timing in HF. The flutter will already have had an audible effect on the sound when monitored directly off the source (the tape), and when sampled at 44.1KHz and 'reconstructed' via a steep c.21KHz low-pass digital filter ....

And to re-iterate; this can be ameliorated to a large extent simply by capturing and playing back with high sample-rate, high bandwidth. But further, if the flutter is then removed and the signal down-sampled, then the resulting Red Book file will be far better than a direct dub of the tape to 44.1, as demonstrated by Plangent.

You have a fundamental misunderstanding. Capturing at 44.1 khz will have all the scrape flutter intact. In fact you wouldn't hear a difference vs the original tape itself. If you captured it at 96 khz, the simple act of using that higher sample rate in no way effects that. You'll still have the same scrape flutter in the recording. The 96 khz or even 384 khz do absolutely nothing to reduce or alter the scrape flutter. Neither will sound different than 44 sampling.

Plangent uses the higher sample rates to capture the bias frequency which also has nothing to do with scrape flutter directly. However, using the bias frequency as a frequency reference they can then do additional DSP to fix the flutter and correct timing as if the original tape had no such flutter. So yes after that additional processing you could downsample to 44.1 and the flutter is gone or mostly gone. Without the additional processing sample rate itself will have no effect on the flutter. None. Zero. It is not happening. Until you understand how this is so, you are imagining something that isn't happening. Which is then leading you to erroneous conclusions about sample rates.

 

[solderdude]

I seem to be experiencing a déjà-vu ....

When they want to capture and use the bias frequency 384kHz is needed as most bias frequencies are > 80kHz.
This is why they MUST use hires. A lower samplerate won't capture the bias frequency. It has nothing to do with the sound quality of RBCD but the need to process at a higher res. Afterwards downsample to RBCD obviously doesn't hurt the corrected recording.
Downsample the result to RB and voila ... de-scraped sound.
To do this it stands to reason to use the bias frequency as that is the only 'constant' on which this process can be based.
Assuming the frequency of the LC oscillator used is constant this technique removes wow and flutter of both the recording and playback decks. Which is neat and quite scientific !

If I 'listen' correctly to Mark S his reasoning is this:

Old recordings converted to 44.1 with scrape-flutter sound 'worse' than the same old recording converted to 96/24 or higher. At least that's what I think he claims. This would be due to timing errors not captured (which is a false assumption I think).

the only way to test this theory is to use a 192/24 or higher digital copy of a tape with the 'sticky tape' problem and downsample to 44/16 and compare. I don't think one can compare a previous 'digital copy' of the same tape (different tape deck ?) at a much earlier date to a later 'digital copy' even at a higher resolution. The conditions have changed...

 

[Wombat]

You have a fundamental misunderstanding. Capturing at 44.1 khz will have all the scrape flutter intact. In fact you wouldn't hear a difference vs the original tape itself. If you captured it at 96 khz, the simple act of using that higher sample rate in no way effects that. You'll still have the same scrape flutter in the recording. The 96 khz or even 384 khz do absolutely nothing to reduce or alter the scrape flutter. Neither will sound different than 44 sampling.

Plangent uses the higher sample rates to capture the bias frequency which also has nothing to do with scrape flutter directly. However, using the bias frequency as a frequency reference they can then do additional DSP to fix the flutter and correct timing as if the original tape had no such flutter. So yes after that additional processing you could downsample to 44.1 and the flutter is gone or mostly gone. Without the additional processing sample rate itself will have no effect on the flutter. None. Zero. It is not happening. Until you understand how this is so, you are imagining something that isn't happening. Which is then leading you to erroneous conclusions about sample rates.

Agree and agree again. However scrape flutter is either audible or not. If it can be addressed without side effects, it is useful in remasters/re-issues for the sake of accuracy(audible or not) but don't let us denigrate the large body of past recordings where it is not a noticeable problem.

I would be interested in seeing if there is processor that can fix that 'slur'(speed uptake?) that occurs in the early stage of the Rolling Stones track, Ruby Tuesday.

Sometimes this audio discrimination thing is a bit like those endeavours to put more of the Bible onto a pin-head. Meaningless from a reading(hearing?) point of view.

 

[Mark S.]

I would have hoped that everyone reading here has at least enough inate curiosity that they've done a little research

@Mark S. it's me again

Our ears are most sensitive to timing variations in roughly the 300-1000Hz range. Our sensitivity to inter-channel variations is orders of magnitude greater than our sensitivity to variations that affect both (or multiple) channels uniformly. I'm not sure whether the kind of variations caused by scrape-flutter falls into the former or the latter category.

Also, still no reason or evidence has been offered as to why these releases would sound better when played back at higher than 44.1Khz.

I found a paper dating back some years concerning rigorous measurements (done in Germany) of scrape-flutter frequencies, and it gave figures by tape type/manufacturer, interestingly. The figures were, IIRC, 720-960 Hz, but I assume tape speed has a bearing, and I don't remember what they used.

Plangent say they find timing errors at frequencies right up to c. 2kHz.

I've made several attempts to explain why the smeared HF (in particular) is made considerably worse by the effects of 44.1's steep, low-pass filtering, close to the audio band, but this is (despite being measurable) "inaudible" according to most respondents here.

 

[andreasmaaan]

I've made several attempts to explain why the smeared HF (in particular) is made considerably worse by the effects of 44.1's steep, low-pass filtering, close to the audio band, but this is (despite being measurable) "inaudible" according to most respondents here.

Fair enough. Not everything that's measurable is audible though.

The logic is: if you can't hear anything at 21KHz, you can't hear transient smearing at 21KHz.

 

[Cosmik]

I would be interested in seeing if there is processor that can fix that 'slur'(speed uptake?) that occurs in the early stage of the Rolling Stones track, Ruby Tuesday.

I'm intrigued! At approximately what time is the effect?

 

[Krunok]

From what I gather, our hearing is acutely sensitive to such micro-variations in timing in HF.

From which technical source exactly did you gather that?

 

[Krunok]

I've made several attempts to explain why the smeared HF (in particular) is made considerably worse by the effects of 44.1's steep, low-pass filtering, close to the audio band, but this is (despite being measurable) "inaudible" according to most respondents here.

And for that reason IMHO it would be extremely usefull if we could draw an "inaudible" line on every parameter we measure, but unfortunately not much research has been done in that direction.