• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

High-res audio comparison: Linn Records Free High Res Samples

Guermantes

Senior Member
Joined
Feb 19, 2018
Messages
428
Likes
457
Location
Brisbane, Australia
I thought oversampling in DSP was in order to avoid aliasing artefacts.

Here's a practical demonstration from FabFilter, makers of VST audio plugins:
 

redshift

Addicted to Fun and Learning
Joined
Jul 11, 2021
Messages
575
Likes
348
Upsampling is converting a digital signal to a higher sampling rate, yes?
This adds no new information and no increase in resolution.

A digital signal could also have its bit-depth changed, say from 16 to 24 bits.
That process could push the noise floor down.

Truncation occurs in the LSB, which leads to quantization distortion.
At what dBFS do you think truncation of the LSB manifests?

It depends on how much processing you do, for example how many filter stages/types you got in the signal processing/feedback control pipeline.
 

redshift

Addicted to Fun and Learning
Joined
Jul 11, 2021
Messages
575
Likes
348
Actually, I'll modify my reply, as I wasn't really addressing the question properly earlier.
If you are going back to the master tape (and just exactly what constitutes a master tape is a whole other conversation) there is a very specific reason to capture at a very high sample rate. You need to be able to capture the bias signal. This is going to be of the order of 150 kHz odd, and depends on the machine and sometimes even choice of tape. The reason for capturing the bias is to allow you to correct for speed variations, and critically, scrape flutter. When the tape passes the heads it scrapes, and just like a finger rubbing the rim of a wineglass you can get very fast changing modulation of the speed of the tape past the head. This leads to a form of intermodulation distortion that is welded into the tape at the moment of recording. If you can see the bias signal you have what amounts to a local clock embedded in the tape that you can use to deconvolve the flutter. If you have the luxury of finding the original tracking tapes you can recover real audio that nobody has ever heard since the day the track was laid down. Even if you only have access to the master tapes, you still have a chance to remove the last generation of scrape flutter.

So this is another issue with these high res releases from old tapes. It would appear that most have not taken advantage of this possibility, and thus have actually lost a real opportunity to crete a better quality release, and have simply been lazy and depended upon woo to justify what they are doing. Proper forensic analysis of the tapes and modern processing would have yielded a conventional 16/48 result that exceeds the real musical information available in the silly lazy money grubbing stuff they are actually pedalling.

Let’s not confuse the technical advantages of hires formats with marketing gimmicks.

If you sport a DSP in your audio chain it is almost certainly internally operating with 192kHz and 24/32 bits.

That is for a reason. Truncation is bad.
 

redshift

Addicted to Fun and Learning
Joined
Jul 11, 2021
Messages
575
Likes
348
I thought oversampling in DSP was in order to avoid aliasing artefacts.

Here's a practical demonstration from FabFilter, makers of VST audio plugins:

Nope, oversampling is for improving resolution. Say going from 8bits @ 10kHz to 12 bits @ 100Hz.

After all, many delta sigma DAC’s is 1bits @ X MHz
 

dc655321

Addicted to Fun and Learning
Joined
Mar 4, 2018
Messages
934
Likes
1,128
If you sport a DSP in your audio chain it is almost certainly internally operating with 192kHz and 24/32 bits.

Um no, DSP can be done entirely in software and at whatever frequency and bitdepth is appropriate for the application.

That is for a reason. Truncation is bad.
Just. Stop.

Nope, oversampling is for improving resolution.
After all, many delta sigma DAC’s is 1bits @ X MHz

In your case, yes, improving the resolution potential of the DAC.
Note, this does not improve the resolution of the signal being converted by the DAC.
That was fixed by the original A2D conversion.
 

redshift

Addicted to Fun and Learning
Joined
Jul 11, 2021
Messages
575
Likes
348
Um no, DSP can be done entirely in software and at whatever frequency and bitdepth is appropriate for the application.


Just. Stop.




In your case, yes, improving the resolution potential of the DAC.
Note, this does not improve the resolution of the signal being converted by the DAC.
That was fixed by the original A2D conversion.

Haha, yeah a DSP is in most cases a programmable device. It’s up to you programming it to do basically what you want with the bits and bytes. Well, up to its rated performance and capacity.

Well of course you can recover more signal than what’s in the original low-res data but then you must know something about the source.

:D
 

Blumlein 88

Master Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
12,449
Likes
18,754
Haha, yeah a DSP is in most cases a programmable device. It’s up to you programming it to do basically what you want with the bits and bytes. Well, up to its rated performance and capacity.

Well of course you can recover more signal than what’s in the original low-res data but then you must know something about the source.

:D
Care to give an example of this magic where you recover more signal than is in the original data?
 

redshift

Addicted to Fun and Learning
Joined
Jul 11, 2021
Messages
575
Likes
348
Care to give an example of this magic where you recover more signal than is in the original data?

6f6e65206b696c6f686572747a2073696e652077617665

Tell me what resolution and bitrate you want that data in.

;)
 

Blumlein 88

Master Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
12,449
Likes
18,754

dc655321

Addicted to Fun and Learning
Joined
Mar 4, 2018
Messages
934
Likes
1,128
@redshift, I was giving you the benefit of the doubt, but am now convinced you haven't got a clue.
Bye.
 
Top Bottom