Jump to content
IGNORED

Some commonsense


Recommended Posts

Conceptually every single method of recording/encoding sound is lossy, because there is no recording method or medium that captures 100% of the sound waves present in the performance/generation of the sound. 

 

But this is a great example of a statement that, while strictly true, is so general as to be meaningless and even misleading or obfuscatory rather than illuminating and clarifying.

 

For one thing, what we call a "sound" wave, as opposed to another kind of wave, is ultimately arbitrary. The difference between a sound wave and a radio wave, for example, is just the frequency. I honestly don't know what the commonly accepted cutoff point is - or even if there is one - but it's interesting to note that according to Wikipedia, the bottom of the AM radio spectrum is about 148kHz, while a 352.8kHz DXD recording technically can reproduce sound at a higher frequency, up to 176.4kHz. So the difference between a sound wave and a radio wave is a made-up convention that doesn't even correspond to human hearing, given that the aforementioned 176.4kHz "sound" frequency is of course far, far beyond human audibility.

 

So IMHO it doesn't make sense to call any digital sample rate "lossy" if it samples at more than 2X the upper limit of human hearing. For that reason, I would argue that it is plausible, but ultimately pointless, to call redbook/CD's 44.kHz sample rate lossy. 

 

The only respect in which 44.1kHz is lossy is that if it is generated by downsampling from a 96kHz (or 192kHz) source, the downsampling is non-integer (44.1 doesn't divide evenly into 96 or 192), and there is some alteration of the source there. But if you downsample, say, a 96k source to 48kHz, or a 176.4k source to 44.1kHz, I would argue that such downsampling is not lossy: because the sample rates divide evenly into each other, you are losing nothing - all audible-range frequencies are preserved. The only difference is that those frequencies are sampled fewer times. But since basic digital sampling theory tells us that any frequency needs to be sampled only twice in order to be properly reproduced, and since 44.1k can sample all frequencies up to 22.05k at least twice, there literally is no difference - nothing is lost.

 

Of course, if you reduce bit-depth, something is lost - but again, the result there simply is an increased noise floor from higher quantization error. This is the hardest part of digital sampling theory for our brains to grasp: "errors" or "missing/lost" stuff in digital sampling manifests simply as the noise floor, not as distortion or "gaps" or "stairsteps" in the waveform.

 

 

 

 

 

Link to comment
2 hours ago, tmtomh said:

So IMHO it doesn't make sense to call any digital sample rate "lossy" if it samples at more than 2X the upper limit of human hearing. For that reason, I would argue that it is plausible, but ultimately pointless, to call redbook/CD's 44.kHz sample rate lossy. 

 

You are missing the point. Whether a compression scheme is lossy or not has nothing to do with the sample rate per se. If a recording were made in 16/44 then it wouldn’t be lossy. Yet when a recording is made in 24/192 and mastered in 24/192 and then down converted to 16/44 or MPEG then its lossy. Simple. I’m not calling redbook/CD sample rate lossy, rather  CDs are themselves often lossy because the recordings often are made at a higher rate/depth.

Custom room treatments for headphone users.

Link to comment
2 hours ago, tmtomh said:

The only respect in which 44.1kHz is lossy is that if it is generated by downsampling from a 96kHz (or 192kHz) source, the downsampling is non-integer (44.1 doesn't divide evenly into 96 or 192), and there is some alteration of the source there. But if you downsample, say, a 96k source to 48kHz, or a 176.4k source to 44.1kHz, I would argue that such downsampling is not lossy: because the sample rates divide evenly into each other, you are losing nothing - all audible-range frequencies are preserved.

 

This is the same logic that JPEG uses to provide lossy compression yet preserving “relevant” image detail. In any case lossy compression is lossy compression no matter how good it is — we can equally argue that MPEG 320 is audibly indistinguishable to a majority of humans from redbook CD.

Custom room treatments for headphone users.

Link to comment
20 hours ago, Ajax said:

Hi Jabbr,

 

I agree that ideally we should replay at the same resolution as the recording (pre digital music was mastered specifically to be replayed on turntables by limiting the base so the needle did not jump out of the grooves. Today we limit the band width of the audio frequency so we don't end up with artefacts).

 

My reasoning of recording in hi-res and playing back at 16/44.1 was simply to achieve more head room during the recording process, which with due care is obviously not essential, especially today with so much compression being added.

 

The point of the article is that mathematically 16/44.1 is adequate for music playback, IF as George points out the recording and mastering has been done with sufficient care. My hearing is limited to 12khz (I'm 63) so for me personally there is no need to record at higher frequency rates (above 44.1), however, maybe there is benefit at recording at 24 bits, however slight.

 

I went to a hi-fi show in Melbourne, Australia about 4 years ago and heard the Devialet ensemble being demonstrated using only CDs and it was truly stunned by the sound, despite the poor acoustic environment of hotel show rooms. There were lots of competing gear with massive power amps and exotic cables but nothing to my ears came close. I bought the demo system, which included Atohm GT1 speakers and it now sits in my living room.

 

However, the best sound I have ever heard was my office system, which consisted of a Benchmark DAC1 (I just purchased a DAC2 second hand for $US900) driving a pair of Adam X7 active monitors listening to Gwyneth Herbert recorded and produced by Peter Gabriel's 'Society of Sound". This was recorded and distributed at 24/48. One night about 7 years ago my then 12 years old son (now an accomplished musician) came into say goodnight and said "that's spooky Dad, it's like she is in the room with us".

 

The point is John states (and my experience confirms) that extremely higher frequency rates aren't required, especially for old buggers like me, and if you don't play your music at pain levels you also don't need more than 16 bits.

 

As John says, it is all in the maths. It is important to remember the reason why we have digital audio in the first place is because a couple of very smart mathematicians, Shannon & Nyquist, developed a theorem that simply put states that if you record at twice the highest maximum audio frequency of the music then that is sufficient for perfect fidelity and no actual information is lost.

 

The other reason for reproducing the article is that the majority of music is available at 16/44.1, whether via CD, downloads or streaming, and manufactures should therefore be concentrating on improving the playback of that resolution, preferable using a combination of software and hardware, not offering us more and more exotic gizmos in an effort to differentiate themselves from other manufactures..

 

 

 

 

Just a question - does your Benchmark internally upsample every input? If so, that is probably the biggest reason why the two files sound alike. My old Benchmark upsampled all inputs to 100khz, which gave the effect of making all the inputs sound the same or at east very similar. .

 

Not sure if the new models do that or something similar.  

 

There are very good sounding DACs that resample everything to DSD and I cannot usually tell the difference between a red book and 24/96k file on those DACs either. 

Anyone who considers protocol unimportant has never dealt with a cat DAC.

Robert A. Heinlein

Link to comment
1 hour ago, jabbr said:

 

You are missing the point. Whether a compression scheme is lossy or not has nothing to do with the sample rate per se. If a recording were made in 16/44 then it wouldn’t be lossy. Yet when a recording is made in 24/192 and mastered in 24/192 and then down converted to 16/44 or MPEG then its lossy. Simple. I’m not calling redbook/CD sample rate lossy, rather  CDs are themselves often lossy because the recordings often are made at a higher rate/depth.

 

With respect, you are missing the point. I was quite careful not to say simply that CDs are not lossy. I said they are lossy but IMHO it's pointless and misleading to call them lossy in the same way an mp3 is lossy.

 

A CD made from (for example) a 176.4kHz digital master is not actually lossy in the sense that we usually use the term "lossy" because every frequency in the audible range is sampled just as accurately by the 44.1k sample rate as by the 176.4k sample rate. There are many good reasons (depending on the situation) to oversample, but if perfect reconstruction (minus quantization error, of course) is the definition of lossless, then a 44.1k CD source made from a 176.4k original will be lossless in the audible range. Yes, the ultrasonics beyond 22.05k will indeed be lost, and in that sense CD is lossy. No disagreement, and that's exactly what I noted in my prior comment.

 

Now, I also agree with you for the most part that a 320k mp3 or AAC file will be indistinguishable from a CD original in many, probably most, listening situations. But IMHO it is important and useful to distinguish between perceptual encoding that throws out data in the audible range, and downsampling that preserves 100% of sample encoding data in the audible range.

Link to comment
1 hour ago, jabbr said:

 

This is the same logic that JPEG uses to provide lossy compression yet preserving “relevant” image detail. In any case lossy compression is lossy compression no matter how good it is — we can equally argue that MPEG 320 is audibly indistinguishable to a majority of humans from redbook CD.

 

No, this is very precisely not the same as JPEG compression. JPEG compression is the visual equivalent of mp3 compression - JPEG compression, when done at middling or low quality settings, will reveal visually detectable artifacts and degradation in the image, just as mp3 will reveal audible artifacts and degradation at lower sample rates.

 

An image compression scheme that would be analogous to CD would be one that discards infrared and ultraviolet light information from a raw digital image capture, but retains all the image-capture information within the human visual range.

 

(And if you take any lossless audio file and losslessly compress it as with FLAC or ALAC, that's analogous to a PNG image, which is basically TIFF with a lossless compression algorithm applied to it.)

Link to comment
27 minutes ago, mansr said:

PCM is lossless in the sense that it captures perfectly any signal within the well-defined bounds determined by sample rate and bit depth. This is in stark contrast to perceptual coding systems that discard a little bit here and a little bit there according to their psychoacoustic models of what is or isn't audible.

Yes, and hopefully this will put an end to the raging semantic argument.

Main System: QNAP TS-451+ NAS > Silent Angel Bonn N8 > Sonore opticalModule Deluxe v2 > Corning SMF with Finisar FTLF1318P3BTL SFPs > Uptone EtherREGEN > exaSound PlayPoint and e32 Mk-II DAC > Meitner MTR-101 Plus monoblocks > Bamberg S5-MTM sealed standmount speakers. 

Crown XLi 1500 powering  AV123 Rocket UFW10 stereo subwoofers

Upgraded power on all switches, renderer and DAC. 

 

Link to comment
3 minutes ago, tmtomh said:

f perfect reconstruction (minus quantization error, of course) is the definition of lossless, then a 44.1k CD source made from a 176.4k original will be lossless in the audible range.

 

The real point however is the mechanism by which the 24/192 data stream is converted to 16/44 ... using a sample rate converter! Are these all the same? Are they all of the same quality?

 

Hell no!

 

Please don't suggest that all sample rate converters produce the exact same result....

 

1 hour ago, mansr said:

PCM is lossless in the sense that it captures perfectly any signal within the well-defined bounds determined by sample rate and bit depth. This is in stark contrast to perceptual coding systems that discard a little bit here and a little bit there according to their psychoacoustic models of what is or isn't audible.

 

See above. PCM *can* be lossless when use to hold the original data. As you very well know, it's the process that converts one PCM format to another that is lossy. Perhaps by y'all definition a perfect sample rate converter would produce lossless CD but I still prefer to have the source audio in as close to the mastering format as I am able to obtain, thank you.

Custom room treatments for headphone users.

Link to comment
On 6/9/2019 at 8:46 PM, esldude said:

Prokofiev’s “Lt. Kije”(Fritz Reiner, Chicago Symphony on RCA Victor)

 

Yes this is what is known as a good recording.  I only have it in the CD, but it is excellent.  

 

 

I hope your CD is a JVC XRCD. Then you would hear how truly great that 1956 Lewis Leyton-made ( I believe) recording really is! It was made before multitrack was used for most commercial recordings. The master was half-track, three mike only, stereo on 1/4-inch tape recorded at either 15 or 30 ips (RCA Victor used both in the early “Golden Age” of stereo and I don’t know which was used for this recording). It sounds stupendous!

 

George

Link to comment
6 hours ago, audiobomber said:

CD's 96dB, seems like it should be enough, yet my experience is that, all else being equal, more bits sounds better. DSD resolution is equivalent to 20 bits in PCM, but DSD recording sometimes imposes a softness that I don't like.

 

My experiences run counter to just about everyone else's - over 3 decades ago I managed to optimise CD replay so that "everything made sense" in the replay - at the time everyone was jumping up and down, bellowing about how bad CDs were in conveying the 'musicality', etc of the performance - and I knew this was total nonsense. because of what I had heard from my own rig.

 

It was easy to understand why people formed that opinion; the standard was uniformly mediocre whenever I listened to a supposedly high end system, irrespective of cost - almost no-one was making the effort to further 'debug' their setups to the point where the sound snapped into shape.

 

6 hours ago, audiobomber said:

When first exploring computer audio, I downloaded a test that provided some familiar tunes, reproduced at 16/44, 24/48, 24/96 and 24/192. The biggest difference was between 16/44 and 24/48. Nevertheless, I believe the quality of the recording far outweighs the importance of 16 vs. 24 bits or sample rates beyond 44.1kHz.

 

I believe that oversampling or transcoding to a higher sample rate is beneficial, as it allows for a more gradual aliasing filter. I prefer to do this in exact multiples of the original signal to avoid interpolation. For 16/44, I use 176.4kHz (4X), or 5.6MHz DSD (128X). My Audiolab CD/DAC/preamp performs integer oversampling and upsamples up to 84.672MHz and 32 bits

 

The vastly most important factor is optimising the replay chain - doing that makes all the number games carry on evaporate completely, as being relevant.

Link to comment
6 hours ago, tmtomh said:

Of course, if you reduce bit-depth, something is lost - but again, the result there simply is an increased noise floor from higher quantization error. This is the hardest part of digital sampling theory for our brains to grasp: "errors" or "missing/lost" stuff in digital sampling manifests simply as the noise floor, not as distortion or "gaps" or "stairsteps" in the waveform.

 

 

 

 

 

 

I've tried a few exercises of reducing 16 bit audio to 8 bits - oh, the horror!! - with nothing else changing, just making sure the best subjectively pleasant dither was applied - and it was just like listening to exactly the same track, in a somewhat noisy environment. I didn't feel that I had lost anything of the music; no wheels fell off in the process ... :).

Link to comment
4 minutes ago, fas42 said:

 

I've tried a few exercises of reducing 16 bit audio to 8 bits - oh, the horror!! - with nothing else changing, just making sure the best subjectively pleasant dither was applied - and it was just like listening to exactly the same track, in a somewhat noisy environment. I didn't feel that I had lost anything of the music; no wheels fell off in the process ... :).

Would be nice to post a few of those dithered 8 bit files for people to hear.  Just a snippet of 30 seconds or so.  I've done the same thing and it can remind you of tape hiss with the right kind of dither. 

And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. 

Link to comment
5 hours ago, Paul R said:

Just a question - does your Benchmark internally upsample every input? If so, that is probably the biggest reason why the two files sound alike.

 Paul

 My old highly modified Musical Fidelity X-DAC V3 that I use with the PC only, upsamples everything to 24/192, yet I have no problem whatsoever with well mastered recordings, hearing obvious differences between the same music in 16/44.1 or 24/96 .

 I did however find that fitting a  .1PPM TCXO with an improved power supply for it, instead of the original 24.576MHZ oscillator noticeably improved the 24/96 version with little difference heard with the 16/44.1 versions.

 Perhaps high res material needs more precise low noise clocking to get the best from it ?

Kind  Regards

Alex

 

How a Digital Audio file sounds, or a Digital Video file looks, is governed to a large extent by the Power Supply area. All that Identical Checksums gives is the possibility of REGENERATING the file to close to that of the original file.

PROFILE UPDATED 13-11-2020

Link to comment
3 hours ago, jabbr said:

 

The real point however is the mechanism by which the 24/192 data stream is converted to 16/44 ... using a sample rate converter! Are these all the same? Are they all of the same quality?

 

Hell no!

 

Please don't suggest that all sample rate converters produce the exact same result....

 

 

See above. PCM *can* be lossless when use to hold the original data. As you very well know, it's the process that converts one PCM format to another that is lossy. Perhaps by y'all definition a perfect sample rate converter would produce lossless CD but I still prefer to have the source audio in as close to the mastering format as I am able to obtain, thank you.

Agreed, I recently tried to downsample a 24/192 file to 16/44 and couldn't without ruining the SQ. 

Link to comment
6 hours ago, sandyk said:

 Paul

 My old highly modified Musical Fidelity X-DAC V3 that I use with the PC only, upsamples everything to 24/192, yet I have no problem whatsoever with well mastered recordings, hearing obvious differences between the same music in 16/44.1 or 24/96 .

 I did however find that fitting a  .1PPM TCXO with an improved power supply for it, instead of the original 24.576MHZ oscillator noticeably improved the 24/96 version with little difference heard with the 16/44.1 versions.

 Perhaps high res material needs more precise low noise clocking to get the best from it ?

Kind  Regards

Alex

 

That is one of the biggest points Alex. It is why on older DACs, a lot of times 88.2k or 96k sounds way better than 176.4k or 192k. 

 

Anyone who considers protocol unimportant has never dealt with a cat DAC.

Robert A. Heinlein

Link to comment
5 hours ago, Rexp said:

Agreed, I recently tried to downsample a 24/192 file to 16/44 and couldn't without ruining the SQ. 

How did you go about doing the down sampling?  Software and if so which one and what settings?

And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. 

Link to comment

 

6 hours ago, Rexp said:

Agreed, I recently tried to downsample a 24/192 file to 16/44 and couldn't without ruining the SQ. 

 

Depends upon what you mean by "ruining the SQ" - if you merely played the resultant 16/44 using the normal method for playing that format you most likely are just showing that the replay chain behaves differently, depending upon format.

 

If I were doing such an experiment, I would use a high quality downsampling, such as that used in Audacity; and go the further steps of resampling the output 16/44 file back up 24/192 again. Using a simple comparison step, I could show that the only difference between the original and this new 24/192 was ultrasonic noise, plus a few tidbits of genuine music-related content above 20kHz, now and again. A final step could be to extract the ultrasonic content from the original file, carefully attenuate any non-noise content so that it never rises above the level of the surrounding noise, and add that to the 2nd 24/192 file, to create a 3rd 24/192 file.

 

So now we have the original 24/192 file; another with identical audio content below 20k, with no ultrasonic content; and a 3rd with something like almost pure ultrasonic noise, derived from the original. Now, in listening to these 3 using any sort of ABC procedure, can you pick them apart?

Link to comment
22 minutes ago, fas42 said:

So now we have the original 24/192 file; another with identical audio content below 20k, with no ultrasonic content; and a 3rd with something like almost pure ultrasonic noise, derived from the original. Now, in listening to these 3 using any sort of ABC procedure, can you pick them apart?

 

Many recent high res recordings can have genuine high frequency musical content to >55kHz, even modern high quality Vinyl recordings too, apparently are capable of an extended frequency response. given the use of suitable high quality Phono cartridges.

 

 Just because we are currently unable to provide a definitive and unchallengeable explanation as to why many people are able to appreciate the audible improvement that the high res formats can make, does NOT mean that they are imagining the improvements !!!

Besides which,whether you like it or not, every format conversion like you are suggesting does result in some audible degradation, and is NOT a LOSSLESS process like many may wish to believe. Even doing the conversions  with different S/W such as Audacity or Sound Forge will result in small audible differences between the conversions.

This even applies to different S/W performing a simple conversion from 24/192 .aiff to 24/192 .wav files.:o

 

How a Digital Audio file sounds, or a Digital Video file looks, is governed to a large extent by the Power Supply area. All that Identical Checksums gives is the possibility of REGENERATING the file to close to that of the original file.

PROFILE UPDATED 13-11-2020

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...