I really like Paul. His stories especially the ones about Arnie are just nice :) (we still have a pair of IRS betas, gammas as well as Kappa 9 (together with ML and Threshold amps) But there are sooo many half true stories when it comes to the production side and to digital topics, that it really hurts me sometimes. Just to explain some background. We are a studio and were several times nominated for the best classic recording for the international classic music award. We have a Steinway D, a Bechstein D piano, a lot of very fine instruments, one of the largest mic collection in Europe and a lot of high end equipment. The first time I really was shocked about a story here in this channel was, when Paul talked about studio gear that is not on par with his hifi listening equipment and that his is the reason why his company wants to create their own preamps... (please look at gear from millennia, forssell, grace and many more... which is at the highest level...) But here we can see a lack of understanding of digital technologies as well. AD and DA converters have come a long way. However already in the 90s we had extremely great sounding converters (think about the Prism ADA 8 XR, or the Lavry gold and quite a couple more...) Today ADDA converters can be completely transparent. Half a year ago I made an official bet in a the largest recording forum: If anybody would hear a difference in a ABX test (double blind AB test) between the original and the converted signal I would donate my 10.000 Euros prism converters. Nobody (professional audio engineers and musicians with absolute pitch) could do it (at least not with a statistical significant result). Today it is possible to hear sometimes different filters on the converters (especially with processed signals because of phase issues when adding operations with the different phase response in the same frequency area, or when using linear phase filters; where preringing can happen (although unlikely especially in this frequency range). Depending on the design of the converter we also can hear the analog section. Now let's take a look as DSD. The file format which is called dsf is in no way better than PCM. In fact it comes with some real shortcomings. Due to its 1 Bit resolution the quantization noise is so bad that we have to use have noise shaping, that effects the original music. This might lead to a difference to the pcm signals but it's not closer to the original. This whole thing reminds me about analog records. As beautiful it is and as much some of us like the sound it is by far not closer to the original than let's say a cd. Cross talk is horrible, dynamic range is pretty bad, and since the record cannot store stereo bass (due to its ms format), bass has to come from the center and has to be equalized with non standardized RIAA eq... but that's another topic). Anyways: dsf was created in a time when the converter chips were real 1 Bit designs. Modern chips like the AKM 4493 or the Sabre ES9028Q2M are nothing like this. They produce 8 bit data streams. And there is a rising fan base of R2R (Multibit converters). Most DSD files are anyways fake because it not possible to edit the music signal in a way that is possible with PCM. So what happens is, that most music is recorded in PCM, then edited and processed and then either converted in the digital domain to dsf or converted to analog and then re-recorded in dsf... At the end: what do I want to say: Listen to the music. The art of high quality recordings starts with the musicians, then with the instruments and the mics, then the preamps and not really with der converters (at least when you are working with professional gear) and for sure not with file format. Already the good old CD format gives us more dynamics and frequency response than what we as listeners need (by the way: this is indeed different on the production side. Here we need the extended frequency response (to avoid aliasing for non linear processing) and the extended dynamics. But ones mixing and mastering is finished, it is not a question of file formats) Again I invite everybody to listen a 786 kHz recording in different sample rates. If the sample rate conversion is done the right way, you are not going to hear it :) ) I hear already a lot of people saying..."no I heard it already) But what you really heard is a different mix or a sloppy mastering engineer. So guys: Don't spend to much time on this topic and enjoy great music, with great sound :-)
Excellent summary and a breath of fresh air. I have this conversation sooooooo many times, I need a recording of it (in any format ;-) ). MIcrophone= analog, mic cable= analog, preamp= analog, next digital conversion, period. Edited, mixed and mastered in the digital format. Only converted while doing it so we can hear the change made, at that time, but still stored in the digital format. 99.9 % of the recordings.
A few days ago I notice they are uploading repeated Paul videos. I miss him, I need new teachings. Greetings from Uruguay!! Paul, you are a great master, thank you very much for sharing your knowledge.
Oh my.... I too was caught up with DSD rates and high resolution PCM's.. I even bought tracks at DSD512 (HUGE FILES) and then realize I can't tell much sonic difference between DSD512 and DSD64. I realize I was losing the objective of this whole exercise which is to enjoy the music !!!
Not just "much" difference. You couldn't tell any difference at all because there is none. Dsd just adds a ton of noise past the human hearing of 22khz. Its literally the same exact music plus noise. It's a scam
I downloaded from NativeDSD Vanessa Fernandez - When the Levee Breaks at DSD 256. Under additional information, They state they re-modulate the original DSD 64 that Bernie Grundman engineered from the analog tapes. Which sounds wonderful on my Holoaudio KTE May DAC that plays native DSD files with out converting to PCM
Love to put all the commenters in a room and watch them argue for hours. NO! I'm correct. NO! This is how it works, not like that. I'll be at the bar enjoying my favorite drinks. Oh, and chuckling the whole time. 😁
Let my DAC get the native master stream and do the optimum conversion to analog. Don't mess up the stream by any transcoding. If the music is mastered digitally, please let me get the PCM output that comes from it rather than a lossy DSD version.
@@TheDanEdwards Of course it is lossy. ANY digitally mastered music is doing all processing in PCM and the output is PCM. When you transcode it to DSD you are recreating the sound waves with a different modulation of the native master PCM stream and will introduce a lot of ultrasonic noise and some quantization errors. For example, if your master file is DXD natively 352.8kHz 24 bits and you convert it to DSD, the stream will be closer to the quality of 96kHz 20 bits PCM. For digital mastering, PCM rules and DSD makes no sense other than as initial recording format per track before mixing (which already needs PCM).
@@martinjohnson2381 for the increase in resolution from WAV to FLAC to DSD, I find the weight and punchiness of the sound goes the other way.... DSD sounded much thinner than WAV
Mp3s have lossless and lossy compresions both at the same time ...the compression is 90%(most lossy) and 75%(less lossy) compared to CD bitrate for 128kbits/s and 320kbits/s respectively
There will be some commenters simply triggered by "DSD" in the title, and by the word "resolution". On the latter I wish folk in audio-land would be more precise with their terminology. The English word "resolution" has been used in many contexts, even within audio, and thus the word has acquired some variations (in the denotations) and certainly connotations ( _especially_ in audiophilia) are all over the place. _Bit depth_ and _total information content_ are concepts that need to be grasped to deal with debates over DSD vs. PCM vs. whatever. The exact same information can be represented in different methods. Furthermore, said information can be stored in different (digital file) formats. Whether one decides to use lossy compression or not is simply a choice made for the purposes of bandwidth (for transmission or storage) or for marketing (to protect a higher quality original from being pirated.) Furthermore, in acoustics the concept of _resolution_ is simply a wave phenomenon, and one can speak of the resolution of a room, for example. For those of us interested in music, *the room resolution along with the transducer (loudspeakers) resolution dominate the listening experience.* Vastly dominates the listening experience. Given that, I suspect that outside of wearing the very best electrostatic headphones (driven by the very best headphone amps), differences in audio information formats are beyond my listening enjoyment threshold.
MP3 music is usually 16/44.1 regardless of the bitrate. Unless it's specifically ultra-compressed to 32kHz, 22.05kHz, 16kHz, or 8kHz, but that can be done just the same with other PCM-based standards.
I use a very high end streamer these days and invariably older tracks are very poorly transcribed from vinyl to digital format. Some sound pretty dreadful as though the engineers have tinkered with the RIAA and lost all the punch and dynamism to the extent that I expect to be dissapointed these days. Also, different streaming platforms sound different and even the SAME tracks on Tidal from two different albums such as the original Steely Dan track from the album "Can't buy a thrill" and the same track on a compliation album sound different. The track on the compliation sounds so much better. I do love the old Hammond organ sound you see! If you want the best sound from vinyl then a good deck and pickup is still the way to go as has been demonstrated to me numerous times by my local dealer. Modern recordings not so much but a back catalogue of older classics will disappoint you on Tidal almost every time ...... Will make noise, but a long way short of good Hi-Fi. Just my tuppenceworth from a 67yr old in Scotland!
@1:38 "People would say: 'If you upsample a 44.1 (kHz) CD, there's no more information. So why does it sound better?" The premise is wrong, because there is more information. The reason an upsampled file (a CD is simply a storage medium for the music file) has more data is because the upsampling process creates more samples. The process does so based on the existing samples. When you upsample a 44.1 kHz file to an 88.2 kHz file, your computer manufacturers an additional 44,100 samples. The upsampling code looks at the adjacent samples, and creates a sample that is the average of the two adjacent samples. Imagine watching a movie, frame by frame. You can see the jumping of the image to each subsequent frame. Now imagine an artist being asked to draw what would be between two of those images. A skilled artist would create such an image that nicely fits between the two adjacent images. Well, a computer does that with audio samples. I do not know why Paul says that there is no more information. He has said that in several videos. There is no more "original" information. But there is more "artificially created" information. In fact, if you upsample a 44.1 kHz file to 88.2 kHz, and then look at the file size of the original file and the newly created upsampled file, you will see that the newly created upsampled file is twice the size. If you use audio editing software that allows you to zoom in and edit individual samples, that software will show you that the upsampled file has twice the number of samples. And each of those additional samples has "information". That information is real. It was created artificially. But those additional samples are real and contain real information. With the above understanding, then let us address why the upsampled file sounds "better": "Better" is subjective. I have listened to a Cary CD player. I believe that it was the 303/300 model (or something similar). It had on-the-fly upsampling (also called oversampling). It could go up to 768 kHz, and it sounded better (to my ear). It was a pleasing sound. But it was not a natural sound. Some folks cannot turn up the bass knob enough. To them, that sounds "better". But it is not natural. Upsampling has its appeal, because it often does make that song sound better, depending on your subjective opinion of what is better. But if the recording was made at that higher sampling rate from the get-go, then that would sound the best, because upsampling can only approximate accurate creations of samples. Whereas, when you record at the higher sampling rate, you are simply creating those samples as they really are. What Paul said about the filters needed for different sampling rates, and how it affects sound quality, is true. But what he said about "no more information" in upsampled files is simply wrong. And if someone in Paul's inner circle is feeding him that wrong information, it begs the question of should Paul be taking their word for the "compression" and "limiting" that he allows them to apply to Octave recordings? That "compression" might sound better, but would the recording sound even better if the compression was not needed in the first place -- if the recording was re-done in a way that captured the dynamics better from the get-go? Is someone feeding Paul an excuse to not have to re-do the recording? I have no doubt that the compression "fix" helps. But is that as good as really nailing down the recording so that the compression fix is not needed? Adding the compression box to the signal chain should be avoided whenever possible. No additional boxes should ever be introduced into the signal chain, even if their settings are all on neutral. Each one acts as an active pre-amp. Each one colors the sound. The coloration is cumulative. And that is why compression should be plan "B" (never plan "A"), and used when no other options are available. Due to Paul's passion for, and understand of, great sound quality, his Octave releases can withstand the compression. Like just taking a shower, stepping out, and getting a pinch of dirt on you. You will still be very clean. But avoiding that compression, entirely, would be best. Paul's closing explanation about the original DSD capture, followed by the mixing, etc, is spot-on. But that is different than his @1:38 remark about no more information.
DSD does not add information compared to the original sound. Adding artificial samples is not the same as adding information to the recording. You could call it removal or distortion of the recording.
@@EnigmaVideos "DSD does not add information compared to the original sound." Are you implying that I stated that DSD does add information compared to the original sound? I am asking, because I did not write that. I put the time stamp to which my comment referred. I also wrote that Paul was correct about his DSD explanation. "Adding artificial samples is not the same as adding information to the recording." Adding is not adding? "You could call it removal or distortion of the recording." Removal of the recording? What does that mean? Distortion of the recording? Yes. Upsampling is 100% pure distortion. Not in the sense of what many people envision (like hard to understand, noisy, unclear, etc). Anything other than the actual recorded sound is, technically, distortion, even when it sounds great. Each artificially created sample is, technically, 100% distortion.
@@NoEgg4u "Are you implying that I stated that DSD does add information compared to the original sound? " No, I think my message is implying what was most likely meant in the video. "Anything other than the actual recorded sound is, technically, distortion, even when it sounds great." So one could call it removal of the authentic recording.
@@EnigmaVideos DSD does add information compared to the original sound. In fact, DSD adds a considerable amount of noise in order to dither the samples. That is why DSD playback requires a low pass filter and why DSD 64 has a lower signal to noise ratio in the upper frequencies.
Paul is correct that there is no more information. The "brick wall filtering" that he mentions is used at 44.1 kHz to prevent aliasing and is applied at the input of the recording device. That means that the recording device does not record any audio signal above the filter limit and therefore there is no further information for an up sample to create.
I really like Paul. His stories especially the ones about Arnie are just nice :) (we still have a pair of IRS betas, gammas as well as Kappa 9 (together with ML and Threshold amps)
But there are sooo many half true stories when it comes to the production side and to digital topics, that it really hurts me sometimes. Just to explain some background. We are a studio and were several times nominated for the best classic recording for the international classic music award. We have a Steinway D, a Bechstein D piano, a lot of very fine instruments, one of the largest mic collection in Europe and a lot of high end equipment. The first time I really was shocked about a story here in this channel was, when Paul talked about studio gear that is not on par with his hifi listening equipment and that his is the reason why his company wants to create their own preamps... (please look at gear from millennia, forssell, grace and many more... which is at the highest level...)
But here we can see a lack of understanding of digital technologies as well. AD and DA converters have come a long way. However already in the 90s we had extremely great sounding converters (think about the Prism ADA 8 XR, or the Lavry gold and quite a couple more...) Today ADDA converters can be completely transparent. Half a year ago I made an official bet in a the largest recording forum: If anybody would hear a difference in a ABX test (double blind AB test) between the original and the converted signal I would donate my 10.000 Euros prism converters. Nobody (professional audio engineers and musicians with absolute pitch) could do it (at least not with a statistical significant result). Today it is possible to hear sometimes different filters on the converters (especially with processed signals because of phase issues when adding operations with the different phase response in the same frequency area, or when using linear phase filters; where preringing can happen (although unlikely especially in this frequency range). Depending on the design of the converter we also can hear the analog section.
Now let's take a look as DSD. The file format which is called dsf is in no way better than PCM. In fact it comes with some real shortcomings. Due to its 1 Bit resolution the quantization noise is so bad that we have to use have noise shaping, that effects the original music. This might lead to a difference to the pcm signals but it's not closer to the original. This whole thing reminds me about analog records. As beautiful it is and as much some of us like the sound it is by far not closer to the original than let's say a cd. Cross talk is horrible, dynamic range is pretty bad, and since the record cannot store stereo bass (due to its ms format), bass has to come from the center and has to be equalized with non standardized RIAA eq... but that's another topic).
Anyways: dsf was created in a time when the converter chips were real 1 Bit designs. Modern chips like the AKM 4493 or the Sabre ES9028Q2M are nothing like this. They produce 8 bit data streams. And there is a rising fan base of R2R (Multibit converters). Most DSD files are anyways fake because it not possible to edit the music signal in a way that is possible with PCM. So what happens is, that most music is recorded in PCM, then edited and processed and then either converted in the digital domain to dsf or converted to analog and then re-recorded in dsf...
At the end: what do I want to say: Listen to the music. The art of high quality recordings starts with the musicians, then with the instruments and the mics, then the preamps and not really with der converters (at least when you are working with professional gear) and for sure not with file format. Already the good old CD format gives us more dynamics and frequency response than what we as listeners need (by the way: this is indeed different on the production side. Here we need the extended frequency response (to avoid aliasing for non linear processing) and the extended dynamics. But ones mixing and mastering is finished, it is not a question of file formats) Again I invite everybody to listen a 786 kHz recording in different sample rates. If the sample rate conversion is done the right way, you are not going to hear it :) ) I hear already a lot of people saying..."no I heard it already) But what you really heard is a different mix or a sloppy mastering engineer.
So guys: Don't spend to much time on this topic and enjoy great music, with great sound :-)
Excellent summary and a breath of fresh air. I have this conversation sooooooo many times, I need a recording of it (in any format ;-) ). MIcrophone= analog, mic cable= analog, preamp= analog, next digital conversion, period. Edited, mixed and mastered in the digital format. Only converted while doing it so we can hear the change made, at that time, but still stored in the digital format. 99.9 % of the recordings.
A few days ago I notice they are uploading repeated Paul videos. I miss him, I need new teachings. Greetings from Uruguay!! Paul, you are a great master, thank you very much for sharing your knowledge.
And he didn't even play the modular Moog 😭
Oh my.... I too was caught up with DSD rates and high resolution PCM's.. I even bought tracks at DSD512 (HUGE FILES) and then realize I can't tell much sonic difference between DSD512 and DSD64. I realize I was losing the objective of this whole exercise which is to enjoy the music !!!
Not just "much" difference. You couldn't tell any difference at all because there is none. Dsd just adds a ton of noise past the human hearing of 22khz. Its literally the same exact music plus noise. It's a scam
I downloaded from NativeDSD Vanessa Fernandez - When the Levee Breaks at DSD 256. Under additional information, They state they re-modulate the original DSD 64 that Bernie Grundman engineered from the analog tapes. Which sounds wonderful on my Holoaudio KTE May DAC that plays native DSD files with out converting to PCM
DSD has inherently a huge amount of high frequency noise which must be removed and contains no added audible benefit over PCM.
Love to put all the commenters in a room and watch them argue for hours. NO! I'm correct. NO! This is how it works, not like that. I'll be at the bar enjoying my favorite drinks. Oh, and chuckling the whole time. 😁
Let my DAC get the native master stream and do the optimum conversion to analog. Don't mess up the stream by any transcoding. If the music is mastered digitally, please let me get the PCM output that comes from it rather than a lossy DSD version.
"rather than a lossy DSD version." - why do you think a DSD copy you have is lossy?
@@TheDanEdwards Of course it is lossy. ANY digitally mastered music is doing all processing in PCM and the output is PCM. When you transcode it to DSD you are recreating the sound waves with a different modulation of the native master PCM stream and will introduce a lot of ultrasonic noise and some quantization errors. For example, if your master file is DXD natively 352.8kHz 24 bits and you convert it to DSD, the stream will be closer to the quality of 96kHz 20 bits PCM. For digital mastering, PCM rules and DSD makes no sense other than as initial recording format per track before mixing (which already needs PCM).
it's quite hard for me to find a good dac to convert DSD files, usually higher khz flac sounds similar if not better than DSD files from the same dac
Maybe that's just what you prefer?
@@martinjohnson2381 for the increase in resolution from WAV to FLAC to DSD, I find the weight and punchiness of the sound goes the other way.... DSD sounded much thinner than WAV
Mp3s have lossless and lossy compresions both at the same time ...the compression is 90%(most lossy) and 75%(less lossy) compared to CD bitrate for 128kbits/s and 320kbits/s respectively
There will be some commenters simply triggered by "DSD" in the title, and by the word "resolution". On the latter I wish folk in audio-land would be more precise with their terminology. The English word "resolution" has been used in many contexts, even within audio, and thus the word has acquired some variations (in the denotations) and certainly connotations ( _especially_ in audiophilia) are all over the place.
_Bit depth_ and _total information content_ are concepts that need to be grasped to deal with debates over DSD vs. PCM vs. whatever. The exact same information can be represented in different methods. Furthermore, said information can be stored in different (digital file) formats. Whether one decides to use lossy compression or not is simply a choice made for the purposes of bandwidth (for transmission or storage) or for marketing (to protect a higher quality original from being pirated.)
Furthermore, in acoustics the concept of _resolution_ is simply a wave phenomenon, and one can speak of the resolution of a room, for example. For those of us interested in music, *the room resolution along with the transducer (loudspeakers) resolution dominate the listening experience.* Vastly dominates the listening experience. Given that, I suspect that outside of wearing the very best electrostatic headphones (driven by the very best headphone amps), differences in audio information formats are beyond my listening enjoyment threshold.
I would care about DSD if recording artists I liked used it.
@@ChiefExecutiveOrbiter
Agreed. For me it's too late. Jimi is dead, among others I enjoy.
No more “Octave Records”? 🥺
3:42
Unless I miss took Paul, I thought he was referring to changing from mixing in analog to possibly mixing in digital?
@@sickjohnson
I hope that’s what he meant. 🥺
Can we ever hear that modulator on DSD?
If high res files are 24/88 to 24/352 . . . and WAV files are 16/44 . . . what is a 128 or 352 MP3 file?
MP3 music is usually 16/44.1 regardless of the bitrate. Unless it's specifically ultra-compressed to 32kHz, 22.05kHz, 16kHz, or 8kHz, but that can be done just the same with other PCM-based standards.
I use a very high end streamer these days and invariably older tracks are very poorly transcribed from vinyl to digital format. Some sound pretty dreadful as though the engineers have tinkered with the RIAA and lost all the punch and dynamism to the extent that I expect to be dissapointed these days. Also, different streaming platforms sound different and even the SAME tracks on Tidal from two different albums such as the original Steely Dan track from the album "Can't buy a thrill" and the same track on a compliation album sound different. The track on the compliation sounds so much better. I do love the old Hammond organ sound you see! If you want the best sound from vinyl then a good deck and pickup is still the way to go as has been demonstrated to me numerous times by my local dealer. Modern recordings not so much but a back catalogue of older classics will disappoint you on Tidal almost every time ...... Will make noise, but a long way short of good Hi-Fi. Just my tuppenceworth from a 67yr old in Scotland!
Poor guy wasting his money on dsd files that are basically cd quality plus noise
You keep talking about dsd yet i have yet to see the format in person
Is that a Moog V1 behind Paul?
Moog Modular IIIP
Makes total sense
Paul, in a reply of yours all is needed you said is 100db (dsd 120db) dynamic range, Paul pcm came out with 32bit DR thats 192db DR ;))
That was excellent Paul!
@1:38 "People would say: 'If you upsample a 44.1 (kHz) CD, there's no more information. So why does it sound better?"
The premise is wrong, because there is more information.
The reason an upsampled file (a CD is simply a storage medium for the music file) has more data is because the upsampling process creates more samples. The process does so based on the existing samples.
When you upsample a 44.1 kHz file to an 88.2 kHz file, your computer manufacturers an additional 44,100 samples. The upsampling code looks at the adjacent samples, and creates a sample that is the average of the two adjacent samples.
Imagine watching a movie, frame by frame. You can see the jumping of the image to each subsequent frame.
Now imagine an artist being asked to draw what would be between two of those images. A skilled artist would create such an image that nicely fits between the two adjacent images.
Well, a computer does that with audio samples.
I do not know why Paul says that there is no more information. He has said that in several videos.
There is no more "original" information. But there is more "artificially created" information.
In fact, if you upsample a 44.1 kHz file to 88.2 kHz, and then look at the file size of the original file and the newly created upsampled file, you will see that the newly created upsampled file is twice the size.
If you use audio editing software that allows you to zoom in and edit individual samples, that software will show you that the upsampled file has twice the number of samples. And each of those additional samples has "information". That information is real. It was created artificially. But those additional samples are real and contain real information.
With the above understanding, then let us address why the upsampled file sounds "better":
"Better" is subjective. I have listened to a Cary CD player. I believe that it was the 303/300 model (or something similar). It had on-the-fly upsampling (also called oversampling). It could go up to 768 kHz, and it sounded better (to my ear). It was a pleasing sound. But it was not a natural sound.
Some folks cannot turn up the bass knob enough. To them, that sounds "better". But it is not natural.
Upsampling has its appeal, because it often does make that song sound better, depending on your subjective opinion of what is better. But if the recording was made at that higher sampling rate from the get-go, then that would sound the best, because upsampling can only approximate accurate creations of samples. Whereas, when you record at the higher sampling rate, you are simply creating those samples as they really are.
What Paul said about the filters needed for different sampling rates, and how it affects sound quality, is true. But what he said about "no more information" in upsampled files is simply wrong. And if someone in Paul's inner circle is feeding him that wrong information, it begs the question of should Paul be taking their word for the "compression" and "limiting" that he allows them to apply to Octave recordings? That "compression" might sound better, but would the recording sound even better if the compression was not needed in the first place -- if the recording was re-done in a way that captured the dynamics better from the get-go? Is someone feeding Paul an excuse to not have to re-do the recording? I have no doubt that the compression "fix" helps. But is that as good as really nailing down the recording so that the compression fix is not needed?
Adding the compression box to the signal chain should be avoided whenever possible. No additional boxes should ever be introduced into the signal chain, even if their settings are all on neutral. Each one acts as an active pre-amp. Each one colors the sound. The coloration is cumulative. And that is why compression should be plan "B" (never plan "A"), and used when no other options are available.
Due to Paul's passion for, and understand of, great sound quality, his Octave releases can withstand the compression. Like just taking a shower, stepping out, and getting a pinch of dirt on you. You will still be very clean. But avoiding that compression, entirely, would be best.
Paul's closing explanation about the original DSD capture, followed by the mixing, etc, is spot-on. But that is different than his @1:38 remark about no more information.
DSD does not add information compared to the original sound. Adding artificial samples is not the same as adding information to the recording. You could call it removal or distortion of the recording.
@@EnigmaVideos "DSD does not add information compared to the original sound."
Are you implying that I stated that DSD does add information compared to the original sound?
I am asking, because I did not write that. I put the time stamp to which my comment referred. I also wrote that Paul was correct about his DSD explanation.
"Adding artificial samples is not the same as adding information to the recording."
Adding is not adding?
"You could call it removal or distortion of the recording."
Removal of the recording?
What does that mean?
Distortion of the recording?
Yes. Upsampling is 100% pure distortion. Not in the sense of what many people envision (like hard to understand, noisy, unclear, etc). Anything other than the actual recorded sound is, technically, distortion, even when it sounds great. Each artificially created sample is, technically, 100% distortion.
@@NoEgg4u
"Are you implying that I stated that DSD does add information compared to the original sound?
"
No, I think my message is implying what was most likely meant in the video.
"Anything other than the actual recorded sound is, technically, distortion, even when it sounds great."
So one could call it removal of the authentic recording.
@@EnigmaVideos DSD does add information compared to the original sound.
In fact, DSD adds a considerable amount of noise in order to dither the samples.
That is why DSD playback requires a low pass filter and why DSD 64 has a lower signal to noise ratio in the upper frequencies.
Paul is correct that there is no more information.
The "brick wall filtering" that he mentions is used at 44.1 kHz to prevent aliasing and is applied at the input of the recording device.
That means that the recording device does not record any audio signal above the filter limit and therefore there is no further information for an up sample to create.
Great information as always Paul!
It is all convoluted. The answer is to just get PS Audio gear and it will all be good.
nice lesson for free, thanks mr paul
THIRD 🥉🏆🍾🥂👏🇳🇱
First 🤭
th-cam.com/video/gCCD40eB-cU/w-d-xo.html
Congratulations on this amazing performance you get 🥇🏆🍾🥂👏🇺🇸
@@RoderikvanReekum thank you sir💪