Octave Music Don Grusin High Resolution Music Analysis (Video)

[Billy Budapest]

Sure it had it's days too, but in my experience not as widespread, I may be wrong and it may have been different in other parts of the world.Maybe what minidisc had that DAT had not is a little bit of a crossover in the consumer world, but the first IPod completely annihilated that in 2001.

Do you remember the Diamond Rio MP3 players? I had one before the batter compartment broke, thereby breaking the rest of the player. They came out way before the iPod, had their time in the spotlight, and now are but a memory.

 

[CtheArgie]

Those are the dsd128 and dsd256 I post in this thread already!

@amirm , I don't think so. I think you may have misunderstood me (or I missed some posts, which I apologize if that is the case). You showed that Octave DSD when transformed to PCM keeps the DSD noise, unless you get the 16/44 version, for obvious reasons. You showed this is not correct. I wanted to know if Blue Coast, who are the other enthusiastic supporters of DSD do the same thing when they transform their files from DSD to PCM.

 

[krabapple]

At the risk deepening this argument...
I was looking at Steely Dan, West of Hollywood, Two Against Nature that I got from a high definition music site.... It has the same issue as this recording, but it also was mastered too loud.

But the thing is....it sounds fabulous (the surround version I mean...have barely ever listened to the stereo).

(and yes, I've looked at its surround waveforms too---it's loud)

 

[krabapple]

That has nothing to do with price of tea in china. Same is true of PCM.

Look, this is very simple: users of SACD were not told that they were being fed boat load of noise. Comment after comment in my video shows people being surprised that this noise exists at all.

Which I frankly find amazing, if we assume users here have been interested in audio for years.

The fact that SACDs have scads of ultrasonic noise content has been written about (and illustrated with graphs) for decades, even in the 'hi end' rags that promote DSD as the superior alternative to PCM.

Same with the fact that 'hi rez' recordings in any format typically have scant musical content above the audible range.

 

[amirm]

@amirm , I don't think so. I think you may have misunderstood me (or I missed some posts, which I apologize if that is the case). You showed that Octave DSD when transformed to PCM keeps the DSD noise, unless you get the 16/44 version, for obvious reasons. You showed this is not correct. I wanted to know if Blue Coast, who are the other enthusiastic supporters of DSD do the same thing when they transform their files from DSD to PCM.

I did understand your question. As i noted, I already post Bluecost tracks. Here is the DSD128 again:

Look on top and see the title of the track: "Blue Coast Collection...."

 

[KSTR]

There is no hope in what I said. Muscscope I used to analyze dsd files closed shop. Should it stop working I am screwed with respect to doing that analysis

There are ways... I used sox-dsd to convert a DSD256 file from 2L Recordings to copy it 1:1 into a 16bit .WAV (which then has a 11,289,600Hz sample rate), with -3dB gain reduction. The result is the original bit-stream as PCM-data which then can be analyzed as usual. To help things, I rescaled sample rate to 1/1000, giving a 11,290 sample rate to avoid display limits of REW.

The spectrum of this raw modulator bit stream then can be fully displayed to fs/2 (5,644,000Hz) with REW -- blue trace. We can see that it has 100kHz of noise-free bandwidth, just as @Miska has stated for DSD256.

Red trace is what the result would be if the output of a logic gate buffer (representing the bit stream in a conceptual hardware) were low-passed with an analog 3rd-order Butterworth filter at 50kHz. The RF noise reduction is tremendous (note that this is a bit hidden with a log frequency scale, linear would show it much better). An analog filter is truly essential here.

Even with that crude approach the noise peak is now at 270kHz and has comparable energy like typical content around 20kHz... low enough to not cause any problems downstream. The analog filter implementation is really tricky, though, with those horribly fast signal edges directly from the digital logic gate... that's why actual solutions can be much different and may show different final analog spectra.

 

[krabapple]

Yeah it bugs me when this happens. Perhaps they are trying to hit the 96 and 192 buzzwords. Or on DVD itself, there was no 88/176, only DVD-A, but audio-only DVDs were ever only a tiny sliver so I doubt anyone is thinking about that any more.

48kHz being the standard in video world, 96 and 192 are the 'high rez' integer multiples of that.

And DVD and its descendants have always been much more popular than DVDA or SACD.

 

[Geert]

There is no way to perform transforms of any sort on a 1 bit format without it overflowing

Great. Confirms how much time I wasted here since yesterday discussing this. Roon has a knowledge base artikel on it, https://help.roonlabs.com/portal/en..._Features_often_require_DSD_to_PCM_conversion, which confirms it implements DSD processing via conversion to DXD, a common approach. DXD is 24 bit 352.8 kHz PCM. This principle is applied when doing volume control incl. crossfades, parametric eq and dsp.

DRC via convolution is done via a different process https://community.roonlabs.com/t/native-dsd-processing/24286/2?u=John, which involves the use of “64-bit DSD” (multi-bit delta sigma) and is what you referred to. According to Roon: “64-bit DSD” uses the same physical representation as “64-bit PCM”. It's what I referred to as "multibit" from the start.

 

[krabapple]

Again, b/c there are people prefer the sound of DSD, at least on their systems. And it works well for multichannel on an SACD. There are also very good DSD masters that are transfers from tape. For whatever reason, it seems to be especially good at that.

Deutsche Gramophone did a transfer comparison years ago and went with high rez PCM; this from a source at Emil Berliner Studios, in a discussion of the philosophy behind their early hi rez multichannel releases (specifically the Kleiber Beethoven 5/7 IIRC). (DG did eventually do releases on SACD too)

 

[krabapple]

Size of DSD64 file is same as 88.2/32 PCM file. It is not particularly large, and I haven't seen any particular issues with the file sizes, especially taking the benefits into account.

88/32. Yeah I use that format all the time.

 

[krabapple]

Yes that's was my original asumption, I do know that WM8742 reference design and development boards have such a filter, but then Amir tested a couple dacs and apparently nothing was done to remove these ultrasonics. So I guess all bets are off. I think in this particular case what was suggested was a 1st or second order simple filter, relatively slow. so I guess more in "protection" of the tweeter purpose than actual getting rid of all ultrasonics. I have no opinion on how important this is for fidelity, or if it matter at all there is obviously a long debate in these pages.

I don't know that the 'required' 50 kHz analog LPF in question was properly considered part of the *DAC* itself, versus a downstream component of the hardware. Perhaps someone more savvy about SACD player design can tell us.

Whatever Amir was playing his SACDs on in this thread, they don't appear to have used such a filter.

 

[Grooved]

$30k amp, $30k speaker, $10k cable, $30k processor. These audiophile won't use DRC unless for AV use. These people will fly, travel and listen before they make such purchase. People using 300B, gigantic horn speaker won't use DRC. DRC is heavily promoted in here, but outside no.

And if they attend a concert of one of their favorite artists, will they ask that the venue system is not corrected?

 

[Kal Rubinson]

No more 'interesting' that the same music recorded in PCM.

If it is. Yes, I know that there is, in general, more available on PCM. OTOH, there's a lot on DSD that is, still, not on PCM and vice-versa. So, I choose the music first and that determines what recording format I buy. That's why I refuse to take a side in this dispute.

 

[kongwee]

And if they attend a concert of one of their favorite artists, will they ask that the venue system is not corrected?

Pretty sure, I don't understand your analogy, I am dump. care to explain?

 

[kongwee]

There is no way to perform transforms of any sort on a 1 bit format without it overflowing. So what Roon does is first convert the 1-bit format to 64 bits. This then becomes PCM data. Transforms are performed on this format. Once done, a DSD modulator is used to convert the signal back to 1 bit. This means that you are likely to get double quantization noise than what you started with.

Are you trying to say DSD file is just a string on 1 and 0 and no way to map a waveform, hence nothing can be done digitally? Unlike PCM can capture in frame and edit?

 

[PeteL]

I don't know that the 'required' 50 kHz analog LPF in question was properly considered part of the *DAC* itself, versus a downstream component of the hardware. Perhaps someone more savvy about SACD player design can tell us.

Whatever Amir was playing his SACDs on in this thread, they don't appear to have used such a filter.

It’s just a recommendation, yes upstream of the DAC chip, but normally DAC chip manufacturers will suggest a « reference design » to help the designer, along with a development board based on that, it’s apretty much a must, of course you can always be creative and redesign the output stage.

 

[Grooved]

Pretty sure, I don't understand your analogy, I am dump. care to explain?

I would better bet that they wouldn't ask because they don't even think that accept a full acoustic concert, every one with even the minimum sound reinforcement will be done with a system that is corrected to the "room".
I've never work or attend a single concert where the system is not corrected, in any venue, even in auditoriums.
Which, if I'm not wrong, turn into a complete joke the fact that an "audiophile" listener who likes an artist music can spend $100k on a system but absolutely wants no correction...
If they want to listening it as close as how it was produced, then they can take their track and go to the studio where it was created, and guess what, the system will be corrected

 

[kongwee]

I would better bet that they wouldn't ask because they don't even think that accept a full acoustic concert, every one with even the minimum sound reinforcement will be done with a system that is corrected to the "room".
I've never work or attend a single concert where the system is not corrected, in any venue, even in auditoriums.
Which, if I'm not wrong, turn into a complete joke the fact that an "audiophile" listener who likes an artist music can spend $100k on a system but absolutely wants no correction...
If they want to listening it as close as how it was produced, then they can take their track and go to the studio where it was created, and guess what, the system will be corrected

I attend concert and work in backstage and control room too. Orchestra, ensemble and other acoustic, they do need the sound reinforcement at time. It is about sound projection. There is a difference between good and bad venue. There is VIP seats where project the sound best. Normal seat doesn't have the sound projection like VIP seat. It just can't be done. What does sound projection have to do with room correction?

For me, if possible I do not want a $100k audiophile system to do any correction too. Counter by placement or a bit of treatment. This is for music listening. For AV, you need DRC. Fly and travel to many audiophile show. You will understand why audiophile avoiding DRC. They get the enjoyment from the system, not from measurement. That DSD survive in audiophile community.

 

[Grooved]

I attend concert and work in backstage and control room too. Orchestra, ensemble and other acoustic, they do need the sound reinforcement at time. It is about sound projection. There is a difference between good and bad venue. There is VIP seats where project the sound best. Normal seat doesn't have the sound projection like VIP seat. It just can't be done. What does sound projection have to do with room correction?

For me, if possible I do not want a $100k audiophile system to do any correction too. Counter by placement or a bit of treatment. This is for music listening. For AV, you need DRC. Fly and travel to many audiophile show. You will understand why audiophile avoiding DRC. They get the enjoyment from the system, not from measurement. That DSD survive in audiophile community.

You have a point that some seats would not need correction like most of the other, I've experience that. And I supposed that you may certainly find this best seat in your listening room, I did a bit too much of generalizing the case
But once in your best seat in your listening room, you're listening to speakers, not the acoustic projections of the orchestra. I'm not saying that you can't enjoy that in a good room, good system and the right setup, without DRC, just that there's great chances that you can improve that with correction, and even with an analog EQ if you want to avoid digital correction.
This what I find a joke, spending already that money and not wanting to add an analog EQ that can you help correct the small problems you can have in your room.
Don't get me wrong, I already enjoyed listening to a system like that without correction, just that I heard on tracks I knew very well that a small correction would have made it even better.

 

[Miska]

Even with that crude approach the noise peak is now at 270kHz and has comparable energy like typical content around 20kHz... low enough to not cause any problems downstream. The analog filter implementation is really tricky, though, with those horribly fast signal edges directly from the digital logic gate... that's why actual solutions can be much different and may show different final analog spectra.

And you can of course adjust the modulator noise slope so the it doesn't have that kind of sharp knee.

So it always makes sense to check the DSP + hardware combination in action. That's why I check with different bandwidths using different analyzers.

Same of course applies even if one uses stock DAC chip (without external DSP) with one's own analog stage.