Octave Music Don Grusin High Resolution Music Analysis (Video)

[sam_adams]

You can still download MusicScope and use it without a license.

YOINK!

 

[Billy Budapest]

And the acoustic noise floor of the room which is probably worse than 16 bit.

Yeah, good call, background noise like fans and air conditioners and how the physics of the room itself boosts/lowers certain frequencies, etc.

 

[pinpoint_oxford]

Well, to be fair to PS Audio, Amir did not like the music selection or the price. So that was two negatives right there before looking at the DSD format failure. So the review "seems" more negative than it really should be.

I think this is important. We can argue DSD vs PCM vs high resolution vs redbook until the cows come home, but you have to like what you're listening to. The latest album they have by Scabaret (to me) is amazing from beginning to end, amazing talent by the trio of women. Also for $29 for an SACD + DVD with digital master audio files is a good deal to me.

I happen think DSD is an interesting format. Does it have a lot of noise in the high-register, yes, but we can't hear it (and doubtfully our loudspeakers could reproduce it anyway), so it likely doesn't matter.

At the end of the day, listen to and buy music you like from artists you like. I support smaller artists and dedication to mastering that think Octave records is trying to do both, whatever the claims of superior fidelity.

 

[Blumlein 88]

I meant in general, can the analog recording equipment be quiet enough to take advantage of the DR afforded by 24-bit digital recording. Not only the mixing desk, you also have to keep in mind the SNR of the microphones, microphone amps, the grounding scheme employed with the recording equipment, etc., even before it hits an A/D converter.

All things considered I am not sure with anything that isn't pure electronic music, that you can get more than about 70-75 db of real range with recordings. That can be somewhat deceptive. Most noise is low in frequency, and our ear breaks up the sound spectrum into 30 or so bands. So some bands of our hearing in the midrange can approach more like 90 db with a recording, but that is probably about it even then. Might be an interesting thread to consider just this topic though we've had a few more or less on this topic before.

Otherwise some electronics get 120 db of dynamic range and squeak out just a few more. To my knowledge the very best doesn't get you past 22 bits worth. When you add in gain and a few stages you are back to likely being limited to 20 bits in the electronics.

 

[Andysu]

Wow... it's like the old joke... "How can you tell when Paul is lying? His lips are moving".

what i want to see is this test with that audio device that amir, used dsd to see if paul sends the dsd into clipping , no seriously i want to see if this dsd can show "the rancor scene" from 'return of the jedi' what the pcm looks like on laserdisc . i'm not fully convinced

 

[Doodski]

I happen think DSD is an interesting format. Does it have a lot of noise in the high-register, yes, but we can't hear it (and doubtfully our loudspeakers could reproduce it anyway), so it likely doesn't matter.

3 dB down, half-power-point, corner-freq, cut-off freq or whatever one calls it it's all more work to do for the amp class(s) with the higher bandwidth.

 

[pinpoint_oxford]

3 dB down, half-power-point, corner-freq, cut-off freq or whatever one calls it it's all more work to do for the amp class(s) with the higher bandwidth.

One could say the same about browsing the modern web with all the extra scripts these days. But some of those sites are useful, if inefficient. Including this site

 

[dadregga]

The problem is that they didn't deliver on the other components either. This is a classic problem with audiophile recordings done this way in that they scrap the barrel to try to find someone who is willing to sign up to such projects for so little sales to audio nerds.

Agree - and agree that that's a common problem with "audiophile labels".

I will say I did enjoy the Zuill Bailey Bach Cello Suites they did, if only because the SACD had "unadvertised" ambient multichannel and I'm a bit of an MCH nerd - but that's a niche of a niche and I agree that buying albums with music you wouldn't otherwise care fore *just* because they're "audiophile quality" is nonsense.

Also yeah, don't retain DSD noise in your damn PCM files that's basic - you can get (inaudible but technically "there") > 20khz content in DSD64, or at least 20khz content with less aggressive filtering, but the DSD "noise hump" is not content.

 

[Andysu]

One could say the same about browsing the modern web with all the extra scripts these days. But some of those sites are useful, if inefficient. Including this site

that was so elusively in the back of my mind far from thought and yes i agree i don't like the layout of the new google main page it sucks so does one with soviet airplane firefox i think its cruel of that company to set a fox on fire .

 

[Doodski]

One could say the same about browsing the modern web with all the extra scripts these days. But some of those sites are useful, if inefficient. Including this site

The tasty angel's share (...and maybe the lion's too.) I suppose although the amplifier power wasted and abused in this sort of power gobbling bandwidth consumption by raw noise data is simply not required and wastes resources. I'm a first person shooter gamer and so I appreciate the high frame rate action, thermal issues and I am familiar with downclocking when quiescent for power savings and longevity. Most of the time it is multiplier reduced down to ~800MHz.

 

[kongwee]

If noise is presented in the recording, I don't mind not being filtered. If you removed a thing in audio file, yes you can. But you will alter the final result of wave file. When yo do video editing, you can remove background noise and hums. But you cannot apply all the way, it will change timbre of the overall presentation. If the DSD is the original recording, the CD quality being filtered. It make no sense to me to listen to the CD.

 

[GDK]

Someone opened a thread about this test on the the PS Audio forum.

There was not much rejoicing.

 

[Madlop26]

Someone opened a thread about this test on the the PS Audio forum.

There was not much rejoicing.

closed after 1 1/2 hours, expected

 

[Michel Forbes]

I'm surprised that the PCM version is delivered at sample rate of 96kHz and 192kHz.
DSD is recorded and mastered using a clock reference at 44,1kHz ( to keep the compatibility with the CD format.
Going from 44.1kHz to 48kHz necessitates an extra sample rate conversion.
Everything is possible since this conversion is done in software but an extra step of processing is involved.
ideally, the delivery should be multiple of 44.1kHz`:
64fs = 88,2 kHz
128fs = 176,4 kHz
256fs = 354,8 kHz

 

[617]

@amirm a good way to illustrate that there's no music in this ultrasonic noise is to pitch shift the. I've done this and it's just noise, maybe with some minor rhythmic pulsations but nothing like upper harmonics of instruments.

I seem to remember I did it in audacity. High pass the file at 22khz with a steep slope if that's an option. Then pitch shift down by 6 octaves or so. 22khz becomes 344 hz.

It becomes quite apparent that this is pure noise.

 

[Blumlein 88]

@amirm a good way to illustrate that there's no music in this ultrasonic noise is to pitch shift the. I've done this and it's just noise, maybe with some minor rhythmic pulsations but nothing like upper harmonics of instruments.

I seem to remember I did it in audacity. High pass the file at 22khz with a steep slope if that's an option. Then pitch shift down by 6 octaves or so. 22khz becomes 344 hz.

It becomes quite apparent that this is pure noise.

Yes, glad I'm not the only to do this. Most of the time there is little there to hear at all. No way can it matter if masked by sub 20khz content.

 

[brandonhall]

I'm a software engineer and would love a side project in the audio world. What software would you all like to see?

 

[Mnyb]

Hey Amir, great work as always, couple of questions.

1) I suppose Octave would make the point that not having filters involved would improve the audio quality, any logic to that?
2) We always talk of the extra over 22KHz but isn't the real point of high res to give finer gradations in the audio band?

That’s not how sampling works re your point 2 you resolve all information there is up to halve the sampling rate , so you won’t get more information because you already have all the information to perfectly reconstruct the signal below 20 kHz .
The video BDWoody posted really tells it all .

Also we won’t be confused by the need to record in something like 24/96 as you need to mix and eq etc . But the end result can be presented to us consumers in CD resolution and it’s fine.

Personally I would be forever happy with 24/48 resolution.

 

[Mnyb]

DSD is such a useless format in that it’s actually not theoretically possible to do any kind of post production in that format . You must convert to PCM and do your stuff and then back to DSD this limits the artistic expression that’s is the point of the whole musical endeavour in the first place ?

Even the industry creates a lot of fud around this and use terms as DXD , DXD is just a very high sample rate PCM format

And to compound the issue if you have room EQ or bass management ( or headphone EQ ), which are necessary for hifi sound , gues what you back to pcm again

So the only practical application for DSD is to record already made analog recordings or the “direct to disc” style one take affair acoustical music with severe limits to its production.
And the data can’t be manipulated during playback?

 

[Mnyb]

No, because Octave Records is recording to DSD but mixing on a 30 year old Studer 900 series mixing desk. Nice desk, but no where close to 24-bit SNR.

Hilarious so they mix in analog to be able to keep everything DSD to avoid using a DAW