Izotope SRC

[pl_svn]

seems to be broad consensus that NOS DACs with 44.1kHz input produce audible artifacts through aliasing/intermodulation.

 

my totally non technical take on (this) NOS DAC: I never felt the "need" to upsample redbook files 'till… no idea about what has changed in Audirvana 1.4.6 but suddenly something, there, was not as right as before (I'm always a little scared when I'm told that "measures are great" :-p and all Damien said is... that he used very sophisticarted equipments to tune this new release)

 

 

 

 

 

 

(I came to the Metrum Octave from a Schiit Bifrost and it was a revelation. the Bifrost sounds great indeed but honestly, imho, more for a very wisely chosen, and very well balanced, pleasant without sounding "artificial", "sound signature" than for intrinsic "technical" merits)

[elcorso]

1)

----------------

2) That's a tough question and I don't know the answer :-) If you ask a filterless NOS DAC designer, they'll say it won't impact sound because we can't hear above 20-22khz anyway (or even lower as we get older), and so it doesn't matter that image energy isn't filtered - our ears filter it. But if you ask someone like Dan Lavry, who argues vehemently and very impolitely against NOS DAC's in his white papers, he says that image energy, even though it can't be heard directly, can still degrade sonic performance because it makes your amp, pre-amp, and speakers work harder and can lead to intermodulation distortion in the audible range. Who's right I suppose is a personal thing depending on your sensitivity to things like pre-ringing and the gear in your chain (for example, Stereophile has an intermodulation distortion test in the measurements section of their amp reviews and some amps do better than others).

 

Then, both could be right and both could be wrong (NOS DAC designers and Mr. Lavry)...!

 

Some explanation, from my ear/brain system, not from graphs.

 

• I recently bought an NOS DAC (Metrum Octave). This unit profit a lot from iZotope SRC settings under A+ at 2X, but I can still play music under no up sampling, then no iZotope, and still enjoy a lot the music.

 

• I own an up sampling DAC, since a lot of time ago, and I'm forced to up sample to 2X, then I also need a forced iZotope SRC settings, mainly on the Pre-ringing slider set to 0.6. The other sliders at default from A+ settings.

 

Then, congratulations to Mr. Alexey Lukin for his incredible filters.

 

Roch

[Michael Cullinan]

Just maxed out the steepness, and minimized the cut-off frequency to see the effect on a bright Pixies recording but it still sounded bright.

 

Are the advanced parameters for up-sampling even working?

I was just going to ask the same thing. (Yesterday) I did 200 steep, .5 cutoff and I hear no difference. I have pretty good ears at least I thought. Lol that should have no treble right? I hit stop in A+. Dont quit out though. I think this is bs. Anyone??

[Freann]

I was just going to ask the same thing. (Yesterday) I did 200 steep, .5 cutoff and I hear no difference. I have pretty good ears at least I thought. Lol that should have no treble right? I hit stop in A+. Dont quit out though. I think this is bs. Anyone??

 

BS – really? You do know the settings are not in effect if upsampling is disabled?

[Jud]

my totally non technical take on (this) NOS DAC: I never felt the "need" to upsample redbook files 'till… no idea about what has changed in Audirvana 1.4.6 but suddenly something, there, was not as right as before (I'm always a little scared when I'm told that "measures are great" :-p and all Damien said is... that he used very sophisticarted equipments to tune this new release

 

(I came to the Metrum Octave from a Schiit Bifrost and it was a revelation. the Bifrost sounds great indeed but honestly, imho, more for a very wisely chosen, and very well balanced, pleasant without sounding "artificial", "sound signature" than for intrinsic "technical" merits)

 

I think that last part in particular is a very interesting way to put it, though the Bifrost has nice parts quality for its price level. Because any traditional DAC uses interpolation filters that cannot even in principle be made perfect, what you are *always* listening to is the designer's judgment of the best "sound signature." (Make no mistake, NOS DACs, even - perhaps especially - those that don't utilize upstream interpolation from software, have their own characteristic imperfections and sonic signature with RedBook files at least.)

[Michael Cullinan]

BS – really? You do know the settings are not in effect if upsampling is disabled?

 

I have upsampling on too. Weird. Do I need to quit out of A+? I think I'm going to just buy a Bose system! Lol

[Jud]

I was just going to ask the same thing. (Yesterday) I did 200 steep, .5 cutoff and I hear no difference. I have pretty good ears at least I thought. Lol that should have no treble right? I hit stop in A+. Dont quit out though. I think this is bs. Anyone??

 

Do you have forced upsampling selected?

 

What DAC are you using?

[Michael Cullinan]

I have a Ressonnence Concero Dac. And I do the multiples of 2 upsampling.... Where 44.1 becomes 176 and 48, 192k

 

What should 200 steep and .5 cutoff with .6 prering sound like. Poop right?

I'm using Patricia Barber as music for testing.

[Jud]

I have a Ressonnence Concero Dac. And I do the multiples of 2 upsampling.... Where 44.1 becomes 176 and 48, 192k

 

What should 200 steep and .5 cutoff with .6 prering sound like. Poop right?

I'm using Patricia Barber as music for testing.

 

If I remember correctly, you do need to quit and restart A+ for new iZotope settings to take effect.

 

Do you hear differences in Redbook files without software upsampling when switching between the Concero's upsampling filters?

[Michael Cullinan]

If I remember correctly, you do need to quit and restart A+ for new iZotope settings to take effect.

 

Do you hear differences in Redbook files without software upsampling when switching between the Concero's upsampling filters?

 

Ahh.... No wonder I hear no difference.

As to the Concero filters, I do hear a difference. The biggest difference, being off vs using either the FIr or apodizing filter. The difference between the 2 filters is much more subtle.

[vk80]

What should 200 steep and .5 cutoff with .6 prering sound like. Poop right?

I'm using Patricia Barber as music for testing.

 

Here's the graphical representation. Essentially anything above 15k is effectively non-existant (-190db). Cymbals, hi-hats, etc. would sound like they are behind a wall / muffled. Redbook is good through 22k, so you're cutting off that whole 7k block of upper frequency / treble [from blue line to white line].

 

[Superdad]

I would say the following just on the basis of reading, and heartily invite comments from those who actually know:

 

- There seems to be broad consensus that NOS DACs with 44.1kHz input produce audible artifacts through aliasing/intermodulation. This can be somewhat ameliorated by reducing filter steepness, but this can give rise to various other problems. NOS DAC builders (at least those who do not rely on upsampling of RedBook through software) have decided any artifacts or other disadvantages are exceeded by the sonic disadvantages of upsampling, which has its own unavoidable imperfections.

 

- There is less consensus whether *exclusion* of low ultrasonic harmonics of various instruments (e.g., trumpet, cymbals) detectably alters the sound of those instruments. (Do these ultrasonics intermodulate with other vibrations produced by the instruments to create audible products, for example?)

 

Jud, thank you for the thoughtful comments.

As someone who has been listening to a custom filterless NOS DAC with very revealing electronics/speakers/room for quite some time, I offer the following:

a) I moved from a very fine Dodson DAC (a well executed oversampling design, but still with BB1704 R2R-ladder DACs), to our filterless-NOS BB1704 design; For the first few years I did not upsample at all (the G4 Mac mini at the time did terrible upsampling and PM, Amarra, A+ could not run on it; well older PM did, but my earlier adaptive USB input board was 16/44 only).

Still the straight 16/44 into the NOS was a revelation, even if I had some inkling that some aliasing artifacts were folding back down into the audible range. I would still put this particular DAC up against most others and for sheer musical immediacy it would walk away.

 

b) Flash forward to my present setup with newer Mac mini, upsampling to 176.4 with A+/iZotope, and the little Romanian WaveIO 24/192 async-XMOS USB>I2S board installed in my DAC (with a beefy dedicated power supply--USB 5V wire cut at Mac end). I can never go back to straight NOS--at least not with Redbook files! My ears (despite age induced high-frequency loss and mild tinnitus) have become very aware of the effect of aliasing artifacts; they even come down into the midrange.

 

c) Now that my iZotope settings are dialed in to please me (natural piano, spooky-real cymbals, overall tactileness), I tried something last night. I have a couple of albums for which I own both the 16/44 CD and the DVDA with 24/96 of the exact same mix. So I compared the Redbook upsampled to 24/176.4 to the 24/96 with iZotope completely off. And wow, they were darn close! Both had an ease about them, not harsh, very present--I could live with either.

 

So my conclusion is that NOS DACs really do need to be fed at least an 88kHz signal to avoid the aliasing nasties.

As has been said here many times before, it comes down to whose upsampling/filtering algorithms you want to listen to: the ones in your DAC or the ones in the software on your computer.

The better DACs where the designers have done their own filters are certainly contenders against the software-only/NOS-hardware combo. (Those better DACs might not come cheap though.) As for the very large number of DACs and (CD players) using the fixed sets of digital filters offered by the chip companies?--no contest.

 

Just my 2 nickels...

[Superdad]

If I remember correctly, you do need to quit and restart A+ for new iZotope settings to take effect.

 

Restart of A+ when adjusting iZotope settings is definitely NOT required. I can just double-click the very same track I was playing and the change takes effect.

 

For changing upsampling setting (from powers of 2 to 2x-only, to off, etc.) I do need to at least switch tracks. My DAC has an LED which changes color depending upon what sample rate it is being fed. So I always know.

[Jud]

Restart of A+ when adjusting iZotope settings is definitely NOT required. I can just double-click the very same track I was playing and the change takes effect.

 

For changing upsampling setting (from powers of 2 to 2x-only, to off, etc.) I do need to at least switch tracks. My DAC has an LED which changes color depending upon what sample rate it is being fed. So I always know.

 

Thanks for correcting my unintentional misinformation.

 

Michael, do you hear any difference with iZotope changes through the Concero now, or still nothing?

[Jud]

Jud, thank you for the thoughtful comments.

As someone who has been listening to a custom filterless NOS DAC with very revealing electronics/speakers/room for quite some time, I offer the following:

 

c) Now that my iZotope settings are dialed in to please me (natural piano, spooky-real cymbals, overall tactileness), I tried something last night. I have a couple of albums for which I own both the 16/44 CD and the DVDA with 24/96 of the exact same mix. So I compared the Redbook upsampled to 24/176.4 to the 24/96 with iZotope completely off. And wow, they were darn close! Both had an ease about them, not harsh, very present--I could live with either.

 

So my conclusion is that NOS DACs really do need to be fed at least an 88kHz signal to avoid the aliasing nasties.

As has been said here many times before, it comes down to whose upsampling/filtering algorithms you want to listen to: the ones in your DAC or the ones in the software on your computer.

The better DACs where the designers have done their own filters are certainly contenders against the software-only/NOS-hardware combo. (Those better DACs might not come cheap though.) As for the very large number of DACs and (CD players) using the fixed sets of digital filters offered by the chip companies?--no contest.

 

Just my 2 nickels...

 

Would be very interested to know what if any differences you hear between the HDTracks 24/192 and MoFi SACD/DSD versions of Pet Sounds, which come from the same stereo remaster.

[Miska]

I think that different SRC settings are a matter of personal preferences and I do not have strong preference for one or another setting. That's why iZotope SRC provides all the settings to tweak.

 

One important aspect, IMO, would be adjustment of the roll-off shape / stop-band parameters.

[Michael Cullinan]

Ill try when I get home from work. I'm going to use some q-tips. Lol.

[Superdad]

Guys:

I just wasted another couple of hours experimenting with settings (I'm supposed to be working, but this is what happens when you are self employed and your office is also your listening studio!).

I tried several of your suggested combos and also adjusted settings individually off both my settings and off of yours. I will now make some generalizations about what I hear (everyone is different), but before I do I should explain how/what tracks I am listening to for these particular tests:

 

Louis Armstrong "St. James Infirmary" from the album "Satchmo Plays King Oliver" (incredible feeling, challenging brass and woodwind)

Brian Bromberg "The Saga Of Harrison Crabfeathers" from the album "Wood" (top to bottom bass, good piano and cymbals)

Bill Evans "Night and Day" and "Peace Piece" from XRCD of "Everybody Digs Bill Evans" (sharp drumstick rim hits, great space, piano, cymbals, etc. on "Night and Day"; slow building piano with lots of heart building into more intense and dense hammering)

Boukman Eksperyans "Mayi A Gaye" from "The Planet Sleeps" (acapella vocal, one woman then joined by another, then a man--primary vocal harmonics extends way out)

 

Those five cuts allow me to hear very quickly what sonic effect the filter parameters are having.

 

Pre-ringing: I find that at anything less that about 0.6 Armstrong's cornet sounds drunk (and more omni-directional than it should). The usually very clear and differentiated large acoustic bass of Bromberg's becomes loose and wooly at low settings.

 

Cutoff Freq.: I previously settled on 1.04. I don't mind a tiny bit lower setting, but below 1 (and depending upon steepness) cymbals and even piano loose air and delicacy for me. Above 1.1 and the supersonic energy interferes with everything.

 

Steepness: Funny, I would have thought this one to be the most critical, but it is mostly just on piano where I find that settings like 20 fail to move me (I prefer about 7). With my other settings where they are (Pre-ring 0.65; Cutoff 1.04; Max Length 1,150,000) I could certainly see varying the Steepness to make some recordings more listenable.

 

Per Dr. Lukin's indication that filter max length should be best at 2,000,000, I gave that a try again, but boy does the air and life go dry at that setting. It's pretty easy to hear and not hard to settle on lower setting. In other words, unlike some of the other parameters I don't feel like I am trading off anything by lowering it. That said, 1,000,000 is too low for me (it gets a little wispy).

 

This has been interesting. And I really want to learn more about what goes on (now and in the past) in the A/D filtering during the mastering process. That could certainly affect the choices we make on the playback side.

 

Lastly, despite the settings trying to replicate Ayre's white paper graphs, I kind of doubt that using those settings is making my DAC sound like Ayre's. That is, I don't think Mr. Hansen would choose the sound that those settings give in my system. Of course his DAC is a different design altogether.

[Superdad]

One important aspect, IMO, would be adjustment of the roll-off shape / stop-band parameters.

 

Dear Miska:

Can you share with us what other parameters of an SRC filter would be, in an ideal world, adjustable? As you probably know, the iZotope SRC engine (as used in Audirvana Plus, Amarra, and I'm sure some Windows players) allows for the following to be adjusted:

 

Steepness, Filter Max Length, Cutoff Frequency, Anti-Aliasing (max. attenuation), and Pre-ringing.

 

Dr. Lukin of iZotope has also been visiting this thread, so a suggestion from you might even give him some idea that could improve flexibility for a great many people around the world.

 

With much respect,

ALEX

[Brault]

... (This also highlights how/why what happens in the mastering process affects what we get during playback--and validates the idea that some recordings will sound best with different playback filter characteristics.) ...

 

ALEX

So we need a method of saving personalized filter settings that are DAC-specific AND source-specific?

[Liam]

My m2tech Young dac has minimum phase filter implementation. "A high performance oversampling filter, custom-made on a programmable logic device (FPGA) implements minimum-phase filters optimized for sound quality which allow for a very low noise floor". What implications does this have for the position of the pre ringing slider, if any? In my ignorance I'm thinking that moving the slider to the left is duplication? Should the slider be to the extreme right? Thanks for any help.

[Miska]

Can you share with us what other parameters of an SRC filter would be, in an ideal world, adjustable?

 

There are many things I consider important regarding SRC filter design, but I'm not sure if all those should be exposed to end users. Because things grow really complex. And lot of those things cannot be expressed in simple numbers, but would require meta-programming language to express as mathematical formulas. Of course all technically doable, but I'm not sure if anybody would actually want that, instead of just buying Matlab. So ideal world has to meet reality somewhere...

 

For example I don't think it is possible to accurately replicate the Ayre's filter with the current set of parameters anyway.

[VandyMan]

My m2tech Young dac has minimum phase filter implementation

 

If I had a DAC that had a filter I enjoyed, I would turn off up-sampling in A+.

[elcorso]

Thanks for correcting my unintentional misinformation.

 

Jud, On A+ you need to hit the "Stop" button for the changes to take effect (as stated by Damien), this will be the same as to reload A+, when the 'device' (your DAC) is released from A+. But, of course, if you re-hit the track after hitting "Stop' has the same effects on iZotope changes.

 

And I must insist, at least for me, iZotope up sampling is much more evident on up sampling DACs (I own several) than the only NOS DAC one I own, the Metric Octave DAC (in his way to be improved by Paul Hynes 8 regulators and LPSU).

 

Then, for now on, I'll take with a grain of salt the general consensus "audible artifacts through aliasing/intermodulation" from NOS DACs.

 

My 15 Colones (about 2 cents),

 

Roch

[Jud]

And I must insist, at least for me, iZotope up sampling is much more evident on up sampling DACs (I own several) than the only NOS DAC one I own, the Metric Octave DAC (in his way to be improved by Paul Hynes 8 regulators and LPSU).

 

Then, for now on, I'll take with a grain of salt the general consensus "audible artifacts through aliasing/intermodulation" from NOS DACs.

 

My 15 Colones (about 2 cents),

 

Roch

 

I would expect nothing less.

 

I will say this much, for whatever it may be worth: Especially with modern NOS DACs, it is very possible whatever artifacts are created are not unpleasant. It is likely also true that having relief from the sonic "signature" of 8x oversampling makes listening to an NOS DAC a pleasure.