Izotope SRC

[Jud]

Just curious, what DAC/system are you using with these settings? I have played around with a number of different ones and there are three I have consistently liked (i) Jud's Suite #3, (ii) Ikkei's, and (iii) SuperDad's (which I believe is also Jud's #2). So I tried your most recent one and in my system at least it is simply too aggressive, making strings and voices to harsh and edgy, but I'm beginning to think that which settings work best are not just taste dependent but also highly equipment dependent.

 

-- This.

 

Miska would be the expert on this (how iZotope settings interact with your DAC's filtering), but I will speculate that if those who've been using cutoffs over 1.00 change to ~.97-.99, other changes in the filter settings will also make a greater difference than they do now.

 

Superdad and I both are using DACs that don't impose any of their own interpolation filtering, so we don't have to take account of interactions between our filtering settings and the internal interpolation filters in the DAC. (Superdad has a true non-oversampling PCM DAC; I have a sigma-delta DAC that allows input at "8x" rates - 352.8/384kHz - and does no further interpolation if the input is at those rates.)

[copy_of_a]

I tried your most recent one and in my system at least it is simply too aggressive, making strings and voices to harsh and edgy, but I'm beginning to think that which settings work best are not just taste dependent but also highly equipment dependent.

Same findings as sdolezalek. In my system to aggressive, upfront and harsh as well.

Hi,

 

well, my DAC has 2 filters: a frequency optimized and a phase optimized filter. I do use the phase optimized filter. Just switched to the steeper (frequ. optimized) filter and with this I also find the settings too aggressive. Also attached another DAC (NAD D1050) and here it is the same. So obviously my settings only work well in conjunction the phase optimized filter. For the frequency optimized filter I would go more like steepness 7 | co 1.02 | pre-ringing 0.72 (so actually superdad's setting with a bit less pre-ringing).

Oh well ... it's really complicated...

[copy_of_a]

...

...

what about...:

steepness: 9

cut-oof: 1.02

pre-ringing: 0.68

?

With the frequency optimized filter of my DAC pretty close to superdads settings ... just a tad more body/attack/grip (what I was aiming for) and still sounding "natural" here.

[sdolezalek]

what about...:

steepness: 9

cut-oof: 1.02

pre-ringing: 0.68

?

With the frequency optimized filter of my DAC pretty close to superdads settings ... just a tad more body/attack/grip (what I was aiming for) and still sounding "natural" here.

 

Still a bit too harsh for my taste (but much better on 24/96 and 24/192 than on 16/44 files) in my system. So far a few of my favorites are as follows:

 

[YashN]

• class-A solid-state output stage & tube output stage (switchable)

• Ultra linear circuitry / fully balanced analogue audio circuitry

• 2 selectable digital filter settings (frequency-optimized & phase-optimized)

• DSD64 & DSD128 (DoP)

(for reference: whenever I play around with software SRC I always use the transistor stage and the phaze-optimized filter.)

 

 

DSD128 (or above) and tube output stage is it.

[odelay]

Fantastic table Sdolezek!!!

(I nicked it!)

 

Thanks everyone for sharing their settings and impressions. Having used the same dodgy settings for months, I've changed thing quite a bit. It's still a work in progress - I wish I could tune it up in 5 minutes like the Master! But I'm enjoying the result.

 

9, 10k, 0,94, 200, 0.30-ish?, max

 

Happy tuning folks :-)

[skatbelt]

So far a few of my favorites are as follows:

 

[ATTACH=CONFIG]16550[/ATTACH]

 

'Own early'? You meant my settings...

[Barcabale]

Rendons à César ce qui est à César !

'Own early'? You meant my settings...

[Jud]

Rendons à César ce qui est à César !

 

[skatbelt]

[Jud]

[ATTACH=CONFIG]16555[/ATTACH]

 

Veni, vidi, audivi!

[Barcabale]

Bravo !

[sdolezalek]

Apologies, in the early days I wasn't keeping track of who suggested what, only what I liked/disliked. ;-)) Updated table attached with the appropriate attribution.

[copy_of_a]

• Double Burr Brown PCM-1792 (dual differential mode)

• class-A solid-state output stage & tube output stage (switchable)

• Ultra linear circuitry / fully balanced analogue audio circuitry

• 2 selectable digital filter settings (frequency-optimized & phase-optimized)

• DSD64 & DSD128 (DoP)

• 24bit/192kHz asynchronous USB (for some reason OS X & Audirvana show 384kHz as max. rate)

(for reference: whenever I play around with software SRC I always use the transistor stage and the phaze-optimized filter.)

DSD128 (or above) and tube output stage is it.

by now I only have DSD64. Only 3 albums. Unfortunately most of the current DSD albums are pretty conservative/old fashioned/boring classical albums (well... for me). So the choice of actual music in DSD format is not really interesting for me personally at the moment.

 

The tube output stage is great! Love it on some albums. However, also the class-A transistor stage sounds extremely nice (and a bit more precise). The DAC shows no harshness at all or "digital" sound whatsoever with the solid state stage. I would say it's also pretty immun to heavy intersample peaks. +3.5db true peak does not exhibit the least sign of distortion ... the music just flows (maybe even with higher true peaks ... but +3.5 was the highest I've seen in a certain track).

[Superdad]

I tried all possible settings and I can't achieve better result than berkeley alpha internal upsampling

 

To do so I'm afraid you will have to try HQ Player and its Poly-sinc filters and mild dither. That's were I ended up. I was skeptical--really liking what I had worked hard to come up with in iZotope--but Miska's Poly-sincs are the best.

[Jud]

To do so I'm afraid you will have to try HQ Player and its Poly-sinc filters and mild dither. That's were I ended up. I was skeptical--really liking what I had worked hard to come up with in iZotope--but Miska's Poly-sincs are the best.

 

It's not at all surprising that filters done by a professional sound better than our amateur efforts. Whether Miska's excellent filters will sound better than the BADA's internal filters alone depends on how Miska's filtering and the BADA's internal filter combine.

[copy_of_a]

To do so I'm afraid you will have to try HQ Player and its Poly-sinc filters and mild dither.

as far as Dither goes you can improve A+ with AU-Plugins. Played around the whole evening with my Dither plugins (PSP Audioware) and I think these sound somewhat better. Not sure whether it's the negative gain (set to -3db, just like HQ Player) of my PSP plugins that work better than the iZotope plugin or if it's due to the shaped Dither filters and their respective settings. Most likely the latter. The cool thing with these plugins is you can select 8bit or 12bit output for pre-listen purposes... to compare how they sound. Differences are minute, though (when set back to 24bit). In any case it was fun to try it out and I will stay with this configuration in A+.

Then again ... since I've just acquired HQPlayer and my DAC sounds really good without upsamling I will most likely use A+ without upsampling (and therefore also without negative gain and without Dither) as my main player and will use HQ Player for more serious listening sessions with upsamling applied since the Poly-sinc filters of HQ Player do in fact sound more traceable. We'll see ...

[levandier]

I've been playing around with izotope settings a fair amount recently and I've come up with two sets of settings I like. The first that I use most often:

 

steepness - 24

cutoff - .95 (or even .9525)

preringing - 1

 

And then the second set I use when I'm willing to trade a bit of air for tighter transients:

 

steepness - 20

cutoff - .83

preringing - 1

[pavement714]

I've been playing around with izotope settings a fair amount recently and I've come up with two sets of settings I like. The first that I use most often:

 

steepness - 24

cutoff - .95 (or even .9525)

preringing - 1

 

And then the second set I use when I'm willing to trade a bit of air for tighter transients:

 

steepness - 20

cutoff - .83

preringing - 1

 

I had to register on this forum just to thank you and the other setting posters for letting me experiment. I went from BitPerfect to Audirvana, and found the SRC upsampling to be similar, but felt something was missing. Thanks to finding this thread, I've gotten a sound from my setup that I really enjoy, with a lot of detail, presence, and warmth.

[copy_of_a]

steepness: 8

cut-off: 1.02

pre-ringing: 0.72

 

"superdad's" (phantastic!) settings optimized for more attack/punch/grip/body & better timing with Rock/Pop/Electronic and the like (well, in my system...)

 

:-)

[Mystic]

steepness: 8

cut-off: 1.02

pre-ringing: 0.72

 

"superdad's" (phantastic!) settings optimized for more attack/punch/grip/body & better timing with Rock/Pop/Electronic and the like (well, in my system...)

 

:-)

 

What about Filter Max Length and Anti- Aliasing? Why do a lot of people not included those 2 settings when posting setting that they like? Is it just that they just keep the A+ default?

[skatbelt]

What about Filter Max Length and Anti- Aliasing? Why do a lot of people not included those 2 settings when posting setting that they like? Is it just that they just keep the A+ default?

 

Good question! I believe it is generally assumed it is best to keep Anti-Aliasing at it's maximum value (=best quality) unless your cpu cannot handle it. But I did find altering the default value for Filter Max Length makes a difference. Actually I just went from 800.000 to 672.000 after upgrading my AES/EBU cable. I had an entry-level Wireworld and upgraded to a Transparent one (see signature).

 

On a side note: I was a non-believer on the subject of cables making a real difference other than just cool looks until I heard how much difference (not subtle) a USB cable can make. The win with the AES/EBU cable upgrade was less dramatic but still substantial. It gave me me a bit more of everything but especially a tad more realness. Maybe also because this cable has the by Berkeley Audio Design advised length of 1.5m (I use their Alpha USB converter) whereas my former one was only 0.5m.

 

Back to SRC, I found that optimal values for steepness, cut-off and pre-ringing were the same with the new cable but that I had to lower the Filter Max Length for best weight and tautness of the lower frequencies and hence the best soundstage.

 

I realize, my findings make things even more complicated. There are so much dependencies.

[Rezseeker]

Thanks to my settings being so far off everyone else's I ended up auditioning some new gear..... lo and behold now I have new B&W CM5 S2 speakers. This is crazy, but WOW now I know how good of a digital system I've been sitting on. So much more texture across the frequency range and such a natural presentation for every genre.

Posted a mini review in another thread, but wanted to thank Ikkei for his detailed analysis and suggested settings that were way on the other end of the spectrum from where I'd been making minor adjustments (especially steepness).

My new settings (still tweaking):

Steepness 88 (was 8 with old speakers)

Filter 1,800,000

Cutoff ,95 (was around 88 before)

Pre-ringing .37 (around 45 before)

[ed71]

Thanks so much for the helpful information in this thread! I have tried a couple of the settings posted in the table, and I am really enjoying the Jud 3 settings. It is hard to describe the improvement, so I won't try other than to say that I plan to stick with these settings.

[Jud]

Thanks so much for the helpful information in this thread! I have tried a couple of the settings posted in the table, and I am really enjoying the Jud 3 settings. It is hard to describe the improvement, so I won't try other than to say that I plan to stick with these settings.

 

Those are the default settings (the same ones you started with on installation of the program) so it's good you like them.