Hi, thank you for the hard work you’ve done and your humility in looking for the best SQ for the selling price even if you had to stop the production.
Can’t wait to read reviews of the new katana and to hear it myself soon after.
Yes bandwidth is further reduced by LPF , peak is almost completely gone.
Yes the roll off is definitely the reason why the "fatiguing" sound was heard. As soon as I changed it . it went completely away (night and day difference for that hi freq sound that was heard on bookshelves)
Further, noise on the analog section from positive to negative rail was reduced.
Rasmussen pre analog filter made quite a difference in my system but before incorporating I want to hear it with the new LPF
Terry thx you for the nice words. I think I learned a good lesson.
Yes the roll off is definitely the reason why the "fatiguing" sound was heard. As soon as I changed it . it went completely away (night and day difference for that hi freq sound that was heard on bookshelves)
Further, noise on the analog section from positive to negative rail was reduced.
Rasmussen pre analog filter made quite a difference in my system but before incorporating I want to hear it with the new LPF
Terry thx you for the nice words. I think I learned a good lesson.
Can you please elaborate into how would a frequency 4 times over the best human hearing capacity can cause fatigue? The best child's ear can only hear up to 20Khz, and for us (adults in the 20s, 30s, 40s) its more close to 16Khz at best. People over 50 probably 12Khz or so. How is an 80Khz peak affecting? Actually, what music has ANYTHING recorded in a 80Khz range?
As far as I know, this is only feedable to a DAC by a tone generator. A CDs maximum frequency will be around half of its 44.1Khz frequency... so I'm confused, and I am doubtful that an 80Khz ring was the reason for this fatigue.
I would appreciate some more info into the matter.
Thanks,
Rafa.
As far as I know, this is only feedable to a DAC by a tone generator. A CDs maximum frequency will be around half of its 44.1Khz frequency... so I'm confused, and I am doubtful that an 80Khz ring was the reason for this fatigue.
I would appreciate some more info into the matter.
Thanks,
Rafa.
Quite a nice review from Amir.
He seems to lack some IT knowledge. That might cause minor
issues on the "apples vs. apples" front.
The PI config matters too. If you e.g. leave the microwave oven on underneath the DAC you shouldn't
wonder about this or that side effect.
Obviously there's more...
I do understand why Allo provides Volumio. Volumio wouldn't be my first choice though.
And I am wondering what DAC filter he's been using.
The choice of filter should also have an impact on the measurements.
I 'd also guess that he used the Allo $11 SMPS for the PI. Hmmh. Many of us experienced that PI power matters too!
Sbooster is OK. I also have it. Not my first choice though.
Great to read that Allo addressed the weak spots (more then just the output filter) already sounds promising.
And then there's the bright future with the upcoming isolator and Nirvana...
...the sky is gonna be the limit.
Then there was this jitter topic.
I also have a theory to the jitter topic that's been discussed over there. Especially the issue around audible jitter or phase noise limits.
I think the point is: To achieve femto second jitter levels ( however that's been measured by each party) you need to produce a highest quality board - on all levels - not just the clock area!
Basically all areas will benefit. And that's what's probably causing this or that enhancement on soundquality - a simple side effect. The argument that
jitter below a certain level is not audible anymore IMO might not reflect the whole situation.
Enjoy.
He seems to lack some IT knowledge. That might cause minor
issues on the "apples vs. apples" front.
The PI config matters too. If you e.g. leave the microwave oven on underneath the DAC you shouldn't
wonder about this or that side effect.
Obviously there's more...
I do understand why Allo provides Volumio. Volumio wouldn't be my first choice though.
And I am wondering what DAC filter he's been using.
The choice of filter should also have an impact on the measurements.
I 'd also guess that he used the Allo $11 SMPS for the PI. Hmmh. Many of us experienced that PI power matters too!
Sbooster is OK. I also have it. Not my first choice though.
Great to read that Allo addressed the weak spots (more then just the output filter) already sounds promising.
And then there's the bright future with the upcoming isolator and Nirvana...
...the sky is gonna be the limit.
Then there was this jitter topic.
I also have a theory to the jitter topic that's been discussed over there. Especially the issue around audible jitter or phase noise limits.
I think the point is: To achieve femto second jitter levels ( however that's been measured by each party) you need to produce a highest quality board - on all levels - not just the clock area!
Basically all areas will benefit. And that's what's probably causing this or that enhancement on soundquality - a simple side effect. The argument that
jitter below a certain level is not audible anymore IMO might not reflect the whole situation.
Enjoy.
Last edited:
when I measured the different filter impulses of my china es9038q2m board I saw an impact on f-curve at high frequency cut-off and though bandwidth - as soundcheck writes above.
Nevertheless this little bandwidth loss of 0,5 dB at 20kHz should not have such big impact regarding fatiguing effect.
I think the amp of cdsgames small speaker setup might have a problem with the 88kHz peak that probably induces distortions of lower harmonics down into lower frequencies in audible f-range. Or the tweeter (is it a metal dome?) has a resonance peak at around 20kHz that is now after reducing bandwidth not stimulated as much as before? Would be interesting to see a measurement as Amir did before and after the new filter update with same SW filter settings on DAC.
Nevertheless this little bandwidth loss of 0,5 dB at 20kHz should not have such big impact regarding fatiguing effect.
I think the amp of cdsgames small speaker setup might have a problem with the 88kHz peak that probably induces distortions of lower harmonics down into lower frequencies in audible f-range. Or the tweeter (is it a metal dome?) has a resonance peak at around 20kHz that is now after reducing bandwidth not stimulated as much as before? Would be interesting to see a measurement as Amir did before and after the new filter update with same SW filter settings on DAC.
If that is true, I would find it almost irresponsible that they are tailoring the DAC to marry a particular speaker and its shortcomings. Let's hope they are testing in a wide variety of equipment: different brands, sizes, styles (ribbon, metallic, soft dome, electrostats).
Otherwise the users for the DAC would be extremely limited and disappointed.
Regards,
Rafa.
Otherwise the users for the DAC would be extremely limited and disappointed.
Regards,
Rafa.
It's not the 88 kHz signal that is audible (it isn't), but modulation of the signal to lower frequencies (for example IM distortion) may result in very audible artifacts.
Hi Rafa
we don't tailor Katana to a specific speaker , but on smaller speakers "fatiguing sound" is heard more pronounced then bigger speakers. With the LPF changes (that are done according to discreet output stage designer ) that sound is removed from smaller speakers and even on best speakers an increase in SQ is apparent .
Quite a nice review from Amir.
And I am wondering what DAC filter he's been using.
The choice of filter should also have an impact on the measurements.
I 'd also guess that he used the Allo $11 SMPS for the PI. Hmmh. Many of us experienced that PI power matters too!
Sbooster is OK. I also have it. Not my first choice though.
Great to read that Allo addressed the weak spots (more then just the output filter) already sounds promising.
And then there's the bright future with the upcoming isolator and Nirvana...
...the sky is gonna be the limit.
Then there was this jitter topic.
I also have a theory to the jitter topic that's been discussed over there. Especially the issue around audible jitter or phase noise limits.
I think the point is: To achieve femto second jitter levels ( however that's been measured by each party) you need to produce a highest quality board - on all levels - not just the clock area!
Basically all areas will benefit. And that's what's probably causing this or that enhancement on soundquality - a simple side effect. The argument that
jitter below a certain level is not audible anymore IMO might not reflect the whole situation.
Enjoy.
Hi Klaus,
I will be discussing the filter choices with Amir.
Nirvana PCB design has started today, we will make 10 samples and we will send you an engineering sample for your own testing. Since the beginning you have been implicated in this design..
Regarding jitter...I fully agree. I think its very hard to achive this level of jitter and yes it shows that board was well engineered and no matter how ,can only help SQ (cannot be a negative if its lower than other DACs)
New isolators were received and we will update the webpage in 24h.
I believe there are 2 different things we are discussing.
1. Digital filter
2. Analog filter
Amir measured the digital filter. The changes we have done are in the analog domain. Small changes in analog filter result in big changes in SQ.
The removal of "fatiguing sound" happened because of analog changes. The LPF was changed to around 100Khz cutoff and R in the RC filter across the opamp was removed . The last change results in a changed rolloff (deeper) and LPF cutoff was done as advised by most articles and books .
The LPF was changed to around 100Khz cutoff and R in the RC filter across the opamp was removed
Your output class A stage dont have feedback RC filter ?
So you are saying Amir was not measuring at the RCA outputs but at the outputs before analugue stage? I dont see this in his report. For me it looks that this is what comes out of the RCA plugs when HF noise signal is played by the DAC...but not sure how they "feed" the DAC in this test.
If so, you should see the impact of digital + analogue filter in this test - pls. correct me if I am wrong.
Yes he was measuring the RCA output. However you have 2 types of filters , digital and analog and both will show up in the final readings.
Dominant one is digital filter since analog filters are less pronounced.
Digital filter is done in the DAC IC (dsp). Analog filter cleans the output of the DAC ic so they serve different purposes
Dominant one is digital filter since analog filters are less pronounced.
Digital filter is done in the DAC IC (dsp). Analog filter cleans the output of the DAC ic so they serve different purposes
One last comment on it...we are always working to make Katana the best DAC on the market and presently working on 2 gen and even 3 gen analog stages.
After extensive testing of THD+N and SQ they might (or not) be released , but (if released) if you have a Katana already it will be as simple as removing the analog stage and replace it with new pcb .
As allo pointed in the volumio forum, is allo really providing customers with this new isolator for free?
Oh boy, wouldn't that be a treat for those customers who are early adopters of Katana..? ;P