IMO, 32 bit floats are inadequate for IIR filters. I heard such an implementation and it produced a subtle brown-like noise as a small proportion of the instantaneous audio level. It was a feature packed tool, which made it all the more annoying that it was slightly broken, and I didn't have the peace of mind to be able to take it seriously enough for creative work.
All these discussions on different DSPs, DAC chips, and bits.
A couple simple questions, that might point you in another direction too.
How does your DSP output it's "Sine Wave" ?
Sadly on budget gear it is often a PWM with some shabby quality caps, trying to clean up your square pulsed "Sine Wave" output , making the output by the PWM only work so well.
Now what is a PWM ? it is a electrical noise generator, pretty much.
Diode rectifiers/PWMs is the main contributor of THD in the electrical grid, and those acceptable 10% of THD, is often not very well handled in the circuit designs.
Another little point, even if you electrical supply is 230 or 110 Volts etc, the power companies are not, required to feed you a balanced 3 phase power for example, it is not uncommon to find different voltages on all 3 if measured.
How is it powered in general?
Again a PWM, with caps for smoothing.
How well is the circuit filtering designed?
And how will your DAC perform in the future as all those crucial, but cheap electrolytic starts to degrade, as they are largely responsible for producing your sine waves.
Circuit design on its whole matters a lot more than any single DAC chip, or bit counts.
And products you can buy finished for 100$, produced in China for 10% of the cost, often is not very well designed in that regard.
Just some food for thought🙂
A couple simple questions, that might point you in another direction too.
How does your DSP output it's "Sine Wave" ?
Sadly on budget gear it is often a PWM with some shabby quality caps, trying to clean up your square pulsed "Sine Wave" output , making the output by the PWM only work so well.
Now what is a PWM ? it is a electrical noise generator, pretty much.
Diode rectifiers/PWMs is the main contributor of THD in the electrical grid, and those acceptable 10% of THD, is often not very well handled in the circuit designs.
Another little point, even if you electrical supply is 230 or 110 Volts etc, the power companies are not, required to feed you a balanced 3 phase power for example, it is not uncommon to find different voltages on all 3 if measured.
How is it powered in general?
Again a PWM, with caps for smoothing.
How well is the circuit filtering designed?
And how will your DAC perform in the future as all those crucial, but cheap electrolytic starts to degrade, as they are largely responsible for producing your sine waves.
Circuit design on its whole matters a lot more than any single DAC chip, or bit counts.
And products you can buy finished for 100$, produced in China for 10% of the cost, often is not very well designed in that regard.
Just some food for thought🙂
Doesn't all of your music come out of a digital/DAC source these days?
Whatever goes into the ASP comes out of a DAC first, no?
So why would it be the DSP xover that would be audible?
Jan
Whatever goes into the ASP comes out of a DAC first, no?
So why would it be the DSP xover that would be audible?
Jan
So how would it work on expensive gear then? Is the PWM somehow different?
I'm also curious what 'shabby quality caps' are?
Jan
For argument's sake let's say little green men from mars exist. How best to defend ourselves against them? It might be fun down the pub.
It might be useful for you to go back and look at the type of person claiming the existence of "digital sheen/artifacts" without of course defining what it is so that it can be measured and assessed. Think about why they might be making such claims. Then look at the type of person that is ignoring such discussions.
How can properties be assigned to something that is unknown (or doesn't exist). Tell us what the digital sheen/artifacts are in a technically intelligible form that can be measured and plenty here will be able to answer your question.
Passive vs active crossovers is one of the rare instances where the old technology doesn't retain one or two technical advantages compared to the new. It does retain marketing appeal though for a shrinking section of the population. Some nontechnical enthusiasts are going to come up with vague technical sounding notions to help justify what they would like to be true. Audiophiles come up with similar notions to help justify valve circuits, record players, DACs, cables, and much else to nontechnical enthusiasts. It doesn't make any of it true in a technical/scientific sense but does that really matter when it comes to a hobby interest? If you want to use a passive crossover for the midrange have some fun messing about implementing it and convincing yourself and perhaps some others that it sounds better.
it is in perpetual move , i still have a BSS 366T modified , after a Mini DSP SHD and a LAKE processor . with a 2 way speaker TAD 1601 / RCF ND850 crossed at 500hz . but i have plan to keep the BSS only for the woofer and goes analog active or passive for the compression driver . the best sounding crossover i had was the First Watt B4 regret to sold it 🙁
You always hear about buyers remorse. Not as much about seller's remorse. I find that often cuts deeper 🙁
How do you find the upper treble to sound on the ND850? I've sort of considered it to be more of a mid range driver than something that can play full range above 500Hz. Of course, I've only played with the 2.0, maybe v1 was different?
Am I wrong to detect some hostility from you Andy? You're right. I haven't been very technical in my explanations. Honestly the "digital sheen" is not something that I've ever detected, so I posted my question in the hopes that if someone on this forum had experienced this oft purported phenomena, they could speak to it.
However, I don't necessarily think it is always right to dismiss someone's impressions, even if they can't show what they hear to be reflected in measurements. Sometimes you can, (dismiss people, that is) and maybe this is one of those instances, but I try to keep an open mind 🙂
Although, I accept that maybe this is just my ignorance speaking 😉
Last edited:
I know this doesn't address digital but analog active systems set up this way are nothing new and many systems have been that way. My own active systems are all set-up this way. They are all hybrids and use active crossovers and passive components.
Rob 🙂
FWIW the DAC used in F8 is ESS9080 (released in 2020). In fact, it uses two of them paralleled for an improved SN.
Sorry for s slight OT.
i have the actual RCF ND850 2inch find it better than my previous 18sound NSD3N as it needs no correction to have a flat respons with my horn and it goes lower you can even cross it at 800/900hz in 6db and it sounds very good . the upper treble is not the most clean i have heard but a 2 way is always a compromise . i have to find the right tweeter to use up 8 /10khz . My faital HF108 with my 1khz horn doesn't do the job..
Attachments
Last edited:
this is both the BSS and LAKE processor inside . not so different sounding after the all the BSS capacitor changed to new ones. but sure a real nightmare for analog lovers with their Smps power supply , opamp stage and electrolyte capacitor coupling 😛
I use a Lake since 10 years now. It even serve as my main mastering converters ( ADC/ DAC)... nothing to complain about processing ( the Dolby eq are pristine) and to find difference in ADC/DAC i had to target units like Lavry Gold, Cranesong, Prism ADA, Antelope,... in other word invest same amount ( or up to x4 that amount of money) to have something 'different' rather than better... the only point that could have been 'bettered' is the clock imo.
Juanitox, apologize but i don't see a Lake in this. Maybe you have Clair's brothers version ( Contour?), if that is the case converters are different from Lake ( treatments are the same though but less powerful, there is no fpga in them).
Juanitox, apologize but i don't see a Lake in this. Maybe you have Clair's brothers version ( Contour?), if that is the case converters are different from Lake ( treatments are the same though but less powerful, there is no fpga in them).
If source digital then i see only benefits with DSP. With DSP crossover you can do something that is impossible with analog one. Digital errors are well below any DAC noise floor with correct math. Analog filters will add more errors. But if source is analog then here could be a debate.
The reason why passive crossovers sometimes sound better is not a matter of passive vs dsp but a matter of implementation by the user. When you use too much dsp you can literally suck the life out of the system. For instance if you measure your speakers REW with an USB measurement mic at some distance of the speaker where all kind of room and defraction artefacts show up and you start dsp'ing like crazy to make this curve flat it won't sound good at all. That's the tricky part about dsp. With passive crossovers one would not even think about making 20+ corrections because it is too complicated and too expensive.
Beg to differ. If you have never heard a DSP'd well-corrected reproduction system without all kinds of room stuff that messes up stereo placement and resolution, you've never heard a top-notch system.
Rooms are like effect boxes, you don't want that, but without a very good DSP system you can't escape.
And with modern tools it's not complicated or expensive at all. I can do it in less than 2hrs. Try that with a passive xover. Heating the soldering iron takes longer ;-) , let alone the calculations and trials and trials and trials ...
Jan
Rooms are like effect boxes, you don't want that, but without a very good DSP system you can't escape.
And with modern tools it's not complicated or expensive at all. I can do it in less than 2hrs. Try that with a passive xover. Heating the soldering iron takes longer ;-) , let alone the calculations and trials and trials and trials ...
Jan
I concur with both @Sjef and @jan.didden .
One can make very good (and bad!) systems with either approach.
Passive tend to be more difficult time and part-wise so IMO the process is more, say, concentrated and dedicated.
With digital DSP - at the beginning - there is a tendency to overdo things simply because one can!
The restrained approach after the novelty excitement phase passes, (hopefully) gives the expected results. Including the room EQ, which is very difficult to get right all-passive.
Anyway, one has to know the ropes whichever approach is used.
One can make very good (and bad!) systems with either approach.
Passive tend to be more difficult time and part-wise so IMO the process is more, say, concentrated and dedicated.
With digital DSP - at the beginning - there is a tendency to overdo things simply because one can!
The restrained approach after the novelty excitement phase passes, (hopefully) gives the expected results. Including the room EQ, which is very difficult to get right all-passive.
Anyway, one has to know the ropes whichever approach is used.