Software DSP

[Matt E]

Is there any software based DSP I can use with XXHighEnd to flatten out the frequency response of my system. I know some may find this herectical, but I fancy giving it a go to see what it sounds like.

Thanks,

Matt

[PeterSt]

In this post such a thing is addressed and how to do it : Re: Setting up the ultimate system!

I don't know anything about it myself.

[Matt E]

Peter,

Thanks for that. I'll check the link out and see if I can make any sense of it all. The thing that spurred my interest in DSP was actually that I was looking for a reasonable cross-over solution. I decided to go for the bi-amping/active cross-over option over passive cross-overs. In terms of analogue active cross-overs most reasonably priced ones seemed to use opamps and add colour to the sound. The external digital ones didn't include NOS DACs (my preference is for no oversampling) and again some not ideal components. The PC based cross-overs didn't seem to implement with XXHighEnd and well to be frank I think XXHighEnd sounds a damn site better than Foobar etc. and am very reluctant to go back to that in order to get a cross-over.

Any ideas/views? How do you approach the cross-over issue?

Cheers

Matt

[PeterSt]

Well, as I see it, the problem with digital cross-overs is that they *always* use DACs you don't like. Or it must be a coincidence.

The problem with the passive stuff is the components added.

The problem with active (analogue) kit is that again it adds components, which indeed as you say will colour as well. Besides, it is not said that all what can be solved with a passive solution can be solved with an active solution (although the very expensive solutions will or will at least come close).

While the above sums it up (but merely repeats your own findings) there is one further solution : be ahead of everything and do it in software.

IMHO this is the most tricky part of it all, because you won't be able to know how much it degrades. So, people like Pedal (from the post I referred to) convolute their whole homes to something they like better, while in the mean time it won't be much different from entering a random "hi-fi" store, listen to their reference system, and me thinking back 10 years, while "you" (hence Pedal etc.) may like it. Yeah, how to explain audio in words ...

So the only thing we can do for now, is stick to some theories, which should be understandable and acceptable, as long as we accept XXHighEnd to sound better than a random other player. Remember, this is important because it is the base for "theories" here;

The only single reason why XXHighEnd sounds better, is because much care is taken about the way the processing is done. I don't say "less" I don't say "lean", but I do say "taken care of". And to keep in mind : one Q or the other, it is and remains bit perfect (leave alone upsampling/volume of course). And so are the other players ...

Now, if you'd apply a "plugin", or other means like from the post I referred to, you will influence that processing again. With Pedal's solution I can't be 100% sure, because it happens in an external device (FireFace400 in his case), but I will put some money on that the influence happens anyway, if I only look at what I do to "get there".
So, it again will go wrong because the processing takes place within XXHighEnd (Plugin solution) or *after* it (FF solution).

But it won't go wrong when it happens *before* XXHighEnd ...

In order to more or less "proove" this, you must first accept that e.g. converting from FLAC cannot influence SQ in XXHighEnd. This is because this happens before playback even starts, while after that conversion, it's a normal .WAV playing. I know, some people even doubt this, and while I tend to believe them, I never saw the proof I asked for (which is easy to do, but never mind this for now). So, that assumed (preprocessing can't influence) ...

When the cross-overs are applied before playback starts, and of course assumed it is done right (which will depend on me by the time), all what is left is just another .WAV to play. No matter what has been done to it, the *processing* of it will be no different opposed to nothing happened to it. And *now* we are getting somewhere ...

All it takes now is the processing in XXHighEnd of enough channels for your XOver, while the processing of that more channels again should not influence SQ. Also, for 4 channels I already know it does not, because I can use that overhere (so yes, 4 channel operation is in XXHighEnd already (throuhout) but you can't utilize it (yet)).
Next, of course, is the necessity of an acceptable means of "DAC", or in other words, that we don't fall back to 4 cheap Behringer whatever devices because otherwise all becomes unaffordable (for -you could well say- everyone). I mean, if I had to use my before DAC which costed 2000 euro only, it would already be not woth it (just because the SQ of it doesn't justify 8000 euro). And so, one of the hidden agenda's of the Phasure NOS1 has been (almost from the beginning) that it would do 8 channels of the quality it does with 2 channels, but now against a relatively small higher price of maybe 500-750 euro (this is still not definitive). This is rather different from 2900 euro x 4.
And mind you, this is for 8 channels, in case you need them; when you can suffice with 4 channels it's actually in the base version, although not 100% (because of additional gain stages needed); maybe another 100 euro above the 2900 does this job.

Of course, all has its "price", and in this case this is about the necessity to use XXHighEnd for it. But hey, I guess this is was we were after in the first place. And so, it shouldn't be a problem.

By means of working on the DAC for such a long time, I know how much it differs when just components can be left out. So, I too have a nest of coils, resistors and caps for my passive cross-overs;
It will be a difference of day and night.
Again.

Peter

[Telstar]

I want to add one thing. I already do some software processing (which i'm not sure comes before or after xxhe plays the file), by means of a vst host and the soundcard drivers routing.

And I compared foobar to XXHE with clear differences in sound. This means that the xxhe processing (or not processing the sound) still has influence when using a software XO.

[Matt E]

Peter,

Thanks again. I will wait to see how your DAC turns out. Any better ideas of dates yet?

Cheers,

Matt

[pedal]

Just for the record:
At the moment I am in digital purist mode. My setup is as simple as can be: XX > Soundcard > DAC. No DSP in between. In the past I used Audiolense digital room correction. It is probably the best of its kind. At the time I started using it, I had a rather untreated listening room, acoustically speaking. Adding Audiolense cleaned up so much of the room problems. It was really “magic”. However, lately I have applied passive treatment of the room. Big fabric covered sitting group. A 26 kilo heavy wool carpet in front of the loudspeakers. 12 pcs of professional build absorber panels from US (UK branch – the name just slipped my mind). Also, I moved my sofa 60cm away from the back wall, further into the room. By passive methods I have managed to cure unwanted bas modes and reflections. Now I have a very clean and distinct direct sound from the loudspeakers, with 1st and 2nd reflections arriving with necessary delay/attenuation. All in all I don’t miss the DSP room correction.
I didn’t retry the DSP room correction AFTER the passive room acoustic treatment. Maybe there is more to gain? (The less it has to correct the better it will perform). -Someday I will have to double check this point.

Here is a hypothesis I have been playing with lately. It might not hold water, but any way here it goes: I think the mathematics of the best DSP today is “perfect”. You can perform EQ, crossovers and attenuation (within a bit depth limit, of course) with no degeneration of SQ. It’s “perfect”, let’s say. Or at least better than what’s obtainable in the analogue domain.

However if you want to do DSP based on a measurement (microphone), then I think the measuring process itself adds a slight margin, or error. The microphone, the cable, the microphone analogue preamp, etc all have a signature. Also the placement of the microphone will be different from your ears. Finally, and probably most important, is the sound level. You measure at a given level. But playback can be 10dB louder or 10dB quieter. I think the room response (the room’s acoustical signature) is different at different sound levels. Also the loudspeakers change character at +/- 10dB levels. So, all in all, while the best room corrections (like Audiolense) can correct a whole lot of acoustical errors, it still leaves its signature on the sound. Not because the program (mathematics) isn’t good enough, but because it’s is difficult (impossible?) to get a "perfect" measurement.

Here, we are talking about margins of course, but that is what this forum is all about. Margins.

Anyway, I wrote this under the influence of 2 dl Cognac after dinner and wine, so I might rephrase myself next Sunday midnight. Cheers! 

PS: If you didn’t bother read all the text above, my conclusion implies that I think the upcoming XX digital XO + Phasure multichannel DAC will outperform my SOTA analogue electronic XO’s in my active driven 3-way loudspeaker system. Cheers again!

[PeterSt]

Thank you Pedal. I think you were still quite understandable after 2lt of cognac. No problem.

Let me add one other thing in the same realm (only tooth paste in my mouth) :

Still assuming XXHighEnd - and its settings - can make a huge difference, we are not *measuring* through XXHighEnd. So, the generator concerned "sounds" different, and therefore alone will create "wrong" correction files. It listens to the wrong player ...

Peter

[leifchristensen]

I listened to PEDAL´s system wiyh AUDIOLENSE and I didn´t like it!
It sounded dried up and lacking in harmonics
when we fiddled with it, the less it was utilized, the better the system sounded
conclusion:
only use digital eq as LAST RESORT solution after all other options have been explored
imo
best
Leif

[PeterSt]

It is a very difficult subject. I never like such things too, while someone like Pedal may "swear" by it. All people with good ears and reputation, but IMHHHO without the experience to hear it all really. It also depends on the further system of course, and while one may not like horns (like yours Leif, or mine), they unveil things a normal speaker can't do. Not that everybody must get horns now, but the more "resolving" a system is, the better anomalies will be heard. Yes, so logic ...

In the very end, plainly nothing is allowed to interfere. This is just my observation in my stribing to that. The more is out of the way, the better it is. The more things are 1:1 with digital as the origin, the better it gets (under the "known" digital restrictions that is, like proper filtering).

Hopefully there will be one day soon, that I can show you / Pedal, that even bass traps and the like are not necessary at all.

Regards,
Peter

[leifchristensen]

HORNS TO THE PEOPLE!
by the way:
who wants to "trap" the bass?

RELEASE IT   

[pedal]

I dont use bass traps. Only some no-nonsens absorbers shortening the RT60 from lower midrange and up. It is needed because I usually play LOUD. My 2 linesource 225cm tall bass towers (7 pcs 12" each channel) goes flat from 80 to 4Hz about. Very clean and tight bass. Almost too tight. Sometimes listening on jazz/rock I long for that extra kick you get from ported cabinets. My cabinets are sealed.

[leifchristensen]

when that occurs; just start your car and go South-East
welcome!
best
Leif

PS the only thing that´s sealed in my house is my destiny!

[PeterSt]

For a short moment I read "decency". So I thought I better read that again. And indeed ...

[PeterSt]

Sometimes listening on jazz/rock I long for that extra kick you get from ported cabinets

Yeah, let's kick some jass.

Pedal, don't start drilling ports in your good speakers. I closed mine too. I just open some doors once in a while.

South East seems the wrong direction to me. South west is better.