Combining a digital dsp with analog crossovers are not always a good idea

[goat76]

Yes, mine are bigger I think and far better designed. I can totally see why they thought they had crap results - that is one hot mess and would take a whole lot of effort to sort out, especially with passive XO in the mix and who knows how they rolled the HP and LP into it for subs. Not going to even look at linked thread.

Edit: added a picure, I've only got 2 subs though. Feeling inadequate....

View attachment 233131

I'm sure your sound system sounds great, and the same goes for the system you now compare it to.

Those speakers are designed by Ingvar Öhman who has visited him many times and been consulted about the room acoustics. I know Amir has great respect for that man and have talked about him here at ASR, the following is a comment from Amir.

"I have known of Ingvar for a decade or more. First through his work with Swedish LTS in blind testing of electronics and then later with accolades about this speakers. At the time the waiting time was one year or more for his speakers? I also heard he has gotten out of speaker making business. True?"

 

[Soundstage]

For the experts out there, here is my question:
if I use perfectly time aligned speakers producing the ideal impulse signal response (like Kii Three) and if I use external DSP for EQ like Roon’s minimum phase filters, am I messing up everything? If yes, then what is the solution then?

 

[voodooless]

For the experts out there, here is my question:
if I use perfectly time aligned speakers producing the ideal impulse signal response (like Kii Three)

They won’t do that. You need a point source to get an aligned impulse response. Any other system will only have alignment within a very limited window.

and if I use external DSP for EQ like Roon’s minimum phase filters, am I messing up everything?

Not everything, it will just not preserve flat phase

If yes, then what is the solution then?

Use phase linear EQ. Room can also do that using convolution.

 

[Soundstage]

They won’t do that. You need a point source to get an aligned impulse response. Any other system will only have alignment within a very limited window.

Not everything, it will just not preserve flat phase

Use phase linear EQ. Room can also do that using convolution.

Thank you.
What about Dirac Live 3.0? Does it use phase linear EQ?

 

[Vacceo]

Sort of counters your "waste of time and money" statement though. As does the experience of many here who have used DSP with "analogue" speakers.

If you have good electronics and passives, I see your point: why spending a lot for marginal improvements?

That said, let me ask something: when the digital signal is processed, do we get a loss in quality from the source? I struggle to find the right technical terms, but in principle, does anything that change the source signal, in principle at least, worsen the final sound?

 

[BDWoody]

4. Why asume they are wrong ?

Because there is no evidence. Why is this so hard?

Why couldn't they do a test with more rigor if they are going to present their info as if it means anything to anyone.

 

[thewas]

Those speakers are designed by Ingvar Öhman who has visited him many times and been consulted about the room acoustics. I know Amir has great respect for that man and have talked about him here at ASR, the following is a comment from Amir.

"I have known of Ingvar for a decade or more. First through his work with Swedish LTS in blind testing of electronics and then later with accolades about this speakers. At the time the waiting time was one year or more for his speakers? I also heard he has gotten out of speaker making business. True?"

The typical appeal to authority argument which doesn't correct the obvious flaws nor replaces the missing measurements.

 

[goat76]

Because there is no evidence. Why is this so hard?

Why couldn't they do a test with more rigor if they are going to present their info as if it means anything to anyone.

Because the test was never meant to mean anything to anyone else, it was just loose talk on a forum about what they thought they heard.

It's possible another forum member will send his unit to LTS for a more rigor test.

This whole thing is blowned out of proportion, first by the thread starter, sorry Tangband, and then by the people responding who clearly haven't read that forum thread which is in the Swedish language.

 

[voodooless]

Thank you.
What about Dirac Live 3.0? Does it use phase linear EQ?

Dirac uses a combination of phase linear and minimum phase filters. It’s probably a very sensible trade-off.

 

[goat76]

The typical appeal to authority argument which doesn't correct the obvious flaws nor replaces the missing measurements.

Hey, I‘m just the guy trying to help you guys understand that all this has blown out of proportion.

I have never heard that sound system, but I know the person behind them have a very deep knowledge about both speaker design, blind testing, and room acoustics.
That's all I ‘m saying, but you sir, seem to be on the hunt for some internet-based fights or something like that, am I right?

I got no eggs in the basket.

 

[thewas]

Hey, I‘m just the guy trying to help you guys understand that all this has blown out of proportion.

I have never heard that sound system, but I know the person behind them have a very deep knowledge about both speaker design, blind testing, and room acoustics.
That's all I ‘m saying, but you sir, seem to be on the hunt for some internet-based fights or something like that, am I right?

Nope, I just stated the lack of objective facts in your post(s) which you try to compensate with the appeal to some names, it seems though you are as you started getting personal.

 

[Tangband]

The owner of the Ino speakers have a very good D/A converter, from Benchmark. Putting the Minidsp out of the chain , still using the Benchmark Dac , three persons thought the sound went better - in the treble area . This is ofcourse subjective thinking from them - but nevertheless a small warning to everyone who really thinks all digital gear always sound the same, or that no degradation can happen with digital gear .

Dont waste money on external digital roomcorrection programes If the sound worsenings in other areas than the bass.

You have to try the gear before buying. Dont get blind by awesome measurement results. Maybe they didnt measure the right things ?
You might also be better of with better active speakers, or different acoustical treatment in the first place.

Every digital roomcorrection ( including GLM and Dirac ) mess up the direct sound from the speaker.

Because we humans always lock our attention to the first arriving sound - the direct sound from the speakers, the correction from the microphone from listening position is often to big to be entirely beneficial soundvise. We dont hear like a microphone.

With that said - I use GLM for my speakers, its a soundadvantage , but the corrections are smaller than the program wants . Lesson learned - no automatic roomcorrection program works in a good way. You have to go in and change things manually.

 

[dualazmak]

To gain real benefits with a dsp, one should instead go fully active with the minidsp - with all the crossovers in the digital domain and no passive coils, resistors or capacitors in the signal path. The source ( for example a computer with a DDC) would then go directly to the digital input of the Minidsp SHD. No A/D:s involved. No external DAC outside the minidsp is needed.

I essentially agree with you.

In my case, I use DSP(XO/EQ/delay) "EKIO" within PC receiving digital signal from JRiver in the same PC, and DSP-ed multichannel (8-Ch) digital signals go into DAC8PRO through ASIO route by single USB cable. Even though I have (almost transparent) protection capacitors for midrange squawkers, tweeters and super-tweeters, the SP drivers are directly and dedicatedly driven by each of the amplifiers; you may find my latest setup here on my project thread; I also established 0.1 msec precision time alignment among all the SP drivers as summarized here.

As for the frequency response of the system, I carefully and intensively measured it at various "stages" in my setup in digital domain and analog domain including before-and-after the protection capacitors validating the "almost no effect" on Fq responses;
- Frequency response measurements by "cumulative white noise averaging": #392
- Where in my multichannel multi-driver (multi-way) multi-amplifier stereo system should I measure/check frequency (Fq) Responses? #393
- Frequency (Fq) responses in the completed system measured by using “cumulative white noise averaging method” under the present standard crossover configurations and relative gains:
Part-1_Fq Responses in EKIO’s digital output level: #394
Part-2_Fq Responses in DAC8PRO’s analog output level: #396
Part-3_Fq Responses in amplifiers’ SP output level before protection capacitors: #401
Part-4_Fq Responses in amplifiers’ SP output level after protection capacitors: #402
Part-5_Fq Responses in actual SP room sound at listening position using one measurement microphone: #403
Part-6_Summary, discussions, and a little step forward: #404, #405-#409

 

[Tangband]

Google translate is your best friend, -from the Swedish forum .
This is what they heard.

”I bought a MiniDSP SHD to test if this could replace CR80es with eq etc. as well as separate PEQ and my DAC.
That would have been a wonderfully flexible solution, I thought!


The testing went on for 3 days, a total of about 16-18 hours of listening and is explained here below very abbreviated.

Day 1:
I re-rigged and wips, the MiniDSP was the brains of the system, it would take care of all division, eq, level and inputs.

I started by configuring the filter and eq to be as close to my previous tuning as possible, for starters!
On the other hand, a little experimentation was done with adjusting the dividing frequency both up and down, but no, the previously established 125Hz was the ultimate.

But something crept in when I was listening to music, a feeling that I want to lower the volume, and that there was something wrong with the treble.
This resulted in a series of tests where I ran the MiniDSP both with its own DAC but also with the Benchmark DAC3 as the DAC and let the MiniDSP handle the digital domain alone.
I configured it so that with the push of a button I could switch between playing from computer to Benchmark or from computer to MiniDSP and Benchmark as DAC.
The levels were calibrated, but I also tested deliberately having a higher level on one and then the other, to hear how the tendency changes with level.
But the same results were found regardless of the sound level on the respective sources...
I tried playing with Roon directly on MiniDSP, and later with USB from computer and ASIO.
The phenomenon persisted, it doesn't sound like I'm used to, something has happened ...
I plug in the Benchmark DAC3 dom DAC and that doesn't completely solve the problem, although I probably imagine something positive happened there, but not enough to be the interesting thing here.

So, finally the Signal Chains look like the following:
Computer with ASIO ->USB to MiniDSP -> digital coax -> Benchmark DAC3 -> final stage to peaks in full range.
Or:
Computer with ASIO -> USB to Benchmark DAC3 -> final stage to peaks in full swing.

I came to the following conclusions:
Although MiniDSP only worked in the digital domain, it is unclear how and why...
But what could be heard was that fast transients in the treble became worse, less "snap".
And that the sound image was experienced more "on" and lacked a bit of the previously convincing naturalness.
This meant that I repeatedly found myself prompted to play weaker with MiniDSP than without, it just wasn't as enjoyable...

This discovery made me so doubtful that I was allowed to give up the testing and sleep a little and let my brain and ears rest.
So said and done, I went to bed so that with new and fresh ears I could listen more the next day.


Day 2:
I woke up at 7, got up and made coffee, turned on the stereo and listened!
Because I was so incredibly eager to try and find what I've done wrong, why does the MiniDSP sound like this?!
Two friends and audio geeks (Ronnie and Juanth) come over on the day and listen, this as they are interested in what this grunka can do.
And they also experience the same impact as me when we switch between the devices....
Hmm nothing is turned on in the software, MiniDSP should just be a digital and completely transparent link between computer and DAC.
But it is not that...

I was so disappointed, I really wanted this to be my modern solution to all the features I need.
I disconnect this and disappointedly I have to email about a repurchase, because unfortunately I can't have this in my system.
It doesn't sound bad enough to be a trashy product, but just not good enough to be included in my facility where I've set the bar for minimal compromise.


Day 3:
I can't give up, I have to go in again and test it, damn I must have done something wrong...

But no, the same result no matter what I do, no matter what I turn off/on in the software....
Yes No, now it can go back, so that's fine.

Dejected, dejected and a little sad, I go back with the box to Hemköp and hand it in for return.

Other experiences:
Dirac!
Funny, can be super good in rooms with worse conditions, with me it was of little use <300Hz compared to manual adjustments.
What it did, it did well at these low frequencies.
At higher frequencies, the sound is bad in all places other than the middle, so nothing for me.
But for listening alone, it can work, where dirac manages to give me the experience that the room is less audible.
However, at the additional expense of clarity in the soundstage, something just doesn't feel "right" here.
Much more detailed than that, I don't care to test Dirac.

I think dirac on multichannel for film could be cool

That was it .

 

[ZolaIII]

@Tangband; mic reading's, speakers position, construction and detailed multi axle measurement's including phase, resistance and distortion. Desired listening level's? Is there really any reason like DRM source for using MiniDSP in the first place instead doing it on a PC (general purpose CPU core's at best using MPC - FPU)? You don't have a large headroom as stated before and that what you do is best used on fundamental level (only a couple of worst offender's). Combine all like; positioning and of axes responses to regulate twitter response towards listening position, cut the lows properly and put a cork in the bass port to lower the ringing fundamentaly simply not making more of it if space refractions are the problem (and they pretty much always are not counting in 1% of assholes living in a large bunkers will well treated everything [large not edgy salon with high ceilings ideally shaped and treated]). Unfortunately finding measurements for sealed ported speakers is usually impossible (because almost no one does them [as they will mesure much worse regarding low end better say will have a roll of higher but will also less compress and will straighten in high uper high mids, waterfalls and response time's will however improve significantly] and in the accustic chamber it doesn't matter but in room it does a lot). Same applies to subs and ideally it would be a pair of them (larger driver's need longer time to settle down in a first place...).
Point is you need to point your approach carefully and use all you can into your advantage best you can.

 

[BDWoody]

This is ofcourse subjective thinking from them -

Right, so why continue to say things like:

but nevertheless a small warning to everyone who really thinks all digital gear always sound the same, or that no degradation can happen with digital gear .

It would need to be more than uncontrolled subjective claims for it to mean anything at all, let alone be a warning.

This is getting old. Claims with no evidence. Again.

Thread closed.