One thing I pity did not have the chance was instant switch comparation between ANALOG xo and those DSPs. That would be quite interesting.
The miniDSP I brought had a gain stage inserted using TL072 op-amp. So the test somewhat provides some insight about its transparency...?
I admit I have strong bias about electronics being generally transparent. Although I had been caught with this as well, when using some "Sherwood" amp that someone gave me. My point is these electronics may provide a "micro difference" where I'm more interested to "macro difference".
Replacing dome tweeters with a compression driver would be a good example of "macro difference".
The miniDSP I brought had a gain stage inserted using TL072 op-amp. So the test somewhat provides some insight about its transparency...?
I admit I have strong bias about electronics being generally transparent. Although I had been caught with this as well, when using some "Sherwood" amp that someone gave me. My point is these electronics may provide a "micro difference" where I'm more interested to "macro difference".
Replacing dome tweeters with a compression driver would be a good example of "macro difference".
Good point.
And is this not exactly the outcome of this test?
Thet they sounded pretty much the same, and will therefore be equally good building blocks in a digitally cross over system.
So use these as the tools for making the macro differences, by using your prefered speaker units.
And as always remember the macro differences made up by the room acoustics, and the way they are treated.
Good thread
PS. And Merry Christmas to all
That is exactly my takeaway from these excercise.
First, that I hear no difference between direct wire vs. DSP (yes, with ADC->DAC)
Second, that it does not matter which DSPs to use. They sound the same.
That comes down to price and functionality, where MiniDSP is superior.
For example DCX can't do Linkwitz Transform which I consider very important.
But DCX don't need a laptop so it may be useful in some situation.
One DSP which was an outlier was the DEQX. I think it is either faulty of there is some missing configuration. They sounded "strange".
new to the digital crossover game...
I'm totally disappointing by the products companies offer and the complexity to have a decent home studio active crossover setup at reasonable price... the market could be udge if a ES9018 minidsp 8channel with FIR and phase correction appear at like 800$... but for now...
can't believe the stupidity..
it's not like it's a new domain... companies should know what is needed like passive volume or 32bit volume not to loose bitdepth or good digital input to bypass tons of conversions that color the sound or a good dac... etc....
to think to have a decent setup i would have to pay 2500$ just for the crossovers more than twice the price of my speakers... kinda freaky..
or diy my way thru hell..
I'm totally disappointing by the products companies offer and the complexity to have a decent home studio active crossover setup at reasonable price... the market could be udge if a ES9018 minidsp 8channel with FIR and phase correction appear at like 800$... but for now...
can't believe the stupidity..
it's not like it's a new domain... companies should know what is needed like passive volume or 32bit volume not to loose bitdepth or good digital input to bypass tons of conversions that color the sound or a good dac... etc....
to think to have a decent setup i would have to pay 2500$ just for the crossovers more than twice the price of my speakers... kinda freaky..
or diy my way thru hell..
This is a very old thread but I need to find the answer to this,,,,,,,,,,,
The signal gos from transport to DAC, then to DSP, then to Digital crossover,,, in some instances. WHICH CONVERSION ARE WE LISTENING TO?
Is it senseless to have a high quality DAC before the other two conversions or does the first conversion the DAC does, stay in the line to the end some how?
Someone please fill me in here,
Thanks so much,
The signal gos from transport to DAC, then to DSP, then to Digital crossover,,, in some instances. WHICH CONVERSION ARE WE LISTENING TO?
Is it senseless to have a high quality DAC before the other two conversions or does the first conversion the DAC does, stay in the line to the end some how?
Someone please fill me in here,
Thanks so much,
Normally the DSP is also the digital crossover (they're all math functions after all) so you definitely don't want to link any DSP to a digital crossover with an analog connection. Assuming here that 'DSP' means some kind of room correction. Ideally keep everything digital until the final D/A, normally before the poweramp.
Going back to your particular question, yes its senseless to convert to analog after some DSP, then feed analog into a digital crossover (which has to do ADC to get into the digital domain again).
<edit> I'm not familiar with tube amps - all my amps have to date been SS so not experienced enough to comment on that question.
Going back to your particular question, yes its senseless to convert to analog after some DSP, then feed analog into a digital crossover (which has to do ADC to get into the digital domain again).
<edit> I'm not familiar with tube amps - all my amps have to date been SS so not experienced enough to comment on that question.
Last edited:
People tend to jump to conclusions on this topic, before doing the hard work of actually doing a real and meaningful comparison. This test was all about testing those assumptions.
If an extra conversion is inaudible, how exactly is it senseless to add one, especially where the alternative may introduce other problems?
All else being equal, I would prefer to have just one digital to analogue conversion. The problem is that this tends to make volume control problematic. Now you need your DSP or a device after it to handle master volume control. If you use digital attenuation as suggested by MiniDSP as a simple solution, and have it all integrated with your remote, does it introduce any kind of sonic degradation? Without doing a proper listening test, why would you jump to this solution over one that has been shown to have (in a couple of examples) no audible artifacts?
The point I'm making here is to check your assumptions and consider finding out if they match reality.
What is a high quality DAC? Do you mean high priced? Do you presume that a high quality DAC is not used in MiniDSP due to the price?
I have been able to identify much more expensive DACs blind. This was not due to their superior sound quality, but rather easy to pick differences between the units like gain and tonal balance.
If an extra conversion is inaudible, how exactly is it senseless to add one, especially where the alternative may introduce other problems?
All else being equal, I would prefer to have just one digital to analogue conversion. The problem is that this tends to make volume control problematic. Now you need your DSP or a device after it to handle master volume control. If you use digital attenuation as suggested by MiniDSP as a simple solution, and have it all integrated with your remote, does it introduce any kind of sonic degradation? Without doing a proper listening test, why would you jump to this solution over one that has been shown to have (in a couple of examples) no audible artifacts?
The point I'm making here is to check your assumptions and consider finding out if they match reality.
What is a high quality DAC? Do you mean high priced? Do you presume that a high quality DAC is not used in MiniDSP due to the price?
I have been able to identify much more expensive DACs blind. This was not due to their superior sound quality, but rather easy to pick differences between the units like gain and tonal balance.
Thanks so much.
I'm running a DAC into a 3way digital crossover because there is no other input on the crossover to take the one cable output of my transport.
I really have a hard time believing the crossover could do DAC conversions as well as my DAC. That's what I thought though, my DAC is being wasted as an input for the digital crossover. I feel better knowing though. I'm amazed at how good this crossover is now. Do comment again if something comes to mind.
Thanks again,
Certainly its a very big 'if'. But if inaudible I would agree with this.
I've not been back to the start of the thread so I'm missing some context here, and no particular desire to catch up. Shown to two others to have 'no audible artifacts' would not satisfy me, I'd need to be sure I heard no artifacts.
Well sure, now check yours. Which assumptions have you assumed I've made?
One that sounds high quality - delivers listening satisfaction to myself.
No, I mean high sound quality. Price is a quantity.
I don't presume it, I observe it.
Sorry Paul, Abrax came on here to answer a question I had about DACs and how many times a signal gets DACed. My DAC has a Saber chip which I think this thread said was one of the latest good chips for digital volume control.
I'm with you Paul, one has to be pretty childish to have UN-blind testing alter ones decisions on how things sound. Yes it happens a bit but you guys are clearly in agreement. I think you did a great job (even kind of blind) and wish I was there.
The best,
I'm with you Paul, one has to be pretty childish to have UN-blind testing alter ones decisions on how things sound. Yes it happens a bit but you guys are clearly in agreement. I think you did a great job (even kind of blind) and wish I was there.
The best,
One way to implement the volume control would be to use a coarse analog volume control built using cheap 8-pole 5-position (8P5T) rotary switch and use the digital control only for fine tuning (or full control below the attenuation of coarse control). Doesn't work with remote, though.
Soviet Grade 1 Ceramic Rotary Switch 8 Pole 5 Positions 8P5T Non Shorting | eBay
The best still is to use a DSP for prototyping and subsequently translate settings into an analog xover with quality opamps. Obviously, it becomes more difficult to implement sharp notch filters and so on, but if you need those, there is something wrong with the drivers or enclosure anyways.
I have a couple of MiniDSP's and they are great, but analog xover beats it on noise and distortion, plus easier to implement volume control without upsetting the gain structure. Added advantage is no latency, important for AV and monitoring use.
I have a couple of MiniDSP's and they are great, but analog xover beats it on noise and distortion, plus easier to implement volume control without upsetting the gain structure. Added advantage is no latency, important for AV and monitoring use.
I have developed a 4-channel volume control (attenuation and gain are possible) that is designed to be used after the miniDSP. It's using the 2164 quad VCA, which has excellent channel tracking. I am getting what I hope to be the final version of the PCB, which includes an on-board power supply.
I'm also developing a 4-channel fixed-gain analog board as well. I should get the first prototype back late this week.
Whether you need volume control or just some more gain after the miniDSP, these should serve as solutions to that problem. As far as the quality of the miniDSP DACs go, I can't help you there! You can always go for the nanoDIGI unit, which has four spdif output channels to which you add your own DACs.
SSM2164 is obsolete though there is the Coolaudio equivalent. THAT VCA's might be better choice (though they are not pin compatible):
THAT Corporation 2162 Dual Blackmer Voltage Controlled Amplifier ICs
They have a word about using 2162 (s) as replacements for 2164:
THAT Alternatives to the Analog Devices SSM2164
I'm using the CoolAudio product, so not obsolete!
It's unique in that it has 4-channels that track very closely (typ. within 0.7dB or less) and you can set the exact gain in dB via the control voltage. This has allowed me to implement a panel control with a volume level readout in dB that is accurate:
http://www.diyaudio.com/forums/mini...-remote-volume-off-control-3.html#post4004082
There are a couple of better performing VCAs but they are 1 or 2 channels only, so you have to work harder on the tracking. This one was easiest for me to implement for my needs. The idea was to put this out until a more modern 4-channel volume control board was made available. For instance linuxworks is working on a CS3318 board... but that is still along way away. It turns out that the 2164 is not expensive and the performance is just about good enough to work very well. I've been doing some extended listening using headphones and I'm not hearing any issues. Measurements show that distortion is mainly 2nd order, which is "tubey" if anything, and the THD is relatively low. For measurements see:
http://www.diyaudio.com/forums/mini...-remote-volume-off-control-6.html#post4109043
I'm currently tweaking various components to maximize performance as much as possible but I am gearing up to build a bunch of these.
It's unique in that it has 4-channels that track very closely (typ. within 0.7dB or less) and you can set the exact gain in dB via the control voltage. This has allowed me to implement a panel control with a volume level readout in dB that is accurate:
http://www.diyaudio.com/forums/mini...-remote-volume-off-control-3.html#post4004082
There are a couple of better performing VCAs but they are 1 or 2 channels only, so you have to work harder on the tracking. This one was easiest for me to implement for my needs. The idea was to put this out until a more modern 4-channel volume control board was made available. For instance linuxworks is working on a CS3318 board... but that is still along way away. It turns out that the 2164 is not expensive and the performance is just about good enough to work very well. I've been doing some extended listening using headphones and I'm not hearing any issues. Measurements show that distortion is mainly 2nd order, which is "tubey" if anything, and the THD is relatively low. For measurements see:
http://www.diyaudio.com/forums/mini...-remote-volume-off-control-6.html#post4109043
I'm currently tweaking various components to maximize performance as much as possible but I am gearing up to build a bunch of these.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.