I happen to have a 2x8 in use at the moment. The unit is dead quiet and providing more detail as its anaolg counterpart.
i'm using it to drive an orion system. The ASP is creating more noise than the 2x8 minidsp
I like the system but my engineers heart is crying that my digital source is converted to analog, send it to the DSP, have A to D and D to A and send the analog outputs to the amplifier.
I don't have good experience with feeding the MiniDSP with a digital signal. Just does not sound right to my ears.
Anyone have the same results?
i'm using it to drive an orion system. The ASP is creating more noise than the 2x8 minidsp
I like the system but my engineers heart is crying that my digital source is converted to analog, send it to the DSP, have A to D and D to A and send the analog outputs to the amplifier.
I don't have good experience with feeding the MiniDSP with a digital signal. Just does not sound right to my ears.
Anyone have the same results?
Ya, it is lower-level DSP:
analog: 48kHz 24bit sampling, processing, DAC (works well together)
digital: resampling somewhere: 44kHz 16bit (CD) to 48kHz 24bit (or feed it 48/24 I2S) [2x4 native 48/24]
So not 96k 24bit sample, 96k 32bit processing like the other miniDSP stuff and other DSPs.
All those connections between devices give also jitters, problems of matchings impedance between boards...
Even with a DiracLive : the links are not SOTA. Spidf is biased, Tolinsk : active devices at each side of the signal... the digital signal of the source is enter with jitter. Another jitter comes between the Dirac device and the DAc device(s) than the measurement system is unable to capture and the dsp to correct... and I just talking about jitter in time domain transmission. Not talking also of the distorsions added each time and the EMC...
It could be more a dream with firewire connections, Masterclock device or async connections, but nobody uses firewire in DIY !
So the inputs & outputs of such devices stay problematic ! Are we resolving the speakers problem but forgett all the work we have done with D/A conversion and clocking time accuracy ! It's a little schysophrenic.
maybe the SQ winned on the speakers are more hearable than the problems given to the convesrsion chain ?
Even with a DiracLive : the links are not SOTA. Spidf is biased, Tolinsk : active devices at each side of the signal... the digital signal of the source is enter with jitter. Another jitter comes between the Dirac device and the DAc device(s) than the measurement system is unable to capture and the dsp to correct... and I just talking about jitter in time domain transmission. Not talking also of the distorsions added each time and the EMC...
It could be more a dream with firewire connections, Masterclock device or async connections, but nobody uses firewire in DIY !
So the inputs & outputs of such devices stay problematic ! Are we resolving the speakers problem but forgett all the work we have done with D/A conversion and clocking time accuracy ! It's a little schysophrenic.
maybe the SQ winned on the speakers are more hearable than the problems given to the convesrsion chain ?
Internally it has 56 bit precision if I remember correctly, it's only IO that is 48/24.
Or we simply add some latency, buffer and reckock the signal and blissfully ignore the problem. This is what both the MiniDSP and DLCP do on the digital inputs.
It shouldn't be. I suspect you have a problem with your ASP and/or PS configuration. Maybe some sort of earth loop?
The miniDSP 2x8 is pretty quiet on its outputs but your ASP should be quieter.
Dave.
If you've rolled the op-amps with something different (maybe LM4562) there have been reports of increased audible noise. (I didn't notice it on mine, so it seems inconsistent.) If you need help with that your best resource is the Orion Users Group.
(Don't want to hijack this miniDSP thread any more than it already has been.)
Dave.
Recloking an already jittered signal at the next input... is it a solution, no you reclok the jitter, the problem is less important but not solved ? Which latency added with the DSP ? How could a minidsp to be better with a random latency injected ?
For example a hardware mode with crystal in a spidf input with a Wolfson 8805 stay problematic ! Is a Minidsp better in its input stage in the digital domain (i don't speak of the worst A/D to D/A or A/D with tranport of the D in an extern DAC)
The problem of the output between the device and the DAC is staying... also
In fact we need than the inputs data to be in another transport protocol like IP... (SqueezeBOX) then convert these datas in one board with sota D/A or transport the digital to the DAC with A SOTA procedure (FIFO buffer ? Masterclock from the DAC or at worst async at the enter of the DAC with the slaved USB device and dac chip from each other in the same DAC). Not links of 2 METERS !
Or you accept than the better results on the speakers woerth it in relation to the degradation of the source to the DAC output !
It's a trade OFF , all those DSP doesn't solve all !
Who works on an device like RasberryPi networked with an embeded dataserver like a Squeezebox clone server and digital treatment on an embeded 32 bits chip or a slaved device to do that with a sota I2S output with matched impedance (uFL connectors and wires) ?
i think buying 1000 euros a DIrac like MINIDSP is gadget... but step by step it goes in the good direction.
What you are talking about has nothing to do with DSP as a tool.
The format is digital, and it sends X samples / second. Lets say we have 96 khz @ 24 bits. Say you say wanted to ignore jitters of up to 1 second ( assuming FIFO property of the stream ) then you could just have a buffer of 96000*24 bits in length, fill the buffer with incoming packets and then use a separate clock and output a packet exactly every 1/96000 th second.
There is no degradation, there is only the question of how long buffer you are willing to use. Say you buy a CD with some music on it. If you ship it by ship or by airplane doesn't affect SQ, even though the ship takes much longer. It just delays your audio a bit.
I assume if good buffers exist in each device like the DAC, only the last device as to be perfectly cloked and slave its input (buffered or not ?) from the DAC clock ? (or common external DAC as in professional recording ?).
Can the time corrections like phase be buffered in digtal domain and stay valid if jitters between devices during the transport exist ? The only perfect async protocol I know with reconstruct async information from source to receptor is TCP/IP !
Assuming the DSP is before the DAC, what happen if with a link like SPIDF, jitter is added when sended to multiple DAC ? Do the informations about phase (time correction) between driver stay accurate and the time variation with the datas not added with the quality of the links. My understanding is the transports of the datas with external devices are poor with MiniDSP...
But I'm not a specialist of this. But when I look at SPIDF with rca plugs... I'm suscpicious !
Can the time corrections like phase be buffered in digtal domain and stay valid if jitters between devices during the transport exist ? The only perfect async protocol I know with reconstruct async information from source to receptor is TCP/IP !
Assuming the DSP is before the DAC, what happen if with a link like SPIDF, jitter is added when sended to multiple DAC ? Do the informations about phase (time correction) between driver stay accurate and the time variation with the datas not added with the quality of the links. My understanding is the transports of the datas with external devices are poor with MiniDSP...
But I'm not a specialist of this. But when I look at SPIDF with rca plugs... I'm suscpicious !
Everything is the same if you buffer it, that's the whole point of digital =)
So yes, if you buffer in the right way then only the clock at the end matters, and as you can usually control that on the same circuit board then all is usually ok as long as the circuit designer knows their stuff.
If I'm not mistaken the reason why so many PA equipment have clocks and such is because they desire low latency, then you can't just buffer all the problems away. You still need some buffers but you want them as small as possible without getting errors. We don't live in that world though so we can do it the easy way =)
I have been playing with my new miniDSP 4X10 Hd.
to me, it sounds at least as good as the Beranger 2496 it replaced.
Its also easier to work with in every way.
Both the digital and analog inputs sounded good, and the unit was dead quiet on 94 db speakers.
Having said that, comparing it to a DAC costing many times its price seems senseless.
If in the future I try for my Ultimate system, I would go for a nanoDIGI board, replace the RCA connectors with BNC. DIY a shunt regulated supply to feed it, and use some high end DAC's to convert to analog.
Just my 2 cents.
Doug
to me, it sounds at least as good as the Beranger 2496 it replaced.
Its also easier to work with in every way.
Both the digital and analog inputs sounded good, and the unit was dead quiet on 94 db speakers.
Having said that, comparing it to a DAC costing many times its price seems senseless.
If in the future I try for my Ultimate system, I would go for a nanoDIGI board, replace the RCA connectors with BNC. DIY a shunt regulated supply to feed it, and use some high end DAC's to convert to analog.
Just my 2 cents.
Doug
Yeah, that's the later, balanced output version, it's not the RevB version of the unit that existed initially. (It's become somewhat confusing.)
But you are correct that the balanced output version does offer 2.0VRMS output limits for an unbalanced interface.
The basic (unbalanced) 2x4 unit is still limited to 0.9VRMS output.
Cheers,
Dave.
I have an unbalanced 2x4 (the most recent version, with the jumper set to allow for 2Vrms input). I have used it for various test/design purposes and currently have it in a system with high efficiency drivers, and at no point has its noise been an audible contribution to the noise floor of the assorted amps that have been used, for whatever that's worth.
The 0.9Vrms output is, of course, only a problem when it's a problem. I have amps with unbalanced input sensitivity from 0.18V to 2V (and I don't even own a lot of amps right now).
I also have a 2x8 board that I haven't gotten around to using yet, mainly because I bought it used and I can't decide whether or not I want to buy the case from minidsp or make one.
The 0.9Vrms output is, of course, only a problem when it's a problem. I have amps with unbalanced input sensitivity from 0.18V to 2V (and I don't even own a lot of amps right now).
I also have a 2x8 board that I haven't gotten around to using yet, mainly because I bought it used and I can't decide whether or not I want to buy the case from minidsp or make one.
Hi Freecrowder,
My apologies for responding to this thread rather late, but I just wanted to add a few comments of my own, since I have purchased 2 of the 2x4 MiniDSP units. My first purchase was just over 2 years ago and the first thing I did when it arrived was to run a few measurements to verify the published specs, which it did with ease.
My system at that time comprised of a 50W/8R/chan. NAD receiver with pre-out and main-in RCA connectors on the back panel that allowed me to connect the MiniDSP into the signal path.
I modified the gain of the power amplifier section in my receiver so that the output of the MiniDSP at 0dBFS = 50W/8R at the power amplifiers output. This optimized the S/N and dynamic range of the system. I’m very pleased with my MiniDSP which is wired into my main audio system, however, I only have one complaint and that is that there is no turn-on/off circuitry in the MiniDSP to stop clicks, pops and thumps coming out of the speakers when powering the system up and down. My solution to this is to sequence the power on the MiniDSP and power amplifier.
I’m currently in the process of putting together a second (bigger) system in my home that will include the second MiniDSP. Overall, I thoroughly recommend the MiniDSP 2x4, especially if you have the ability to optimize the signal levels in your system.
Peter
My apologies for responding to this thread rather late, but I just wanted to add a few comments of my own, since I have purchased 2 of the 2x4 MiniDSP units. My first purchase was just over 2 years ago and the first thing I did when it arrived was to run a few measurements to verify the published specs, which it did with ease.
My system at that time comprised of a 50W/8R/chan. NAD receiver with pre-out and main-in RCA connectors on the back panel that allowed me to connect the MiniDSP into the signal path.
I modified the gain of the power amplifier section in my receiver so that the output of the MiniDSP at 0dBFS = 50W/8R at the power amplifiers output. This optimized the S/N and dynamic range of the system. I’m very pleased with my MiniDSP which is wired into my main audio system, however, I only have one complaint and that is that there is no turn-on/off circuitry in the MiniDSP to stop clicks, pops and thumps coming out of the speakers when powering the system up and down. My solution to this is to sequence the power on the MiniDSP and power amplifier.
I’m currently in the process of putting together a second (bigger) system in my home that will include the second MiniDSP. Overall, I thoroughly recommend the MiniDSP 2x4, especially if you have the ability to optimize the signal levels in your system.
Peter
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.