miniDSP kits, our answers to your technical questions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
98% of music data is 44.1.
I think that you may have missed the point I was making about 44.1kHz I2S being for DAC devices, not DSP, so let's move on...

And again 48dB slopes are also a must
Just happens that we're already working on it as we've already mentioned in many other posts. But let me say it once again. Yes, we are working on it.

Serious audio equipment should support native sample rates.
Please guys, do your homework before making such radical statements or trying to take our product down, it helps. Basics of the DCX2496 is that converters (ADC/DAC) & DSP IC run at internal 96kHz clock rate (not 44.1kHz). If you can find a DSP that runs its core at an off shoot clock rate like 44.1kHz let me know, i'd be interested to see that. Why: Because It would mean having to reload a different binary code/depending on the digital input sampling rate. So what everybody does (like us) is to include a SRC like the DCX2496 does (check out their spec sheet for more info) for you to be able to get any digital audio source to a generic DSP core. I'd also hope that as a startup that tries hard to please DIYers, spends time answering threads and make a flexible platform, you could either give us a little break before shooting us down or maybe let us prove that we can provide a more DIY oriented solution than a rack mount product from Behringer. Anyway, can we close that issue of 44.1kHz once & for all? Didn't I say earlier on that we would release a 44.1kHz plug-in? :confused:

I do understand that going the x*48khz route is the more easy way for you guys at the current stage
Not really actually, 48kHz multiples just happens to be the standard for most DSP audio platform sampling rate for many years now. We didn't invent anything... :)

. Using a PC as DSP (at 64bit!) would be by far more efficient in most cases.
Alright, let's summarize what we're building here: A low cost, low power board, that runs low latency digital signal processing for flexible configurations. Trying to get the same flexibility, latency on a PC isn't going to be as easy as it sounds. Trying to fit your PC inside your DIY amp/DIY preamp, DIY active loudspeaker isn't going to be as easy either.. :)

Alright, I'll stop here. I do understand you're trying to get constructive, but I think that your technical knowledge on DSP might need to fill some gaps. As said in other post, we're always happy to educate, but please don't try to patronize us too much with radical statements on a field (digital audio/digital processing) where we do know what we're talking about. You'll get much better results at maybe asking few things you may not know about DSP... :)

Best Regards

miniDSP DevTeam
 
Since this product is intended for DIY

How would one go about hooking things like Clip meters, RMS meters or remote volume controls, stereo/mono mode?

Feature requests:
external program select(extra bass mode, external sub mode, any user preset) would be useful for DIY active speakers
Rectifier on DC input to make stupid proof. After all, we're not professionals :)
 
Last edited:
Since this product is intended for DIY

How would one go about hooking things like Clip meters, RMS meters or remote volume controls, stereo/mono mode?

Feature requests:
external program select(extra bass mode, external sub mode, any user preset) would be useful for DIY active speakers
Rectifier on DC input to make stupid proof. After all, we're not professionals :)

Remote volume controls for master volume is currently the only supported feature. Other things like displaying RMS meters for each channel would require quite a bit more I/O so it will need another board. We have ideas, but nothing concrete at this point. All these controls being software based at this point.

External program selection is indeed a very good idea which we've been toying around, but building presets (i.e. saving in memory many settings, if not all) requires some redesign of our current micro controller firmware.
 
Shaun,

I may have to clarify that "remote control" is actually not "IR remote control". By volume control, I meant the fact that you could connect a potentiometer on the miniDSP kit to control the master output of that kit.

If you have multiple kits to control and want to have a single pot, just tie the pot pins together to a single potentiometer. Tricky/Smart diyers may realize that they could do the same with a DAC/PWM+LP, an IR receiver and a simple ATMEGA/AVR code to control the voltage (0 to 5V) remotely via an IR remote this time. :)

Question: Wouldn't controlling the level at the input source(pre-crossover) achieve the same feature for you?

Anyway, hope this clarifies, and let me know if you have further questions.

DevTeam
 
Question: Wouldn't controlling the level at the input source(pre-crossover) achieve the same feature for you?

Yes and no. And dunno.

Yes: it will work.

No: I think, basically, it seems desirable to keep the signal at the highest amplitude in the digital domain, so as to take maximum advantage of the available dynamic range, and in an effort to avoid loss of low level detail. So, logically, the volume control should be after the DSP section.

Dunno: I don't really know how much this matters (seems to satisfy those who use the Behringer DCX2496 crossover). Perhaps you can enlighten me otherwise?
 
Ex-Moderator R.I.P.
Joined 2005
Dunno: I don't really know how much this matters (seems to satisfy those who use the Behringer DCX2496 crossover). Perhaps you can enlighten me otherwise?

With regards to Behringer, are not certain issues related to that specific product
Maybe something about noise and overload, that some people try to deal with in a certain way of placing attenuation
 
With regards to Behringer, are not certain issues related to that specific product
Maybe something about noise and overload, that some people try to deal with in a certain way of placing attenuation

I speak under correction, but I got the impression that it had to do with the effects of attenuating (i.e., volume control) before the DCX input vs. after the analoguye outputs. Overload may be yet another problem, but I have not paid attention to that (I don't own a DCX2496).
 
From what i can tell this is a very cool board, happy i stumbled across this thread. Would appreciate if you could answer a few questions that come to mind:

1) You mentioned in an earlier post "use our SPDIF I/O card with Sample Rate Converter embedded to get the I2S to 48khz" - does this card already exist or it is forthcoming?

2) Website mentions Room Correction plugin. I assume this is also forthcoming. Could you tell me if it will be possible to run two plugins at once? For example Room Correction + 4 way digital crossover at the same time? This seems like a reasonable combo. I assume there will be a microphone input?

3) For the time being would it be possible for me to use a DEQ2496 for room correction (maybe upsampling) and digitally input this to the miniDSP for crossover duties, taking analogue outs from the miniDSP into the amps? (for an active speaker setup).

For what it's worth, i think the prospect of building the miniDSP into a speaker cabinet having a single digital input is far more attractive than having an external DCX2496 with a bunch of analogue cables. On that basis i find this an inspiring product, at least when a little more of the H/W interface + plugins come out. Nice work.
 
Yes and no. And dunno.

Yes: it will work.

No: I think, basically, it seems desirable to keep the signal at the highest amplitude in the digital domain, so as to take maximum advantage of the available dynamic range, and in an effort to avoid loss of low level detail. So, logically, the volume control should be after the DSP section.

Dunno: I don't really know how much this matters (seems to satisfy those who use the Behringer DCX2496 crossover). Perhaps you can enlighten me otherwise?

Shaun,

Your question raises a very good point about how much what's so called "gain structure" matters in audio. (any audio system as a matter of fact). I intended to write a longer FAQ about this topic, so please bear with me if I don't provide all the details/answers to your questions. Maybe they will follow up in my written up article.

The gist of gain structure, is that in order to obtain the best Signal to Noise Ratio, you'll need to feed a signal to your ADC, strong enough to be away from the quantization noise floor of the device (last bits at bottom). In other words, if you feed a tiny signal (close to the noise floor) to any audio device, then amplify it greatly with an amplifier (with volumes turned too high), you'll most likely hear noise that you shouldn't (hissing). So indeed, your no answer is close to what we would typically say. Don't make radical level changes on your audio chain, until the amplification stage. (Strong signal all the way through the audio chain)

Best way to tune an audio system IMO, is what's called unity gain structure, close to the ceiling of the ADC input stage, leaving some headroom for the crest factor of your signal. In general, 12dB being more than enough. Then control your level at the amplifier itself. This way, you're sure to maximize ADC/DAC of your systems.

Is Attenuation better in the analog or digital domain? It depends to what extend..Digital level control is really handy and works perfectly fine for few dB of attenuation/boost.. However just always bear in mind that the same problem I described above will happen at the DAC.

Bottom line, no single recipe for miracle. Everything depends on your connected equipment, signal strength and processing. Do a bit of thinking on your system and as Tinitus mentioned, make sure that you don't overload your input stage.

Hope this gives you the information you looked for.
 
From what i can tell this is a very cool board, happy i stumbled across this thread. Would appreciate if you could answer a few questions that come to mind:

1) You mentioned in an earlier post "use our SPDIF I/O card with Sample Rate Converter embedded to get the I2S to 48khz" - does this card already exist or it is forthcoming?

2) Website mentions Room Correction plugin. I assume this is also forthcoming. Could you tell me if it will be possible to run two plugins at once? For example Room Correction + 4 way digital crossover at the same time? This seems like a reasonable combo. I assume there will be a microphone input?

3) For the time being would it be possible for me to use a DEQ2496 for room correction (maybe upsampling) and digitally input this to the miniDSP for crossover duties, taking analogue outs from the miniDSP into the amps? (for an active speaker setup).

For what it's worth, i think the prospect of building the miniDSP into a speaker cabinet having a single digital input is far more attractive than having an external DCX2496 with a bunch of analogue cables. On that basis i find this an inspiring product, at least when a little more of the H/W interface + plugins come out. Nice work.

Thanks for your words of encouragements! :) In answer to your questions:
- Yes the SPDIF card does exist and is already running in a couple of beta sites over in HK. Manufacturing of our boards stalled 2 weeks ago because of Chinese New Year, but will start back next week. We hope to have the I/O cards on the market on the 2nd week of March if all goes well.
- We won't have a "room correction" plug-in per say. All our plug-ins come with either graphic EQ or Parametric EQ, so one could do the correction you need by looking at a spectrum analyzer. If your DEQ already did autocorrection on your signal, i'm guessing that it would have already configured a couple of PEQ where the algorithm guessed something wrong. You could always take those and copy them manually into the miniDSP if you want a single board solution.
As for the auto-room correction, we intend to make it work as a combination Room EQ Wizard (REW) software making FFT measurements + import feature to all our plug-in. More to follow lately when details become real.
- And Yes, I don't see any issues with the setup you just described.
We indeed believe that one of the strength of our product is how small of a DSP crossover solution you can easily put together for active speaker configuration. Something that isn't so common compared to large rack mount units.

Hope this information helps,
 
Yes, very helpful thanks.

One more quick thing, can you make the miniDSP (+ spdif I/O board) pass through one of the channels (left or right) to an SPDIF output, whilst processing the other channel locally to the analogue (crossover) outputs? I am thinking to put one miniDSP into each speaker cabinet. One will take spdif at its input, process say just the left channel for crossover duties but passing through the right channel (digitally) to an spdif output which will then connect across to the other speaker (which contains the second miniDSP for crossover duties). I hope it makes sense.
 
I have been following this product and thread since it started. I am anxious to try it but am unable to at this time.

So far the manufacturer support has been incredible. Very similar to the "original" Hypex support. (I stopped following that once my units were complete so I don't know the current state.)

Reason for the post:

As "mrwireless" hit on, there are MANY uses and implementations for this board. Has a "builders thread" started (edit: by anyone, manufacturer OR customers) with the different projects these are being used for? I'm sure the stated / documented projects would inspire a host of new, different and exciting uses.
 
Last edited:
Yes, very helpful thanks.

One more quick thing, can you make the miniDSP (+ spdif I/O board) pass through one of the channels (left or right) to an SPDIF output, whilst processing the other channel locally to the analogue (crossover) outputs? I am thinking to put one miniDSP into each speaker cabinet. One will take spdif at its input, process say just the left channel for crossover duties but passing through the right channel (digitally) to an spdif output which will then connect across to the other speaker (which contains the second miniDSP for crossover duties). I hope it makes sense.

Yes, it makes complete sense and you will be able to link multiple digital boards. This capability being one of the initial features we designed into our board.

We're still working on putting together all the documentation to clarify how it will work, but put it simply, you will be able to "link out" an incoming SPDIF/Toslink L&R from one crossover + SPDIF board to another SPDIF+Crossover board. The only step involved being to enable link out on one of the board such that it accounts for the very short latency of going through the first board (time alignment). All this will be detailed a lot more in our SPDIF board user manual, so please bear with us a bit more if you have more questions. Documentations should clarify most of your questions.
 
As "mrwireless" hit on, there are MANY uses and implementations for this board. Has a "builders thread" started (edit: by anyone, manufacturer OR customers) with the different projects these are being used for? I'm sure the stated / documented projects would inspire a host of new, different and exciting uses.

Thanks Troy, we always appreciate words of encouragement/suggestions to build a better community! :)

As for your question, we did start a builder/showcase forum on our website here but our community members, working hard at prototyping their little projects have yet to populate with all their pics. I'm sure that it will happen over time. I could certainly start a similar thread in this DIY audio forum...
 
Hi,

I'm also intrested in your DSP system, it looks really nice.

- Like Soundexcel, I'm also looking for a 4-way mono crossover plug-in with PEQ controls, will this be available in the near future ?

- Can you link 2 miniDSP boards as master/slave ( the master has a pot-meter and controls the slave ) ? Cause I want a "dual mono" DSP system, but only use one potentiometer to control both boards, is this possible ? I know, it's already possible with a stereo pot-meter. But I prefer a mono-potentiometer, cause stereo pot-meters have 2-3dB difference between L & R channel.

- To program 2 boards, you need 1 or 2 plug-ins ?
 
When will be a plugin available with 4way cross-over and the PEQ? I am planning to use one board per side for a full active system.
Is one plug-in sufficiant to set multiple boards?

4way crossover with PEQ is in the works but will require more time as once again, it's a complete rework of the UI/firmware compared to the current one. With the upcoming release of miniDIGI and miniAMP, our hands quite full these days but we'll try to release this in the next few weeks.

- Can you link 2 miniDSP boards as master/slave ( the master has a pot-meter and controls the slave ) ? Cause I want a "dual mono" DSP system, but only use one potentiometer to control both boards, is this possible ? I know, it's already possible with a stereo pot-meter. But I prefer a mono-potentiometer, cause stereo pot-meters have 2-3dB difference between L & R channel.

- To program 2 boards, you need 1 or 2 plug-ins ?
Yes, you can use a single potentiometer (mono) for 2 boards (or multiple) as a master volume control. Either ways work fine since they are just controlling the ADC of the micro-controller.

As for the plug-in, just so you know that you can only control a single board at a time. But to answer your question, a single plug-in is enough for 2 boards.
 
Shelving filter for loudspeaker baffle step correction - available?

Wow, I am pretty excited after finding this thread and your company. As a hobby, I build loudspeakers. As you surely know, the Behringer DCX2496 is very popular with DIY speaker builders because it has crossover and equalization capabilities that make it a perfect match for multiway loudspeaker crossovers. I see your products filling a similar niche, and with some interesting additional potential.

But...

One capability that the DCX has, that I have not seen on the datasheets for the PEQ or crossover modules, is the ability to implement shelving low-pass or high-pass filters. The typical loudspeaker has a drop in output between 100Hz and 1kHz that generally follows a 6dB/oct shelving filter response - using the mirror image shelving filter (e.g. with boost) compensates for this almost exactly. This is typically available as part of a PEQ, but I did not see it mentioned in the PEQ module datasheet. The shelving filter results in much fewer phase issues than a graphical equalizer does and so is preferable to that.

Are 6dB/oct and 12dB/oct shelving filters available in your products? Without them, I believe that your products will be missing a critical puzzle piece for loudspeaker crossovers.

Thanks,

-Charlie
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.