Welcome, Guest
Username: Password: Remember me

NOTE: This is a "Community" forum. Please be mindful that community members are here to help as part of a community effort. We therefore appreciate your effort in keeping this forum a happy place!

If you have a specific issue (e.g. hardware, failure) and want help from our support team, please use our tech support portal (Support menu - > Contact Us).
Thanks a lot of your help in making a better community.
  • Page:
  • 1
  • 2

TOPIC: Looks like OpenDRC has some new competition...

Looks like OpenDRC has some new competition... 3 years 11 months ago #21836

  • Richard
  • Richard's Avatar
  • Offline
  • Expert Boarder
  • Posts: 127
  • Thank you received: 68
  • Karma: 13
Looks like there’s some competition out there for the OpenDRC…
Anybody seen this product?



www.aplaudio.com/conc2/products/apl1s/

Here’s how they compare…



OpenDRC is far more affordable - less than a third of the price!!! And on paper, OpenDRC still outperforms the APL1S, because is has 1.5x times more taps available. OpenDRC 2x2 plug-in also adds 6 parametric EQs, IIR crossovers and compressor for each channel. ((Not sure what ALP1s firmware offers besides FIR, or whether you have to also purchase their 320 Euro software…?)
I certainly wouldn’t buy the APL over the OpenDRC for woofers, but for tweeters, probably 4096 taps would be enough and 96kHz is very tempting too…

Anyway, though it looked interesting…
APL have certainly put more useful professional features into their hardware box.
I only wish OpenDRC had digital i/o AND analogue i/o in just one device!!!
Not to mention pro level +20dBu outs, front panel LED level meters, bypass switch (or even mute button?) and 16 presets to play with. Pleeeeeeese…

Wonder if this will spur MiniDSP into upgrading the beloved OpenDRC platform further and launch the ProDRC range we’ve all been dreaming about…?

This message has attachments images.
Please log in or register to see it.

The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21839

  • sly
  • sly's Avatar
  • Offline
  • Gold Boarder
  • Posts: 211
  • Thank you received: 119
  • Karma: 30
Richard wrote:
I only wish OpenDRC had digital i/o AND analogue i/o in just one device!!!
With a little modification you can add this yourself. The miniSHARC inside the OpenDRC-AN already has SPDIF inputs and outputs. These are at 3.3v. Consumer gear is at 0.5V and will need to be buffered to work with the miniSHARC's input... but it is doable. Twisted pear audio makes an SPDIF buffer to provide the 3.3V needed for the minisharc. It's labled as a "Single S/PDIF Level Converter Kit•Converts one consumer-level S/PDIF or AES-EBU input to TTL-level S/PDIF"

This will give you your input to the miniSHARC. To get an output, you need to build a resistor divider to reduce the 3.3V back down to 0.5.

You also need to use the miniSHARC 4x8 plugin. The SPDIF input is routed to the SPDIF input section on the MiniSHARC plugin. The 2 channels of SPDIF output are routed to channels 7 and 8 of the plugin.

Now you can use any combination of analog input and output. All you need to do is spend about $30 on parts.

I will add that miniDSPs unit doesn't do everything in one box. Not everyone needs everything. Some just need digital, some need analog. I like that they have different models to choose from, rather than asking people to spend ~$1000 on features they may not need. At least they let you purchase whichever unit you need and make it customizable for those who want more.
Last Edit: 3 years 11 months ago by sly.
The administrator has disabled public write access.
The following user(s) said Thank You: Richard

Looks like OpenDRC has some new competition... 3 years 11 months ago #21849

  • Richard
  • Richard's Avatar
  • Offline
  • Expert Boarder
  • Posts: 127
  • Thank you received: 68
  • Karma: 13
"...rather than asking people to spend ~$1000 on features they may not need..."

Their DIGI-FP board is $45, and the AN-FP board is $45.
When you're buying one version included anyway, the other option is only $45 extra... That isn't $1000's, that's small change, and MiniDSP should really modify their products to offer both analogue and digital all in one for eg. $50 more, rather than having three different partial solutions…(eg. still can't get analog > digital from any of those three) …and save people the headaches, hassle and frustration.

Don't know about you, but I'd happily pay the extra $45 = approx £29 GBP to have the second board installed alongside the first board (perhaps even on side of the device) if there was just a simple method to switch between outputs as desired, like a front panel button or in the 2x2 software plug-in. They haven’t really designed them in a way that allows that easily, otherwise why don’t they just do it?! I don't want the MiniSharc 4x8 plug-in because it reduces FIR taps from 6144 per ch down to 1200 per ch. Only the OpenDRC-DA8 runs that plug-in and has silly consumer RCA connections anyway, no XLRs at all.
If they made a professional 1U rack version with that plug-in running 6144 taps per channel, I'd buy it obviously.

On other hand, APL-1S is too expensive for what it is - you could buy all three versions of OpenDRC for the same cost!
The APL-1S is effectively like an OpenDRC-DI digital version with a stereo AD/DA converter built-in for $889 extra (£577 GBP approx) Silly money, I agree. Buying three of these boxes for 3 way FIR crossover would effectively be like £577 x 3 = £1731 (approx $2667) spent just on those APL built-in converters! You could buy a lovely Mytek 8x192 professional converter with 10 picosec jitter for less than that!!!!! Also, APL has fewer taps anyway, with only 4096 taps per ch (coincidentally the same as DEQX HDP-5) whereas OpenDRC has 6144 - that's why it's the winner. But for sheer convenience, not having converters built-in is still a major issue, especially if you can't afford the Mytek or three Lynx Hilos, or are simply using your FIR boxes for various alternate audio setups – not just one fixed pair of speakers in your living room. If adding the S/PDIF connection is really so easy as you describe above, why don't MiniDSP just put the facility in themselves, or make a universal I/O machine. People have been asking for 2 -3 years on similar subject, not just me.

The underlying problem here is conception.
OpenDRC isn’t conceived as a loudspeaker FIR crossover solution, it’s “conceived” as a stereo in / stereo out box for “Digital Room Correction”, which may be a lofty audiophile goal for some, but misses the much wider market as an active FIR linear phase crossover - because OpenDRC power has simply enormous potential in this area - far more than anything else on the market vs price - except it's just rather awkward to implement your 3-way setup with OpenDRC due to the stereo box hardware limitation, and you’ll need 3 or 4 stereo boxes to do your 3- or 4-way crossover in stereo, plus complicated routing with external DACs and 1>3 AES/EBU splitters etc.

MiniDSP could make it so much simpler to get a 6144 tap FIR multichannel crossover as they have all the technology already, but they’ve focused on broadband stereo or 5.1 room correction concept instead. Digital Room Correction is kinda pointless IMHO without having the speaker crossover anywhere near linear and correct in the first place, because you’ll be trying to correct for too many overlapping systematic problems inherent in the loudspeaker and erratic problems in the room acoustics at the same time, rather than getting the speaker output perfect first, then worrying about how good or bad the room is, as measured from a known linear output speaker.
Many more people in the world are trying to do crossovers than room correction. It’s a much bigger market. Look at the vast number of DSP loudspeaker processors for sale, at all price levels, vs the small number of dedicated room correction systems, which are mostly hi-fi market only. The DIY speaker builders, recording studios, and even car audio markets would probably buy custom 2x6 FIR crossover solutions if they existed all-in-one box. Currently they’d have to pay about $5K for something.
MiniDSP should address the crossover market more seriously and make a 2x6 or 2x8 or 4x8 version of the OpenDRC with 6144 taps per channel and professional XLR analogue and AES/EBU ins and outs, or ADAT multichannel optical. There's professional users out there who are paying $3K or $4K for so-called linear phase crossovers like the Danley SC48 with 24 dB/oct linear phase slopes onboard, or the Lake LM series, or Xilica or whatever which have very little real DSP power for FIR with maybe only 512 taps per ch or less. Many of these can only apply global FIR to the inputs and can't fully linearise the phase of drivers individually. Having the crossover slopes linear phase (eg. Danley, Lake) doesn't make the overall mutiway speaker linear phase at all. Each driver needs specific correction too. The Four Audio HD2 is probably the best professional solution and it costs $6K and the consumer DEQX is about that too. People will start paying more money if the hardware gets better. MiniDSP is determined to make everything great value and affordable - I'm not complaining about that! - but they'd be better putting more features / connections / multichannel horsepower and making a much more flexible powerful professional friendly machine - eg. 8 channels of 6144 tap FIR with analog and digital I/O in a 1U or 2U rack would be super great value and super affordable at eg. $1500, compared to all the competition. It's frustrating that they have the know-how and the technology to make that, but they decide not to, and in fact push the broadband stereo Dirac room correction range - yet still leave the basic problem of linearizing the multiway loudspeaker drivers themselves not really catered for in any one specific product.

Bear in mind, not everybody reading here is a seasoned electronics DIY expert with the time, skills and desire to modify hi-tech products at home using those Twisted Pear boards you rightly suggest, etc. I'm certainly not one myself, and I take my hat off to those who can do this clever stuff. (Otherwise I'd love to fit ADAT i/o board, LCD screen, level meters, and chain multiple SHARCs together to get more horsepower...) The rest of us are here because MiniDSP sell off-the-shelf, ready-made products, in a box, ready to use, with such HUGE potential, that they're too good to ignore. I know hours and hours, and years and years of searching on Google hasn't brought me any closer to a perfect FIR corrected 3 way stereo crossover setup than what I'd get running 6 discrete rePhase generated custom driver FIR corrections on 3 separate stereo OpenDRC boxes. I can't get more than 6144 taps running anywhere else, short of putting a Windows computer running BruteFIR in realtime into the audio chain.
Last Edit: 3 years 11 months ago by Richard.
The administrator has disabled public write access.
The following user(s) said Thank You: Thunderstruck

Looks like OpenDRC has some new competition... 3 years 11 months ago #21858

  • john.reekie
  • john.reekie's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 3481
  • Thank you received: 1498
  • Karma: 137
Do you really need 6k taps on every channel?
I am not miniDSP support.

"You must ask the right questions." - Dr. Alfred Lanning's hologram.
-> Have you read the User Manual??
-> Have you drawn and posted a diagram?
-> Have you posted a screenshot?
-> Have you posted your config file?
The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21861

  • sly
  • sly's Avatar
  • Offline
  • Gold Boarder
  • Posts: 211
  • Thank you received: 119
  • Karma: 30
I'd say no. The more taps you have, the more delay you have. 6k taps would make video unwatchable. A lot of people may not realize it, but the best systems I have heard make a compromise between FIR and IIR filters. IIR filters can influence phase without adding any delay (minimum phase). Proper implementation of an IIR filter can fix most phasing issues without having to resort to using FIR linear phase filters.

EQ problems are usually equated to phase problems. You fix one, you fix the other. Use rePhase to adjust the system with minimum phase IIR filters. Then fine tune it with linear phase FIR filters. You don't need that many taps.

Personally, I don't like that so many people are requesting that miniDSP remove IIR filtering to add more taps to FIR filtering. IMO, it shows a fundamental misunderstanding of how phase response and amplitude response are interrelated. Taps are useful for linear crossovers and for correcting minor room effects that affect phase. I tune most of my response with the traditional minimum phase (no delay) IIR filter. It can solve most response problems without adding any delay. Then I use FIR filtering to create the crossovers, correct for box port phase shifting and minor phase correction due to room and driver response errors.
Last Edit: 3 years 11 months ago by sly.
The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21876

  • Richard
  • Richard's Avatar
  • Offline
  • Expert Boarder
  • Posts: 127
  • Thank you received: 68
  • Karma: 13
Re: John Reekie
“Do you really need 6k taps on every channel?”

Obviously yes, to get enough resolution at low frequencies, given that the OpenDRC doesn’t include downsampling.
But - I wouldn’t need that much on tweeters if I’m combining IIR highpass crossover with FIR for EQ and phase correction, (otherwise if highpass is done with FIR filter alone, its precision down to ultra low freqs is essential to protect the tweeter.)

Re: Sly
Myself, I’m only interested in perfecting audio reproduction quality, eg. rather than watching videos, so latency isn’t a problem, per se, but a known side effect. Latency is usually half the FFT filter length for phase linear filters with central spike so can be calculated easily enough. For mastering studio / monitoring work, a loudspeaker’s accurate reproduction of sound is the most important criterion to me. If I wanted to do video work professionally, I would look into getting a video delay sync unit, and for home cinema type use, (or timing sensitive audio applications, such as live music PA system) I would have to probably accept compromising the absolute phase linear accuracy of the audio, as you say,and generate a best compromise IIR + FIR type filter setting with more acceptable latency, but that doesn't mean I'm accepting those same compromises for normal audio only music listening. I'd always want that to be as correct as possible. Wouldn't you too? So that's where the long FIR taps are needed.

Combining IIR filters with FIR can add improved resolution accuracy at low frequencies, as you say, and I think that's why MiniDSP includes them in the OpenDRC plug-ins. It would be a grievous mistake to not include them. However phase correction to at least 20Hz still requires a fairly long FIR with lots of taps, unless your speakers are already pretty spot on and closed box with very low resonant freq woofer. eg. My 4-way Dynaudio Finale modified closed box 360 litre cab is much better in that respect than my 21 inch subwoofer vented at 24Hz , or my 15 inch 3-way vented at 21 Hz.

I haven't been asking for MORE than 6144 taps per channel. OpenDRC is already the BEST horsepower spec convolution engine type product on the market anyway. I'd obviously be happy if they did upgrade it one day of course! I have asked them to put more than 2 channels in one box though, for routing convenience. But what I was talking about here was simply the input / output connectors, that's all. The APL has done things right IMHO regarding I/O. Also other features like LED meters would be nice to see.
Last Edit: 3 years 11 months ago by Richard.
The administrator has disabled public write access.
The following user(s) said Thank You: Thunderstruck

Looks like OpenDRC has some new competition... 3 years 11 months ago #21877

  • Richard
  • Richard's Avatar
  • Offline
  • Expert Boarder
  • Posts: 127
  • Thank you received: 68
  • Karma: 13
I’ve been thinking about these subjects in more detail…

NUMBER OF TAPS vs FIR LOW FREQUENCY RESOLUTION / INTERVAL SPACING

Consider a basic industry standard third octave 31 band graphic equalizer, which has been popular for many decades. It’s two bottom sliders are “20Hz” and “25Hz”, which gives control to the nearest 5Hz interval in the lowest bass range. Some audiophiles might turn their noses up at third octave graphic equalisers, but remember, to match its 5Hz resolving power at the same low frequencies, a 44.1kHz sampled digital FIR filter would need 8822 taps !!!!!

If there are N coefficients (taps) there will be (N / 2) – 1 intervals between 0Hz and Fs /2.
So for 6144 taps in the OpenDRC running at 48000 Hz sampling rate (Fs) there are 3071 intervals between 0Hz and 24kHz (the Nyquist limit). These are equally spaced, on linear freq axis, so they occur every 7.815 Hz interval.



Looking at all these other FIR hardware processors, ranked in order of the best they can do in terms of frequency resolution interval, none of them can match the performance of OpenDRC. It’s the best device on the market AFAIK, for raw DSP horsepower, and price-wise too!!!!! Only software convolution, eg. BruteFIR can really do a lot better. Even machines with downsampling like Four Audio and HOLM (not included, Outline iP48 also uses “Warped FIR” proprietary downsampling method but no published specs???) can only outperform OpenDRC using 8x downsampling or more, which specifically restricts you to woofer processing, not fullrange audio correction with that accuracy. So for that, OpenDRC rules, basically!!!
In theory, with 16x downsampling, the OpenDRC could resolve to 0.488 Hz intervals, if MiniDSP ever added this downsampling option in a firmware update - which seems fairly straightforward, since you just process 1 out of every 16 samples, and discard the other 15. That would only work for woofer applications below 1500Hz Nyquist limit.


NUMBER OF TAPS / FIR FILTER LENGTH vs HANDLING LOW FREQUENCIES

Assume output is 24 bit digital audio feeding into 24 bit DAC.
So anything less than 24 x 20 log 2 = –144.5dBFS is below the quantization limit.
ie. Dither noise masks the signal at this point anyway – I know, maybe lower level signals can be detected using noise-shaped dither, but for now, let’s say we don’t care about stuff below –144.5dBFS. That’s 0.00000630957% of fullscale level.

The symmetrical shape impulse of a linear phase 24dB Linkwitz Riley high pass filter at 100Hz, would have side skirt ripples stretching out about 19 milliseconds either side of central spike before their level was below –144.5dBFS.
It would therefore require a 38 millisecond FIR filter length to capture it on both sides of the central spike, if you wanted full 24 bit resolution down to –144.5dBFS.

I’ve included the relevant "FIR filter length" on the chart above. Anything shorter than 38 milliseconds wouldn’t even be able to handle this linear phase 100Hz HPF at full 24 bit resolution! At 100Hz, the wavelength is only 10 milliseconds, but the side skirts of the linear phase HPF filter require approx 38 milliseconds of elbow room to be encapsulated accurately.

If you regard 16 bit accuracy as okay, rather than 24 bit accuracy, you could measure the skirts to -96dBFS instead and get away with a bit less perhaps, but by the time you get down to 20Hz the side skirts are much longer, and (for 24 bit) you might need about 95 milliseconds either side of the central spike which is 190 milliseconds. So that's more than OpenDRC can handle, with only 128ms of FIR capacity. In other words, 6144 taps (128ms duration at 48kHz) isn’t enough for certain filter shapes at very low frequencies, and there must be a necessary, but unwanted compromise in performance.

Deciding exactly which arbitrary filter shapes you can safely fit inside your FIR filter with N of taps really depends on what particular shape the filter is, the rolloff levels, etc, and how wide its side skirts are, but at the 20Hz end of the scale it very likely will exceed the FIR length, so you’ll be left with either truncating the filter abruptly or smoothing with windowing functions, but ultimately inaccuracy of the filter at very low frequencies. So having more taps possible will be more desirable.

This message has an attachment image.
Please log in or register to see it.

Last Edit: 3 years 11 months ago by Richard.
The administrator has disabled public write access.
The following user(s) said Thank You: Thunderstruck, sly

Looks like OpenDRC has some new competition... 3 years 11 months ago #21878

  • john.reekie
  • john.reekie's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 3481
  • Thank you received: 1498
  • Karma: 137
Richard wrote:
Re: John Reekie
“Do you really need 6k taps on every channel?”

Obviously yes, to get enough resolution at low frequencies, given that the OpenDRC doesn’t include downsampling.
But - I wouldn’t need that much on tweeters if I’m combining IIR highpass crossover with FIR for EQ and phase correction,

It's just that if 2 x 6k is the limit per processor, then a three-way would require three processors, which is getting up there in complexity. If you can use (say) 2k, 4k, and 6k per channel, then you only need two processors, which is a much simpler proposition.

I don't know if there are faster processors that can be used, but given that all the units you list have similar limitations, I suspect perhaps not? (Impressive list, by the way :) )
I am not miniDSP support.

"You must ask the right questions." - Dr. Alfred Lanning's hologram.
-> Have you read the User Manual??
-> Have you drawn and posted a diagram?
-> Have you posted a screenshot?
-> Have you posted your config file?
The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21906

  • Richard
  • Richard's Avatar
  • Offline
  • Expert Boarder
  • Posts: 127
  • Thank you received: 68
  • Karma: 13
john.reekie wrote:
It's just that if 2 x 6k is the limit per processor, then a three-way would require three processors, which is getting up there in complexity.

Hi John,
Your first sentance is correct...
Yes, three-way stereo speakers (2x tweeters, 2x midranges, 2x woofers) requires 6x audio channels of processing, which, if using OpenDRC stereo boxes, requires buying three machines and spending $349 x 3 = $1047
Conclusion, it's definitely more complex (annoying) having a three-box setup than an all-in-one-box solution (how I wish MiniDSP would just make an all-in-one-box solution of equivalent power I could buy instead!) - but at least you know you've bought the best FIR solution on the market, for a still very reasonable price in fact. The Four Audio HD2 is another viable solution, if the downsampling Nyquist limits and tap ranges they result are workable for your particular drivers. If you wanted linear phase midrange from 100Hz HPF to 4500Hz LPF you might be out of luck. Also, that thing costs about $6K and also requires their expensive Monkey Forest software too, which is an intimidating nightmare of DOS prompt windows, although if one did spend hours reading the manual, one might be able to figure it out eventually (I haven't yet...) OpenDRC is totally user-friendly and actually outperforms HD2 for tweeter and midrange FIR duty, but HD2 does have it beat for woofers - though with much greater latency, of course. Everything has compromises! But again, it all depends on exactly how much filtering is needed. You can get a lot achieved with 6144 taps on woofers. I have a feeling that MiniDSP might release a new OpenDRC 2x2 plug-in one day with 2x and 4x and 8x downsampling, and that would put the OpenDRC miles ahead of the competition anyway. It's something they've talked about 2 or 3 years ago. It's not a hardware modification, just new firmware, so it's probably doable... Let's hope so ;)

john.reekie wrote:
If you can use (say) 2k, 4k, and 6k per channel, then you only need two processors, which is a much simpler proposition.

Second part of your the sentance is either wrong - or sorry, I really didn't understand what you meant....????
"Two processors" - does that mean two OpenDRC boxes? That would give only 4 channels of audio processing paths, so not enough for 6 channels of loudspeaker driver signals. Also there's also no way to divide the full 6144 taps per channel between more than one FIR processing block. If you don't use all 6144 taps, the rest are just digital zeroes at the end of the biquad file, sitting in an armchair with their feet up, not contributing to that convolution. You can't route another new audio channel through them independently too with another custom FIR filter assigned to them. There's no way to split the one FIR impulse into two isolated pathways because every incoming audio sample is convolved with all 6144 samples in the file, all the way through start to finish, to produce the audio output sample stream. That' s how convolution works. Any blank zeroes just won't contribute anything. Hardware wise, the OpenDRC only has one physical input and output connector patch per channel anyway, whether XLR analogue or digital format. You can only get 2 in 2 out channels max. So you would always need 3 stereo machines for the necessary 6 channels. I can't economise or avoid this. I'd need three stereo machines and I may as well use all 6144 taps for each of the 6 audio paths, because tweeter and midrange latency will need to match woofer latency to time align anyway.
Last Edit: 3 years 11 months ago by Richard.
The administrator has disabled public write access.
The following user(s) said Thank You: Thunderstruck

Looks like OpenDRC has some new competition... 3 years 11 months ago #21907

  • john.reekie
  • john.reekie's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 3481
  • Thank you received: 1498
  • Karma: 137
Richard wrote:
"Two processors" - does that mean two OpenDRC boxes?

No, it means two processors (in one box).
I am not miniDSP support.

"You must ask the right questions." - Dr. Alfred Lanning's hologram.
-> Have you read the User Manual??
-> Have you drawn and posted a diagram?
-> Have you posted a screenshot?
-> Have you posted your config file?
Last Edit: 3 years 11 months ago by john.reekie.
The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21913

  • Richard
  • Richard's Avatar
  • Offline
  • Expert Boarder
  • Posts: 127
  • Thank you received: 68
  • Karma: 13
"No, it means two processors (in one box)."

Which product is that??? Any specific weblink you've found? There's nothing that MiniDSP make with two processors in one box, to do what you're suggesting. Do you mean custom-build something myself - a DIY project with two miniSHARC processors in one box? It wouldn't be any help if I can't get the required OpenDRC 2x2 plug-ins to run on it, because it doesn't support that setup. In any case, I don't think building anything DIY is ever a simpler proposition than just buying a set of three nice matching ready-built OpenDRC boxes which work perfectly already. That's what I've gone for.
Last Edit: 3 years 11 months ago by Richard.
The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21914

  • john.reekie
  • john.reekie's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 3481
  • Thank you received: 1498
  • Karma: 137
I am not miniDSP support.

"You must ask the right questions." - Dr. Alfred Lanning's hologram.
-> Have you read the User Manual??
-> Have you drawn and posted a diagram?
-> Have you posted a screenshot?
-> Have you posted your config file?
The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21915

  • john.reekie
  • john.reekie's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 3481
  • Thank you received: 1498
  • Karma: 137
Richard wrote:
There's nothing that MiniDSP make with two processors in one box

Right. I know that. But you keep talking about a one-box solution with more horsepower that you think miniDSP should build....
I am not miniDSP support.

"You must ask the right questions." - Dr. Alfred Lanning's hologram.
-> Have you read the User Manual??
-> Have you drawn and posted a diagram?
-> Have you posted a screenshot?
-> Have you posted your config file?
The administrator has disabled public write access.

Looks like OpenDRC has some new competition... 3 years 11 months ago #21916

  • Richard
  • Richard's Avatar
  • Offline
  • Expert Boarder
  • Posts: 127
  • Thank you received: 68
  • Karma: 13
Hmm, well yes that is a SHARC processor, but I can't just buy that and easily assemble a working multichannel FIR convolution engine with digital and analogue I/O myself. I can't compile any custom software and I can't manufacture any circuit boards. I know some clever people out there can do this stuff,and obviously companies like MiniDSP, but sadly not me. I said exactly this in final paragraph of my 3rd post here on this thread.
Sorry, I was only looking to buy / compare products that are available to buy ready made. ( At most, I might attempt a minor modification if others had done it successfully and there were clear instructions online....)

My points originally were simply...
First, it is a shame there isn't a more powerful all-in-one box with more horsepower. MiniDSP could probably build something running their MiniSHARC 2x8 plug-in adapted to allow 6144 taps per ch instead of 1200. I could see that as a feasible product for them to sell. Certainly, I would buy it, and I'm sure a lot of other professional users would buy it too. I suggested something similar back in 2013 on this forum thread here. Others have asked for multichannel versions too. Maybe one day...

Second, it's a shame, even buying a three-box solution of OpenDRC, (as I have), that their connections still restrict you to digital or analogue, not both - when other similar products like APL-1S and most pro crossovers don't restrict you. Their -FP boards are relatively cheap so it wouldn't be impossible for MiniDSP to amalgamate them in one machine yet still have an affordable product. A better product for many.

I'm sure most OpenDRC users here will strongly agree on both points... although those interested in broadband stereo room correction only (rather than crossover duty) might not feel they need multi-channel, but they may want it one day for 5.1 video playback ,etc.
Last Edit: 3 years 11 months ago by Richard.
The administrator has disabled public write access.
The following user(s) said Thank You: Thunderstruck

Looks like OpenDRC has some new competition... 3 years 11 months ago #21918

  • john.reekie
  • john.reekie's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 3481
  • Thank you received: 1498
  • Karma: 137
I'm not talking about what you should build, I'm talking about what you think miniDSP should build :)

As in: "MiniDSP should address the crossover market more seriously and make a 2x6 or 2x8 or 4x8 version of the OpenDRC with 6144 taps per channel". I asked if you really need 6k taps per channel, for the simple reason that more taps means more processors (all other things being equal), and more processors means more complexity.

I had written more about that, but it's occurred to me that maybe miniDSP already have right solution: cascade or parallel multiple OpenDRCs. As you've done. If they just brought the SPDIF connections on the miniSHARC board out to the rear panel (and maybe buffered the output) and modified the OpenDRC 2x2 plugin a little bit, it would be a neat solution. (The plugin change would be to allow input selection SPDIF or I2S, and the SPDIF output switchable between processed output or a pass through. That way units can be dasiy-chained.)

Hm, and they need a blank front panel for units without a Vol-FP. (Only the first one needs it.)
I am not miniDSP support.

"You must ask the right questions." - Dr. Alfred Lanning's hologram.
-> Have you read the User Manual??
-> Have you drawn and posted a diagram?
-> Have you posted a screenshot?
-> Have you posted your config file?
The administrator has disabled public write access.
  • Page:
  • 1
  • 2
Moderators: devteam