Mmmh.. That's a first for us, but I guess there is a beginning for everything.
I don't believe we ever got that kind of language over here and I'm not ready to set a precedent in being treated like dogs... For the time being, I'd politely recommend that you chill out a bit and be considerate of the language you're using. When you've chilled out and are ready to actually have a proper discussion as normal people communicate, we'll certainly be ready to answer your post and let you know what we're about to release next week. Other than that, I certainly have no reason to answer to this kind of post or even entertain that kind of attitude from anybody.
Tony
In have a great interest for MiniDSP and I'm glad to know there are people in MiniDSP, carefully reading diyAudio, and surely, taking notice of some ideas or feelings that get expressed over there. Sorry if you felt being treated like a dog. This was not the purpose of my post.
I'll be more careful with the wording next time.
Glad to read there is something new, planned for next week. Considering buying one MiniDSP system for evaluation.
The sole and only strong message I want to deliver, is that MiniDSP has a great potential, but quite curiously, at the moment, I don't see myself using it because it can't run some uncomplicated 2nd order IIRs I need. Those are the facts. This is the only "detail" preventing me buying a MiniDSP at the moment. You need to agree this is a stupid situation. Because it is notorious that, when dealing with IIRs, programming is a trivial task, consisting in uploading coefficients in memory, chaining the IIRs in a structured way, and pressing "start". It is so obvious, so simple and so trivial. One 2nd order IIR is six instructions, in DSP56K assembly ! And MiniDSP can't (or doesn't want to) cope with "random" IIRs, custom-made IIRs that depart from the trivial low-pass / high pass patterns.
What else does MiniDSP, more, better, that would explain why it is so limited in this particular aspect ? What do we get, as compensation ? Maybe the ability to change the IIR coefficients on the fly without "zipping" noise ? Yes or no ? What's the zipping noise floor, in case of on-the-fly changing IIR coefficients ? I definitely know that de-zipping a DSP application, with slowly changing IIR coefficients is not a trivial task.
But there are applications where you don't need the de-zipping feature. You set the filter, and that's it. You don't change it during playback. This is the point I was expecting you would raise. And this allows you to considerably simplify the wole software. And make it open. Using IIR coefficients info upload and IIR chaining/routing info upload. Is it the first time you get somebody querying about this particular point?
Regards,
Steph
I'll be more careful with the wording next time.
Glad to read there is something new, planned for next week. Considering buying one MiniDSP system for evaluation.
The sole and only strong message I want to deliver, is that MiniDSP has a great potential, but quite curiously, at the moment, I don't see myself using it because it can't run some uncomplicated 2nd order IIRs I need. Those are the facts. This is the only "detail" preventing me buying a MiniDSP at the moment. You need to agree this is a stupid situation. Because it is notorious that, when dealing with IIRs, programming is a trivial task, consisting in uploading coefficients in memory, chaining the IIRs in a structured way, and pressing "start". It is so obvious, so simple and so trivial. One 2nd order IIR is six instructions, in DSP56K assembly ! And MiniDSP can't (or doesn't want to) cope with "random" IIRs, custom-made IIRs that depart from the trivial low-pass / high pass patterns.
What else does MiniDSP, more, better, that would explain why it is so limited in this particular aspect ? What do we get, as compensation ? Maybe the ability to change the IIR coefficients on the fly without "zipping" noise ? Yes or no ? What's the zipping noise floor, in case of on-the-fly changing IIR coefficients ? I definitely know that de-zipping a DSP application, with slowly changing IIR coefficients is not a trivial task.
But there are applications where you don't need the de-zipping feature. You set the filter, and that's it. You don't change it during playback. This is the point I was expecting you would raise. And this allows you to considerably simplify the wole software. And make it open. Using IIR coefficients info upload and IIR chaining/routing info upload. Is it the first time you get somebody querying about this particular point?
Regards,
Steph
Last edited:
Thanks for your understanding and change of attitude, we appreciate.
In answer to your questions:
I guess that I'd maybe highlight a couple of details of the miniDSP project. The first thing is that its modular aspect makes it of interest to a very wide range of users. From very knowledgeable (Engineers like you) to not at all (complete newbies that are just discovering what is a PEQ). The important fact is that we have no intention to change that direction, we just want to build a DSP community, with whoever wants to give a try to using these fun little boards. (Advanced + Newbies)
Why am I highlighting this? Because it gives you context as to why initial plug-ins were designed in a certain way in the first revisions.
From day1, we always set our eyes on building a set of more advance plug-ins, which will indeed provide access to all biquad/IIR coefficients and is what I believe you're after.
- Would that plug-in be what newbies look for? Certainly not.
- Do we have time to "educate" newbies? Certainly not, we already get enough inquiries. We're not running a school, we develop products and that's where we want to spend our time. A lot of info is as you mentioned freely available out there.
- How long before you see some first signs of this firmware coming out? Very soon since we're already in QC stages of a first rev. More to follow in the next few weeks.
- Are we listening to users? Staffing multiple forums, answering emails within 24h and from listening to what all our community members say, we certainly believe we do. Maybe you should do a bit more reading of minidsp forum or DIYaudio to see that we actually implemented a lot of suggestions from our community.
- How to suggest ideas? Certainly not by beating on us. I can tell you right now that it just won't work.
We tend to listen and add to our "feature list" ideas when a) they financially make sense (engineering time = cost) b) we have time to implement them in our schedule c) they will benefit the community as a whole and not just one user.- Is there any reason why we don't have it by now? Rome wasn't built in a day and while things may look on your scale they take time, we've been very busy with other projects. That's all...
Hope this makes sense and as mentioned earlier, I'd suggest that you do a bit of homework before making statements. A plain simple question (aka: When will you release an advance plug-in?) is all it takes. All the other accusations regarding the work of our staff or direction we're taking with our startup company is definitely not the way you're going to get our attention.
Tony
Last edited:
You need to make a difference between people that :
1- won't be able to benefit from the MiniDSP advantages whatever the plugins you do, whatever the pedagogy you throw in (10%) - they may be happy but they don't see the benefits, actually.
2- want to use MiniDSP as a hi-quality equalizer, not caring about the fact it is DSP instead of analog (30%)
3- are DSP-litterate, and want MiniDSP upgrading their DCX2496 (30%)
4- are DSP-experts, and want MiniDSP to run some advanced experiments (30%)
If you want to keep the stuff simple, you need to drop the dezipping feature for profile #4.
I can understand that it is difficult to have profiles #1, #2, and #3 properly dezipped. There you need to apply strategies that depend on the IIRs in use.
Steph
1- won't be able to benefit from the MiniDSP advantages whatever the plugins you do, whatever the pedagogy you throw in (10%) - they may be happy but they don't see the benefits, actually.
2- want to use MiniDSP as a hi-quality equalizer, not caring about the fact it is DSP instead of analog (30%)
3- are DSP-litterate, and want MiniDSP upgrading their DCX2496 (30%)
4- are DSP-experts, and want MiniDSP to run some advanced experiments (30%)
If you want to keep the stuff simple, you need to drop the dezipping feature for profile #4.
I can understand that it is difficult to have profiles #1, #2, and #3 properly dezipped. There you need to apply strategies that depend on the IIRs in use.
Steph
Last edited:
Hoi Steph - or salut, depending on which region you're from
I think you should calm down a little, because you're probably not the only dsp engineer here around, and because MiniDSP is offering a product that, if it fulfills your requirements well use it, otherwise make your own stuff.
If you're an expert in 56K programming and need to get to the biquad level, then get yourself an evaluation board (Freescale's board is at 150 US$ now) and have fun with it. By the way, MiniDSP is based on an AD chip, not a Freescale one, it's a 28-bit path and probably no guard bits - MiniDSP may tell us.
Regarding the 'state variable', i guess you're talking about state-space structures. As far as i know, these structures are mostly effective for 16-bit implementations because they can produce a gradiant of pole density, with more poles in the bass than in the treble. For 24-bit or 28-bit implementations they're less useful, and probably completely unuseful if you perform properly your fix-point operations in double precision, which is what MiniDSP is claiming to do. Also, state-space structures require more instructions than regular direct form to perform - for sure more than 6 56K instructions.
Hope to see you soon on these pages presenting your own project of a fully custom dsp filter!
Best!
chaparK
I think you should calm down a little, because you're probably not the only dsp engineer here around, and because MiniDSP is offering a product that, if it fulfills your requirements well use it, otherwise make your own stuff.
If you're an expert in 56K programming and need to get to the biquad level, then get yourself an evaluation board (Freescale's board is at 150 US$ now) and have fun with it. By the way, MiniDSP is based on an AD chip, not a Freescale one, it's a 28-bit path and probably no guard bits - MiniDSP may tell us.
Regarding the 'state variable', i guess you're talking about state-space structures. As far as i know, these structures are mostly effective for 16-bit implementations because they can produce a gradiant of pole density, with more poles in the bass than in the treble. For 24-bit or 28-bit implementations they're less useful, and probably completely unuseful if you perform properly your fix-point operations in double precision, which is what MiniDSP is claiming to do. Also, state-space structures require more instructions than regular direct form to perform - for sure more than 6 56K instructions.
Hope to see you soon on these pages presenting your own project of a fully custom dsp filter!
Best!
chaparK
Hoi Steph - or salut, depending on which region you're from
I think you should calm down a little, because you're probably not the only dsp engineer here around, and because MiniDSP is offering a product that, if it fulfills your requirements well use it, otherwise make your own stuff.
If you're an expert in 56K programming and need to get to the biquad level, then get yourself an evaluation board (Freescale's board is at 150 US$ now) and have fun with it. By the way, MiniDSP is based on an AD chip, not a Freescale one, it's a 28-bit path and probably no guard bits - MiniDSP may tell us.
Regarding the 'state variable', i guess you're talking about state-space structures. As far as i know, these structures are mostly effective for 16-bit implementations because they can produce a gradiant of pole density, with more poles in the bass than in the treble. For 24-bit or 28-bit implementations they're less useful, and probably completely unuseful if you perform properly your fix-point operations in double precision, which is what MiniDSP is claiming to do. Also, state-space structures require more instructions than regular direct form to perform - for sure more than 6 56K instructions.
Hope to see you soon on these pages presenting your own project of a fully custom dsp filter.
Best!
chaparK
New Advanced miniDSP plug-ins
Great news,
Following my previous comments mentioning that we would eventually release advanced versions of our plug-ins allowing greater programming freedom of the Biquad filters, it just became reality!
For the past few weeks, we've been working on new sets of plug-ins and following extensive testing, they are about to be released tomorrow if all goes well.
Please read the following thread for more info : http://www.diyaudio.com/forums/minidsp/168164-minidsp-new-plug-ins.html
Have fun and let us know if you have further questions.
DevTeam
Great news,
Following my previous comments mentioning that we would eventually release advanced versions of our plug-ins allowing greater programming freedom of the Biquad filters, it just became reality!
For the past few weeks, we've been working on new sets of plug-ins and following extensive testing, they are about to be released tomorrow if all goes well.
Please read the following thread for more info : http://www.diyaudio.com/forums/minidsp/168164-minidsp-new-plug-ins.html
Have fun and let us know if you have further questions.
DevTeam
minidsp, your product looks great for the application I have intended for it. and I am planning on purchasing one within the next week or so, however, I have a few questions regarding the unit.
1) stated dc voltage is 4.5-24V DC, I am thinking of running this unit off 2 x 12V batteries in series, can the unit cope with 25.4 V?
2) what is the current estimated shipping time to the UK?
Chhers.
Jez
1) stated dc voltage is 4.5-24V DC, I am thinking of running this unit off 2 x 12V batteries in series, can the unit cope with 25.4 V?
2) what is the current estimated shipping time to the UK?
Chhers.
Jez
Hello Jez,
Thanks for your comments,
- 25.4V will be fine. Are you going for a portable system? Could be a neat application... tell us more.
- As for shipping, it takes about 3-4days to reach UK typically. Shipments are by EMS, as fast as UPS/Fedex but for 1/2 the cost.
Hope this information helps,
Thanks for your comments,
- 25.4V will be fine. Are you going for a portable system? Could be a neat application... tell us more.
- As for shipping, it takes about 3-4days to reach UK typically. Shipments are by EMS, as fast as UPS/Fedex but for 1/2 the cost.
Hope this information helps,
Yes I am going to be using this in a (semi) portable enclosure. There is quite a few people over at speakerplans who have made, and are making portable PA rigs which run off 12V batteries.
Most have used either pasisve designs or an active car x-over, both of which aren't very efficient.
The benefits of using a DSP over these are massive, and i'll be buying one very soon.
Jez
Hello,
We're indeed planning to release an enclosure on day, but just been busy with other tasks and our custom box supplier delayed a couple of prototype ideas we had. We'll try to push it the release on our schedule.
DevTeam.
Banned
Joined 2002
Sorry no we don't have an app note and explanation on how to. There is a fair amount of info online and using our advanced Biquad programming module, it just requires you to figure out the biquad coeff.
We can always see in the future how we can implement an easy applet maybe inside our site to suggest the correct values.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.