The answer is: it depends!
Are you a speaker builder who is always wanting to tweak your crossover or wants to be able to re-use the crossover for several projects? DSP is the better choice.
Do you want the flexibility of being able to implement all kinds of filter functions, not just the ones that the analog circuitry provides? DSP is again better here.
Do you want the lowest possible noise, and an output level that can easily work with any power amplifier? Ah, the better one in this case is the analog crossover.
Are you an expert, or want to become one, when it comes to designing crossovers for loudspeakers and know what terms like "baffle step" and "power response" mean? If yes, then maybe DSP is better for you. If not, and you think that the Linkwitz Riley 4th order crossover is the bomb, stick with a good quality pro audio crossover (and maybe a PEQ unit) and you will be happy as a clam.
Are you a speaker builder who is always wanting to tweak your crossover or wants to be able to re-use the crossover for several projects? DSP is the better choice.
Do you want the flexibility of being able to implement all kinds of filter functions, not just the ones that the analog circuitry provides? DSP is again better here.
Do you want the lowest possible noise, and an output level that can easily work with any power amplifier? Ah, the better one in this case is the analog crossover.
Are you an expert, or want to become one, when it comes to designing crossovers for loudspeakers and know what terms like "baffle step" and "power response" mean? If yes, then maybe DSP is better for you. If not, and you think that the Linkwitz Riley 4th order crossover is the bomb, stick with a good quality pro audio crossover (and maybe a PEQ unit) and you will be happy as a clam.
One other thing a DSP is very useful for is driver protection, I have just finished a sub and it uses peak limiting for mechanical excursion protection and rms limiting for thermal protection, quite simple in a dsp, much harder in discreet circuits. Also very high order filters (I use 48dB/Octave) are much simpler to implement without requiring precision components. However noise floor is higher and dynamic range more limited then with the best discrete designs and there is a considerable investment in knowledge and tools. Good fun though if you like trying out new things 
Ohh I should add I built my DSP from scratch, i dont know about the flexibility of commercial products such as you mention.
Ohh I should add I built my DSP from scratch, i dont know about the flexibility of commercial products such as you mention.
Last edited:
The miniDSP may not be the best device to compare to a really good analog XO -- it is cheap and its analog performance is inline with that.
Certainly most DSP XOs are much more versatile than an analog XO, but not without their downsides … and many have analog parts that are crap. DCX2496 a good example, used digital in/digital out it is quite good, but if not allof the analog circuitry needs replacement.
Is your source digital or analog? Then a digital XO using the digital in makes sense (look at the nanoDSP and use your own DACs), if you have analog source you'll need to digitize that to use a DSP.
dave
Certainly most DSP XOs are much more versatile than an analog XO, but not without their downsides … and many have analog parts that are crap. DCX2496 a good example, used digital in/digital out it is quite good, but if not allof the analog circuitry needs replacement.
Is your source digital or analog? Then a digital XO using the digital in makes sense (look at the nanoDSP and use your own DACs), if you have analog source you'll need to digitize that to use a DSP.
dave
I am using ADI and Sigma Studio has a combined compressor
very effective it is too no clunking noises or nasty smells from the JBL under sustained abuse BUT I have yet to let my friends take it gigging so time will tell Hello Dave.. my source is mostly digital.
I have a minidsp 2x4 and is cheap.. noisy and with a lot of jitter.
I built and use the ESP P09 xover. It is a very good basic design that is flexible, provides high quality signal if proper components are used (decent opamps, resistors, and caps), and is very quiet. The boards are inexpensive and not the highest quality, as pads can be lifted rather easily from desoldering parts if you need to make adjustments. I suggest using sockets to establish the proper configuration, then solder everything when you're satisfied.
The boards are designed for it, but you do NOT need to build L-R 4th. If you care to learn about different filter types and orders, you can easily customize the various filter stages to ANY type and order from 1st order to 4th or even higher with just a little creative thinking. The P09 is an excellent platform. You can learn a lot and truly DIY. You'll need a power supply, and the one he sells works fine.
DSP is for those who don't care to learn much about filters, but already know a bit about digital tools and just want a quick and easy way to start playing music. I doubt that it can provide equivalent signal quality. It might be a good way to establish the proper xover configuration, then transfer that to a hardwired analog xover.
Pro analog xovers are not really very flexible and sound awful in a home stereo. Avoid them.
Peace,
Tom E
The boards are designed for it, but you do NOT need to build L-R 4th. If you care to learn about different filter types and orders, you can easily customize the various filter stages to ANY type and order from 1st order to 4th or even higher with just a little creative thinking. The P09 is an excellent platform. You can learn a lot and truly DIY. You'll need a power supply, and the one he sells works fine.
DSP is for those who don't care to learn much about filters, but already know a bit about digital tools and just want a quick and easy way to start playing music. I doubt that it can provide equivalent signal quality. It might be a good way to establish the proper xover configuration, then transfer that to a hardwired analog xover.
Pro analog xovers are not really very flexible and sound awful in a home stereo. Avoid them.
Peace,
Tom E
Thanks Tom.. I have done some ESP projects ... I have a P3A, P101 and a P88..
And now I will be going with the P09... crossovers...
Do not skimp on parts quality. What Elliot writes about components not making a difference in sound is nonsense, and I've proven it to myself many times. WIMA MKP caps are the best generally available, but if you can cram a small value polystyrene with non-magnetic leads into the circuit, they are much better, especially for high frequencies. Opamps are also important. The LME49720 and OPA2134 are good choices. I use a combination of the two to avoid a specific "voice". Resistors are not as variable, but I believe non-magnetic parts such as Vishay CMF are excellent value.
Think about soldering sockets into the caps and resistors pads so you can easily swap out values until you find the proper configuration. You will NOT get it right on your first try, so it's fun to experiment and learn. Wish I had followed my own advice there, as it would have saved a lot of time and frustration, along with more than a few pads. I eventually had to buy a couple more boards and start fresh.
There are some useful on-line calculators to help select component values if you want to deviate from the L/R 4th. Remember that you are trying to sum the drivers acoustically, not electrically. The former is more difficult than the latter, but ultimately far more satisfying and accurate when you get it right.
Peace,
Tom E
Think about soldering sockets into the caps and resistors pads so you can easily swap out values until you find the proper configuration. You will NOT get it right on your first try, so it's fun to experiment and learn. Wish I had followed my own advice there, as it would have saved a lot of time and frustration, along with more than a few pads. I eventually had to buy a couple more boards and start fresh.
There are some useful on-line calculators to help select component values if you want to deviate from the L/R 4th. Remember that you are trying to sum the drivers acoustically, not electrically. The former is more difficult than the latter, but ultimately far more satisfying and accurate when you get it right.
Peace,
Tom E
So far, the "analog" folks say they doubt that digital can sound as good as analog, but they don't test reality. Digital folks talk about flexability, but they, too, don't test to see if there are sonic differences.
The only comparison that I know about, that I would put any credit on is the one Linkwitz did for his "digital challenge" DSP_challenge Worth reading for his requirements and for the results. Made me move from analog to digital myself.
The only comparison that I know about, that I would put any credit on is the one Linkwitz did for his "digital challenge" DSP_challenge Worth reading for his requirements and for the results. Made me move from analog to digital myself.
I would be interested to know how you went about testing and setting up your P09 after constructing it.
I'm not sure what you mean by testing and setting up.
I did not do any testing with instruments. I listened to frequency sweeps through individual speakers to determine if the xovers were functioning.
Setting up was attaching them to preamp outputs and power amp inputs. What else is there? They were mounted directly into the amp enclosures and hardwired.
Please reform your question, if necessary.
Peace,
Tom E
I did not do any testing with instruments. I listened to frequency sweeps through individual speakers to determine if the xovers were functioning.
Setting up was attaching them to preamp outputs and power amp inputs. What else is there? They were mounted directly into the amp enclosures and hardwired.
Please reform your question, if necessary.
Peace,
Tom E
Thanks for your reply.
I meant the process of verifying that the board is performing correctly, and that the variable components had been set correctly (although you might not have used potentiometers for gain depending on your version of the design). I was particularly referring to testing with instruments and, since you did not do any of this, this now seems irrelevant.
It seems to me that setting the gain on the pots (which are included in the version I have) is crucial to a flat response, also that verifying the crossover frequencies and slopes would be reassuring. These were the things I was particularly interested in.
It seems that these things haven't been done on this product before.
I would be interested to know your impressions of the quality of this crossover.
I meant the process of verifying that the board is performing correctly, and that the variable components had been set correctly (although you might not have used potentiometers for gain depending on your version of the design). I was particularly referring to testing with instruments and, since you did not do any of this, this now seems irrelevant.
It seems to me that setting the gain on the pots (which are included in the version I have) is crucial to a flat response, also that verifying the crossover frequencies and slopes would be reassuring. These were the things I was particularly interested in.
It seems that these things haven't been done on this product before.
I would be interested to know your impressions of the quality of this crossover.
Elliott posts graphs of the xover's response curves on his site in the project description. The filters rely on established formulas and can be assumed to operate accurately if the proper value components are used. The slopes are fixed by the xover configuration (number of stages) and cannot deviate unless you've really botched it. The corner freq's are set from the component values. If there is minor variation from the target freq, you would be hard pressed to hear it. There really is no need to measure if you've built the thing correctly. Unlike passive xovers, there is zero interaction between them and the speakers to distort the freq response.
The quality of the boards is okay. They are laid out very well, but the pads are a bit weak and can be lifted if you desolder too often. The xovers are very quiet with a decent power supply. As stated previously, the quality (not accuracy) of the sound depends somewhat on the opamps and passive components.
The gain is a bit tricky to get right and, if you've done everything correctly, will have the most influence on the final sound. I use the pots only on the low freq path and run my mids/highs full on. It takes a bit of experimentation to get it right. If you have the proper equipment, you could measure the response from your speakers. I don't. It requires some patience and careful listening over a time span of days to adjust by ear. That is the ultimate criterion for whether the xover works or not: if music sounds good, you've done it right.
Of far greater importance, is how accurately the xover integrates the drivers to each other. If that's what you're asking about, there are plenty of other threads that discuss measurements in great detail. I am not qualified or inclined to repeat it all here.
Peace,
Tom E
The quality of the boards is okay. They are laid out very well, but the pads are a bit weak and can be lifted if you desolder too often. The xovers are very quiet with a decent power supply. As stated previously, the quality (not accuracy) of the sound depends somewhat on the opamps and passive components.
The gain is a bit tricky to get right and, if you've done everything correctly, will have the most influence on the final sound. I use the pots only on the low freq path and run my mids/highs full on. It takes a bit of experimentation to get it right. If you have the proper equipment, you could measure the response from your speakers. I don't. It requires some patience and careful listening over a time span of days to adjust by ear. That is the ultimate criterion for whether the xover works or not: if music sounds good, you've done it right.
Of far greater importance, is how accurately the xover integrates the drivers to each other. If that's what you're asking about, there are plenty of other threads that discuss measurements in great detail. I am not qualified or inclined to repeat it all here.
Peace,
Tom E
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.