Two words: "not digital".
What I mean by this is that with the DSP the sound was a little harsh, and long listening sessions were a bit tiring. Not so with the passive filter. It was a bit like going from a cheap amp to a very good amp. The overall balance is the same (bass, mids, treble), but it "sounds a lot better".
Passive filters can be a pain sometimes. I don't know that i am using the right DSP or what, but i didn't heard my passives as better in any respect even when the responses where very similar.
Thanks for the reply! I haven't experienced such "tiring" things. BTW I and my friends tried to figure out what people in the full analogue chain like, but we couldn't figure it out.
Just a single case. I believe that it is easier to construct a bad passive speaker than with dsp. I also believe that there is no audible difference if response, timing and crossover slopes are same for both.
I don't want to say that designing a multi-way dsp speaker is easy, but I can handle that and don't even want to start learning to design passive speakers! DSP and separate amp channels for each "way" give more freedom for choosing drivers and getting good timing, and are more amplifier-friendly. With dsp one can easily try and save/use different xo topologies and response curves, and also compensate some boundary- and room interferences in different rooms, do and undo without any extra cost!
Member
Joined 2009
Paid Member
I see a challenge for the DSP route as related to the quality of the DACs (not the chip sets but the whole implementation). With multiple channels it's tempting not to invest too much in parts etc. And there's also the expectation for very good measurable performance. So a company like miniDSP is under some pressure to deliver products with multiple DACs where it gets difficult to justify 'fancy' implementation. When you shop for a single DAC you can get very choosy because you are buying only 1 stereo unit, not multiple channels. In my subjective experience, modern DACs measure very well, are clean and flawless but more than half those I've heard produce fatigue in the treble for my ears. fyi - the miniDSP streamer unit has their best DAC implementation but I haven't heard it.
So DSP active shifts the expense from boutique caps and coils and lots of labour-of-love for fine tuning over to expensive DACs if you want the best results. I haven't designed and built a multi-way myself but am getting more and more interested in the idea.
So DSP active shifts the expense from boutique caps and coils and lots of labour-of-love for fine tuning over to expensive DACs if you want the best results. I haven't designed and built a multi-way myself but am getting more and more interested in the idea.
What about people that can't hear differences between DACs with errors well below conventionally accepted audibility thresholds?
Caps and coils are a bit different because they are pretty nonlinear things but the lack of measurements and conventional audibility experiments is something of a red flag if boutique caps and coils are to be sold to engineers rather than audiophiles.
Bigun, this thread is for beginners and learning to diy multiway speakers! Not about the ultimate performance achievable. Only when reaching the ultimate, ideologies like purely analog path and components or the best dac in the world may be relevant.
A beginner must be prepared to use the first cabinet to warm the globe by burning it! Drivers, xo parts or the dsp can be recycled over and over again!
A beginner must be prepared to use the first cabinet to warm the globe by burning it! Drivers, xo parts or the dsp can be recycled over and over again!
My first design in the middle of 80s' was 4-way passive. Today everything related to pre-engineering, measurements, simulation and QC is about 100x easier than 30+ years ago. So I don't buy any explanation that this is difficult. Low motivation or zero resources for this hobby. In addition, design process of passive, active linear and active digital is the same. Only difference is how you set filter parameters; with soldering iron or clicking/rotating some wheel in application.
Member
Joined 2009
Paid Member
sorry - I was responding to comments that DSPs sound harsh.
Anyhow, for multi-way I do consider myself a begginer as I've only ever made single driver full range. As a beginner I would really consider the miniDSP option as the most beginner friendly option compared to learning the black art of passive XO design and fine tuning. Why ? because passive approach benefits from a full junk box of parts for some trial and error for fine tuning which beginers do not have; because there are interactions between driver and finite impedance of the XO network that add complexity along with corresponding interaction between drivers, because it's not so simple to 'fix' driver artifacts. In DSP you can make drivers 'ideal flat' before you even begin and active approach means each driver is under direct control of an amplifier and can not interact with the XO or with each other. You can fine tune with a mouse, allowing quick comparisons between options in terms of subjective sound - no waiting for soldering iron.
Last edited: