My main system is comprised of DML panels that can cover everything above ~150Hz or so. However they require heavy EQing to achieve a relatively flat response and are pretty inefficient. I am not super bothered.... the system is in my kitchen so the noise floor is often super high anyway (the dishwasher is currently running) but I was wondering what I'm missing from a more "pure" speaker. I have a "pure" build in the works to compare with this but I was just curious. I know nothing is free but to my ears they sound OK.
For music reproduction the power response patern is at the top of a loudspeaker priority, understand it as loudspeaker in room.
Then come little things that make a system sounds better if talking about digital gears : quality of the clock, quality of the analog output (never here with multi output dac, or at least never subtile; the noise floor as you say too; dynamic behavior till the woofer midrange band, i.e. from 40/60 hz... many modern ADC and DAC are weak there, at least most of the ones I have heard.
A flat response is not what you want at listening post, you want more a gentle dying spl curve from the low till the end. Imho first thing to correct. Anyway a noisy room is not what you want too and a kitchen most of the time cause the hard surfaces are just a nigthmare with hars sounding resonances/ringing.
Perhaps this thread belongs to the acoustic room section ?
Then come little things that make a system sounds better if talking about digital gears : quality of the clock, quality of the analog output (never here with multi output dac, or at least never subtile; the noise floor as you say too; dynamic behavior till the woofer midrange band, i.e. from 40/60 hz... many modern ADC and DAC are weak there, at least most of the ones I have heard.
A flat response is not what you want at listening post, you want more a gentle dying spl curve from the low till the end. Imho first thing to correct. Anyway a noisy room is not what you want too and a kitchen most of the time cause the hard surfaces are just a nigthmare with hars sounding resonances/ringing.
Perhaps this thread belongs to the acoustic room section ?
"Pure" dosn't have anything to do with DSP, or anything digital.
Pure is plain analog.... from source to speaker.
Because processing an audio sound, in particular with a digital device, becomes maddeningly complex and erases any original form of the original sound.
Pure is plain analog.... from source to speaker.
Because processing an audio sound, in particular with a digital device, becomes maddeningly complex and erases any original form of the original sound.
the simplier is to avoid digital filter but still use oversampling then. Imo ADC should be avoided in home environment, which call for digital libraries and DAC only in order the EQ, phase matching are made in the digital domain before numeric to analog conversion.
This is what I meant; I generally have some fall off programmed in after about 1kHz
And if I posted in the wrong forum my apologies; mods please move as necessary. I just figured this was the place as I am discussing DSPs but it is a broader philosophical question.
I meant you would have more answers in that forum section. You should try the 10 dB smooth dive from the beginning of your low end till 2ok.
Only measurements would tell you where are the bad ringing peaks of the kitchen, many harsh sound can comes above 1k hz (mainly from 4k to 9k to be large) but be carefful with some 600/to700 hz peaks too...
Only measurements would tell you where are the bad ringing peaks of the kitchen, many harsh sound can comes above 1k hz (mainly from 4k to 9k to be large) but be carefful with some 600/to700 hz peaks too...
The last place I'd worry about dealing with some 'super perfect tweaked response' from speakers is in my kitchen.
It's a kitchen!
It's a kitchen!
To answer the OP's original question, DSP becomes too much when the rounding-off errors due to the processing become too much. This is also one of the reasons as to why many people try to use fewer filters by combining the responses of as many as possible.
@diyiggy I will def (re) measure with the V2 of my panels. I will have to check the current filters on my V1 panels to see how bad it is. I think there were some big notches in the 3-4K range, and another one around 9K, with big boosts above 10K.
@wiseoldtech problem is that's where I listen to the majority of my music! Def a lemonade out of lemons situation. I am satisfied though.
@newvirus2008 hmm this is a concern. The filters I need are pretty numerous as well which prompts more complexity. I may try a set of simple filters to just deal with the most glaring issues in the response. Once I have some free time I will take some measurements
@wiseoldtech problem is that's where I listen to the majority of my music! Def a lemonade out of lemons situation. I am satisfied though.
@newvirus2008 hmm this is a concern. The filters I need are pretty numerous as well which prompts more complexity. I may try a set of simple filters to just deal with the most glaring issues in the response. Once I have some free time I will take some measurements
Notches where they are are not a problem imho...peaks above the average power response may be. I will tame the 10k till 15k if in your shoes to avoid fatiguing sound. But just my 2 cents.
Does latency matter for FIR if all is compute before the async reclocking node like usb to multichannel card...HD Streamer from minidsp for illustration ? But the measurement mic in the Nuc that is latency harmfull if the further Dac after is sending the impulse responses, It should nevertheless be ok for music playbacks as all is buffered between rhe Nuc and external Async DAC no ?
Dear diyiggy, there should be no problem with music playback, as the only thing that matters is the arrival time at the speakers. With streaming video, it may be necessary to delay the entire video stream to ensure that all computations are already complete and ready in time for sample output delivery. The memory required for holding back so much video information would be significant, especially for high resolutions like 2k, 4k etc.
However, if the video is already stored on a HDD/computer, it is enough that the video read be delayed, avoiding the memory requirement. Also, if the latency is small (say 10-20ms) it is easier to ignore things altogether as it may not be very noticeable anyway.
However, if the video is already stored on a HDD/computer, it is enough that the video read be delayed, avoiding the memory requirement. Also, if the latency is small (say 10-20ms) it is easier to ignore things altogether as it may not be very noticeable anyway.