There is no option to eq the sound on the NDX2 or any Naim streamer source.
If using Roon there are some parametric digital eq controls as well an option of creating a custom filter response for your room using a tool such as REW on the Roon server (āCoreā)
However for max sq itās best to avoidā¦ and address your room nodes by changing speaker locations or possibly furnishings.
Sorry but this is not alway practical or solves the solution. I would also argue that in my experience it improves sq if done well by someone who understands how to create the proper filters.
I use Roon for room eq using custom convolution filters created for me by Home Audio Fidelity his service and knowledge is second to none. Itās only improved what my system sounds like. Without them my Room dominates and this has been with different speakers and positioning. I canāt get rid of my rooms resonance or modes and there are quite a few in the bass region due to its dimensions without having seating and speakers in a completely impractical positions for the room to function as a living space. I have been trying again without DSP recently to see if I can not use it and itās just so much better with it.
I realise itā may not not be possible, but as somebody who has experience with DSP and digital filtering (even going right back to my undergraduate final year project) I can confidentially say filtering and then driving regular loudspeakers is simply not a totally reliable way of removing room resonance issues whilst retaining fidelity in a regular listening environment.
Filters are not magic, and will necessarily produce phase shifts at various points across the frequency response of the speakers as well, and any adjustment will, depending on the degree of adjustment required, will typically be optimum over a small restrictive listening areaā¦
So yes if you have really no choice it can be a last resortā¦ but where possible prevention is better than cure and the OP now suggests they have started experimenting here with positive outcomes.
Any sort of digital filtering for room optimisation is best done when only subtle correction required in my experience, if there are significant room resonances, overhangs and reflections, smearing will occur and obviously when the signal is distorted through smearing it canāt typically be recovered. So filtering almost by definition canāt improve SQ, but can help match a sound to the limitations of the listening environment with a cost of a loss of overall quality / accuracy but with a gain of overall aesthetic or sound impression.
Yeah ā¦ I remember the days of Graphic Equalisers and Tone Controls (Treble/Mid/Bass), which weāre all considered āwork of the devilā back in the day - simply doing nothing but degrade the sound.
I guess these modern versions of DSP arenāt dissimilar, though I know absolutely zero about them.
Being able to place loudspeakers anywhere to fit in with domestic room layouts at home, without compromising sound quality, would be excellent. Itās not always possible for everyone to have speakers equidistant from walls and avoid furniture etc.
Music is for enjoying at home and not in a sterile studio type setting ā¦ sat in a single seater chair, facing a pair of speakers and an array of expensive hifi equipment isnāt the most social of activities.
Much better to have the whole family in a comfortable surrounding doing whatever, with high quality music playing away
The Linn Space Optimisation concept sounds good in theory (from all their promotional material), but perhaps itās the same as DSP and should perhaps be avoided?
If Naim could see the benefits, I guess they would incorporate such technology in their products too.
There is no such thing as a āconvolutionā filterā¦ convolution is a method of number series multiplication, that is a DSP processing technique ā¦ it can be used in DSP to multiply a filter response kernel; ie a set of specific sample values shaped in response to function over a window of continuous samples. The actual filters would be FIR filters, and you convolve their kernel response with the sample stream over a window sizeā¦ the size of the window is sometimes referred to as the number of taps.
The advantage and disadvantages of FIR filters over IIR filters has been well discussed on the forum. There is no magic source.
Simplistically the main advantage of FIR filters over IIR filters is usually linear frequency/phase response at the possible cost of increased processing / arithmetic noise.
Naim use IIR filters in their DACs currently. chord Electronics currently use FIR filters.
In Roon the parametric Eq uses IIR filters, whilst the Room optimisation Eq uses an FIR filter.
For those that would like to understand And not rely on marketing misnomers, here is a link to Dr Steven W Smithās superb DSP guide designed fir scientists and engineers; ie not too mathematically heavy. Smith has kindly made his book available on the web for allā¦
Convolution [in DSP] http://www.dspguide.com/ch6/2.htm
Your nitpicking Simon itās a set of filters thatās use in the Convolution engine not in the standard peq and what is acheivable is way beyond what peq can as the filters also help with impulse response, phase alignments. I am no expert and donāt claim to be but my ears tell me its makes a substantial positive impact and more so than any other tweak. DSP is the future if you ask me. Most high end actives these days utilise in some form to compensate for room placement my next amp purchase will need to have it. I am done with the old ways hifi needs to fit into my life not the other way round.
I do appreciate the detailed explanation of this Simon thanks. For my room in particular I use Roonās capabilities for importing convolution filters that were generated by a botique outfit (HAF) for subtle room optimization. For my less than optimal listening environment, itās made a very very nice improvement.
Can you explain what you are wanting to state here. What are you referring to as a āpeqā. If you are meaning parametric eq, that is precisely what can can be achieved by FIR or IIR filters, and typically would be undertaken this way in DAW (digital audio workstation) software.
Sorry, quite frankly as an engineer I donāt see this as nitpicking more like being accurate rather than stringing a whole series of consumer marketing terms together in to me a none sensical way that really doesnāt mean much. I do apologise if I appear rudeā¦ but hopefully you can understand itās hard to have a discussion with nonsensical terms.
Hiā¦ yes using Roonās FIR filter kernel import function can be usefulā¦ and as you say the secret isnāt keep the adjustments subtle, try and focus of frequency attenuation rather than frequency gain, keep any troughs to only a few dBā¦ and it should keep any artefacts to a minimum.
Itās a shame with Roon one cant increase the kernel size (number of taps) to match available processor power or adjust the windowing function, but at least it does correctly support different filter responses for different sample rates.
Peq is parametric EQ as you know. From my basic understanding peq uses irr filters limiting what it can do in the area of phase and impulse corrections where when using convolution itās fir based which allows these adjustments. This is what I am referring to and
they are often referred to maybe righlty or wrongly referred to as convolution filters as they will only work in a convolution engine. Just using peq in Roon is not enough for better room eq, The filters supplied by HAF use convolution for this reason, that and I guess its easier to package it up as one file. Yes it takes more grunt but the results are worth it in my opinion and it really does take the room out of the equation as much as possible and letās the music flow.
Ok, so parametric eq can use digital filters of either FIR or IIRā¦ The IIR (infinite Impulse Response) filters can be more easily programmed using algorithmic and or recursive functions in softwareā¦ and are in theory limitless to the impulse response, but donāt require taps and long convolutionā¦ by contrast the FIR (Finite Impulse response ) filters are created by creating a kernel responseā¦ ie a set of samples which represents the finite response to an impulse of infinite power of infinite narrow slice of time ( which is the same impulse idea as used in the sample theorem ). This set of samples can be adjusted by software such as REW to create a filter response such as parametric eq, or room optimisationā¦ or follow a particular responseā¦ this method however is finite ā¦ that is it is only calculated over a window of samplesā¦ this window traverses the stream of input samples to create an output signal streamā¦ this is the process of convolution.
Both IIR and FIR filters have their pros and consā¦ however just about any filter will add distortion to the output signalā¦ so itās not a free lunch.
FIR is computationally intensive and doesnāt really support real-time adjustments, but provides a constant phase response
IIR is less computationally intensive, supports more readily real-time modifications, but in real-time applications doesnāt offer linear phase.
Sure DSP is easier and cheaper to do than in the analogue domain, and is typically more accurate than analogue filtering at the cost of reduced speed and dynamic range.