Monitor Field Trip to Guitar Center

Nevermind, I can apply the calibration to the measured files (EDIT: Previous post have been updated with calibration applied). Not much of a difference IMO:

LSR305 No Calibration - Blue
LSR305 Calibration Applied - Red



Possible the GC listening room impacts the monitors tone. I know my MR5s sound good with 0 dB HF in my living room, but need -2 dB HF on my bedroom... Dunno. At GC, the LSR305s did sound a bit hot (not annoyingly so though).

 

Compensated all plots, and changed dBSPL scale. Sorry about any confusion. Hope these plots and brief impressions are useful.

 

Yes, room does seem to affect things up a bit.

There are differences between the monitors of course, but...



That said, most of the big issues seem to be in the bass, sub-bass, and perhaps low mids. A small room also creates a somewhat "noisy" response probably due to reflections. If I apply the -2 dB on my MR5s I almost don't absolutely need to do much on the mids and treble.

The cheapo miniDSP 2x4 however only includes parametric equalizers and programmable crossovers (IIRs). The more expensive miniDSP stuff includes more powerful FIRs. BTW most analog passive crossover are effectively IIRs I think.

If I ever go crazy high end, I might consider DIY 3-way or 2.5-way speakers (stupid expensive drivers), with active crossover using a set of high-end miniDSP boxes in the mix and some dedicated driver amplifiers. If using a computer, add a Sonarworks solution to handle fine tuning.

It still not going to be perfect though. Room issues are a 3-D affair me thinks. But I bet things will be awesome.

 

BTW, even in sub-optimal rooms, and with their relative limited low end extension, 5" entry level monitors hold their own well against headphones:

http://www.superbestaudiofriends.or...-field-trip-to-guitar-center.2484/#post-65296

This is somewhat surprising to me, specially with full size headphones, considering headphones don't have to deal with room interactions.

I maybe looking at things wrong but, it's almost as if upper mids and treble are better with certain monitors, and lower mids and bass are better with certain headphones. I think most of the soundstage and cross-channel information is in the mids and treble too.

 

I think decay may be seen in the frequency response. It's perhaps not very obvious though. Comb-filtering can also be seen in the frequency response, just not the smoothed out one.

Materials and environment conditions are perhaps other considerations, but difficult to take care off practically.

 

Yes. I have their cheapo deal-io $100 one, but the new box is very attractive. It does have a 400 MHz 32/40-bit processor (ADSP21489) according to their site. Full FIR support!

Furthermore, it supports 2 Vrms outputs instead of just 0.9 Vrms. And this new one can do 192 kHz audio streaming through TOSLINK and USB, w ASIO support. In the earlier unit I have, USB is only for control, not for streaming. The damn thing now even support IR remote for volume control, input selection and so forth. It is a very attractive solution. Specially if one wants to do a high end DIY three/four-way reference/programmable speaker with stupid expensive drivers (More$l, $can$peak, $ea$...)

Side Note: As far as FIR vs IIR. I feel IIR is much more powerful than FIR assuming the two are same order. In fact, a FIR is a digital filter with only feedforward and no-feedback processing. IIRs have both feedforward and feedback processing. The deal is that a biquad is a 2nd order IIR filter (one of many such IIR configurations possible). One can cascade a number of biquads and get an equivalent higher order IIR filter, but not necessarily crazy high. A FIR such as the one Schiit is using is about 18000 order. One could create a programmable 18000 order IIR that would likely kick the living shit of any FIR of similar order, because it can be configured to behave as an FIR among other very crazy things. Also, most analog front ends don't go past the equivalent of 7th order IIRs, because otherwise they can go unstable due to component tolerances. In digital they can go higher, but not crazy much, because otherwise they can go unstable due to arithmetic precision limitations (try designing a 100th order Butterworth - one may start juggling with different structures to keep things from blowing up). That said one can have high order on the zeros (FIR) and low on the poles (feedback) and do awesome shit... Anyhow, it's just a side note. Feel free to ignore.

 

Just read they offer this not so cheap upgrade for the 2x4HD also:

https://www.minidsp.com/products/dirac-series/2x4hd-to-ddrc-24

And this alternative version of the gadget:

https://www.minidsp.com/products/dirac-series/ddrc-24

I don't know what 24-bit ADC and DAC they are using. Possible it's the AD1939?

Re-read a few times after reading the miniDSP site. This is very awesome!