Dipping into DSP

I had experienced a lot of trouble trying to get a pair of REL subs (Tx5) 'tuned in' with my little Harbeth P3ESRs.  Using the REL recommended process left me constantly 'twiddling' with the crossover and gain to try and control the bass so that it added but didn't overwhelm.  In the end I decided to bite the bullet and measure the room response to see what was happening.  That has lead onto 6 months of learning.

Room Measurements:

Setting up REW (Room EQ Wizard) free software with a calibrated UMIK-1 mic proved relatively easy with a lot of on-line advice available. Understanding the results was a different matter and after 6 months I'm certainly no expert, probably using 15-20% of REW capabilities.

At first sight the room response from all speakers were shocking and only made sense after learning about applying smoothing more in tune with hearing capabilities. I should mention that my listening space is small, effectively a bedroom over a garage about 2.3x3.4m with lots of 'stuff' which limits speaker locations to nearfield about 2m apart.

Starting with the main speakers the initial measurements showed the Harbeths recording bass down to 40Hz, impossible for these small boxes. This started a learning process about room reflections and modes.

Repositioning the speakers made a huge difference, after much trial and effort testing incremental movement allowed the bass 'hump' to be controlled. Further research taught that a 'Rule of 1/3s' position to be similar to Cardas for my room dimensions and started to give much improved results leaving the Harbeths delivering down to 90Hz only. Research for sub woofer positioning suggested multiple options and the one that appeared to work best has the RELs at 1/4 positions across the front wall.  The combined results were reasonable but still 'boomed' at about 60Hz as heard and seen on the measurements.  This appeared to be an effect of the room shape with a sloping roofline up to the rear where a mezzanine room created a massive reflection which could either be adjusted with room correctiong (not an option with SHMBO) or maybe controlled with DSP.

DSP:

The different DSP suppliers offered mainy different options, configurations and prices which basically came down to either a) modify my Majestic to give variable phono output for a DSP feed to the subs or b) a DSP with both balanced and unbalanced output but could effectively replace the Majestic. I took the 2nd option, slecting the miniDSP SHD on the basis that I knew I could resell a s/h SHD and with a bit of work with cables it could give the flexibility to test direct comparison with the Majestic.

Setting up a SHD required computer knowledge and patience to configure it so that I could have either Majestic or SHD feeding the Proprius / Harbeth main speakers while only the SHD output to the REL subs. Once working with REW the immediate DSP results adjusting cross over and PEQ (Parametric) filters were impressive and after testing and making measurements I found the response below 200Hz could be completely smoothed.

One finding was that direct comparison of the SHD(with no filters applied) and Majestic using exactly the same input and feeding the Proprius did both measure and sound different.  I'm not saying one is right and one wrong but they were different with the miniDSP 'brighter' above 8KHz.

At this point there was a dilema, does the miniDSP replace the Majestic as there is duplication? Extended listening gave a clear answer, although with my untrained, old ears I might not hear a major difference I thought the SHD fatiguing after a few hours.  Why, I don't know but even if though the SHD gave better 'measured' response it didn't have the smooth, unfatiguing GS sound.

However, I still wanted that bass 'hump' removed.  The answer was to go back to option a), have John C modify my Majestic, sell the MiniDSP SHD and replace it with a cheaper and simpler unit miniDSP DDRC-2x4. 

The end result with Majestic balanced output to Proprius  / Harbeths and phono output to DDRC-2x4 / RELs may not be as technically precise as the alternative SHD but it sounds right, to my ears.  Integration of the Harbeths below 90Hz is controlled by the DDRC-2x4 while everything above 90Hz is unaltered, pure GS path. Okay so this means the DDRC 2x4 is applying additional ADC/DAC conversion to the REL feed but I certainly cannot hear any impact on these lower registers and the DSP filters provide a smooth fade of bass down to 40Hz.

Apologies for the long post but it condenses 6 months learning curve and ends up with an excellent result, in my view.