
First, I adjusted the 250 Hz peak when combined with the lower cabinet. However, I only had to add a little sound-absorbing material in the middle of the duct.
Since I did not have a proper sound absorbing material on hand, I substituted aquarium filter mat. The material was probably polypropylene, 8mm thick. The only place to put it in was at the first fold, and I just pushed it in from the duct entrance.
First, I compared each frequency response. The results of the measurements with the
"Audio Frequency Analyzer" are as follows (without sound absorbing material = upper figure, with sound absorbing material = lower figure).

As expected, the one with sound absorbing material was about 2dB lower at 250Hz. If the amount of sound-absorbing material is increased, it should decrease even more.
However, it might block the sound path, so I left this one piece alone.
I finally got around to listening.
Usually I play CDs that I have on hand and listen to them as I go along, but this time I had to hear how the different methods would change the sound.
I used a sample CD from the January 2011 issue of Stereo magazine.
This time, I chose classical and jazz from these recordings because I wanted to hear the difference especially in the low-frequency direction.
12. Saint-Saëns: Symphony No.3 "Organ Symphony"
13. Ornette Coleman: Bird Food/Trissonique
The left-right spacing was about 1 m, and the listening position was about 2 m away from the SPs. Sealed and both bass reflexes were performed with TAOC stands, and the acoustic maze was placed on the lower cabinet.


The sealed system is designed to be used on a desktop, with the ducts blocked by a board. For the bass reflex (upward-facing) system, the board was shifted so that the duct would be open.
The feeling of each method was as follows.
- Sealed box:

The timpani percussion was light, and it was difficult to hear the descending part of the organ scale. Although I could hear the pitch of the bass, the sound of the strings hitting the fingerboard seemed louder.
- Bass reflex (upward-facing):

The timpani percussion was rather audible. The organ scale was also not well resolved. The sound of the fingerboard still seemed to resonate.
- Bass reflex (downward-facing):

The timpani percussion became more prominent and the tutti was more powerful. The volume of the wood bass could be felt.
- Acoustic Maze:

The scale of the orchestra sounded a whole lot larger. I could hear the pitch of the organ. The bass guitar sounded more powerful.
Since this is just a subjective impression, I measured the frequency response of each piece.
I set up the iPhone at the listening position (1.8m, about 15° to the loudspeaker) with the left channel only, played a song, and checked the frequency response with peak hold.
Before that, I examined the characteristics of the music source itself with
"Music Frequency Analyzer". This application can digitally analyze ripped music data. The following result is a peak-hold analysis of the frequency response of the entire music source.

"Organ Symphony" was nicely smooth over the entire frequency range.

"Bird Food" was recorded at a fairly high level in the 63-250 Hz range.
If the frequency response of this sound source could be reproduced well in a listening environment with speakers, it would be the reproduction of the original sound. However, in practice, I think it is quite tough.
The actual measurements for "Organ Symphony" were, from top to bottom, sealed box, bass reflex (upward facing), bass reflex (downward facing), and acoustic maze.

With the sealed box, 125-200 Hz had a reasonable sound level. With the bass reflex (upward), the same band seemed to be emphasized.
With the bass reflex (downward), the 160 and 200 Hz frequencies dropped slightly and the energy shifted to the 100 and 125 Hz frequencies.
In the acoustic maze, 100 Hz rose further and 80 Hz rose by 6 dB.
The same trend was observed with the "Bird Food." The 160 and 250 Hz frequencies in the bass reflex (upward) were at more stronger level.

# Ignore the frequencies of 20Hz and 31.5Hz as they are external noises such as automobile noise.
# A line was drawn later between 80Hz and 100Hz to make it easier to see.
I expected the acoustic maze system to improve down to about 63Hz, but as far as the characteristics are concerned, it came up a little short.
I was hoping for a little more at lower frequencies, given the combined system size of the top and bottom. I thought it was necessary to either extend the sound path further or switch to a double bass reflex system.
This time, I focused on how the low-frequency characteristics changed depending on the type of system.
I thought it was not necessary to emphasize bass so much for a desktop installation. I found that this unit has sufficient reproduction ability even with sealed or bass-reflex systems.
After that, this system should be flipped upside down or duct blocked depending on the sound preferences and the genre of music. For serious listening, the unit should be placed on the lower cabinet. It may be a little troublesome.
.s