HiBy RS8 flagship Darwin R2R Android DAP -- Class A -- news and impressions thread
Dec 3, 2022 at 7:17 AM Post #1,186 of 3,436
One of the great things about the RS8 imaging and sound stage capabilities is now doing Convolution with it.
I am still learning about convolution, but so far I have acquired a collection of impulse responses for the Promenadikeskus concert hall in Pori, Finland.
The RS8 is impressive as I virtually move around the seating locations in the concert hall.
Pitch, Timing, Reverberation, Perspective, Primary and Secondary Reflections, Harmonics all of these aspects are simulated. It is quite sophisticated.

Just for fun, I am now sitting in the mezzanine.

1670069590417.png
1670069600863.png



Concert Hall Impulse Responses — Pori, Finland: Reference Juha Merimaa1, Timo Peltonen2, and Tapio Lokki3 juha.merimaa@hut.fi, timo.peltonen@akukon.fi, tapio.lokki@hut.fi 1Laboratory of Acoustics and Audio Signal Processing Helsinki University of Technology P.O.Box 3000, FI-02015 TKK, Finland


125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz 8 kHz T30 (s) 2.7 2.5 2.4 2.4 2.1 1.7 1.2 EDT (s) 2.5 2.4 2.4 2.3 2.0 1.6 0.9 G (dB) 8.9 8.7 8.9 9.6 9.9 8.2 4.3 C80 (dB) −0.3 −3.6 −3.1 −1.8 −0.6 1.6 6.6 1 − IACCE 0.10 0.26 0.74 0.74 0.74 0.73 - 1 − IACCL 0.12 0.31 0.87 0.89 0.94 0.94 - LFP 0.24 0.24 0.28 0.37 0.34 0.41 - LFSF 0.37 0.33 0.34 0.37 0.48 0.59 - SNR (dB) 66.1 65.7 67.3 72.0 74.3 74.2 70.5


"The measurements were performed in the concert hall of Promenadikeskus located in Pori, Finland. The 700 seat hall was designed by Architect Company Guttner and built in 1999. The acoustic ¨ consulting was provided by Akukon Oy Consulting Engineers. The hall is illustrated with architectural drawings in Fig. 1, photographs in Figs. 2–4 and average room acoustical parameters in Table 1. For a description of the parameters1 and the computational methods, as well as further analysis of the individual source-receiver combinations, see the accompanying document (Merimaa et al., 2005). The hall is roughly shoebox shaped with dimensions of 33 x 23 x 15 m (length x width x height) yielding a total volume of approximately 9300 m2. There are balconies on both sides, and the floor rises towards the rear part of the hall. A variety of diffusers has been installed on the walls, and the hall has a set of hanging reflectors above the stage (see Figs. 3 and 4) projecting more early sound both to the audience as well as to the musicians. The upholstered seats have a folding seat cushion with perforation on the bottom plate, yielding a relatively small difference between the absorption of an occupied and an unoccupied seat. The first three rows of seats can be removed but they were present during the measurements. The measurements were conducted in an empty hall with three source positions on the stage, four receiver positions in the audience area, and three receiver positions on the stage. The positions were chosen according to recommendations of Gade (1989), and they are illustrated in the floorplan in Fig. 5 and listed in Table 2. Each stage position is specified with the distance to the foremost point on the stage (on the midline, denoted front of the stage) in the direction of the midline, as well as with the distance of the position from the midline. The receiver positions in the audience area are specified with row and seat numbers, where row 1 is the first (removable) row in front of the stage and seat numbers are given relative to the midline of the hall."

The processor demands are not bad at all. I am surprised that the convolution calculations are not processor intensive. I am able to use a Surface Pro 7 i7 16Gb computer to run the convolution with about 20% CPU and 2% GPU utilization. It does take more than 5Gb of RAM however for just the convolution.


If anyone cares, I did find a site that seems to have a large collection of impulse responses for purchase.

https://hofa-plugins.de/en/plugins/...MIntzTz7Ld-wIV5S2tBh3bdwOCEAMYASAAEgIQRPD_BwE

@Joe Bloggs

Concert Hall Impulse Responses - Pori, Finland
----------------------------------------------

The impulse response database described in this document is available
for download from http://www.acoustics.hut.fi/projects/poririrs/.

Contents of the document:
1. Introduction
2. Microphone and channel specifications
3. List of files
4. References
5. Copyright


1. Introduction
---------------

The responses included in the database were measured in the
Promenadikeskus concert hall located in Pori, Finland. The responses
are provided as 24-bit wav files including 2-5 channels depending on
the type of the response. Each wav file includes the combined response
to an omnidirectional dodecahedron loudspeaker and a subwoofer. The
responses have been measured with 3 source positions on the stage
(S1-S3), 4 receiver positions in the audience area (R1-R4), and 3
receiver positions on the stage (P1-P3). All provided responses are
system compensated and denoised.



2. Microphone and channel specifications
----------------------------------------

Microphone directions are specified in a Cartesian coordinate
system. The positive X direction (front) for each receiver position in
the audience area is defined as the direction towards the stage
perpendicular to the backs of the seats in the corresponding
position. For the receiver positions on the stage, the front is
defined as the direction perpendicular to and pointing towards the
front wall behind the stage. Positive Y and Z directions are defined
as left and up, respectively, when facing the front.

The utilized microphones and channels are defined as follows:

DPA 4006
Pair of DPA Type 4006 omnidirectional microphones facing the front
with a distance of 10 cm between them in the left-right direction.
L = leftmost microphone, R = rightmost microphone.

HATS
Brüel and Kjær HATS dummy head custom fitted with DPA Type 4053
microphones, facing the front. L = left ear input, R = right ear
input.

HATS, d.f. EQ
Same as HATS, but diffuse field equalized [1].

Pearl
Pearl TL-4 stereo condenser microphone with cardioid directivity
patterns. L = cardioid facing left, R = cardioid facing right.

SIRR
SoundField B-format signals processed with the SIRR method [1] for
reproduction with a standard 5.0 loudspeaker setup. Loudspeakers for
the corresponding channels should be located in the horizontal plane
equidistant from a listener with the following azimuthal angles:
left (L): 30, right (R): -30, center (C): 0, left surround (LS):
110, and right surround (RS): -110.

SoundField
B-format signals from a SoundField MKV microphone system. W is an
omnidirectional signal and X, Y, and Z have figure-of-eight
directivity patterns with positive directions aligned with the
corresponding positive coordinate axes.



3. List of files
----------------

The measurements with different microphones have been packed into
individual zip files. Furthermore, measurements with receivers on the
stage are available in separate packages. The full database includes
the following response packages:

File Contents Channels Src Rcv Microphone
---- -------- -------- --- --- ----------
binaural.zip s1_r1_b.wav L,R S1 R1 HATS
s1_r2_b.wav L,R S1 R2 HATS
s1_r3_b.wav L,R S1 R3 HATS
s1_r4_b.wav L,R S1 R4 HATS
s2_r1_b.wav L,R S2 R1 HATS
s2_r2_b.wav L,R S2 R2 HATS
s2_r3_b.wav L,R S2 R3 HATS
s2_r4_b.wav L,R S2 R4 HATS
s3_r1_b.wav L,R S3 R1 HATS
s3_r2_b.wav L,R S3 R2 HATS
s3_r3_b.wav L,R S3 R3 HATS
s3_r4_b.wav L,R S3 R4 HATS

bin_dfeq.zip s1_r1_bd.wav L,R S1 R1 HATS, d.f. EQ
s1_r2_bd.wav L,R S1 R2 HATS, d.f. EQ
s1_r3_bd.wav L,R S1 R3 HATS, d.f. EQ
s1_r4_bd.wav L,R S1 R4 HATS, d.f. EQ
s2_r1_bd.wav L,R S2 R1 HATS, d.f. EQ
s2_r2_bd.wav L,R S2 R2 HATS, d.f. EQ
s2_r3_bd.wav L,R S2 R3 HATS, d.f. EQ
s2_r4_bd.wav L,R S2 R4 HATS, d.f. EQ
s3_r1_bd.wav L,R S3 R1 HATS, d.f. EQ
s3_r2_bd.wav L,R S3 R2 HATS, d.f. EQ
s3_r3_bd.wav L,R S3 R3 HATS, d.f. EQ
s3_r4_bd.wav L,R S3 R4 HATS, d.f. EQ

cardioid.zip s1_r1_c.wav L,R S1 R1 Pearl
s1_r2_c.wav L,R S1 R2 Pearl
s1_r3_c.wav L,R S1 R3 Pearl
s1_r4_c.wav L,R S1 R4 Pearl
s2_r1_c.wav L,R S2 R1 Pearl
s2_r2_c.wav L,R S2 R2 Pearl
s2_r3_c.wav L,R S2 R3 Pearl
s2_r4_c.wav L,R S2 R4 Pearl
s3_r1_c.wav L,R S3 R1 Pearl
s3_r2_c.wav L,R S3 R2 Pearl
s3_r3_c.wav L,R S3 R3 Pearl
s3_r4_c.wav L,R S3 R4 Pearl

omni.zip s1_r1_o.wav L,R S1 R1 DPA 4006
s1_r2_o.wav L,R S1 R2 DPA 4006
s1_r3_o.wav L,R S1 R3 DPA 4006
s1_r4_o.wav L,R S1 R4 DPA 4006
s2_r1_o.wav L,R S2 R1 DPA 4006
s2_r2_o.wav L,R S2 R2 DPA 4006
s2_r3_o.wav L,R S2 R3 DPA 4006
s2_r4_o.wav L,R S2 R4 DPA 4006
s3_r1_o.wav L,R S3 R1 DPA 4006
s3_r2_o.wav L,R S3 R2 DPA 4006
s3_r3_o.wav L,R S3 R3 DPA 4006
s3_r4_o.wav L,R S3 R4 DPA 4006

omni_p.zip s1_p1_o.wav L,R S1 P1 DPA 4006
s1_p2_o.wav L,R S1 P2 DPA 4006
s1_p3_o.wav L,R S1 P3 DPA 4006
s2_p1_o.wav L,R S2 P1 DPA 4006
s2_p2_o.wav L,R S2 P2 DPA 4006
s2_p3_o.wav L,R S2 P3 DPA 4006
s3_p1_o.wav L,R S3 P1 DPA 4006
s3_p2_o.wav L,R S3 P2 DPA 4006
s3_p3_o.wav L,R S3 P3 DPA 4006

sirr.zip s1_r1_sr.wav L,R,C,LS,RS S1 R1 SIRR
s1_r2_sr.wav L,R,C,LS,RS S1 R2 SIRR
s1_r3_sr.wav L,R,C,LS,RS S1 R3 SIRR
s1_r4_sr.wav L,R,C,LS,RS S1 R4 SIRR
s2_r1_sr.wav L,R,C,LS,RS S2 R1 SIRR
s2_r2_sr.wav L,R,C,LS,RS S2 R2 SIRR
s2_r3_sr.wav L,R,C,LS,RS S2 R3 SIRR
s2_r4_sr.wav L,R,C,LS,RS S2 R4 SIRR
s3_r1_sr.wav L,R,C,LS,RS S3 R1 SIRR
s3_r2_sr.wav L,R,C,LS,RS S3 R2 SIRR
s3_r3_sr.wav L,R,C,LS,RS S3 R3 SIRR
s3_r4_sr.wav L,R,C,LS,RS S3 R4 SIRR

sndfld.zip s1_r1_sf.wav W,X,Y,Z S1 R1 SoundField
s1_r2_sf.wav W,X,Y,Z S1 R2 SoundField
s1_r3_sf.wav W,X,Y,Z S1 R3 SoundField
s1_r4_sf.wav W,X,Y,Z S1 R4 SoundField
s2_r1_sf.wav W,X,Y,Z S2 R1 SoundField
s2_r2_sf.wav W,X,Y,Z S2 R2 SoundField
s2_r3_sf.wav W,X,Y,Z S2 R3 SoundField
s2_r4_sf.wav W,X,Y,Z S2 R4 SoundField
s3_r1_sf.wav W,X,Y,Z S3 R1 SoundField
s3_r2_sf.wav W,X,Y,Z S3 R2 SoundField
s3_r3_sf.wav W,X,Y,Z S3 R3 SoundField
s3_r4_sf.wav W,X,Y,Z S3 R4 SoundField

sndfld_p.zip s1_r1_sf.wav W,X,Y,Z S1 P1 SoundField
s1_r2_sf.wav W,X,Y,Z S1 P2 SoundField
s1_r3_sf.wav W,X,Y,Z S1 P3 SoundField
s2_r1_sf.wav W,X,Y,Z S2 P1 SoundField
s2_r2_sf.wav W,X,Y,Z S2 P2 SoundField
s2_r3_sf.wav W,X,Y,Z S2 P3 SoundField
s3_r1_sf.wav W,X,Y,Z S3 P1 SoundField
s3_r2_sf.wav W,X,Y,Z S3 P2 SoundField
s3_r3_sf.wav W,X,Y,Z S3 P3 SoundField



4. References
-------------

[1] J. Merimaa, T. Peltonen, and T. Lokki, 2005: "Concert hall impulse
responses - Pori, Finland: Reference." Available at
http://www.acoustics.hut.fi/projects/poririrs/.

[2] J. Merimaa, T. Peltonen, and T. Lokki, 2005: "Concert hall impulse
responses - Pori, Finland: Analysis results." Available at
http://www.acoustics.hut.fi/projects/poririrs/.



5. Copyright
------------

SIRR processed data are Copyright (C)2005 by Helsinki University of
Technology (TKK). All other data are Copyright (C)2005 by TKK, Akukon
Oy Consulting Engineers, and the authors. The data are provided free
for noncommercial purposes, provided the authors are cited when the
data are used in any research application. Commercial use is
prohibited and allowed only by written permission of the copyright
owners.
 
Last edited:
Dec 3, 2022 at 8:03 AM Post #1,187 of 3,436
There are some nice background documents on the project.
 

Attachments

  • poriref.pdf
    614.4 KB · Views: 0
Dec 3, 2022 at 8:47 AM Post #1,188 of 3,436
New firmware 1.04 and I see HibyCast is installing. Anyone upgrade from 1.02 to 1.04 :)
 
Last edited:
Dec 3, 2022 at 8:55 AM Post #1,190 of 3,436
Hi there,

I have good news today.

I asked Dmitry who is the developer of Neutron Music Player (NMP) for better support on the RS8, and he made it!
NMP will support up to PCM1536/DSD1024 playback on the RS8 in the next update release.

The dev version is available now at the following link.
http://neutronmp.com/download/dev/NeutronMP_ARM64_DEV_301122.apk

Many thanks to Dmitry!! :beerchug:

Note:
- As you know, NMP is a paid app. You need to purchase the license to use it.
- I got permission from Dmitry to share the above link here.


Left: PCM1536 (96kHz x 16 oversampling) / Right: native DSD1024 playback
PCM1536&DSD1024onNMP.JPG
Neutron might have a tricky UI but from my experience is the best music player.
 
Dec 3, 2022 at 9:01 AM Post #1,191 of 3,436
Ooh, is there a change log for 1.04?
I'd rather not be the UAT. I was trying to play with the internal convolver when the option to download 1.04 came up.
I started downloading it, but then cancelled out of it just in case.
 
Last edited:
Dec 3, 2022 at 9:22 AM Post #1,192 of 3,436
New firmware 1.04 and I see HibyCast is installing. Anyone upgrade from 1.02 to 1.04 :)
Thanks for the heads up, updating!

I was running 1.03...

6BFBA4E1-A3BF-41E6-A421-A724D1510D72.jpeg
 
Last edited:
Dec 3, 2022 at 11:00 AM Post #1,194 of 3,436
fw 1.04 update, Amazon Music HD doesn't work anymore and Tidal app still has a hiccup where it takes a split second to unfold MQA so you see in notification 88.2k and then 705.6k. Also, streaming Tidal still has issues when it switches from 705.6k (44.1k based track) to 768k (48k based track) and there is no sound. Tidal issue was the same in 1.02 and 1.03 (was there even 1.03? I upgraded from 1.02 to 1.04, so not sure), while Amazon Music HD streaming issue used to be only not being able to play ultra HD files (only showing playback at 44.1k or 48k instead of 192k), and now with 1.04, it doesn't play any tracks, just fast forwards through the song without any audio output.

This is most likely app compatibility issue which Hiby needs to resolve with Amazon and Tidal. But, I haven't noticed anybody else mention these issues. So, anybody else noticing the same behavior with Amazon Music HD and Tidal streaming on their RS8?
 
Last edited:
Dec 3, 2022 at 11:13 AM Post #1,195 of 3,436
I can't comment on Tidal or Amazon but Apple Music works fine with 1.04 so far
 
Dec 3, 2022 at 12:22 PM Post #1,196 of 3,436
fw 1.04 update, Amazon Music HD doesn't work anymore and Tidal app still has a hiccup where it takes a split second to unfold MQA so you see in notification 88.2k and then 705.6k. Also, streaming Tidal still has issues when it switches from 705.6k (44.1k based track) to 768k (48k based track) and there is no sound. Tidal issue was the same in 1.02 and 1.03, while Amazon Music HD streaming issue used to be only not being able to play ultra HD files (only showing playback at 44.1k or 48k instead of 192k), and now with 1.04, it doesn't play any tracks, just fast forwards through the song without any audio output.

This is most likely app compatibility issue which Hiby needs to resolve with Amazon and Tidal. But, I haven't noticed anybody else mention these issues. So, anybody else noticing the same behavior with Amazon Music HD and Tidal streaming on their RS8?
I've run into the same issue with Tidal. It does the jump from a lower value to 705.6K with MQA files .. and sometimes it will play, but there is no sound. I also sometimes get crackling/static on top of the music at the start of the track after this switch, the static only lasts for a few seconds and all is normal afterwards. This occurs with both streaming and downloaded Tidal files.

I've also had the "no sound" issue also happen using Hiby music, when a new track starts. Its a rare occurrence though.
 
Last edited:
Dec 3, 2022 at 10:14 PM Post #1,198 of 3,436
I've run into the same issue with Tidal. It does the jump from a lower value to 705.6K with MQA files .. and sometimes it will play, but there is no sound. I also sometimes get crackling/static on top of the music at the start of the track after this switch, the static only lasts for a few seconds and all is normal afterwards. This occurs with both streaming and downloaded Tidal files.

I've also had the "no sound" issue also happen using Hiby music, when a new track starts. Its a rare occurrence though.
Its the same also with me. I thought it might be a problem on my unit and I had the same issue on my dx320 and had to send it back for repair. If I recall I think I experience it on Tidal app also. Hopefully this is just an rs8 firmware issue. Nevertheless its very rare and minor and I just replay the song and sound comes back.
 
Dec 3, 2022 at 11:52 PM Post #1,199 of 3,436
fw 1.04 update, Amazon Music HD doesn't work anymore and Tidal app still has a hiccup where it takes a split second to unfold MQA so you see in notification 88.2k and then 705.6k. Also, streaming Tidal still has issues when it switches from 705.6k (44.1k based track) to 768k (48k based track) and there is no sound. Tidal issue was the same in 1.02 and 1.03 (was there even 1.03? I upgraded from 1.02 to 1.04, so not sure), while Amazon Music HD streaming issue used to be only not being able to play ultra HD files (only showing playback at 44.1k or 48k instead of 192k), and now with 1.04, it doesn't play any tracks, just fast forwards through the song without any audio output.

This is most likely app compatibility issue which Hiby needs to resolve with Amazon and Tidal. But, I haven't noticed anybody else mention these issues. So, anybody else noticing the same behavior with Amazon Music HD and Tidal streaming on their RS8?

Anybody else noticed when going from 1.02 to 1.04 the treble gain back some of the sparkle and the soundstage is a little bit wider? I do have over 150hrs on RS8 already, so it's not a matter of additional burn in. But, I can definitely hear more sparkle/air in treble after the fw update. I know, the topic of fw vs sound changes is always controversial in every DAP thread, but this is how I hear it.
 
Dec 4, 2022 at 12:15 AM Post #1,200 of 3,436
Musicteck is sold out of the RS8 :)
Have you tried this on your RS6? Thanks for this, it's really interesting. I hope that I can run it on RS6.
I don't know all the details. Some Impulse Response .wav files only work on one side or the other. It may depend on the DAWS it was designed for. Meaning I assume the Impulse Response was designed to add effects the sound engineer would apply to a multi channel system. Full surround sound plus subwoofers for example so the convolver would need to be rather sophisticated. The convolver on the RS6 and RS8 needs an impulse response that will work on two channels. I think the RS6 and RS8 would work the same. Finding a bargain library of impulse response is a challenge. Antiverb 7 seems to have a nice plugin and major concert halls around the world, but it costs €500+. It has Syndey Opera House, LA Disney, and other venues. Really want it, but not sure if I also need external hardware. It requires some usb key. Not sure if vst will plug into JRiver....
Screenshot_20221203-211721.png


Tomorrow I will check, I think HQP had a more feature rich convolver. Either way, I need to write some emails to see it these IR are only for recording or if I can use for rendering playback.

Unlike a lot of claims that music is out of your head... Convolution really is scientificaly out of your head recorded binarual from actual locations. And not some harmonic distortion inherently in the headphone ala HD800.
 
Last edited:

Users who are viewing this thread

Back
Top