Smyth Research Realiser A16
Sep 10, 2019 at 12:13 PM Post #6,302 of 15,989
Sep 10, 2019 at 12:30 PM Post #6,303 of 15,989
I’m aware that the ONE procedure might take a lot of work for plugging and unplugging cables. However, if there’s a not quite acoustically well-behaved room, I think that the ONE procedure by rotating in the chair might get a decent compromise.
Yes I think so. I just wanted to point out that it's not as convenient as described in the A8 manual. At least at the moment.
One more difference with the A8: the A8 used every individual sweep response more effectively: for example the sweep response for looking straight at the center would also be used as the sweep response for looking at the left front speaker, and for looking straight at the right front speaker. So you didnt have to do 8 x 3 but only 12 looking angles. Currently with the A16 you would have to do 3 lookangles for every virtual speaker (to have the most basic head tracking). That is really a missed chance: with more channels and the possibility of more lookangles for 360 degrees head tracking it would have been possible with those same 12 lookangles to have 12 ear level speakers with full 360 degrees head tracking measured! To accomplish this now you would have to do 10 x 12 sweep measurements! Crazy!
 
Last edited:
Sep 10, 2019 at 1:09 PM Post #6,304 of 15,989
Sep 10, 2019 at 1:12 PM Post #6,305 of 15,989
That is really a missed chance: with more channels and the possibility of more lookangles for 360 degrees head tracking it would have been possible with those same 12 lookangles to have 12 ear level speakers with full 360 degrees head tracking measured! To accomplish this now you would have to do 10 x 12 sweep measurements! Crazy!
Yes, this is exactly what I thoght myself about this. The problem is implementation and that is what the Smyths totally underestimated.
They always said "Yes we could do this and that, that's just configuration" etc. but underestimated embedding this in the GUI etc.
For example I have an email from Stephen where he told me that the hardware is capable of having 32 bands of PEQ per channel (meaning 512 PEQs active at one time!).
But is it implemented? No. Why? Because it's a lot of work!

(I do a little programming myself at work. At the moment I write a Python script, it has now about 700 lines and I'm working on it now for about 4 weeks... Testing, Debugging etc. )
 
Sep 10, 2019 at 2:13 PM Post #6,306 of 15,989
Just found a website with some Atmos trailers and Test tones in different formats (7.1.4, 9.1.6 etc.) for download as mp4 files.
It's in german though:
https://www.heimkino-atmos.de/Sonstiges/Trailer,-Downloads-und-Testtoene/
Thanks for the link. I’ve tried the test tone 5.1.4. It works fine. Among the trailers, conductor is only in stereo, however, the other two, leaf and amaze are in Dolby Atmos.
NqbN6En.jpg
 
Last edited:
Sep 10, 2019 at 5:56 PM Post #6,308 of 15,989
Has anyone tried to send RIR measurement sweeps through USB?
I have just implemented a FIR filter for my room using REW, Rephase and Equalizer APO and I would like to measure my room through that filter that is convolved by my computer and the sent via HDMI to my TV.
If you can record the test signals from analog in, run your convolutions and output to your speakers in near real time (i.e. only a small fraction of a second of delay) that should be no problem at all. If the delay introduced by Equalizer APO is too big, you may have to switch to an ASIO based tool. Voicemeeter Banana may be suitable, but it's apparently also limited to at most 8 channels.

It is possible, but it takes a lot of work and plugging and unplugging stuff.
Alternatively, just start new PRIRs, possibly name the speaker accordingly between chair rotations, and then reassemble one Room from each of the rotations. The advantage would be to that you don't have to get up and risk moving the microphones in your ears between measurements.
 
Sep 10, 2019 at 6:13 PM Post #6,309 of 15,989
I think I found a bug. The spdif input can’t lock to an 88.2khz pcm signal. I just get a lot of jittery sound. Other sample rates seem to work ok.
And 44.1 kHz? I was already wondering because of the following specs in the A16 manual, that only mention 48, 96, and 192 kHz, I would find it really bizar if they did not implement 44.1 kHz input capability because then we could not connect a simple CD player directly to the A16 (not that I need to do that):
2.1.3 Realiser A16 specifications
SVS loudspeaker virtualisation with integrated headtracking (16ch 32bit floating-point processing@48kHz sampling rate, processing latency 32ms, maximum reverb length 750ms). All source signals above a sampling rate of 48kHz are down-sampled to 48kHz.
Virtualisation sources:
1. HDMI inputs (1-4): Dolby Atmos decoded bitstream (16ch), 8ch LPCM (24bit@48/96/192kHz)
2. SPDIF inputs (coaxial and optical): 2ch LPCM (24bit@48/96/192kHz) and Dolby Digital bitstream
3. USB 2.0 input: 16ch LPCM (24bit@48/96/192kHz)
4. Analogue line inputs: 16ch (24bit@48kHz)
5. Stereo line inputs: 2ch (24-bit@48kHz)
[Edit: actually it isn't a bug if it is not in the specfifications, even if it bugging us.]
 
Last edited:
Sep 10, 2019 at 7:55 PM Post #6,310 of 15,989
If you can record the test signals from analog in, run your convolutions and output to your speakers in near real time (i.e. only a small fraction of a second of delay) that should be no problem at all. If the delay introduced by Equalizer APO is too big, you may have to switch to an ASIO based tool. Voicemeeter Banana may be suitable, but it's apparently also limited to at most 8 channels.

Alternatively, just start new PRIRs, possibly name the speaker accordingly between chair rotations, and then reassemble one Room from each of the rotations. The advantage would be to that you don't have to get up and risk moving the microphones in your ears between measurements.

I actually did try it both ways when I received my A16 and I found both methods to sound basically identical.
 
Sep 10, 2019 at 8:06 PM Post #6,311 of 15,989
I see nobody mentioned the missing extra head-trackers on the A16 kickstarter comments page yet. Maybe someone can do that, also mention the fact that it happened to many people? I can not because I am pre-order.
(I am not really hopefull the Smyths ever look there anymore, but who knows, and it won't hurt to mention it there.)
 
Sep 10, 2019 at 8:52 PM Post #6,312 of 15,989
Their facbook page and messaging might be the way to query the head trackers - they seem to respond to messages there and you can see when it's been read
 
Sep 11, 2019 at 1:02 AM Post #6,313 of 15,989
And 44.1 kHz? I was already wondering because of the following specs in the A16 manual, that only mention 48, 96, and 192 kHz, I would find it really bizar if they did not implement 44.1 kHz input capability because then we could not connect a simple CD player directly to the A16 (not that I need to do that):

[Edit: actually it isn't a bug if it is not in the specfifications, even if it bugging us.]

Like really? I can not connect my CD player? That makes no sense for me.
 
Sep 11, 2019 at 1:16 AM Post #6,314 of 15,989
I’m aware that the ONE procedure might take a lot of work for plugging and unplugging cables. However, if there’s a not quite acoustically well-behaved room, I think that the ONE procedure by rotating in the chair might get a decent compromise.

I think the easiest thing to do is place one speaker in the center of the room, optimize that speaker with any ad hoc room treatments you can muster, and the rotate your head and body. That way, you'll get the best in room performance of the speaker. Even better, if you have a DSP unit with A to D analogue inputs (such as the miniDSP 2X4 HD, or the SHD) you can run something like Dirac Live on that speaker, after doing room treatments, and get the triple benefit of getting the effect of a speaker playing in a bigger room (b/c its farther away from room boundaries than it would otherwise be), ad hoc room treatment (b/c you may not like what they do to your room decor if used on an everyday basis), and DSP (b/c the miniDSP can accept the Realiser signal on an analog basis, digitize it and then DSP it before doing its own D-A conversion).

That's what I plan to do with my in house KEF LS 50/SB2000 system when I get a Realiser.
 
Last edited:

Users who are viewing this thread

Back
Top