[jazzfan]

The thread in my sig please, thanks.

Thanks for mentioning your thread. I've removed my threads from Realizer A16 product page and added a link your thread. I also submitted a request to remove the two redundant threads I created:

• Smyth Research Realiser A16 Hints & Tips
• Smyth Research Realiser A16 Impressions

I look forward to referencing your thread as the definitive A16 repository. Cheers!

 

[sander99]

Not a high priority, but as a future enhancement request, it should be possible to replace the current head-tracking system with a small front-facing camera combined with facial recognition and motion detection software. I believe such a system would not only be more aesthetically pleasing than the current head-tracking system but also could have the potential to accommodate an unlimited range of three-dimensional look-angles. Is this idea worthy enough to add to the feature request list?

"Head tracking" consists of 2 things:

1. Detecting the movements/position of the head
2. Counter rotating the virtual sound image relative to your head movements, such that the virtual sound image seems to be stationary

The A16 head tracker can already detect head rotation over 360° over 3 axis, in other words: it can already detect all rotational movements of the head. So that is not a problem. It's step 2 where the limitations are. Smyth already planned vertical head tracking, or rather: vertical counter rotation (that's what's missing now).

Apart from that your idea is nice of course but not very high priority indeed because they have like a zillion other things to improve!
By the way Smyth already once said they planned to introduce a wireless head tracker for the A16, but as we know they planned so many things...
Your idea would of course also be wireless and have the advantage that you could switch between many different headphones without having to re-attach the head tracker to the other headphones.
I would like a wireless head tracker to use with wireless headphones. (If that is possible without adding so much latency that the counter rotation isn't convincing anymore.)

On a side note: I will maybe create something to easily detach and attach a head tracker with one click, and won't bother to join the cables of the headphones and the head tracker together. That way I can easily switch headphones. I had a sennheiser HD650 with a broken headband and I 3d printed a new headband, various designs in fact (made by a friend who wanted to practise his solid works skills) until I was happy with the size and the clamping force (cost me less then 1 Euro each, an original Sennheiser headband costs 89 Euro's!). Then I got the idea I could extend the design with some sort of "mechanical connector", and create a matching part to permanently attach to the head tracker. For other headphones I could do domething similar or just make a seperate "connector part" that I permanently attach to the headphones.

Edit: I would say: for the time being don't bother Smyth with low priority requests yet, especially if it is something completely new and elaborate like the above. Maybe later when they have implemented most of the things they originally planned/promised.

 

[Eich1eeF]

Not a high priority, but as a future enhancement request, it should be possible to replace the current head-tracking system with a small front-facing camera combined with facial recognition and motion detection software. I believe such a system would not only be more aesthetically pleasing than the current head-tracking system but also could have the potential to accommodate an unlimited range of three-dimensional look-angles. Is this idea worthy enough to add to the feature request list?

This is one of the few improvements where you don't have to wait for the Smyths, you can do it yourself.

The Realiser doesn't have the CPU horsepower nor the interface to connect a camera. In the meantime, you could reverse engineer the interface of the head tracker and get a PC based solution running, with an arduino or similar board to interface with the A16, or use a Raspberry Pi for everything. You can start with something like the good old Wiimote based headtracking as shown in the well known Johnny Lee video below. An alternative would be a "proper" camera based solution using marker tracking e.g. with https://github.com/artoolkit/ (just print a small 2D barcode and attach it to the headphones, 6 on the faces of a cube for full 360 degree sensing). It's not obvious to me if any of those solution, much less full face tracking would have the required low latency and angular resolution. Doing that in hardware on a small production run device like the realiser will probably drive up the price considerably.

Such a solution would provide 3 axis rotation and 3 axis position sensing - we'll probably have to wait for the Realiser A32 with even more elaborate PRIP measurements to make use of that...

 

[dsperber]

I tried playing CD and mp3 via HDMI in. The 10s mute is now down to about 0.5 s I'd say. Much better but not perfect. Still misses the beginning of some songs. I also encountered the "stutter" (as dsperber described it "Ta-Ta") after skipping etc. when the stream starts again.
Then I tried SPDIF (optical) and there it seems you don't have that problem. Just 0.5 s mute time and "soft" un-mute, as you expect!

So I recommend for 2ch PCM sources use SPDIF!

Had my own chance to reproduce your CD/SPDIF results today. I agree with your observation on the "stutter" (as you call it, which is definitely more accurate and professional sounding than my "ta-ta" which was the first thing that came to mind when I wanted to verbalize it). Definitely isn't there using PCM via optical input, whereas it is there with PCM via HDMI input. So I agree that for now, using optical input for CD audio seems the best choice and we'll have to live with the 0.5s mute time until Smyth eliminates the use of the decoder on PCM input.

So this was also my first opportunity to use an A16 input (optical) other than HDMI. And it also was yet another opportunity for me to interact with the level of user-friendliness (or not so much) currently built into the user interface. Others may feel this to be relatively unimportant on any to-do list when contrasted with true important promised functionality (e.g. DTS decoding) that is currently still not available. But my own feeling is that user interface is experienced 100% of every time you use the device, no matter what you're doing. And so any inconvenience which you run into EVERY TIME YOU WANT TO DO ANYTHING is really quite important.

Specifically, with regard to today's playing around with CD via optical vs. CD via HDMI, it is once again obvious that not having a power-on default VOLUME is a nuisance. At power-on I happened to be sitting on my preset 01 (5.1) which is perfectly usable for 2-channel CD audio, so that's where I was working. But my volume was still up around 79 from yesterday's usage. CD audio is much hotter than ordinary movies and HDTV, and reasonable listening volume for me needs to be around 63. So obviously I had to down-the-volume right away.

I also had set the INPUT for preset 01 to optical. That required going BACK to main menu, navigating down from the preset/UserA line to the INPUT line, and then sequentially ADJ + from HDMI2 up to coaxial (which I discovered is the very last input available). With headphones on while doing the ADJ + on INPUT, I noticed noise/pop when passing through the rest of the inputs, none of which had any input cable connected. I think it might have been the USB input which produced the "pop". Anyway, it took real time to complete all the finger-presses needed to finally get to optical input, and only then could I get back to preset 01 using PA.

And of course, when I wanted to repeat the testing again for HDMI I had to repeat the complete sequence of lots of finger-presses, but this time reversing the input with ADJ - multiple times to come down from coaxial to HDMI2. Once again, passing through USB produced the pop noise in my headphones.

And naturally, when I really want to watch HDTV tonight using that same 5.1 preset 01, I'm going to have to up-the-volume back to 79, because 62 is too low.

So, what would have been very convenient would have been to be able to set up two specially purposed for 2-channel PCM CD audio presets for A/B comparison. I shouldn't be playing around with my preset 01 which I should reserve for 5.1 HDTV, coming from HDMI1 and with default power-on volume of 79. One new special preset for CD audio testing would have HDMI2 as input, the other would have optical as input. Both would have say 62 as the power-on initial default volume.

Also, I would then want the ability to do an "instant random/direct selection" of either preset ideally through one or two two keys. Obviously it is of maximum convenience to the user when doing A/B comparisons to have fixed preset-selecting buttons on the remote like P1-P4 with the A8. Given that there are no hard-wired preset-selecting buttons on the A16 remote what might we come up with instead? I now propose one more A16 modification to add to the wish list of changes and improvements.

==> I PROPOSE THAT WHEN NOT IN "TEST" MODE (speaker solo with sound loop) THAT THE NUMERIC KEYPAD AREA REVERTS FROM SPEAKER-SOLO FUNCTIONALITY TO DIRECT PRESET-SELECTION PURPOSE.

It seems a real waste to permanently allocate the 12 numeric keypad buttons at the bottom of the remote to the truly relatively rarely used purpose of soloing Atmos speakers. The 11 engraved speaker-solo buttons in the top part of the remote are probably the most commonly used speaker-solo buttons because they apply to 5.1/7.1 sound systems. And the 12 numeric keypad buttons can be brought back into use as speaker-solo buttons when the TEST button is pressed. Until TEST is pressed again to get out of speaker solo mode and back to normal mode, the numeric keypad can be the assignable speaker-solo buttons as they are today.

But when operating in normal mode (i.e. WHEN NOT IN TEST), and really as soon as the PA/PB buttons are pressed, the 12 keys of the numeric keypad instead BECOME THE DIRECT-SELECTION buttons for presets. In other words they are now P1-P12 for the A16. (don't have an answer yet for presets P13-P16, but I'll leave that as of really really minor concern to me). This does lose today's capability to do "live solo" of speakers, using actual live source sound on that channel (as opposed to the music loop from TEST). But I'm very willing to give up that really really rarely needed "demo" use of Atmos speaker soloing (for speakers not in the 7.1 plane) to instead have direct preset keys P1-P12.

PROBLEM SOLVED!!! Our remote now has preset-selecting keys P1-P12!!! The 12-key numeric keypad area will be re-purposed primarily for direct preset selection, unless in TEST mode when it will revert to today's speaker-solo function.

And, if each preset can also have its own NVRAM power-on default value for both VOLUME and INPUT, well now we're really making progress. Zero need to push keys zillions of times to navigate sequentially using BACK, arrow up/down, ADJ +/-, because direct selection isn't possible for the very primary one or two variables central to what we're really A/B comparing. All we have to do is properly set up one or two (or more) presets to reflect not only VOLUME and INPUT but also the assorted significant variables we're probably really wanting to compare (including the true main ones, like PRIR, HPEQ, etc.) and then we can do an INSTANT ONE-CLICK A/B COMPARE by simply pressing the numeric keys.

I suspect there will still be a several second delay when switching presets (as there is today, to populate all the variables needed to present the speaker map display) so it won't really be an "instant" A/B comparison. But it will at least be a single-click A/B comparison. WHAT A RELIEF!!

Sure would have simplified my testing today with optical vs. HDMI input for CD audio.

 

[jazzfan]

"Head tracking" consists of 2 things:

1. Detecting the movements/position of the head
2. Counter rotating the virtual sound image relative to your head movements, such that the virtual sound image seems to be stationary

The A16 head tracker can already detect head rotation over 360° over 3 axis, in other words: it can already detect all rotational movements of the head. So that is not a problem. It's step 2 where the limitations are. Smyth already planned vertical head tracking, or rather: vertical counter rotation (that's what's missing now).

Apart from that your idea is nice of course but not very high priority indeed because they have like a zillion other things to improve!
By the way Smyth already once said they planned to introduce a wireless head tracker for the A16, but as we know they planned so many things...
Your idea would of course also be wireless and have the advantage that you could switch between many different headphones without having to re-attach the head tracker to the other headphones.
I would like a wireless head tracker to use with wireless headphones. (If that is possible without adding so much latency that the counter rotation isn't convincing anymore.)

On a side note: I will maybe create something to easily detach and attach a head tracker with one click, and won't bother to join the cables of the headphones and the head tracker together. That way I can easily switch headphones. I had a sennheiser HD650 with a broken headband and I 3d printed a new headband, various designs in fact (made by a friend who wanted to practise his solid works skills) until I was happy with the size and the clamping force (cost me less then 1 Euro each, an original Sennheiser headband costs 89 Euro's!). Then I got the idea I could extend the design with some sort of "mechanical connector", and create a matching part to permanently attach to the head tracker. For other headphones I could do domething similar or just make a seperate "connector part" that I permanently attach to the headphones.

Edit: I would say: for the time being don't bother Smyth with low priority requests yet, especially if it is something completely new and elaborate like the above. Maybe later when they have implemented most of the things they originally planned/promised.

I agree. This should be a later enhancement. I didn't mean to imply that the current head-tracking system is incapable of detecting head rotation in a 3D space. All I'm suggesting is an alternative approach that would accomplish the same goal, but without the need to physically attach an external device to the headband. The vertical head tracking that is missing now can of course still be implemented with the proposed alternative solution.

I must confess the other reason for suggesting an alternate solution is to avoid the laughter that will ultimately ensue from friends as they observe me enjoying the Realiser while donning a head-tracker.

 

[jazzfan]

This is one of the few improvements where you don't have to wait for the Smyths, you can do it yourself.

The Realiser doesn't have the CPU horsepower nor the interface to connect a camera. In the meantime, you could reverse engineer the interface of the head tracker and get a PC based solution running, with an arduino or similar board to interface with the A16, or use a Raspberry Pi for everything. You can start with something like the good old Wiimote based headtracking as shown in the well known Johnny Lee video below. An alternative would be a "proper" camera based solution using marker tracking e.g. with https://github.com/artoolkit/ (just print a small 2D barcode and attach it to the headphones, 6 on the faces of a cube for full 360 degree sensing). It's not obvious to me if any of those solution, much less full face tracking would have the required low latency and angular resolution. Doing that in hardware on a small production run device like the realiser will probably drive up the price considerably.

Such a solution would provide 3 axis rotation and 3 axis position sensing - we'll probably have to wait for the Realiser A32 with even more elaborate PRIP measurements to make use of that...

All excellent points. If you don't want to wait for an "A32", it sounds like you already have the beginnings of a new third-party Realiser accessory. So, if you can produce an affordable solution, I'll buy it and ditch the supplied head-tracker.

 

[hadron70]

I PROPOSE THAT WHEN NOT IN "TEST" MODE (speaker solo with sound loop) THAT THE NUMERIC KEYPAD AREA REVERTS FROM SPEAKER-SOLO FUNCTIONALITY TO DIRECT PRESET-SELECTION PURPOSE

Being able to listen to each speaker in SOLO mode is one of the strongest features of the Realiser. I often use it when listening to real world surround tracks (i.e., not in "TEST" mode) for three purposes: 1) to check whether the signal is correctly received and interpreted by the A16, 2) to analyze the soundmix (where the sound engineer put the dialogue, music, and effects), and 3) for the simple "Wow" effect when experiencing the exact localisation the A16 is capable of. I was very glad to see that the A16 allows the SOLO mode also for the overhead speakers by utlilizing all buttons on the remote. Especially because the overhead speakers are even more difficult to identify without the SOLO mode. Reassigning the numeric keypad on the remote would be a huge step back. So my vote regarding such a proposition is "NO".

Of course it would be nice to have the remote freely configurable, but folks keep in my mind that there already is a HUGE backlog of features and there might only be 1-2 people working on this device.

 

[hadron70]

After 1-2 hours of listening to a 5.1 Dolby Stream sent to a 7.1 preset and without entering any menus, my A16 started his machine gun (the "BUZZ"). The firmware 1.60 worked as intended and sent the A16 into an error state which shut off the BUZZ after ~1 sec, and after restart resumed normal operation.
This happened on two days in a row. Kind of concerning that it happened during simple playback.

 

[jackhamm]

On a different topic, the headtracker 4 conductor cable has been a pain in the neck to keep from tangling, even with the clips that Smyth provides. I’ve been able to acquire 3/4” 3:1 20 ft. heat shrink, fish both the headphone and head tracker connectors through it and still have a semi flexible cord that’s a lot easier to manage. I was able to get all the supplies I needed through Amazon and eBay.

 

[jackhamm]

Sorry about the double pictures

 

[J.Villain]

Facial tracking sounds nice but has many issues. The first is that for people who watch movies in the dark it will be hard for a system to track a face in the dark. Also if people aren't sitting right in front of the camera it just won't work. The next is that the A-16 probably doesn't have the processing power to add facial recognition. The solution to that would be to build an external box that does the tracking and processing then outputs a signal that the A-16 understands. With enough horse power you might get the latency down to a level where it actually is usable. In any case this could be done by a third party. There is also the issue that you would have to be sitting in front of the camera for it to work. Personally I would prefer they spent more time working on shipping me my F'ing unit than futzing with this.

"I felt sorry for I had no shoes until I met a man that had no feet."

 

[dsperber]

Being able to listen to each speaker in SOLO mode is one of the strongest features of the Realiser. I often use it when listening to real world surround tracks (i.e., not in "TEST" mode) for three purposes: 1) to check whether the signal is correctly received and interpreted by the A16, 2) to analyze the soundmix (where the sound engineer put the dialogue, music, and effects), and 3) for the simple "Wow" effect when experiencing the exact localisation the A16 is capable of. I was very glad to see that the A16 allows the SOLO mode also for the overhead speakers by utlilizing all buttons on the remote. Especially because the overhead speakers are even more difficult to identify without the SOLO mode. Reassigning the numeric keypad on the remote would be a huge step back. So my vote regarding such a proposition is "NO".

I agree with you, that it would be a genuine functional loss to lose "live speaker solo" for other than the 9.1 speakers and OH(T) speaker (all of which have their own nine fixed and dedicated keys in the upper part of the remote). And obviously that would be the cost of re-purposing the bottom user-assignable 12 numeric keys as direct P1-P12 preset-selector buttons.

We have a conundrum, as both uses are meaningful and important. I personally would find P1-P12, or at least P1-P6 for the bottom 6 numeric keys, much more useful to me..

Of course it would be nice to have the remote freely configurable.

Doesn't take a complicated design to come up with a reasonable solution, even if there are some compromises made to minimize the burden on Smyth programmer(s).

The 12 numeric keys are, today, already user-assignable as speaker solo keys, either for "live content" or for TEST mode.. There are some factory delivered non-empty assignments, and there are some factory delivered open keys with no default assignment. But all 12 keys are user-assignable. Seems that it should be reasonably easy and simple to build and support a new system setting for further controlling these 12 keys, adding to the already existing user-customizable nature of them for speaker solo use. How about a user-assigned designation of these 12 keys in "4 blocks of 3" just as they appear on the remote? For normal use each row of three keys could be toggled for use either as (a) "live" speaker solo use for whatever user-assignable speaker gets set, or (b) one of the preset groups P1-P3, P4-P6, P7-P9, P10-P12. There could even be a choice provided for P13-P15, though obviously you can't have 15 direct preset keys all available at the same time, and also there's no P16 available directly.

So, for you the setup might be to pick the first two rows for P1-P3, and P4-P6, which is intuitive because those are the actual NUMBER keys with those numbers engraved. So using them as the associated direct preset-selecting keys is intuitive. And then you would still have the bottom two rows available for user-assigned six additional speakers for "live solo". You already have 9.1 + OH(T) speaker solo capability in the upper keys of the remote, and these additional 6 user-assignable speaker solo keys down at the bottom of the remote would provide full 9.1.6 + OH(T) speaker solo capability (both "live" and TEST). Seems sufficient. And of course you can always change the assignments of these 4x3 rows of keys if you had a special temporary need.

For others, who have many more presets to cycle through for different listening room situations and would therefore MUCH PREFER to avoid the long tedious process of ADJ +/- sequentially through unwanted presets (each one at a cost of 5 seconds) and have a DIRECT way to select any one of up to 12 presets, they could choose to allocate their 4 rows of 3 numeric keys for normal (non-TEST) use as one or more of P1-P3, P4-P6, P7-P9, or P10-P12. I would probably find myself in this group of users.

Anyway, the remote has the keys that it has. There are obviously no "dedicated" preset keys like the P1-P4 on the A8 remote. So some other approach is required, if we agree that direct selection of preset is a very useful function, and a way to avoid spending 5-seconds per passing sequentially through each unwanted preset just to sequentially reach a target preset.

I myself would find everyday utility to being able to switch presets directly, in my normal everyday use of the A16 for its now newly flexible numerous different source inputs. In contrast, for me "live solo" of overhead Atmos speakers is something I will likely very rarely need or even want to demonstrate, but if I wanted to I could always re-configure temporarily and then go back to preset use when my "live wow demo" is finished.

So:

(1) revise INPUT to be an in-memory preset-specific value (rather than a global system value applicable to all presets, as it is right now), so that you can configure any preset to be for a specific input by default. And, like VOLUME, being in-memory preset-specific its current setting (which you can change in-memory dynamically) in each preset persists even as you might navigate to other presets and then return back.

(2) provide NVRAM stored initial starting power-on default values by preset for both INPUT and VOLUME,

(3) facilitate use of the 12 numeric keys (in groups of four rows of 3 keys each) on the numeric pad so that each row of 3 keys is independently assigned to either be for (a) any three user-assignable "live" speaker solo, or (b) direct preset selection of P1-P3, P4-P6, P7-P9, P10-12, or even P13-P15 (if that's what you'd like one of the rows to be). Note that to keep this implementation reasonably simple and intuitive, this use and configuration of the 12 numeric keys persists no matter in "live solo" or TEST mode. You can always temporarily re-configure the keys if you want to solo additional speakers ("live" or TEST), temporarily setting them to be for speaker solo use rather than for direct preset use, restoring them back for preset use after you're done with your special situation.

I think this is a reasonable design.

 

[dsperber]

How about making it even more flexible... and probably more simple and straightforward and intuitive.

Today, each of the 12 numeric keys is user-assignable for any speaker, for solo purposes. The assigned speaker value for each button is designated by using ADJ +/- to arrive at the desired value.

So, how about simply adding 16 additional "values" to what gets presented as you ADJ your way through the "dropdown list" of values to be selected from. In other words, just add P1, P2, P3, etc., to also appear in addition to all of the actual speaker names which can be selected from?

That way all 12 keys are individually assignable by the user, to either be (a) speaker solo, or (b) any preset P1-P16!

And of course if it's a speaker solo assignment then it works for both "live" as well as for TEST. If it's a preset direct selection then it does NOT function as speaker solo, no matter "live" or TEST.

How simple is that? I would still almost certainly use the 1-6 numeric keys for P1-P6, since the keycaps are engraved with those numbers. That leaves the remaining lower six keys for soloing the overhead speakers, which also makes intuitive sense. Or, if using six of the outer keys for soloing six overhead speakers makes more intuitive sense to you and this is a strong need, then you can do that and still use the middle column buttons plus bottom three keys for whatever preset numbers you'd like to assign.

The point is that extending the existing user-assigned speaker solo functionality of the 12 numeric keys can easily be understood to also include direct preset functionality, simply by adding another P1-P16 set of values to the "dropdown list" you pass through with ADJ +/-. Intuitive.

 

[audiohobbit]

You read my proposal for easier preset selecting??

At first all those solo buttons seemed so important to you and now you want to reassign them?
I'm against it.

Anyway, I think all those proposals are of no use. I don't even now if they got my email or even read it.

 

[bcwang]

I've got another idea on this whole solo button vs preset discussion.

How about the numeric keys normally act as preset selection. However, once one of the other "solo" buttons are pressed and it's in "solo" mode, then the numeric buttons become the overhead speaker buttons. Once the "all" button is pressed and we are no longer soloing speakers, the numeric keys are back to preset modes.

The only drawback I can think of right now is you can't change presets directly when soloing a speaker. Not sure if this is a use case anyone has. But even if so, it's still relatively few presses to workaround this if given this control scheme.