[Richter Di]

No USB DAC. What does that even mean. No USB out of processor to a DAC? No USB input to be processed, and if not a USB input what then HDMI and SPIDF?

Fair question. My understanding was that you could output digital signals from devices like a Laptop, an iPad (via the camera adapter), Sony NW-ZX2 with the Sony cradle... and the A16 would act as the DAC.
Not all the digital devices offer a SPIDIF out, but many offer USB out.

 

[illram]

Did it drop USB input? I never remember seeing that mentioned.

 

[Richter Di]

In the interest of everyone, as in to avoid getting smyth company loose focus, I have the honor to decide by myself to reply on their behalf:
1. Because this was a challenging project.
2. You have listened to an A16 prototype. Certification and assembly process on the production units is ongoing, thank you for your patience.
3. Because of unforeseen complications, technical and / or legal blocks. Please be assured that we are doing out best to maintain all key functionalities in the shipping product while some functionality may become available after future firmware update.

Best regards,
A smyth impersonator.

Why @arnaud would Smyth loose focus if the French dealer answered my questions.

But I guess your answers are political correct. I congratulate you on an well done impersonator job.

 

[462429]

---
I congratulate you on an well done impersonator job.

And why his job could be done so well? Maybe he 'copied' Smyth well because he was able to 'measure' their info capabilities well and maybe that was possible because their input was of a very low quality? If their info was of high quality, he certainly wouldn't be able to 'measure' and imitate it.
Did he unconsciously discover the hidden miracle of the Realiser A16?!?
It seems to me that his ironic joke could be the answer to my biggest concern, which I recently formulated as a question for Gilles Gerin.
In short - does Realiser A16 'capture' speakers on the lossy level of Dolby Digital (or mp3) and later bring all the input content to that level?
In the other words - is it enough for speakers and headphones (while going through Realiser) to sound the same?
Do headphones sound the same as the speakers would sound when the speakers wouldn't go through Realiser?

 

[sander99]

Fair question. My understanding was that you could output digital signals from devices like a Laptop, an iPad (via the camera adapter), Sony NW-ZX2 with the Sony cradle... and the A16 would act as the DAC.
Not all the digital devices offer a SPIDIF out, but many offer USB out.

I think there is some confusion about the USB connection(s).
The A16 has 2 usb connections, one on the back (B-type) and one on the front (A-type). The one on the front was added to the design later (new prototype). The functionality of the one on the front has not been (fully) implemented. Amongst other things there was an idea to enable the front usb connection to drive an external USB DAC (for the headphones output of one user I guess). The one on the back is supposed to be connected to a computer, to enable sending 16 channels of PCM to the A16, and 2 channels of PCM the other way. They never said that the one on the back is not fully functional yet. However I don't know if other devices than a computer would (in all cases) be able to handle the specific communication protocol to use this (they mentioned ASIO drivers, I don't know what that means exactly).
Edit: Also beware of the difference between driving a USB DAC and just acting as a USB memory device... When you say many devices offer a USB out I think in most cases it is just about acting as a USB memory device.

 

[Eich1eeF]

I think there is some confusion about the USB connection(s).
The A16 has 2 usb connections, one on the back (B-type) and one on the front (A-type). The one on the front was added to the design later (new prototype). The functionality of the one on the front has not been (fully) implemented. Amongst other things there was an idea to enable the front usb connection to drive an external USB DAC (for the headphones output of one user I guess). The one on the back is supposed to be connected to a computer, to enable sending 16 channels of PCM to the A16, and 2 channels of PCM the other way. They never said that the one on the back is not fully functional yet. However I don't know if other devices than a computer would (in all cases) be able to handle the specific communication protocol to use this (they mentioned ASIO drivers, I don't know what that means exactly).

The back USB "sound card" is based on a ploytech device: https://www.usb-audio.com/jct2/, the ASIO driver should reduce the latency, which should not be relevant for playing back "canned" audio. The device appears to support usb class standard audio as well, so the A16 should work as a sound card for practically all players that support such use.

I don't know if if has been announced, but I hope that one should be able to play back audio from a computer, and re-record the "realised" 2 channel version via the same sound card at the same time.

 

[raband]

Got a reply from Gilles for the question I asked.

Was basically:

Why did you send me this? I don't have anything to do with Smyth's production.

All I do is try to keep the local French backers informed.

Am tipping someone forgot to tell him that they'd volunteered him to act as a go-between?

 

[sander99]

It seems to me that his ironic joke could be the answer to my biggest concern, which I recently formulated as a question for Gilles Gerin.
In short - does Realiser A16 'capture' speakers on the lossy level of Dolby Digital (or mp3) and later bring all the input content to that level?
In the other words - is it enough for speakers and headphones (while going through Realiser) to sound the same?
Do headphones sound the same as the speakers would sound when the speakers wouldn't go through Realiser?

Have you read the A8 thread? There have been some serious audiophiles participating there. Some were sceptical in the beginning when they just heard about the Realiser A8. Some also worried about processing in 48Khz sampling rate. And about the extra DA and AD conversion with the initial version without HDMI inputs. But when they got their A8's they became super enthousiastic. There is no reason to believe the A16's core functionality is gonna be on a lower level than the A8.

 

[462429]

Have you read the A8 thread? There have been some serious audiophiles participating there. Some were sceptical in the beginning when they just heard about the Realiser A8. Some also worried about processing in 48Khz sampling rate. And about the extra DA and AD conversion with the initial version without HDMI inputs. But when they got their A8's they became super enthousiastic...

Yes, I have read it and I haven't noticed much audiophile enjoyment reporting. I mostly noticed preoccupation with correcting something.
Besides, what does it mean 'an audiophile'? A person who likes strong bass? Besides, who has recommended the Yarra 3DX, Dolby Digital based device, to us? Yes, serious audiophiles.

... There is no reason to believe the A16's core functionality is gonna be on a lower level than the A8.

I see at least one reason - A16 has twice as much channels and more than twice as much work to do.

 

[m8o]

I see at least one reason - A16 has twice as much channels and more than twice as much work to do.

I have to assume there is a significant increase in processing power of the A16 to handle this.... given Moore's law and the years between the devices.

 

[pcfi]

Yes, I have read it and I haven't noticed much audiophile enjoyment reporting. I mostly noticed preoccupation with correcting something.

Given the A8's price I think you'll find that all A8 owners are, by definition, audiophiles. Or professional audio engineers. Does this help? https://www.computeraudiophile.com/forums/topic/12112-review-smyth-realizer/

I see at least one reason - A16 has twice as much channels and more than twice as much work to do.

I really can't see your logic here. Are Atmos receivers worse than stereo receivers because they have more channels to deal with and more work to do?

 

[Eich1eeF]

@camillem:
1. Is the true internal processing level of the Realiser A16 'lossless' or 'lossy'? I mean, given the enormous processing requirements of the real time de-rotation of 16 channels?

Doing those things lossy like MP3 or Dolby Digital would actually be harder than just doing the processing lossless. Most complex audio processing is done in frequency domain, i.e. not on the PCM signal as such, but on a converted version after a Fast Fourier transform. That transform is essentially lossless. It'll take more compute time to "compress" the signal than to perform all other computations on the full signal, since the psychoacoustic model required for compression would be rather complex and introduce additional latency.

2. When Realiser A16 'captures' high-end speakers, does it 'capture' them in their high-end totality, or it 'captures' them on the level when they play some simpler content like, for example, Dolby Digital? Quite concretely - what is the kHz and bit quality of the measurement 'sweeps'?

The A16 constructs a model of how your "high-end" speakers interact with the room and your head and torso. The fidelity of that model is probably limited due to the signal to nose ratio of the in ear microphones used to record them. The same applies for the corrections for your headphones. However, this does not manifest as a noise floor added to all signals, instead, any signal played back is "off" by exactly the same relative amount (e.g. 1/10000) and in the same direction. It's as if the room you measured your speakers in was a few millimeters off in size. There's probably absolutely no point in going beyond 16 bit for the measurements, since the microphones will be a lot worse than that. I believe the HRTF measurements are currently sampled at 48kHz and 96kHz support will come in a later update. My speakers are all limited to ~23kHz, and my ears to about 20kHz, so I don't think it'll make any appreciable difference.

I'd say the relevant "resolution" characteristic of the HRTF would probably be the number of distinct frequencies for which the measurements are stored, comparable to a graphical equalizer, if you will. The other limit would be the maximum reverberation time that can be measured, but that has already been documented somewhere.

 

[Eich1eeF]

I have to assume there is a significant increase in processing power of the A16 to handle this.... given Moore's law and the years between the devices.

The A16 uses Analog ADSP21489 according to the block diagram.
https://www.analog.com/en/products/adsp-21489.html#product-documentation
The A8 comes with a ADSP21369: http://signalsurbruit.blogspot.com/2011/03/svs-smyth-realiser-a8.html with a datasheet at https://www.analog.com/en/products/adsp-21369.html#product-overview.

The 21489 is binary code compatible to the 21369 according to the data sheet. I had assumed that it would be multiple times as fast, but all I can find is 2.4 GFLOPS peak for the A8 and 2.7 GFPLOS peak for the A16. There's also https://www.analog.com/en/design-center/landing-pages/001/sharc-benchmarks.html with 27.43 us for a 1024 point fft on the A8, and 20.44 μs on the A16, and similar ratios for other common DSP tasks.

 

[J.Villain]

In the interest of everyone, as in to avoid getting smyth company loose focus, I have the honor to decide by myself to reply on their behalf:
1. Because this was a challenging project.

Best regards,
A smyth impersonator.

So this is what it has come to, some one posing as a Smyth because the real company doesn't give a damn. How about because they are spending their time working on the project for sE Electronics and basically dropped this project because Zou wasn't stupid enough to give them the money up front? They have our money and don't care if this project ever ships. They have spent a year working on the exchange web site. $500 and a high school kid would have had it finished in a week. There is no way they should have asked for our money beyond the deposit when they did. In my opinion their actions have been fraudulent. And no I will not shut up.

A seriously ticked off customer.

 

[arnaud]

So this is what it has come to, some one posing as a Smyth because the real company doesn't give a damn. How about because they are spending their time working on the project for sE Electronics and basically dropped this project because Zou wasn't stupid enough to give them the money up front? They have our money and don't care if this project ever ships. They have spent a year working on the exchange web site. $500 and a high school kid would have had it finished in a week. There is no way they should have asked for our money beyond the deposit when they did. In my opinion their actions have been fraudulent. And no I will not shut up.

A seriously ticked off customer.

I don’t know about your line of work but, in general, one does need some cash to operate and work on something to pay the bills. I have no clue if what you insinuate about other project is true or not. But if it is, I find that reassuring as it means they have some source of income to keep them afloat.

The way i see it, us early backers have got a tremendous deal and it’s only fair to sit tight until this thing ships. Again, if you are that bitter, why not just give up your pledge and ask for yor money back rather than trying to keep throwing oil on fire?

With some of the posters, it feels like the main item on their agenda is to paint the company in the worst possible light, proclaim that the thing won’t ever ship (e.g put it in the same bag as ossic and these other vaporware kickstarters).