|
|
Post by kcatthedog on Sept 4, 2015 15:39:21 GMT -6
the Apollo is a digital out that does not truncated the digital word.
|
|
|
|
Post by NoFilterChuck on Sept 4, 2015 15:45:19 GMT -6
SPDIF protocol demands maintaining a 32-bit word length, so you're gonna have to elaborate a bit more, as SPDIF devices don't allow word length truncation...
|
|
|
|
Post by Johnkenn on Sept 4, 2015 16:06:18 GMT -6
I had the Thunderbridge with my Symphony for a while...I think it would be a good investment for you, NoFilterChuck...maybe they have a refurb or a used one somewhere? There were just many less bugaboos when connected via Tbolt. |
|
|
|
Post by Johnkenn on Sept 4, 2015 16:07:37 GMT -6
BTW - I have no idea about truncation and trunk size...but this thing sounds pretty damn good here  |
|
|
|
Post by NoFilterChuck on Sept 4, 2015 17:04:27 GMT -6
Definitely no used ones on ebay/reverb/craigslist. *sigh*
|
|
|
|
Post by svart on Sept 4, 2015 19:04:44 GMT -6
SPDIF protocol demands maintaining a 32-bit word length, so you're gonna have to elaborate a bit more, as SPDIF devices don't allow word length truncation... The subframes are 32 bit, but the audio data is a maximum of 24 bits per subframe and whether that is filled with 24 valid bits or not, the encoder still creates 24 bits in the stream. It'll just fill the empty ones with zeros.. The rest of the bits are preamble, status and control bits. One 32 bit subframe is left data, the next is right data to for a complete frame. 192 of these is one "block" with one collective 192 bit word for block status made up of a single bit per subframe. However! None of this matters beyond academics because SPDIF/AES encoders/transmitters are almost always serial data PCM input.. (sometimes I2S) That's right, they generally use the same digital data streams that any D/A would use! In other words, the level could be controlled at the driver level before it ever gets to the SPDIF/AES encoder and the encoder would simply encode the level it was given without any extra signal loss besides what would normally happen on any DAWs faders. I don't know that any DAW does straight truncation on faders, and I highly doubt it. It's most likely a mixture of algorithms working for both speed and precision. In any case, the digital stream would be processed the same way a fader on a DAW would be, with no differences. So I don't know why they say they can't offer a fader in the Symphony, but they do in the Apollo. I have a feeling that it's a case of the symphony having limited processing resources available, or some other limitation that they are spinning as some kind of positive.. Bits are bits, they are just being moved from one protocol to another and back. |
|
|
|
Post by Johnkenn on Sept 4, 2015 21:56:49 GMT -6
SPDIF protocol demands maintaining a 32-bit word length, so you're gonna have to elaborate a bit more, as SPDIF devices don't allow word length truncation... The subframes are 32 bit, but the audio data is a maximum of 24 bits per subframe and whether that is filled with 24 valid bits or not, the encoder still creates 24 bits in the stream. It'll just fill the empty ones with zeros.. The rest of the bits are preamble, status and control bits. One 32 bit subframe is left data, the next is right data to for a complete frame. 192 of these is one "block" with one collective 192 bit word for block status made up of a single bit per subframe. However! None of this matters beyond academics because SPDIF/AES encoders/transmitters are almost always serial data PCM input.. (sometimes I2S) That's right, they generally use the same digital data streams that any D/A would use! In other words, the level could be controlled at the driver level before it ever gets to the SPDIF/AES encoder and the encoder would simply encode the level it was given without any extra signal loss besides what would normally happen on any DAWs faders. I don't know that any DAW does straight truncation on faders, and I highly doubt it. It's most likely a mixture of algorithms working for both speed and precision. In any case, the digital stream would be processed the same way a fader on a DAW would be, with no differences. So I don't know why they say they can't offer a fader in the Symphony, but they do in the Apollo. I have a feeling that it's a case of the symphony having limited processing resources available, or some other limitation that they are spinning as some kind of positive.. Bits are bits, they are just being moved from one protocol to another and back. Yeah...exactly what I was going to say... |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Sept 4, 2015 22:21:19 GMT -6
Uhm. Yes. The supporter from UA (Edit: I meant Apogee!) is talking, uhm, BS about this so called reason IMHO, sounds somehow logical if you have no clue about DSP, but isn't. How do you explain a missing feature in an expensive machine? Well, with something like this...
The only real advantage of an analog volume control or monitor controller is a limitation of volume in case something went wrong with the computer/DAW/driver/whatever, so this can prevent possible damage to expensive monitoring. Mute button is nice for similar reason...
|
|
|
|
Post by kcatthedog on Sept 5, 2015 5:22:42 GMT -6
SBF if you are talking about me I don't work for UA I just use the apollo. I am just talking about the apollo master volume control which I believe controls all volume and doesn't truncate the Digital word as it is adjusted.
I haven't used the svart box yet but with my dbox you used to set this control to max and mirror spdif to main outputs. I know Martin used the SB with his apollo with no problems as done JohnKenn as well.
Every input to apollo goes through its gui ,Console ,and then normally to the main outs. You can select that spdif mirrors the main outputs so I thought in effect the master rotary control then by default is also controlling the output on spdif ?
This is what I was referring from current 8 hardware manual:
(22) Monitor Level & Mute Knob
This rotary encoder serves two functions. Rotating the knob adjusts the monitor output level, and pressing the knob mutes the monitor outputs.
Monitor Level
Rotating the knob clockwise increases the signal level at the Left & Right Monitor Outputs on the rear panel. If ALT monitor outputs are configured and active, this knob controls the signal level at the ALT monitor’s line outputs.
Although this is a digital control, the Left & Right Monitor Outputs volume is attenuated in the analog domain, after D/A conversion (digitally-controlled analog volume). This method provides the utmost monitoring fidelity, in contrast to digital volume controls that reduce levels by truncating the digital word length (aka “dropping bits”).
I honestly don't care what interface people use, I just know about the Apollo so I share info about it as I am up on the ua forum a lot. The formal UA rep there and here is Gannon (Universal Audio).
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Sept 5, 2015 8:14:35 GMT -6
I hit up Apogee about this and asked them if they could add this feature, and this is what they told me: Oh, sorry...i did not mean you, but the guy from APOGEE who wrote this statement.... My fault! |
|
|
|
Post by NoFilterChuck on Sept 5, 2015 10:50:26 GMT -6
I could see how the Apollo would be using the audio driver to create the SPDIF stream that goes to the SPDIF chip, as opposed to a dedicated hardware chip, because of the increased bandwidth offered by using Thunderbolt. Since the Symphony is older, they are probably using a hardware chip to automatically convert the packets sent over the USB cable to the DAC chip. meaning that the Driver can't talk directly to the SPDIF chip, like they can with the Apollo. I could see this being the reason that they can't implement volume control of the SPDIF outputs. It's probably a similar situation as how you can't put thunderbolt on those old Mac Pros, because the motherboard just doesn't have the functionality.
|
|
|
|
Post by dandeurloo on Sept 5, 2015 23:52:59 GMT -6
My guess is they handle it how motu handles it in the 16a. Which adds an additional not so great 8 channel op amp that is controlled digitally. It is a bottle neck. It took a
Iot of modding to clean it up. The only way to know is to demo a real monitor controller and see if you hear a difference. Do you know anyone with one you can demo. Scott Liebers has a killer one that he should sell. It sounds amazing with all the right features!
|
|
|
|
Post by kcatthedog on Sept 6, 2015 4:41:58 GMT -6
no probs SBF !
|
|
|
|
Post by svart on Sept 6, 2015 8:50:28 GMT -6
My guess is they handle it how motu handles it in the 16a. Which adds an additional not so great 8 channel op amp that is controlled digitally. It is a bottle neck. It took a Iot of modding to clean it up. The only way to know is to demo a real monitor controller and see if you hear a difference. Do you know anyone with one you can demo. Scott Liebers has a killer one that he should sell. It sounds amazing with all the right features! We're talking about the digital stream before the spdif conversion. Most devices with fader for digital out levels do so in software level operations before sending digital stream the spdif encoder chip. |
|
|
|
Post by svart on Sept 17, 2015 12:03:05 GMT -6
Updated the first post with specific specs.
|
|
|
|
Post by Johnkenn on Sept 17, 2015 12:33:34 GMT -6
Have you sold all the first batch?
|
|
|
|
Post by Johnkenn on Sept 17, 2015 12:34:39 GMT -6
You should update the pic in the first post too.
|
|
ilok
New Member
Posts: 2 
|
Post by ilok on Sept 21, 2015 10:26:54 GMT -6
|
|
|
|
Post by svart on Sept 21, 2015 16:31:55 GMT -6
I've seen that site. It's an interesting take on using a PCM1794 DAC chip. However, I don't understand the desire to run the chip without digital filtering. Lets take a second and think about it.. So say you are running your DAC clock at 44.1K, and your highest frequencies you want to reproduce are at say 22khz, then: 44.1K-22K= 22.1K 22.1K is where your first "image" of the signal will start without digital filters. That's only 100hz away from your signal of interest! What you'll have here is a noisy mess up in that range as your beat products start to fill the spectrum in that area. This is also going to eat up headroom on opamps if the images are allowed to continue without filtering. Most people who experiment with DACs and poorly designed reconstruction filters actually hear these image/mix products and think it's "detail", much in the way an aural exciter works by creating harmonics in the upper frequencies. The problem is that these are not true to the source. It may sound euphonic, but it's not real. Because of that, I have chosen to use the digital filters, which are much more efficient at getting rid of images and harmonic trash and allows me to use the least amount of analog filtering. I suppose one could create a high order low pass filter network, but then you'd run into issues with group delay, phase issues over frequency, etc, much like folks run into issues building higher order crossovers for speakers.. It's pretty much the same thing, with the same issues. Honestly it's a lot of work for barely any benefit, but a lot of potential problems. |
|
ilok
New Member
Posts: 2
|
Post by ilok on Sept 23, 2015 10:38:07 GMT -6
NOS just sounds better, more analog like sound, and people will pay for it |
|
|
|
Post by svart on Sept 23, 2015 10:48:26 GMT -6
NOS just sounds better, more analog like sound, and people will pay for it Well, "analog" sound typically means harmonics. Soft compression from tape and tubes causes light harmonics, and a euphonic sound.. Same as running without digital filters, which as I mentioned, will allows images and artifacts to mix down into the audible region causing distortion, and thusly an analog sound. So, it may be pleasing, but it's not what I was after with this design and I think people appreciate an uncolored design. |
|
|
|
Post by Martin John Butler on Sept 23, 2015 10:54:11 GMT -6
At this point, that makes sense to me. Every Slate, UAD, Waves, et al product I can think of trumpets "color" or "mojo" or "analogue" or "tube" or something like that.
With such an avalanche of "color", I'd think most people want the cleanest reproduction of the sound they colored the way they want it, not more color. That is unless you absolutely love the particular color a converter adds, like the way a Burl does.
|
|
|
|
Post by Johnkenn on Sept 23, 2015 16:52:33 GMT -6
I kind've see the AD conversion as the foundation that you build everything else on top of. I like the idea of having a clean canvas to paint on.
|
|
|
|
Post by kcatthedog on Sept 23, 2015 17:16:42 GMT -6
You ought to now, seeing as how you have one |
|
|
|
Post by Johnkenn on Sept 23, 2015 19:12:05 GMT -6
And I think it should speak to what I think of it. I've had Lavry, Symphony, Apollo, Butl, RM, etc.
|
|
