Splitter Studio Crack

[Nguyet Edmondson]

This is not how I wish it would look like but basically it gives an idea of the features.

- I often need Volume & Pan automation, for example (and this is only one of many) if you are running two parallel treatments on your Main Bus and need to morph from one to the other at one point of the song? I believe this would be more elegant to have those parameters set per channel instead of having to create a Mixtool (which doesn't go down to -inf), rename it "Gain for Chain1" etc..., or the Binaural Pan (or Dual Pan). The Pan part would adjust its function wether you are using "Normal" or "Channel Split".

I believe the concept of the splitter is powerful and can be just as complete as Ableton's racks & chains. I love the idea to use one big window as a patch to route effects rather than creating several sends, busses etc... which would emulate real world console mixing but would take a lot of real estate in your mixer.

Another approach would be to really see this Inserts pages as a mini mixer in itself, you could even add internal FX Sends & VCAs within the page, all routing to the output. It already kind of is a modular mixer after all.

That's exactly what I am trying to do right now! I'm surprised this is not implemented ... seems like an oversight. So now I've gotta figure out how to do this without the splitter ... it's frustrating when you're trying to be creative and end up having to fight with your DAW!

If you're using an HDMI splitter that goes to two TVs, will the Mac recognize both TVs or will it only identify one of them? I DO NOT want 2 TVs showing up as separate displays on my Mac.. All I'm trying to do is clone/mirror one TV to another but can't seem to figure out how to do this. Thank you!

My guess is that the Mac won't detect the TVs, or show them as displays anywhere in System Information. It will see the HDMI splitter as a display, and this will only count as using one of your supported video outputs.

Thanks for any help. Chris said to ping you on this.

I am new to sigma studio (ADAU1452) and try to get something super simple working.

I have a single I2S input at 48KHz, 24 bit data, 32 time periods, MCLK is an input going to pin 21 (XTALIN/MCLK) along with BCLK. LRCLK, and DATA.
I simply want to output four copies of this on the four I2S outputs.

I have tried multiple things and followed what I thought was a simple input and output, compile and download but I have seen different things on all 4 ports. Best I have been able to get is after playing with the GUI, to get serial output 1 to reflect serial input 1 however the sample rate was 1/4.

Any chance you could just create a simple project with one in and 4 out and let me give it a whirl. It would take the least amount of your time and probably teach me a lot as I would be able to go thorough it.

I doubt that you set the PLL source register as 'Direct from MCLK pin', but you must choose "PLL clock' in that register. That register is not quite correctly named, it's not source to the PLL but the source to the core. If your MCLK is 12.288MHz and you choose 'direct from MCLK pin' then you are running the core about 24 times lower than its nominal clock speed. Its nominal clock speed is 294.912MHz. Below is where you have to set PLL.

Is your MCLK and serial clocks from the same clock domain? If your MCLK and serial clocks are from some different devices, then you should use the ASRC, or you should set the start pulse rate to be the incoming serial port rate. By doing this, you can simply have the core start the frame processing after a new sample has arrived.

The below project is set by assuming your clock is divided from the same domain. I assume you are using serial input port 0 and used that port to receive the signals. Change the same setting if you are using any other port.

Internal to the DSP the output of the PLL will be divided down to create the different BCLK and LRCLK rates. since they are divided down from the same clock they will not drift apart and they will be exactly the same frequency. However, the clock edges will NOT be aligned and cannot be made to be aligned. If there was a way to zero the counters that divide down the clock during initialization then they would be the same and not drift but this is not possible with this part. I have tried to work around this.

In addition, when you select which clock to slave a port from, this clock is NOT available for other serial ports to output. The other ports can slave from it but not output the clock from it. The best solution is to use an external clock buffer/driver chip.

By the way, the clock busses for slaving, clock inputs 0-3 are not the same as the clock outputs0-3. There is no bridge internally between the serial input clock domains and the serial output clock domains with the exception of the internal rates of course. But if you want to synchronize a serial input and output ports to an external clock source, then you need to run the clocks to both sets of pins. As an example, if you run a clock to serial input port zero and set the serial input port to slave from clock pins 0. Then you set the serial output ports to slave from clock 0 domain you will get nothing out of the serial output port because it is looking at the serial OUTPUT port clock pins, not the serial INPUT port clock pins. Again, I wish there was a bridge between the input and output ports but here is not. It would have been a timing issue anyway. So you have to externally run the clocks to both sets of pins.

Awesome, thanks very much.
Yes the clocks are all from a single source (same domain) coming in to the I2S input port 1 and to MCLK.

I think I am inferring from what you are saying is that the PLL is locking to BCLK?

Is the MCLK input even used in this design?

The input port IN_0
is 0/1 the left / right inputs from the I2S in? I am guessing that is documented somewhere obvious but I never saw it.

This part looks super cool but I am blown away that anyone could just jump into this without some help.

The tutorials are neat but (to me) assume you have an understanding of what you are doing. Some videos of the simplest things to get you started would go a long way.

Thank you very very much for responding.

The data can easily be routed out of the four serial outputs but the clocks you cannot. They must be connected externally or you use the internal clocks. The possible solutions will depend on what is important to your application.

1) You can do as Harish suggested which is to use the ASRC for the input to convert the sampling rate to the internal rate then set the serial output ports all to be masters then you can drive the rest of the system that way.

2) Which is similar. Since you are using the same master clock as the serial input source is using, you are locked together. So just use the internal clocks to drive the serial output ports and they will be locked but the edges will not be exactly aligned but that might not matter for your application.

I think what Harish provided meets exactly what I needed. I provide the MCLK, BCLK, LRCLK and data and get a locked (x4) output of BCLK, LRCLK and data. Exact alignment is not important.

At the moment I think that is all I need.

Thanks again for all your help.

I suspect there's something wrong with how I'm using dup_count in the precode. Can someone please tell me how to resolve this? Using %count didn't do me any good either, though perhaps I was using it incorrectly.

It's always worth to look at the generated code (code tab) of a transformation. If you can understand the code generated (sometimes it takes some effort I must admit) then you get normally the explanation why something is not working as expected.

In your case: The splitter works on row level. It doesn't help that you define a macro variable as pre-code. The macro variable will simply resolve and then the code the SAS interpreter sees is 'count(var_id)>1'. The count() function counts how many times a substring appears in a string - so it needs at least 2 parameters. I assume when you run your job it's not only not doing what you want but even throws an error.

You cant' use a Splitter Transformation to write duplicates to a separate dataset. I wouldn't know right now if there is any standard transformation which allows you to write duplicate records to one table and the remaining unique records to another table. You might need to implement this as a user written transformation.

Below the SQL code you could use in a user written transformation to get the duplicates. Alternatively you could also use a SQL Join transformation to set up such a SQL (and using a standard transformation is what you should do if you can).

The main reason I made this is so that when working in a studio where the engineer and recording are taking place in the same room, it enables anyone else in the room such as producers, band members, management etc to hear what is going on in my (the engineers) headphones. It is also a good a means to split a signal in other applications.