[lag0a]

I am wondering if you can add RAM and a RAM interface into your DAC to store the music file there temporarily to play it then automaticly delete it.
 
This idea came to me because I think the CPU imprints it's own sound signature while processing and buffering the music file into your DAC. I don't think you can do much with adding a separate CPU and motherboard interface to your DAC but a RAM interface seems possible. This is to stop the computer from consistently buffering, streaming, and communicating with the DAC through usb, toslink, or coaxial spdif.
 
If the computer can first buffer and send the whole music file into the DAC's RAM then stop communicating with the DAC until the music file is done playing could improve the sound.
 
If there is already such technology out there then please let me know or any other similiar and better storage instead of RAM.

 

[Head Injury]

Do you have any evidence that suggests this is necessary at all? That the CPU really does "imprint its own sound signature"? That RAM in the DAC would fix this if it were true?

 

[lag0a]

I have tried undervolting and underclocking my CPU and it sounds a lot better that way. Some CPUs might not boot up but if you can undervolt it to 0.8v and 1ghz then it should sound better than higher volts and higher clock. I have also read about people using linear power supplies at the P4 (CPU) connector and P24 (Motherboard) connector, and linear power at the P4 connector drastically improved the sound but lesser improvement at the P24 connector. Sound improvement at the P24 is about 1/3 compared to the P4.  I don't know if it will fix it but it will stop the computer from communicating with the DAC before the music plays so less interference from the computer. I just feel something is wrong with the computer having a link with the DAC and, consistently buffering and streaming with the DAC while the music is playing.

 

[John In Cali]

That would be amazingly hard. But if you can accomplish it, all the best to you.

 

[lag0a]

I don't have the know how to build something like this or software engineer skills to program it so I'm pretty much looking for something already made or being developed. I think this RAM/RAM interface should come after the USB/toslink/coaxial spdif receiver in your DAC to receive the data then store it before ASRC or sent directly to the DAC chip. I don't care if after receiving the whole music file the RAM/RAMinteface disable the USB/toslink/coaxial spdif receiver in your DAC to stop communication with the computer since you already have the music file to play.

 

[Head Injury]

Quote:

I have tried undervolting and underclocking my CPU and it sounds a lot better that way. Some CPUs might not boot up but if you can undervolt it to 0.8v and 1ghz then it should sound better than higher volts and higher clock. I have also read about people using linear power supplies at the P4 (CPU) connector and P24 (Motherboard) connector, and linear power at the P4 connector drastically improved the sound but lesser improvement at the P24 connector. Sound improvement at the P24 is about 1/3 compared to the P4.  I don't know if it will fix it but it will stop the computer from communicating with the DAC before the music plays so less interference from the computer. I just feel something is wrong with the computer having a link with the DAC and, consistently buffering and streaming with the DAC while the music is playing.

What CPU is this? Do you have any power-saving features enabled? Are these only subjective impressions, or have you done objective tests?
 
This all sounds like typical audiophoolery. Develop a way to fix something that isn't a problem, then charge extra for it.
 
No modern CPU should struggle with data streaming. If they did, we'd be in trouble.

 

[lag0a]

AMD Athlon 64 3700+. Cool n Quiet lowers it to 1.1v 1ghz so I had to disable it to try 0.8v and 1ghz. I don't have any equipment for objective tests but my ears could hear the difference immediately.
 
Well, this doesn't have to be a fix for anything since some people may want something like this as a small temporary storage device inside their DAC instead of streaming from an external source.
 
I'm not talking about the CPU struggling to do data streaming or anything else. I'm talking about breaking the connection between the computer and DAC after the DAC receives the whole music file or else the CPU needs to be actively streaming to the DAC.

 

[Head Injury]

Quote:

 
I'm not talking about the CPU struggling to do data screaming or anything else. I'm talking about breaking the connection between the computer and DAC after the DAC receives the whole music file or else the CPU needs to be actively communicating with the DAC.

Why wouldn't the CPU be able to stream data to the DAC without error? It's only digital information, 1s and 0s. A CPU can't color that just by sending it. You need to do so purposely, with plugins that analyze and alter the data.
 
If you're getting errors, you're getting drop out. That's not going to be coloration, that's going to be static and missing sound. Jitter could be an issue, but there's many DACs on the market that already have jitter reduction mechanisms.
 
I guess I just don't see what the problem is or why reducing the CPU's voltage helps. And with only subjective impressions there's no guarantee you're hearing something real at all.
 
Maybe your power supply isn't very good, and is generating noise at high CPU voltages. This probably isn't going to affect the digital signal enough to matter, and it won't be solved with different buffering settings.

 

[Gatepc]

I'd say that the CPU effecting sound is impossible. However the power supply could be horrible and could perform better when using less power (when you under volt it). So it could be the power supply that just sucks.

 

[lag0a]

It could be the voltage regulators in the switching power supply and the voltage regulators in the motherboard suck so undervolting the CPU changes the sound.

 

[mikeaj]

Is your DAC powered by the computer power supply?  If yes, then above two posts may possibly be relevant, though I'm still very skeptical of such a change.  Invest in a better computer power supply in that case.
 
To answer the original question, if somehow you buffer the audio data in the extended-functionality DAC (this is somehow sounding more like a separate digital music streamer computer), what about the microcontroller that feeds the data from the buffer to the DAC?  Oh no that's going to color the sound too?

 

[lag0a]

No. My Twisted Pear Dual Mono Opus DACs are powered by their own linear power supplies. I connect my computer's sound card's coaxial digital output to my DAC's coaxial SPDIF receiver.
 
The CPU controls everything in the motherboard so I am pretty sure it is the cause aside from the power from the outlet -> computer switching power supply -> motherboard that could produce power line noise from sharing power in the neighborhood -> computer power supply voltage regulators -> motherboard voltage regulators.
 
I haven't thoroughly thought out what this device is. It's just an idea. It can't be worse than my 170+ watt computer if it only has RAM and a RAM interface.
 
If your current computer source setup isn't smart enough to stream the data to this unknown device (RAM/RAM interface) then after sending all the data stop communicating with it then I don't know what else can so you don't actually need a separate music streamer. This can even solve dropouts since you won't need to buffer it anymore.
 
The key point is storing the data  in RAM at the DAC side instead of buffering from the computer so ASIO for low latency, and real time streaming from the OS won't matter.