DAC crowdfunding now in progress

[Hank Murrow]

Beautiful!! Thanks Doc.

[Woodenboat]

Yes, Thank you Doc!  This is my first DAC, and a bit of a stretch to my budget, so I am VERY interested in what's going on with the development.  Would it be too much to ask for an update, say, every two weeks or so until the first units ship?  Thanks again!

[Doc B.]

How about if I agree to give an update when there is something to share? The way this works is we send a bunch of money, wait a while, then a big box shows up and we start doing all sort of interesting things I'd be happy to share. Right now we are somewhere in the middle of the wait a while part.

The next thing I will be doing is working wth PB to show him how we put the carbon fiber look on the cabinets. Peebs is under the weather this week, but we might start on that next week before we shut down for the holiday.

[Woodenboat]

Thanks for getting back again Doc, what you said is exactly what I was hoping for and looking forward to.  Hope you all have a great holiday!

[Paul Birkeland]

I can say that while we are waiting for the boards to be printed, stuffed, and tested, we will be working on prepping the cabinets and procuring all the other random parts from our vendors.  I'll also be sneaking into the office this coming week to thrash around with the programming software. 

We expect things to move fairly quickly once the boards arrive, and it does seem reasonable to pop up some photos on this thread of our progress.

-PB

[John Swenson]

I also did a tweak to the filter yesterday which gets just a little more "air" and "openness" out of your files.

I was also tweaking the gain slightly. You want the highest gain you can get out of the DAC chip to get the best SNR, but if it's too high you can get clipping on some files. This is called inter-sample peaks. Lets say the original analog waveform coming from the mics has a sharp peak that is in-between two samples, if those two samples are set to maximum digital level, when the filter in the DAC is computing what the analog signal did in-between the samples it can get a signal higher than maximum. My filter is particularly prone to this since it is optimized to faithfully reproduce the original transients.

I thought I had this dialed in just right, but yesterday I was listening to some acapella choral music and noticed some clips, so I had to tweak it down just a little bit more and then gave it just a little more margin. It now handles everything I have just fine. It is strange that just singing has the steepest transients I have in my recordings, you would not expect that, but I have seen it many times before. Not piano or drums, but good old singing, the human voice can do amazing things.

(BTW with the new filter that recording is out of this world)

John S.

[Paul Birkeland]

I worked a bit today on applying the carbon fiber vinyl to the DAC enclosure covers.  Doc B. showed me some of his mysterious techniques, and these will be keeping me busy for a while!

[Johnnycopy]

I worked a bit today on applying the carbon fiber vinyl to the DAC enclosure covers.  Doc B. showed me some of his mysterious techniques, and these will be keeping me busy for a while!

thanks for the updates to you and to John S.  Exciting times.

By the way, A month ago i saw a bmw sport utility totally reskinned in carbon fibre vinyl.  Strange sight, took the installer three days.

[Doc B.]

The CF vinyl was inspired by my CBR project. I was happy with the look and when we wanted to do something to the DAC cabinet that was a little different a light bulb went off.

[Natural Sound]

I was also tweaking the gain slightly. You want the highest gain you can get out of the DAC chip to get the best SNR, but if it's too high you can get clipping on some files. This is called inter-sample peaks. Lets say the original analog waveform coming from the mics has a sharp peak that is in-between two samples, if those two samples are set to maximum digital level, when the filter in the DAC is computing what the analog signal did in-between the samples it can get a signal higher than maximum. My filter is particularly prone to this since it is optimized to faithfully reproduce the original transients.

John, I have experienced the clipping you are talking about on my current DAC. Quite frankly it drives me nuts when I hear it. I'm glad that you have taken the time to address this phenomenon before the first run ships out. 

[pdxgrampa]

     
Does this mean an Ohlins rear shock and SHOWA forks on the the DAC and Yoshimura rearsets on the battery kit? 

[johnsonad]

Does the DAC in current form run with OSX Yosemite?

[Johnnycopy]

The CF vinyl was inspired by my CBR project. I was happy with the look and when we wanted to do something to the DAC cabinet that was a little different a light bulb went off.

Stealth!

[kumasan]

I thought I had this dialed in just right, but yesterday I was listening to some acapella choral music and noticed some clips, so I had to tweak it down just a little bit more and then gave it just a little more margin. It now handles everything I have just fine. It is strange that just singing has the steepest transients I have in my recordings, you would not expect that, but I have seen it many times before. Not piano or drums, but good old singing, the human voice can do amazing things.

It makes me think if the same problem will be found in even more dynamic recordings. Will it be possible to write a small application that runs through a ripped CD collection and log any problems regarding inter-sample peaks? In that way we could all run the app and send your the result hence having a much larger sample of CD's analyzed to avoid problems.

[John Swenson]

It makes me think if the same problem will be found in even more dynamic recordings. Will it be possible to write a small application that runs through a ripped CD collection and log any problems regarding inter-sample peaks? In that way we could all run the app and send your the result hence having a much larger sample of CD's analyzed to avoid problems.

Such software would have to actually implement the filter and run it for real on the data, checking for overflow as it went. This is not a trivial little piece of software and of course there would have to be versions for at least windows, OSX and Linux.

Not really something I want to deal with right now. I added a little bit of extra margin so it should be all right for anything out there.

Strange as it may seem it is compressed recordings that have the problem. Think of it this way, you have an analog waveform whose peak analog voltage is exactly equal to the maximum of the ADC. If the sample points fall on both sides of a sharp transient, those sample values will be less than full value, so the reconstruction will be right at maximum value, no clipping.

The problem happens when the digital stream is then compressed and the whole thing is moved so the highest sample value is now at the maximum possible sample value. Thus when reconstructed the inter sample peak is greater than the max and it clips.

John S.