John Curl's Blowtorch preamplifier part III

Status
Not open for further replies.
All of AD's embedded ASRC's are probably based off of the AD1896 and while not perfect, is lower distortion than everything else in most systems.

Very true. Problem is the distortion and or the resampling filter is still audible to some extent or other. There may be a few choice of filters, but it always seems to be a zero sum game where one can trade off one audible defect for another. But, I am very picky when it comes to SQ. It doesn't take much ear fatigue to mask defects, either. Brains also accommodate to many smaller SQ shortcomings typically over a few days of listening, probably the main effect going on behind many claims of component burn-in.

EDIT: Anyway, if too many little defects are allowed to exist I hear them right away when I listen to a new system. In fact since we are talking about a particular application, I have been over to Richard Marsh's house and listened to his main system and his headphone system at the computer. I want him to be happy with the sound and while I think there is probably room for improvement from my perspective, Richard has his own preferences and priorities. I don't want to be a pain about it. But, since we are talking about what he might do, I would encourage some solution I would like on the assumption he would too once he heard it. 🙂
 
Last edited:
I don't know how audibly transparent (or not) the AD1896 is. If I had to guess I would say it's probably close based on the fact that it was incorporated into a number of well-regarded components including the original Benchmark DAC1. The current DAC3 uses TI SRC4392 which is conceptually similar but is probably a little better.
 
I'm not clear on the concept of THD in isolation with a purely mathematical process. There are compromises at any point in time to account for computational limitations to achieve real time results. I usually see these as a total artifact/noise floor rather than concentrated in harmonics. I might be wrong, the data sheets these days are very sparse and don't offer much insight.

Yeah, the AD1896 datasheet is actually the best of the group. I found the characterization a little strange at first but I think it's described in this paper:

http://www.analog.com/media/en/tech...-articles/5148255032673409856AES2005_ASRC.pdf
 
EDIT: Anyway, if too many little defects are allowed to exist I hear them right away when I listen to a new system. In fact since we are talking about a particular application, I have been over to Richard Marsh's house and listened to his main system and his headphone system at the computer. I want him to be happy with the sound and while I think there is probably room for improvement from my perspective, Richard has his own preferences and priorities. I don't want to be a pain about it. But, since we are talking about what he might do, I would encourage some solution I would like on the assumption he would too once he heard it. 🙂

I am not here. I am over there. Hanging with my girl(s). I will be back on Aug 19 and soon will have two more of the stereo amps completed. They will be starting on mine next week and expedite their build because I said Mark has to hear them 🙂 Then, all those things I had in the mental pipeline will get done. We can even listen to HPA off the DAC3, too. We will also use HD downloads for music source material.

I know you are a critical and experienced listener and so I will bet you a dollar you are going to be really happily surprised at what it sounds like. Its going to be fun! , 🙂


-Richard

36572346_1161687830650721_6247201881047695360_n.jpg 37335347_1177593162393521_3307524743799767040_n[1].jpg

Who wants to work with this next gen EE? Raise your hand.
 
Last edited:
Yeah, the AD1896 datasheet is actually the best of the group. I found the characterization a little strange at first but I think it's described in this paper:

http://www.analog.com/media/en/tech...-articles/5148255032673409856AES2005_ASRC.pdf

An alternative is to use all material with the same sample rate, then no on-the-fly conversion is needed. You'd probably have to convert all the source material offline as it were, but that process could be highly optimised as it doesn't matter how long the processing takes?
 
That does work, but you'll find very few people that want to make a DAC or DSP system that operates at only one sample rate. ASRC are most useful to bridge clock domains, not only for sample rate conversion. Without an ASRC you will need a VCO + PLL, FIFO, or similar when interfacing between two clock domains such as an external DAC and recovered SPDIF clock (if you don't want to clock the DAC directly from the recovered clock, that is). So, it's quite useful if you have multiple sources like a USB interface, SPDIF interface, etc.

On a PC you can do transparent (IMO) SRC in real-time though anyway. I find Foobar2000's PPHS resampler to be fine but there are many good options.
 
Last edited:
I am not here. I am over there. Hanging with my girl(s). I will be back on Aug 19 and soon will have two more of the stereo amps completed. They will be starting on mine next week and expedite their build because I said Mark has to hear them 🙂 Then, all those things I had in the mental pipeline will get done. We can even listen to HPA off the DAC3, too. We will also use HD downloads for music source material.

I know you are a critical and experienced listener and so I will bet you a dollar you are going to be really happily surprised at what it sounds like. Its going to be fun! , 🙂


-Richard

View attachment 694821 View attachment 694823

Who wants to work with this next gen EE? Raise your hand.
Richard, she is beautiful, other girls too. Who thy are?
 
Richard, she is beautiful, other girls too. Who thy are?

All live in Nepal. The single photo is my girl. Her name is Sarada. But I call her Sara. And, she is the tall one in the middle of the gang of girls. 19 years old. I have been Godfather to her and family since I met her at about age 5-6. I have supported her education ever since and visit her every year and we talk on Facebook daily. She is from subsistence farmer family in very remote region -- no roads... I never been to their village because it is a 3 day and night walk one way. Not fun and no facilities in village. I have educated her and brought her to Bangkok about 8 times... she is in my condo sleeping right now. She gets straight A grades and has been accepted to many colleges... all that she applies for in several countries. She reads, writes and speaks fluent English. She is great in mathematics and science. She will be going to either Canada or Australia in a couple months for university education as an EE. She will get MS also. I hope my kidney holds out long enough to see that day for her. She will now have a good life and future and she will help her poor family also. Especially, her sister and young brother.

The girls are all just girl friends she has known a long time. Also, so talented and so beautiful inside and out. Sensitive, loving and smart. So many people are poor in the world and have to live boring dreadful lives because they are locked in bad conditions. They live in quiet desperation. You can make a difference.

Anyone wishing to go to Nepal should contact me and I will give you info on her father who is a guide.

<3

-Richard
 
Last edited:
That does work, but you'll find very few people that want to make a DAC or DSP system that operates at only one sample rate. ASRC are most useful to bridge clock domains, not only for sample rate conversion. Without an ASRC you will need a VCO + PLL, FIFO, or similar when interfacing between two clock domains such as an external DAC and recovered SPDIF clock (if you don't want to clock the DAC directly from the recovered clock, that is). So, it's quite useful if you have multiple sources like a USB interface, SPDIF interface, etc.

On a PC you can do transparent (IMO) SRC in real-time though anyway. I find Foobar2000's PPHS resampler to be fine but there are many good options.

It's really not that difficult to make a system that loads the right coefficients when the sample rate changes. It can usually happen in milliseconds. No need for sample rate conversion. Yes, need for good system and software design. I know its possible, we did it in a headphone.
 
Agreed. Two cups or avg less of tea per day each containing conservative amount of Nettle leaves (level teaspoon) four maybe five days per week for now is ample in my recent experience.....additional glasses of pure water are mandatory. In my recent experience this kick starts getting junk out of kidney/liver system and generally....there may be temporary adverse reactions. There is much cut/paste 'info' and claims on the net, only some is sensible it seems....some interesting old wisdom though. Like any such 'herbal', caution and time are involved, and be prepared to take intermittently.

Dan.
 
Richard, I wish that I had something to contribute on curing your kidney problem, but I don't, unfortunately. All I can do is wish you well, and to read what others recommend. I personally appreciate any unorthodox approaches as much as the doctors approved ones. I too have found that some additions to diet can make a big difference with various chronic problems.
 
Last edited:
Status
Not open for further replies.