What makes a good DA?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Mbira

Well-known member
Joined
Jun 4, 2004
Messages
2,422
Location
Austin, TX
I’ve been getting more in to mastering and I currently have an older Apollo 16 FireWire unit. I’m debating whether I should upgrade my DA. The Apollo has AES out and I was considering running that in to a Lynx Hilo. It would also be really nice bacause I don’t have a hardware monitor controller on my desk and so I could just use the Hilo for that. I hear great things about the Hilo, but I’m wondering if people could recommend what objective specs I could or should be looking at to see whether this would be something that I would notice a difference with.

Anyone seen research on what really makes a “reference quality” DA?
 
There’s a lot of great knowledge about converters and sampling on Dan Lavry’s website.

 
AusTex64 said:
There’s a lot of great knowledge about converters and sampling on Dan Lavry’s website.

I remember reading one of his white papers years ago about antialiasing filters. I remember that one focusing mainly on sample rate. I should go and look again to see if he has anything more recent.

Btw are you here in ATX, AusTex?
 
There are things to look for,  high dynamic range,  flat frequency response,  high S/N ratio,  low distortion,  low jitter.

But even if you have specs you still need to listen to the actual unit.  A good DAC imo needs to be accurate,  revealing ,and detailed, so that you can make correct decisions.
 
john12ax7 said:
There are things to look for,  high dynamic range,  flat frequency response,  high S/N ratio,  low distortion,  low jitter.
The problem is today most converters have just about the same specs, determined by the chip manufacturer. The devil is in the details, as in analog path, power supply arrangement, RFI.EMI protection...

But even if you have specs you still need to listen to the actual unit. 
I agree. Unless you have access to a unit and serious measurement equipment, the best assessment one can make today is by living with the product.
 
It's complicated. Modern converters spec great in terms of THD (as measured with sinewaves), frequency response and noise.

To my ears there's something wrong with the current sigma delta converters. To varying degrees (it's astonishing how different even higher end converters with great specs sound next to each other) they sound artificial, loosing impact in the bass and render an unnatural high end (IMHO, of course).

I found the perfect affordable DAC for my ears in a DIY project:

http://soekris.dk/dam1021.html

It's a (conceptually) simple PCM DAC using high precision resistors. THD is not as good as some modern sigma delta chips, but because of the sign magnitude principle THD stays really low regardless of amplitude. In any event, THD is lower than most mic preamps, analog console signal paths and obviously every analog tape recorder ever built.

You can customize the oversampling filters as you like in that DAC (or even go fully NOS - non oversampling, which I would not recommend). I realized that much of the sound of a DAC, at least at 44,1 khz, is actually the filter and how seemingly small changes at the very top end of the audio band change the whole presentation drastically.

With the best filters that DAC really lays bare how good some, but also how bad most recordings really sound. Good recordings have the effortless dynamics and natural top end quality of vinyl.
 
living sounds said:
It's complicated. Modern converters spec great in terms of THD (as measured with sinewaves), frequency response and noise.

To my ears there's something wrong with the current sigma delta converters. To varying degrees (it's astonishing how different even higher end converters with great specs sound next to each other) they sound artificial, loosing impact in the bass and render an unnatural high end (IMHO, of course).

I found the perfect affordable DAC for my ears in a DIY project:

http://soekris.dk/dam1021.html

It's a (conceptually) simple PCM DAC using high precision resistors. THD is not as good as some modern sigma delta chips, but because of the sign magnitude principle THD stays really low regardless of amplitude. In any event, THD is lower than most mic preamps, analog console signal paths and obviously every analog tape recorder ever built.

You can customize the oversampling filters as you like in that DAC (or even go fully NOS - non oversampling, which I would not recommend). I realized that much of the sound of a DAC, at least at 44,1 khz, is actually the filter and how seemingly small changes at the very top end of the audio band change the whole presentation drastically.

With the best filters that DAC really lays bare how good some, but also how bad most recordings really sound. Good recordings have the effortless dynamics and natural top end quality of vinyl.

That project looks interesting.  I didn't see much info there on the website about the DIY aspect-Do you know if it's possible to just buy the board?  It looks like it's fully assembled (and sold out everywhere I can find).
 
living sounds said:
To my ears there's something wrong with the current sigma delta converters.
you are not alone.
there have been a few folks in mastering that share your opinion.,
they would not dare express thoughts in public for fear of being treated like a pariah. 
back in olden days, a tweaked PCM-1630 (dual slope conversion) would compare with the analog tape source. 
same for the earlier 1610 (SAR) with GIC filters replacing the passive units. 
has been said to me that delta-sigma sounds warm and slow. 
I know of several collectors that seek out early CD's:  Zeppelin 1 that Barry Diamatne mastered has a cult following.

 
Mbira said:
That project looks interesting.  I didn't see much info there on the website about the DIY aspect-Do you know if it's possible to just buy the board?  It looks like it's fully assembled (and sold out everywhere I can find).

There are several versions. You can get the DIY version that includes all power regulators and a clock. There's also an OEM version lacking clock and power regulation. The main thread is here:

https://www.diyaudio.com/forums/vendor-s-bazaar/259488-reference-dac-module-discrete-2r-sign-magnitude-24-bit-384-khz.html

And a thread for a more recent version:

https://www.diyaudio.com/forums/digital-line-level/291645-building-soekris-dam1121.html

 
gridcurrent said:
you are not alone.
there have been a few folks in mastering that share your opinion.,
they would not dare express thoughts in public for fear of being treated like a pariah. 
back in olden days, a tweaked PCM-1630 (dual slope conversion) would compare with the analog tape source. 
same for the earlier 1610 (SAR) with GIC filters replacing the passive units. 
has been said to me that delta-sigma sounds warm and slow. 
I know of several collectors that seek out early CD's:  Zeppelin 1 that Barry Diamatne mastered has a cult following.

There are a lot of Hi-Fi people who agree, too. And not the audiophools.

I am building a 32 channel converter based on the DAM1021 to feed my console. The hard part is to get rid of S-D on the AD side of things. There are very few options, which are ancient and/or extremely expensive.

Current SAR chips are good enough to beat any SD-based audio AD converter in the specs department, but not a single product on the market.

I hope this project will eventually result in an AD converter usable for audio:

https://www.diyaudio.com/forums/equipment-and-tools/335005-osva-source-versatile-analyzer.html

Soekris has also announced his intention to build a SAR AD converter, but so far nothing has transpired. Mastering people would take it out of his hands I think.
 
living sounds said:
Good recordings have the effortless dynamics and natural top end quality of vinyl.
There is nothing natural in the HF behaviour of vinyl. The lathe is fighting with the HF limiters and the cutter resonances, the cartridge as well, and the distortion due to the stylus' radius is there, combined with the retraction of vinyl. How much degradation comes from playing a record just once?
 
I recall thinking that some DACs appear to handle playback of real dense mixes better than others.  A bit more focus , or not giving up and just throwing it all together,  I guess in regards to this ..... What spec would be considered mostly responsible for this?
 
abbey road d enfer said:
There is nothing natural in the HF behaviour of vinyl. The lathe is fighting with the HF limiters and the cutter resonances, the cartridge as well, and the distortion due to the stylus' radius is there, combined with the retraction of vinyl. How much degradation comes from playing a record just once?

Or tape. Or synths directly from the console. Most modern records sound dynamically dead and fuzzy to my ears. I think this also has to do with ITB processing, especially bad SRC, but I've tried a lot of converters that impart this quality to the signal, too.

 
scott2000 said:
Are there any references you could give?

For example, listen to
Lionel Richie's "All Night Long" from the 1983 CD Album Can't Slow Down
A-ha's "Take On Me" from the 1985 CD Album Hunting High And Low
Nirvana's "Nevermind" Album MFSL version (recorded via a Pacific Microsonics converter). Not the best EQ choices, but better in terms of pure sonics.
Ry Cooder's  "Paris Texas" Album from 1985

Auditioning via a proper PCM converter the hihats and cymbals sound nice and defined despite quite high amplitudes - and not like modulated noise. No blanket on the drum attacks, and basses appear to move the air. No fuzzyness on the acoustics guitar strings, you can feel the guitar's body. It's hard to put in words, but much more fun to listen to IMO.

Not all "old digital" sounds good, of course.
 
abbey road d enfer said:
I would guess you've heard about the Loudness War...

Sure, but I'm talking about the feel of the music, even modern remasters that aren't that much louder than the original ones sound broken in the same way to my ears.

 
living sounds said:
Sure, but I'm talking about the feel of the music, even modern remasters that aren't that much louder than the original ones sound broken in the same way to my ears.
If you compare CD with vinyl, it's an almost impossible task, because CD has an absolute upper limit, when for vinyl it's a moving target since it changes with frequency, so you never know where you are in the dynamic range.
Remastering old recordings for CD is a human task, so error prone. I would suspect the human factor before the equipment or technology.
 
living sounds said:
For example, listen to
Lionel Richie's "All Night Long" from the 1983 CD Album Can't Slow Down
A-ha's "Take On Me" from the 1985 CD Album Hunting High And Low
Nirvana's "Nevermind" Album MFSL version (recorded via a Pacific Microsonics converter). Not the best EQ choices, but better in terms of pure sonics.

Ok...So we are talking about cds.....

I could handle having a Nirvana cd.... I'm more a fan of vinyl sound when it's good but, it's tough weeding through all the sellers... I actually have that Richie lp for some reason...lol

I've started collecting some cd's and it's fascinating to hear how different they are from one pressing to the next...and how nice some  sound since revisiting the format..

Thanks!

 

Latest posts

Back
Top