The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Audio interfaces and their AD/DA chips LISTED Audio Interfaces
Old 3 weeks ago
  #301
Lives for gear
 

But there are still only a handful manufacturers, I mean not that many.

32-bit sounds (pun intended?) logical for computations, historically data processing is based on 8, 16, 32, 64... etc. bits. Typically DSPs and algorithms are optimized to handle data of 2^n bits length. Processing 24-bit data doesn't make sense, I expect it to be handled internally as 32-bit with 8 padded bits. Also it helps avoiding rounding error issues.

Noise levels will ultimately be limited by physical laws and the only way to reduce them would be to lower temperatures. Now I could imagine cryogenically cooled electronics for audiophools.


Still wondering about the probably-not-so-outstandingly-breathtakingly-magical ADCs and DACs used high end professional (not audiophool) equipment like (in no specific order) Sonosax, Nagra (Audio Technology Switzerland), Sound Devices, Aaton...

My bet is that recent designs mostly rely on very good common off-the-shelf (COTS) audio chips, possibly carefully individually hand-picked based on extensive lab screening tests.

Older designs are possibly based on non-audio-specific semiconductors, typically parts used more for test and measurement equipment as well as some higher-end industrial (non-audio) instrumentation.
Of course some parts used in top notch instruments are custom designed, which contributes to high prices.
Old 3 weeks ago
  #302
Gear Maniac
 

The chips are the same in many brands

The Ak4413 and AK5388 are popular now. There are 32 bit versions available but have not seen them.
The differences are in layout, firmware, and drivers. This is how RME and Apogee do so well with the same components as their lower priced competition.
This is a good thread.
Old 3 weeks ago
  #303
Gear Head
 

I just read about Focusrite doing something a little different with thier Red range than what others would do with thier conversion lay out by using summing. Here's an interesting video on using the two AK4413 DACs summed together in parallel. The Rednet converters have just the standard layout like Lynx and everyone else. They also go into how they designed the ISA AIR modeled impedance circuit improving on from the Clarett range. YouTube
Old 3 weeks ago
  #304
Gear Maniac
 

AK5388

The AK5388 are ADC’s. That is why there are maybe four in your 8Pre. Eight channels for Mic input and eight for line inputs. Total of 16 channels of inputs.
There are only two channels of monitor output. So a single CS stereo dac should work.
The paralleled channels give a 3 dB drop in S/N ratio. Some of the dacs using ESS chips combine four dacs to drop noise by 6 dB.
Old 3 weeks ago
  #305
Gear Head
 

Quote:
Originally Posted by Panelhead View Post
The AK5388 are ADC’s. That is why there are maybe four in your 8Pre. Eight channels for Mic input and eight for line inputs. Total of 16 channels of inputs.
There are only two channels of monitor output. So a single CS stereo dac should work.
The paralleled channels give a 3 dB drop in S/N ratio. Some of the dacs using ESS chips combine four dacs to drop noise by 6 dB.


In the Clarett 8 Pre its actually two Ak5388 4ch ADCs for the 8 mic/line ins, 2 AK4413 4ch DAC for the line outs 3-10 and the CS4398 2 ch for the Monitor out 1 and 2. I just posted a photo of my 8 Pre mainboard on the other page. Line out 7-8 shares with Headphone phone 1 while, Line out 9-10 shares with Headphone 2. I'm not sure what codec is used for the ADAT and Spdif, but it's on the board some where.
Old 3 weeks ago
  #306
Lives for gear
 

Does someone know if the current best off-the-shelf audio chips or chipsets, if sorted carefully in the lab, can match audio quality of high-end recorders if the overall design is done very carefully?

As often mentioned, it's not just about ICs with amazing datasheet specs (measured under ideal test lab conditions with high-end equipment), overall performance is limited by the weakest points of the design. For example the best audio CODEC IC is useless if there are power supply noise or sampling clock jitter issues.

Sadly many manufacturers using per se great ICs don't allow them to perform optimally because they just wanted to save a few $$$$ manufactruing costs. Stating that DAC model XYZ is used in some device is not a very useful statement as it all depends on how the device performs, not about the datasheet specs of some single components.

And BTW datasheet data must be read carefully, especially for parts operated over a wide temperature range. Some specs are mentioned over the whole temperature and power supply voltage range while others are only "typical" or sometimes it's not even clear which test conditions apply.
Old 3 weeks ago
  #307
Quote:
Originally Posted by Schoeller View Post
Does someone know if the current best off-the-shelf audio chips or chipsets, if sorted carefully in the lab, can match audio quality of high-end recorders if the overall design is done very carefully?
What "high end recorders" did you have in mind? And what did you imagine they're built from? Everyone has access to the very same chips. It's all about what you do with them.
Old 3 weeks ago
  #308
Lives for gear
 

I meant Nagra, SD, Aaton, etc.

Was mostly wondering if current state-of-the-art (i.e. not necessarily referring to somewhat older designs like the Nagra VI) dedicated COTS audio chips or chipsets are used or if it is still meaningful to rely on non-audio-specific high-end ICs like those used for industrial or scientifc instrumentation purposes, which also supposes that the integration design is more demanding as relying on COTS specialized audio ICs.
I suppose the Nagra VI with its older (though not obsolete) design still outperforms many more recent designs, referring to audio quality.
Old 3 weeks ago
  #309
Gear Maniac
 

If digital is considered

The golden age of analog has passed. But it really seems digital gear continues to improve and get cheaper.
The boards are 4/6/8 layer. Grounding and shielding improve. The quality of the onboard clocks improves with time. The power supplies also can be made quieter, lower impedance, wider bandwidth.
The chip performance also gets better in each new generation. I doubt this has plateaued yet. The golden age of digital recording and playback is still ahead.

Last edited by Panelhead; 3 weeks ago at 04:17 AM.. Reason: Spelling
Old 3 weeks ago
  #310
Lives for gear
 

I fully agree.

Purely analog signal paths are also expensive, they require a high amount of high quality both active and passive components. Both development and production are more expensive and require very specific audio know-how.

That said, until we get digital mics we still require an analog frontend which means that analog audio R&D know-how is stil required. Further, going digital requires knowledge in the design of digital circuits as well as very good programming experience (which is typically a problem, see how many devices have oddly designed firmware, very common are user interface issues which clearly show that software developers don't know how the device is used in real life).

Like you said, I expect that performance of audio ICs will still improve. Indeed I even suspect that mostly price constraints rather than technological limitations dictate the current audio IC specs.
Typically audio IC manufacturers like AKM and others (I say manufacturers even if most don't operate their own foundry) may focus more on volume markets and will propose ICs costing a couple of CHF/USD/EUR ea. rather than developing high-end ICs which would maybe cost 10 or 20 times more and remaining only interesting for high-end niche products like Nagra, Sound Devices, Sonosax, Zaxcom, Aaton,...

The other solution without audio-specific ICs is much more expensive, from R&D to production and also way more complex from an engineering POV (even if possibly higher performances could be achieved????).

That said, I expect that recent audio chips and chipsets have reached such a high performance level than in most cases they won't be the limiting factor of the whole design even for fairly demanding applications. Interestingly (AFAIK) they're all relatively inexpensive.
Old 1 week ago
  #311
Here for the gear
This is a great thread. I design high frequency interfaces for radios (ADC/DAC). These designs operate at GHz sampling rates. Many of the same rules apply, however. Quiet power supply rails, low jitter clock source (phase noise), good layout and gain staging can make or break a design. The number of units that are sold is tiny compared to so many other markets in audio ADCs. There's also a huge amount of consolidation taking place in the IC industry (just noticed TI bought Burr Brown...) as well. I think given that every stereo, phone, car radio has a bluetooth or usb interface there is a ton of audio DACs out there compared to ADCs.

You're never going to get rid of analog interface. At some point, a speaker needs to push air (or your jawbone as my VMODA M100's do, they are amazing too). The whole chain needs quality IMO.

I'd like to hear opinions on what the biggest cause of distortion and noise is in a signal chain. A 24 bit ADC has a theoretical dynamic range of 144 dB (6 dB/bit). Nobody gets close to that in their design. Is it the analog gain in front that's doing it?

Update: I pulled the data sheet for the AK5388A and they do have a noise floor very close to -140 dBFS. The harmonics are responsible for the higher noise floor so it's more of a SFDR. The eval board looks familiar to higher frequency designs too.

Last edited by shooten; 1 week ago at 10:33 PM..
Old 1 week ago
  #312
I have a DAC here that is a bit unique. It doesn't use an audio designed DAC but instead a multibit industrial DAC chip. It uses the Analog Devices AD5547 DAC. It has current output and a very linear dynamic range. It is 16 bits but an honest 16 bits. It's made by Schiit Audio. It's the Modi Multibit and sells for $249.

It uses a servo-ed Analog Devices AD8512 dual opamp like the Bricasti M7. Me, I swapped that for the ADA4898-2 and WOW. Huge clear sound with remarkable detail. I use it in the TV room for the home theater experience and it's pretty darn good for that. The low end makes you take bathroom breaks.
www.schiitt.com
Old 1 week ago
  #313
Lives for gear
 

Interesting. The AD5547 DAC belongs already to the more expensive ICs, at least compared to the average audio ICs used in most devices.

Wondering how something like an AD5791 would perform as audio DAC.

I'd also be curious to see how top notch industrial/scientific ADC, DAC and programmable (pre-)amplifiers would perform if used for audio (while normally used for data acquisition and test & measurement and not optimized for audio).

Overall the ADC or DAC ICs are only the beginning of the story. For very high performances the whole circuit design is very demanding as zillions of small design inattentions can negatively affect the analog signal quality.
Old 1 week ago
  #314
Quote:
Originally Posted by Jim Williams View Post
I have a DAC here that is a bit unique. It doesn't use an audio designed DAC but instead a multibit industrial DAC chip. It uses the Analog Devices AD5547 DAC. It has current output and a very linear dynamic range. It is 16 bits but an honest 16 bits. It's made by Schiit Audio. It's the Modi Multibit and sells for $249.
I don't really understand why someone would use this DAC for audio conversion, though it would make a rather nice digital volume control. (It's a multiplying DAC.) As a gain control, it has very high bandwidth, but not when you're changing the actual DAC value. For the later mode, it appears that you could update it at 2 MHz and get maybe 10-bit performance. But Schitt doesn't like oversampling so they have time for it to settle to higher resolution. Used in that way, there are no AC plots or specs at all. When used as a gain control, they quote THD at a rather unimpressive -104 dB, and only at 1 kHz.

Have you put that Schitt product on the bench, Jim? What are you seeing in the spectrum?

David L.Rick
Old 1 week ago
  #315
THD at below the 16 bit 96 db dynamic range isn't an issue with this design. It's intended for PCM audio playback like CD's and DVD's, the consumer market. Designed by industry legend Mike Moffet who designed the highly rated Theta Digital DAC's 30 years ago.

Shiitt also makes a high end converter with that DAC chip, the Yggdrasil for $2300. It's apparently quite good and unique. It won the 2017 DAC of the year in The Absolute Sound mag against some very top notch contenders. You can see the build quality on their web site as they post pics of the innards...
Old 1 day ago
  #316
Here for the gear
 

Presonus quantum converters chips

Does someone know DAC and ADC converters of the presonus QUANTUM?, I wrote to Presonus but they answered me a generic AKM...someone could kindly photograph the quantum motherboard please?
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
gramps / So much gear, so little time
13
Andy D / Gearslutz Secondhand Gear Classifieds
0
pschniddy / Gearslutz Secondhand Gear Classifieds
0
scoob / Gearslutz Secondhand Gear Classifieds
2

Forum Jump
Forum Jump