Why do so few DACs support 88.2 and/or 176.4 kHz?

[blessingx]

Because of some Refence Recordings, HDTracks, etc. files, I started off my recent DAC hunt by eliminating any model that didn't support 88.2 or 176.4 kHz. Unfortunately, that eliminates a ton of DACs (MiniMax, Neko, etc.). I know several use the same chip, but not all. Some, like Lavry, expect even less standard kHz, so curious why these two 44.1 x X points aren't more widely supported?

[Eric5676]

Best guess: Demand vs. R&D costs.

[grawk]

Another guess: they support them, but don't advertise them.

[Nebby]

I know some DAC chips require software mode to enable the higher sample rates, so that would increase the complexity and costs of the DAC.

[The Monkey]

Ric, FWIW, I know the PWD supports it, and I believe the Bryston BDA-1 does, too.

[Torpedo]

The Lavry DA11 supports 88.2 through USB. Not sure about 176.4, don't have any file at such sampling freq.

[blessingx]

Thanks guys. I think I've found my solution (kinda downgrading back to previously owned gear for some reason - maybe the JH influence), but been curious about this for months and kept wanting to ask. BTW, Lavry claims up to 200 mHz in wide PLL mode on both DA10/11, though I'm not sure of the specifics over USB specifically.

[The Monkey]

Well, Ric, don't make us guess--what are you going with, the Lavry?

[blessingx]

Well, I'm backtracking to safe grounds lately with the full-sized. Becoming cowardly in my old age. Been plenty happy with the three Parasounds, but want to handle 24-bit stuff on at least one and go much smaller if possible. The DA10 was the only DAC I've owned superior to the Paras generally, especially when messing around with the modes. Dan Lavry would probably cringe, but throwing the switch to wide mode and upsampling the tracks (which of course counters my original question - but I'd like both options) pushed the DA10 above the D/AC 1600 (although my DA10 was sold a long time ago, kind Al let me his to A/B last year). As for as I can tell, the DA11 loses this manual override, so coupled with used going half price and not needing USB, put a bid on the DA10 this morning. Still tempted by PWD and WyredStallyns4Sound, but cash going elsewhere. We'll see.

[The Monkey]

Nice. I've always wanted to try the Lavry.

[skullguise]

Hmmmm, interesting observations on the Lavry, thanks for them. I owned the DA-10 for a while, guess I didn't fiddle with it enough!

I'm enjoying the Parasound 1600HD now, though (along with a couple Sabre32 implementations, DIY Buffalo32 and EE Minimax DAC); so I may not be tempted to bite the Lavry apple again too soon.

Oh, and BTW, nice Bill & Ted reference!

[blessingx]

Well, you'll never hear me badmouth the Parasounds. Love bang for buck and I picked up the D/AC 1600, 1500 and a 1100 for about the same as I'm getting the used DA10. Lavry has a similar signature and only adds a some air and a little better placement if memory serves. Well that and the nasty bitrate piece of mind.

Posted it elsewhere, but a camera kit and a $20 UCA202 USB/toslink adapter coupled with one of the Parasounds and Pico Slim makes an incredible, oddly proportioned, ultra-unportable iPad/JH13 rig. Dinny if you have an adapter you should give it a shot. Interesting to compare against the much more usable Pico DAC.

[NekoAudio]

I support 88.2kHz but not 176.4kHz. It's a limitation of the WM8804 S/PDIF receiver chip that it cannot figure out on its own if something is supposed to be 176.4kHz or 192kHz. I don't exactly know why, but maybe it has something to do with its reclocking logic. It's outgoing clock isn't a function of the incoming clock. Anyway, it requires you to tell it what mode to use via software control, if you want to support 176.4kHz.

I didn't consider it that much of a problem because it's rare for anyone except sound engineers to be working with 176.4kHz material. And even then, it's probably perfectly fine for your software to be downsampling to 88.2kHz if you're feeding into a DAC.

[Duggeh]

^ Direct from the Cats mouth.

I've never actually seen a file or disc with [176.4] on it. Just how rare is such a beast? I get the numbers being neat multiples of CD 44.1, which presumably has some very sound technical merit with regard to something. That being the case, why did 96 and 192 become the ones to go for? Just because they're respectively higher and have a 9 in them?

[Duggeh]

Indeed.

[nikongod]

I've never actually seen a file or disc with [176.4] on it. Just how rare is such a beast? I get the numbers being neat multiples of CD 44.1, which presumably has some very sound technical merit with regard to something.

Going from sample rate to sample rate on perfect multiples is said to be "better" in some way. I think Dan Lavry wrote a paper about this.

That being the case, why did 96 and 192 become the ones to go for? Just because they're respectively higher and have a 9 in them?

video has historically used (I think 12)/32

32*3=96

96*2=192

When they started making the recording junk to record at faster rates because it was blatantly obvious that 44.1 wasnt enough it was probably not much harder to build ADC's and DAC's that did their thing at 96Khz VS 88.2Khz (they are only 10% apart) and once you have something that can run a little too fast its very easy to slow it down. The recording engineer apparently never got the memo about even multiples which left "more speed" vs "sooooooooooooo slllllloooooooooooooow"

[Grahame]

Redbook was 44.1Khz , so they could record digital audio on "Video Recorders".

48 Khz was for DAT, Digital Audio Tape - Wikipedia, the free encyclopedia - so you couldn't easily make a digital copy of a CD using DAT

2-35] Why 44.1KHz? Why not 48KHz?

My how technology has changed things.

The using integer multiples for sample rates is also known as "not making up stuff that wasn't part of the original signal" , or "better" if you prefer

You'll need some sort of crystal and clock divider to get a clock signal for your DAC, hence the need for multiple crystals as the clock rates are not integer multiples.

[Dusty Chalk]

I actually have several 88.2 recordings because it downsamples to 44.1 nicely. Would have gone 176.4 if my Masterlink went that high.

48kHz wasn't just the standard for DAT and video, it was pretty standard in professional recording (in the early days of digital, when we couldn't get much higher than CD). I don't know if it was DAT -> rest of professional industry, or rest of professional industry -> DAT.

[deepak]

^ Direct from the Cats mouth.

I've never actually seen a file or disc with [176.4] on it. Just how rare is such a beast? I get the numbers being neat multiples of CD 44.1, which presumably has some very sound technical merit with regard to something. That being the case, why did 96 and 192 become the ones to go for? Just because they're respectively higher and have a 9 in them?

Reference Recordings uses 24/176, Keith Johnson prefers that sample rate. Give me good old 16/44 any day, tens of thousands of CDs to choose from and CDs are now cheaper than ever

[blessingx]

I didn't consider it that much of a problem because it's rare for anyone except sound engineers to be working with 176.4kHz material. And even then, it's probably perfectly fine for your software to be downsampling to 88.2kHz if you're feeding into a DAC.

Hey Wes, thanks for the response. I'm curious though why it's fine for 176.4 to be downsampled to 88.2 and (assuming based on DAC market) not equally so for 192 down to 96? Is it just the number of tracks out there (I may be an oddity as I have a bunch of 176.4s and zero 192s), is it something inherently different about the two kHz groups and stepdowns or is this only about feeding the DAC, not later upsampling?

[Asr]

Hey Wes, thanks for the response. I'm curious though why it's fine for 176.4 to be downsampled to 88.2 and (assuming based on DAC market) not equally so for 192 down to 96? Is it just the number of tracks out there (I may be an oddity as I have a bunch of 176.4s and zero 192s), is it something inherently different about the two kHz groups and stepdowns or is this only about feeding the DAC, not later upsampling?

This has something to do with sync vs async sample rate conversion and odd order vs even order. I was reading about this recently when I discovered Micromega's CD players. They have a "white paper" PDF about the subject here, start on page 3: http://pro.micromega-hifi.com/pdf/5166_wp_cd30_uk.pdf

[NekoAudio]

Hey Wes, thanks for the response. I'm curious though why it's fine for 176.4 to be downsampled to 88.2 and (assuming based on DAC market) not equally so for 192 down to 96? Is it just the number of tracks out there (I may be an oddity as I have a bunch of 176.4s and zero 192s), is it something inherently different about the two kHz groups and stepdowns or is this only about feeding the DAC, not later upsampling?

I don't think I fully understand your last sentence, but I'm viewing it from the perspective of the analog reconstruction. From a signal theory standpoint, 88.2kHz is fully sufficient for reproducing the audible analog waveform. Since 88.2kHz is half of 176.4kHz, it is perfectly fine to drop every other audio sample and you will end up with the exact same analog result as far as your audible spectrum is concerned because you haven't lost any of the information you care about.

I would consider doing the same to go from 192kHz to 96kHz equally okay.

[blessingx]

Wes, thanks again for your input here. Definitely appreciated. I guess my confusion could be boiled down to the maybe overly simplistic question - if it's perfectly fine to downsample 176.4 to 88.2 and 192 to 96, and you end up with the exact same analog output with the lower mHz, why do people build 192 (or higher) DACs? What's the upside? Less likely downsample errors? Better marketing?

[Voltron]

Nice. I've always wanted to try the Lavry.

Kind Al's Loaner Service is still in effect if you want to re-ignite your DAC desires. Just let me know and it will be boxed and sent.

[NekoAudio]

Wes, thanks again for your input here. Definitely appreciated. I guess my confusion could be boiled down to the maybe overly simplistic question - if it's perfectly fine to downsample 176.4 to 88.2 and 192 to 96, and you end up with the exact same analog output with the lower mHz, why do people build 192 (or higher) DACs? What's the upside? Less likely downsample errors? Better marketing?

Hm. It's one thing for me to say what signal/information theory says, and another thing for me to get into (non-Neko) product design/marketing decisions. While wearing my Neko Audio hat at least. ;P

In terms of measurements, feeding different sample rates into a DAC chip can result in a different output. I have measurements posted of 44.1kHz versus 96kHz. So technically what I said earlier isn't 100% true in reality. But it is true in terms of math. For all the DAC chip datasheets I've looked at, the measurements are better at lower sample rates.

FWIW, I only listen to 16-bit / 44.1kHz content. I have a few DVD-A discs but they're inconvenient to play.