Last night a DJ stole my kHz...

January 28, 2009 6:05 PM

Please help me figure out if there is a problem with my new digital studio monitors (speakers.) They are supposed to decode 192 KHz streams, but I can only get them to do 96 KHz... plus I don't know what any of this really means...

I just bought a pair of Behringer MS-40 digital monitors. They supposedly have "Ultra-high resolution 24-bit/192 kHz D/A converters for an incredible dynamic range" but I can't seem to get that work.

Here's my setup:

Speakers are connected via optical (TOSLINK? S/PDIF?) to my mother board. The board is a MSI K9N SLI Platinum with a Realtek ALC883 codec. I've upgraded to the latest drivers, etc.

The speakers work fine, but only when I've selected 2 channel, 24 bit, 96000 Hz (Studio Quality). I get only silence when I choose the 2 channel, 24 bit, 192000 Hz (Studio Quality) setting.

The thing is I'm not 100% up to speed on S/PDIF, DACs and the like, so I don't know if the behavior I'm witnessing is a problem or not.

I've gone through all of my settings and tried different things, I'm beginning to think that the problem is in the DAC in the speakers. But it seems like if there was a problem, they wouldn't work at all.

Anyway, I'm at my wits end, so any advice is welcome.

My OS is Vista 64. Usage is for straight stereo sound only, no DTS, Dolby, yadda, yadda...

Please no responses saying I'm not going to miss the extra 96 KHz, that only parakeets can hear the difference or whatever... I'm trying to figure out if the speakers are working as advertised, that's all...

Thanks in advance!

posted by wfrgms to technology (11 answers total)

As far as I know, toslink cannot carry 192kHz audio, it's a bandwidth limitation. In addition, DRM restrictions in Windows mean that 192Khz audio is only permitted to be output over an analogue connection.
posted by Mwongozi at 6:15 PM on January 28, 2009

The original TosLink standard was limited to 48kHz; the standard was later extended to accept up to 192Khz. Support on audio devices, especially sound cards & extra-especially on-board hardware, is spotty.

And, as mentioned, Vista's handy DRM limits high bitrate (>96KHz) audio sources to analogue outputs only :-(
posted by Pinback at 7:44 PM on January 28, 2009

S/PDIF doesn't really "max out" ... the standard, to my understanding, doesn't have a fixed maximum bit rate. It's more a function of what the devices on either end support.

Support for more than 96kHz/24b PCM is pretty slim, and so that's been a sort of de facto max for a while, but going beyond that is not prohibited. There are starting to be more devices around that speak 192kHz/24b, although they're a lot more rare than 96kHz.

I tend to wonder if Realtek isn't lying in their specs a bit, and if that chipset really is capable of 192kHz/24b/2ch PCM, or if it can only do 192kHz in some compressed format that maybe your speakers don't support, or maybe even at a lower bit depth. It wouldn't be the first time I've seen some sneaky stuff done in specs — are you sure (in the sense of reading anyone else's successful reports) that it's even possible? If not, then you're dealing with two questionable devices that you're trying to test, rather than just one.

Also, I'd give coax a shot instead of TOSLINK. Except in prevention of ground loops, the optical TOSLINK is inferior to S/PDIF in pretty much every way. (And since its max length is so short, it's not even that useful for preventing ground loops...typically both devices are plugged into the same ground reference anyway.) If your devices have S/PDIF coax instead of optical, you might try them. I've read that the copper transceivers max out at a higher line speed than the optical ones.

Threads talking about S/PDIF "max" bitrate issues:
S/PDIF max bitrate [AVS Forum] - this forum typically has very good information, in my experience
diyAudio Forums - this one mentions that some cards advertising 192kHz only do it on analog, not SPDIF
SPDIF sample rates - call this a 'dissenting voice', but I think it's wrong and the 96kHz/24b is not a hard limit, but obviously there is disagreement
posted by Kadin2048 at 7:45 PM on January 28, 2009

Ugh, I always notice something after I hit submit:

Except in prevention of ground loops, the optical TOSLINK is inferior to coax in pretty much every way.

Both the optical (TOSLINK) and electrical (coax) connections carry the same digital datastream, called S/PDIF. But despite optical connections typically being assumed to have more bandwidth than electrical ones, TOSLINK is based on very simple technology (it's like the very best of 1982!) and uses very crappy plastic optical "fiber" in most cables, limiting its maximum length to less than coax, which is where you'd expect it to be useful.

The only places I'd use the optical connector are: Where you have a bad ground-loop problem despite physical proximity (ex.: a laptop running on battery, and hence floating, connecting to speakers that are grounded), where you have a lot of electrical interference to the point where it's causing problems, or where there's no coax connector available (some Apple gear, Sony Minidisc players).

[end derail]
posted by Kadin2048 at 7:53 PM on January 28, 2009

(Ooops, sorry, that's 48KHz per channel = 96KHz overall, later updated to 192KHz overall. The bitrate was upped from ~3Mbps to 125Mbps)
posted by Pinback at 7:58 PM on January 28, 2009

And, on preview: TosLink does have a hard maximum sample rate, as I mentioned above. I don't think it requires fixed predefined sample rates though; afaik, in TosLink, 100KHz is as valid a sample rate as 96KHz, as the TosLink / S/PDIF specs say the clock should be derived from the bitstream, not just locked to it.

Whether any devices support such odd sample rates is a different question :-)
posted by Pinback at 8:06 PM on January 28, 2009

Well, I just fished a 75ohm coax cable out of my big bin of cables and tried it in place of the optical - and it works! I can now select the 192 kHz setting! (Not that I really hear a difference, but it's good to know that the speakers work as advertised.)

Thanks everyone! I read in several places that there were no difference between TOSLINK and digital coax, so I wouldn't have thought about that on my own!

Cheers!
posted by wfrgms at 11:19 PM on January 28, 2009

You couldn't hear the difference between 96 and 192 kHz anyway. And that's not snark, it's physics:

The highest pitch such a system can reproduce is 1/2 the sampling frequency (see http://en.wikipedia.org/wiki/Nyquist_frequency'>Nyquist frequency). So a 96 kHz audio stream could faithfully represent 48 kHz sound.

Humans can hear roughly 20 - 20,000 Hz, maybe a little higher for some. Anything higher would be inaudible, like a dog whistle. This means that you only need about 40 kHz sampling to achieve the entire range of human hearing.

It might be useful to have samples at 96 or 192 when you're processing sound, but for speakers it's pointless.
posted by qxntpqbbbqxl at 5:06 AM on January 29, 2009

It's probable your sound card is unable to output 192 on the optical side. It thinks it is, but the hardware just isn't doing it.

re: derail: If optical carries the same datastream, then it should be identical in performance. The problems you site aren't in the format, they are in the implementation. IE, use a better cable. However, another place where optical might be inferior is that it is doing two extra d-d conversions. It has to take the electrical signal and convert it to light, and then from light back to electrical. If those circuits are flaky, you'll get bit loss. Again, however, it's not the format, it's the implementation. It's probably just as likely you'd suffer similar problems in the conversion from the digital signal to the coaxial digital signal. Which is to say, rare. I have pretty cheap equipment, and use cheap cables (including using a standard RCA patch cable for coaxial), and have never had an issue either way... Both mediums are so much better than analog it's not even a contest.

second derail: even though we can't hear higher spectrum sound (and chances are, our equipment can't reproduce it and the original microphone couldn't capture it), if there WAS some on the signal, the effects of cutting the bandwidth might be apparent via harmonics and/or mathematical inaccuracies. You wouldn't be hearing the actual sound, rather, odd effects of its loss. Example- suppose the original sound wave is sort of square-like. It's easy to reproduce most of it, but right at the angle of the square, it takes a lot of bandwidth to reproduce it faithfully. Less bandwidth means the wave gets rounded off, and in that process, data is lost.

Here are visual examples: This is a 20000 square wave at generated at 192, zoomed way in. This is that same wav, converted to 96. This is the same thing, generated at 96. (The dots are the actual samples, the wave is what the d-a converter would try to turn the samples into. Look at how the conversion affected the wave- it had to approximate what it thought the right spots for the samples were.

There IS a difference, but we're talking about the razor's edge of sound reproduction, and this assumes the sound was ever there in the first place. And it's still better than analog, where the physical effects of the recording will introduce more distortion anyway...
posted by gjc at 6:50 AM on January 29, 2009

You couldn't hear the difference between 96 and 192 kHz anyway.

Yes, and if you had read my entire post you would have seen where I specifically stated:

"Please no responses saying I'm not going to miss the extra 96 KHz, that only parakeets can hear the difference or whatever... I'm trying to figure out if the speakers are working as advertised, that's all..."

There is one in every crowd, isn't there?
posted by wfrgms at 10:49 AM on January 29, 2009

Interesting thread.

Somewhat off topic, but YouTube has videos* that will play a pure tone from very low Hz to very high. I know, not a perfect clinical test, blah blah, but it's interesting. I found I could hear from about 25 Hz to about 15,000 Hz. I'm in my late 20s, so the drop off at the high end is likely just natural hearing loss, maybe a bit of loss from heavy headphones use.

* http://www.youtube.com/watch?v=4G60hM1W_mk
posted by wastelands at 9:35 PM on January 29, 2009