[shlomo]

I'm wondering if any of you with the HagUSB and ASIO4ALL have been able to get 24 bits from USB to SPIDF.  Is this even possible?  I would like to use HagUSB with a 24-bit DAC.  I think 24-bit is desirable to allow for lossless digital volume control in foobar2000, and possibly future SACD/DVD-A playback.  Thanks.

[hagtech]

The standard audio output via USB is stuck at 16 bits.  For now.  You could write a custom software driver to make your own USB interface using a different USB chip, but that leads to a lot of compatibility headaches and operating system issues.  It can be done.  I believe there are some pro audio (recording) equipment that do this.  Also Gordon at Wavelength is trying it.  He will likely be the first in the hi-end world to have such capability.

Keep in mind that you need 24 bit source material.  Do you have any?  Have you ripped some DVD-As to hard drive and now want to play them back via USB?  If you're just using CDs as source, then making a pipe of 24 bits with only 16 of them active is silly. 

jh

[shlomo]

Keep in mind that you need 24 bit source material.  Do you have any?  Have you ripped some DVD-As to hard drive and now want to play them back via USB?  If you're just using CDs as source, then making a pipe of 24 bits with only 16 of them active is silly.

hey, thanks for the reply.  my main goal here is having the option of lossless digital signal attenuation.  i heard that 6dB of attenuation needs 1 bit.  i think this means that a 16-bit cd ripped to FLAC and played back with 12dB attenuation would actually be 18 bits.  so a 16-bit pipe would be throwing out information, whereas a 24-bit pipe would not.

my conception of this issue may be totally misguided, as i am no engineer.

[hagtech]

6dB of attenuation needs 1 bit

This is true! 

However, you only lose information if you use software sliders for volume control.  This is highly NOT recommended.   Use a normal linestage or some other analog volume control.  Then you lose zero bits of information.  Even if you do calculations at 24 bits you will still end up messing a bit with the sonics.  Pro audio tools are into 32 and 64 bit floating point DSP for equalization and processing.  And then they always spit out FS (full scale) when written to a CD. 

I don't believe you can actually achieve "lossless" digital attenuation, except in 6dB steps (bit shift right).  For me, I define lossless as the ability to reverse steps.  That is, if you run a song through a volume control three times, it is possible to run it again three times with the opposite gain and obtain the exact original bits.  Or for example, start with a WAV file, convert it to AIFF, then to FLAC, and the back to WAV.  As long as you can get the exact same bits you started with, then it is lossless.  Floating point calculations may not always be perfect enough (that's why they're using 64 bits!).

When a studio does a mastering job, they start with the raw tracks.  These are then mixed, panned, faded, and equalized to hell in order to get the "right" mix.  This can take hours or days.  Many, many passes through the DSP.  However, each pass always starts with the original bits, so the mix is only processed once.

jh

[shlomo]

ok, good insights!  to summarize, digital attenuation should sound better with 24 bits than with 16 bits, but it still wouldn't be lossless.  an analog volume control method is always preferable.  thanks! 

information of general interest...  as far as 24-bit usb-spdif transports go, i haven't seen any under $400 that are "audiophile grade", and these are mods of m-audio hardware and use m-audio drivers.  then there is the m-audio stuff itself which is a bit cheaper.  i'm guessing that the hagusb is better than the stock m-audio equipment when dealing with 16-bit source material.

random other question:  does the half kit come with the black box? 

[hagtech]

does the half kit come with the black box?

Nope.  Half-kit is blank board with the surface mount PCM2704 chip pre-installed.  Nothing else but a manual.  The box is cheap, less than $3 I think. 

jh

[Kane Williams]

After reading some great user reviews, I've just sent for one of these HagUSB converters and am looking forward to trying out my laptop as my main music source with my Xindak DAC-5 or Denon AVC-A1SE.

However, I was also hoping to be able to pipe quality 24 bit high sample rate audio out of my music software packages like Reason and Cubase etc, but see that it's limited to 16 bit. The higher bitrate is more important to me that higher sample rate so 48kHz would be fine. Is there any chance this thing can work with 24 bit word lengths?

If not, do you guys think getting something like an EMU 0202 and using it for DAC duties instead of the XindaK or may be this unit on Ebay in conjunction with the Xindak and running my music productions at high bit rate will produce better sound quality than using the very well respected HagUSB with the Xindak and keeping my productions at 16/48?

Also, does anyone have an opinion which will give better results, the DAC's in my Denon or the Xindac? I know the Denon uses multiple DAC's in Dual differential mode, but I'm no expert.

As I intend to connect the Xindac to my studio monitors via balanced XLR's and don't believe software volume control is a good idea, is there such thing as a good, clean, transparent balanced in/out box with simply a volume control?

I'm no DIY guy at all, but love to read about what you guys all do and I love to support small companies such as Hagerman.

Here is the Ebay item I mentioned, does it sound any good parts wise etc?

http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=220248986681

[hagtech]

Sorry, the HAGUSB only does 16-bit processing.

jh

[Kane Williams]

I have sent an email to Lessloss Audio asking about their USB to SPDIF converter. They say on their website that it will not allow 88.2Khz or above, but it does not say if it is restricted to 16 bit?

24 bit 48Khz would still be more useful to me than the HagUSB. If their device will work at 24 bit word length, then I may have an unused HagUSB for sale. I will of course try out the Hag with CD playback before I sell it in case it's better than the Lessloss.

[BradJudy]

Since you are using pro audio applications that can do ASIO output, you can use an EMU 0404 USB to do 24/96 digital out to your DAC.  It's only $200.  You can also play with using it as a DAC as well and see if you like it. 

Apparently it will not output digital for non-ASIO software though, so it might be a little odd if your computer is multipurpose (e.g.  you want to play games and have the sound output through the same path). 

[Theo]

It's no news that the trend for PC audio is higher bit rate and sampling, although 44/16 may be sounding good enough for most people.  Following trends is what many, if not most, people do.  I hope Jim comes up with a new version of HAGUSB that supports higher bit rate and sampling.

[Kane Williams]

Just had a reply from Lessloss and they say that their USB-SPDIF converter is also limited to 16/48, so pretty much the same unit as the HagUSB probably.

[Kane Williams]

I've been considering the EMU 0404 for my laptop "Pro audio" purposes. I already have a nice EMU 1212M card for my desktop and the quality is excellent. This will be my main PC for studio use anyway.

However, I'm hoping the Xindak DAC-5 I have on order will be better or at least equal for DA. The Xindak has 3 outputs, which will be of great use. I can have the balanced outs to my M Patch 2 passive attenuator and then onto my 2 sets of monitors, another set of Tube RCA outputs to my headphone amp (I've got some nice tubes on their way too) and the last set of RCA's to my Sub (has it's own 80Hz crossover and volume control).

I wonder how much more jitter will be added going SPDIF from the card to the DAC-5 than just using the card? I've been reading a lot about clocking and jitter etc. On the Lessloss site, they say that Wordclock is decent but not as good as Superclock. fair enough, but thay say that certain soundcard manufacturers claim that Wordclock achieves about 10ns of jitter, which is APPAULING! If this is true, I now see why EMU boast about for the 500ps of jitter that the 0404 is capable of.

I am trying to find out if Lessloss can add a Superclock output to my Xindak. I may then mod my Toshiba SD900E to allow Superclock input (the CD being slaved to the DAC) as shown on their site. This apparently should achieve about 10ps of jitter for CD playback! Nice!

BTW, my Toshiba has pretty low jitter, about 185ps, which is pretty good for a DVD player. This is from an old Home Cinema Choice review (they measured it). Do you think this would have been measured at the SPDIF output? If so, how much more jitter will be added (via cable etc) before the singnal reaches the DAC?

On another note, Lessloss do claim that USB-SPDIF-DAC from a PC does achieve better sonic results than practically ANY CD player connected to a DAC via SPDIF (ie with the DAC slaved to the CD player). This appears to be what has been witnessed by the many HagUSB users too. So, the only improvement I can see is to slave the CD player to the DAC and the only way I have seen this done is the Lessloss Audio way. Anyone other companies clocking the CD transport to the DAC?

[Kane Williams]

I'm now checking out the inexpensive Musiland Lilo III. It's a small USB soundcard that is capable of 24/96 USB top SPDIF conversion. Musiland have written a new USB protocol for low jitter. The only thing is it has an optical output rather than coaxial. 

The DAC/ADC, headphone/Line out/Mic In and DSP functions are a bonus for me as a musician.

[hagtech]

You have to be careful when comparing jitter specifications.  It all depends on who measures it and how.  There are several methods, all reasonable, but they report widely different results.  Then there is RMS or PP.  Also the frequency span of measurement.  Most importantly is not the jitter number itself, but the spectral content.  A small amount of jitter with coherent spurs is way more audible then a large amount of random jitter.  That's the issue with SPDIF, the embbeded clock information gets mixed with the data causing coherent noise spectra (hence the JTEST).  With perfect transmission it is not an issue.  With bad transmission (small bandpass, impedance mismatches, receiver nonlinearity) the problems show up.  Make the transmission good, and the problems are minimized (HAGUSB is a very good broadband transmitter, far better than most). 

jh

[slwiser]

I compared my HagUSB with my M-Audio USB Audiophile using CD source material; i.e., 44kHz at 16 bit depth and can't here any differences using my Lavry DA10 DAC.  I use my M-Audio when I am using high reference material though; i.e., upto 96kHz and 24 bit depth.  The M-Audio uses the Lavry's SPDIF input and the HagUSB uses it's XLR into the Lavry.  This makes a real nice to use with two transports.