[Budley007]

Well, There were several people that asked about RMAA test for the Headroom MicroDAC. The only problem is, is that this is my first attempt at RMAA testing. I have no idea if I did it correctly or not. Is it possible to examine the following setup and results and determine my accuracy? Is the source a detrimental factor? Can external DACs be tested this way?

I realize I may be setting myself up to look like a moron by posting this, but with the level of expertise I read from time to time on this forum, constructive criticism is inevitable.

Setup: USB->MicroDAC->1212M(1/4" analog L/R)
16-bit, 44 kHz USB Input

Setup: 1212M Optical SPDIF out->MicroDAC->1212M(1/4" analog L/R)
16-bit, 44 kHz Optical S/PDIF Input
24-bit, 48 kHz Optical S/PDIF Input
24-bit, 96 kHz Optical S/PDIF Input
32-bit, 96 kHz Optical S/PDIF Input

EDIT(11-25-05) New results posted with -10 output signal, +4 input signal, (USB +4 input signal)

For those that are curious as to interconnects:

USB cable - 2m Belkin Pure
Optical cable - 1m MiniPlug to Toslink Premium Optical Cable from Sys. Concepts, Inc.
Audio cable - 5' (1/4" mono)x2 Pro Co Lifelines Instrument -> Planet Waves 1/8" Stereo/mono 1/4" splitter.

 

[Glassman]

the USB connection only allows for 16bit at 32-48kHz samplerates, so the measured performance will indeed be limited by the 16bit precision, which is at most 98dB.. you should measure it using S/PDIF input instead.. also make sure the inputs are set to -10 in PatchMix session preferences..

 

[Budley007]

Thanks Glassman!

edit: Okay, if I understand you correctly, the USB 16-bit, 44kHz is the only test I posted that is accurate for USB?

I'll retest the S/PDIF with -10 in the session settings and repost the results.

 

[thomaspf]

Thanks, I greatly appreciate this!

I am waiting for your new results. I assume you apply no volume control whatsoever on the output?

The harmonics to the 1Khz test signal look pretty high especially for the 24/96 case.

What is particularly interesting is the huge difference in quality between the identical DAC chip in the 1212m itself and the MicroDAC. The analog stage on the 1212M must be dramatically better.

Cheers

Thomas

 

[Glassman]

now everybody claiming USB is clearly superior to S/PDIF, look at those graphs! especially those claiming that some silly modified USB soundcards will beat the crap out of even the best transports in existence

sorry, just couldn't resist..

 

[thomaspf]

Interesting to compare to the Firestone audio spitfire.

http://www.firestone-audio.com/rmaa/SPITFIRE.htm

Cheers

Thomas

 

[SnoopyRocks]

The test results don't compare favorably with those Cirrus specifies for the 4398. Headroom doesn't make it easy to see tell if it's the test setup here or their implementation that's the probelm since they take the same attitude as cellular handset manufacturers in regards to specifications - physical dimensions only.

Yet they prominently display a photo of the PCB?

Why is the linearity (THD) 10.7 dB worse going from 24/48k to 24/96k? Note that the 24/48 test is limited by the 3rd harmonic for SFDR whereas it's the 5th at 24/96. Something is wrong here, especially if the output is really 3dB below full scale on the 4398. If the results are represenative, then this is stong motivation not to run the MicroDAC at 24/96.

 

[AdamWill]

Quote:

Originally Posted by thomaspf Thanks, I greatly appreciate this! I am waiting for your new results. I assume you apply no volume control whatsoever on the output? The harmonics to the 1Khz test signal look pretty high especially for the 24/96 case. What is particularly interesting is the huge difference in quality between the identical DAC chip in the 1212m itself and the MicroDAC. The analog stage on the 1212M must be dramatically better. Cheers Thomas

Erm. Am I misreading the test? I thought it was comparing the USB and optical input methods to the Micro DAC. The Micro DAC is doing the D-A in both cases. ?

 

[maarek99]

Where are the measurements?

I couldn't test an outboard dac with my emu 0404 properly because it goes into a loop (sound passes through the dac, back into the 0404 and goes again through the dac) and introduces quite a lot of unpleasent things into the measurements.

RMAA is not the end all, be all of measurements anyway. The 0404 gives excellent results all around but I don't consider it high end at all.

 

[Budley007]

Quote:

Originally Posted by Glassman .. also make sure the inputs are set to -10 in PatchMix session preferences..

Ok Glassman, I reset the I/O Card output to -10 and completed another set of optical test.

I couldn't use the signal if I switched the I/O Card inputs to -10 because of massive distortion, so I left it at +4. (I assume these are the parameters as you meant them to be used. If anyone can get a clean signal with both I/O Card in and out set at -10, I'd like to know how they did it.) So for testing purposes -10 Output, +4 Input was used. I also retested the USB at 1644 if for whatever reason, it was tested under the same conditions as the optical SPDIF and at the same time.

Couple of questions/comments you folks had:

Yes, SnoopyRocks, the 24 bit, 48 kHz appears to produce better results than the 24 bit, 96 kHz. I honestly have no clue why it would be this way, but I can tell you that if I left the PatchMix set for 48 kHz and set RMAA to generate a 96 kHz test signal, then the results were better than what you see posted. Assuming this would render the test inaccurate, I generated a new PatchMix Session to follow the RMAA test signal frequency at 44.1, 48 and 96 kHz respectively. Why the results were less favorable? Dunno, but I still believe that the results posted are more accurate. It just makes sense. Maybe I'm wrong.

AdamWill, the test was done in an attempt to see how well the MicroDAC handled the digital signal conversion with both optical SPDIF input and USB input and at various sampling rates. The fact that the 1212M can still present a very clean recordable signal from an external source may not be the focus of these test, but is definitely a plus to those of us that use the card in this capacity.

 

[thomaspf]

@AdamWill

This test only looks at the MicroDAC but the 1212M/1820M are well tested and also happen to have a CS4398 as DAC.

http://www.digit-life.com/articles2/...820/index.html

Over on the amp forum the QA manager from Headphone has responded to my initial request for measurements and has announced that they will publish AP measurements of all their gear.

That should help clarify whether their gear is as limited as it looks from these initial measurements or whether there is a problem in the setup.

@maarek99
Quote:

RMAA is not the end all, be all of measurements anyway

If you have any specific ideas for better measurements that characterize sound quality please go ahead and add them to the thread over in the amps forum. The Headphone guy explicitely asked for input.

Cheers

Thomas

 

[maarek99]

Quote:

Originally Posted by thomaspf If you have any specific ideas fo...a better measuring system than RMAA....right?

 

[thomaspf]

I have a better test system at work and made a couple of suggestions on what I would like to see. Just stating that there are better ones does not seem very helpful. RMAA is as good as the sound card you use for testing and the 1212M is pretty good.

In fact I was most grateful Budley007 took the effort to perform these test and the publish the numbers to begin with. I understand on this forum its easier to sell boxes without any specs and quality bars but before I quit with my money I want to know that the equipment is functional even at a basic level.

Cheers

Thomas

 

[Tim D]

I think it did pretty decent (actually far better than I would have thought given the mini-portable design). The usb setup shows ground loop/noise problem that isn't present with the optical. Not sure about the IMD swept freq. Also it does well considering it really is not fair to compare it with a strict loopback of most combined DAC/ADC units. For one there is absolutely no grounding problem that could occur in those simple loopback instance compared to having two devices in play. Is the DAC running off batteries? Also some freq's are showing weird IMD reponse vs others so I'm wondering if there is some mismatch in the test setup.

My USB audio sounds fine, you just need to make sure you get rid of the grounding problems which will likely come into play when you start connecting it with other devices, etc.

*edit* nm I just looked it up and the spitfire isn't a combined DAC/ADC that I can see. But they seem to neglect to mention the test setup. Still a few things...the Spitfire is not portable and has a 24V DC power supply, and the RMAA tests were produced by manufacturer at liberty to choose whatever production unit measures best (and not in a situation to make simple mistakes in their testing either) vs a customer measuring a random production unit. The difference aren't so great that you'd ever *easily* be able to hear the extra noise and harmonics in the -100 to -120 nether regions. Most people don't even know they have usb ground noise issues in their setup. But yes the spitfire is quite impressive if they all come off the lot that way (someone should ask if they use RMAA to test and QA each and every unit and which soundcard). Most marketing mumbo jumbo stating that usb audio removes the audio from internal 'grunge' inside a computer is absolute crap unless the customer is knowledgable and precautions are taken to make sure the crap isn't just carried via usb ground. In this instance it does seem rather 'slight', but is made obvious when you compare its performance with optical(which doesn't carry computer ground).

Finally too many people kid themselves with these graphs and spec's when it comes to the 'higher-end'. 99% of your source material is 16-bit 44khz! And the Micro-DAC does absolutely brilliantly there...now if only more portables actually supplied an optical out! There is really no point to go ga ga over 24-bit performance until you can convince Sony or other record industry executives to throw away more money towards making that 24-bit bump in recordings which mainstream customers are still just not ready for in both ears and equipment. You can find 24-bit audio in DVD's mostly for movie soundtracks, but hey most consumers just dumb it down with dynamic compression anyhow so even when they *are* given huge dynamic range they purposely limit it for practical purposes! Which is why the benchmark DAC-1 for all its near 20-bit performance glory, really doesn't matter how well it does in graph and specs (which I'm sure it does very well), when all you should really care about is how well it plays back your source material. Course people probably shouldn't buy gas guzzling high performance sport cars for A-B bumper to bumper traffic either, but what do I know? Anyways if it matters, that MicroDAC measure's better than the Zhaou? DAC I tried from China.

 

[Budley007]

Quote:

Originally Posted by Tim D Is the DAC running off batteries?

No Tim, the test results posted were all with the supplied A/C block. I ran some initial test using batteries and the comparison was barely negligible and in some cases, the A/C supply produced better results, albiet very miniscule.