Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Thomas W. Bethel]

Fifthcircle wrote on Wed, 02 November 2005 14:24

FWIW, using my RME Digiface (and the 96/8 PAD when I owned it), I found the sound considerably better in ASIO.  It has been awhile since I checked, though.  It was a big enough difference that I wondered if the data integrity was being maintained.  Perhaps it was something else, but I did in fact hear a rather substantial difference. Plus- in most programs, the monitoring choices in ASIO are better anyways... --Ben

Same here that is why I was wondering what is going on and which one is messing with the sound the least? Thanks for the reply.

[bobkatz]

Here's one for Henrik:

You have a Masterlink and an external A/D converter. You record to the Masterlink's hard disc via AES/EBU and then again via S/PDIF and then again via Toslink from the external A/D (if the Masterlink has a toslink input). What's the difference?

BK

[Bob Olhsson]

Samplitude 7.1 and later checks out identical on ASIO and MME.

[Bill B]

And probably WDM as well.

[trevord]

At first glance i would say there should be no difference
BUT
RME's claim to fame is that they implement a significant portion of the ASIO interface in a FPGA (programmable hardware).
So
it may be safe to assume there are major differences in the "sound path" using the two audio devices.
"sound path" would be correctly called "the path the digital data takes going to the d-to-a converter"

again
at first glance there should be no reason why the sound would be different.
Unless the latency specified for the ASIO is so low as to cause problems when the same latency is used in the Windows driver.
Because the ASIO "driver" is in hardware, they have extremely low latencies.

Can you also check if your windows driver output is defaulting thru some "enhancement" (like surrond or something like that)

trevor

[ammitsboel]

bobkatz wrote on Wed, 02 November 2005 23:30

Here's one for Henrik: You have a Masterlink and an external A/D converter. You record to the Masterlink's hard disc via AES/EBU and then again via S/PDIF and then again via Toslink from the external A/D (if the Masterlink has a toslink input). What's the difference?

I've done enough of this to have a basic understanding of what happens sound wise. Maybe the ones that have a Masterlink(or a burner) available could to this test and post what they hear. Provided they don't get hypnotised by "the powerful voice" of "THERE IS NO DIFFERENCE"...

I once wrote to Lynx about why the new AES16 ASIO drivers sounded bad compared to the old AES16 ASIO drivers, they of course totally denied it could ever happen. I had upgraded the drivers in the morning and really looked forward to give them a go, but after 2 hours of session work I had to stop because for some reason everything sounded a lot worse than it used to. After checking all my gear and found nothing wrong I tried installing the old ASIO driver and surprisingly everything sounded good again!

I actually liked the sound of the RME digi cards better...

MME or WDM gives more of that infamous plastic sound doesn't it?? I can almost hear some engineer saying that it's actually useful for something...

Regards
Henrik

[carlsaff]

So if you export a file using the ASIO driver and the same file using the MME-WDM driver, do they sound different? Also, if you reverse the phase of one export, do you get a null file when you sum it with the other?

If the answer to both questions is "yes," I'm curious about the scientific rationale that explains how two bit-identical files can sound different.

Either there is a reason or there isn't.

[jdg]

how is someone 'exporting'

are you bouncing in the box?
or running in realtime out to another DAW/Recorder?

i would be interested in that test, not the bounce ITB test.

but i run OSX, so maybe im not all that interested after all

[trevord]

carlsaff wrote on Fri, 04 November 2005 20:52

So if you export a file using the ASIO driver and the same file using the MME-WDM driver, do they sound different? Also, if you reverse the phase of one export, do you get a null file when you sum it with the other? If the answer to both questions is "yes," I'm curious about the scientific rationale that explains how two bit-identical files can sound different. Either there is a reason or there isn't.

two things:
1) don't confuse identical files on a disk with identical sound output.

2) file export has nothing to do with the driver.

There are many examples where identical data sounds different at the output because of drivers/hardware.
A hardware example:
If you had two cards and one was a creative you definitely would get different digital data out from each card. Because the early audigy's operated in one frequency/bit width and "fudged" the other frequency/bit width combinations. (Meaning the card accepted it but truncated/mangled to suit)
For software:
many cards do this in the driver. When you select a bit-width and frequency in your application, it doesn't mean the card has all of a sudden got resources it didn't have before. It means the driver is mangling the data to suit the card and allowing you to continue work. For example an audio app might support 32 bit audio export. But you are not going to hear the identical audio from any two apps because of how they treat 32 bit data for a non 32-bit card.

I am not coming down either way on this particular situation. For example if someone said there is a difference between cubase and nuendo with otherwise identical setup, they would have to do a lot of explaining.
Even if they said they noticed a difference between WDM and ASIO for sonar.. it would be hard to swallow.

But
if ever there was a possibility where digital data is handled differently on its way to the converter its RME cards with ASIO vs WDM.

Companies make decisions and their use of technology reflect those opinions.
Sonar has 100% backed the windows camp and they probably handle MME-WDM best of all. But the ASIO/VST support is a hacked up kludge.
It probably the same for RME. RME is the considered by many to be the standard in ASIO implementation. They have (or used to) a strategic relationship with steinberg. It would not surprise me if the WDM implementation was less than perfect (or less than  the ASIO driver)

[trevord]

I think i should add some explanation here.
When ASIO is selected for RME cards, there is NO ASIO driver.
OR
no code is executed by the processor to get data from memory to the RME card.
The RME card fetches data from memory via the PCI bus.
Even the type of PCI access is different (32 vs 16 bit)

Again - theoretically it should make no difference if everything in the WDM driver is perfect (or matches what the RME hardware did)
but
you can see possibilities for issues

[Bill B]

ammitsboel wrote on Fri, 04 November 2005 13:21

MME or WDM gives more of that infamous plastic sound doesn't it?? Regards Henrik

What???? Henrik, try eating your lunch instead of smoking it.
Can you provide an audio clip of this effect?

Also, MME and WDM are separate driver models- I see them being lumped together often.

[Ronny]

ammitsboel wrote on Fri, 04 November 2005 13:21

I actually liked the sound of the RME digi cards better... MME or WDM gives more of that infamous plastic sound doesn't it?? I can almost hear some engineer saying that it's actually useful for something... Regards Henrik

I run the RME busmaster cards, they'll do MME and ASIO and I can't tell any audible differences. Certainly MME is not plastic sounding compared to ASIO. What physical element could be causing that? In fact the only thing you should notice is the meters in Digicheck refresh at a different rate on MME than ASIO, that's it, visual display differences yes, but no audible differences from my experience.  

[dcollins]

Ronny wrote on Fri, 04 November 2005 20:36

Certainly MME is not plastic sounding compared to ASIO. What physical element could be causing that?

The "audible" cortex, of course!

DC

[ammitsboel]

Ronny wrote on Fri, 04 November 2005 20:36

Certainly MME is not plastic sounding compared to ASIO. What physical element could be causing that?

It's an interesting question. Maybe it's just W****** way of doing it and the larger buffer??

Somebody should really make a DAW that is no PC nor MAC but instead a dedicated computer running it's own software with it's own dedicated digital ins/out. It will newer be PCM1630 but it could be almost as good and still be able to do easy editing.

Regards
Henrik

[jlarcombe]

ammitsboel wrote on Sat, 05 November 2005 11:51

Somebody should really make a DAW that is no PC nor MAC but instead a dedicated computer running it's own software with it's own dedicated digital ins/out. It will newer be PCM1630 but it could be almost as good and still be able to do easy editing.

What's so special about the PCM1630? I've never used one myself but as I understand it they're not exactly the first choice in most mastering houses these days...

What advantages do you feel a PCM1630 has over more modern equipment?