Print

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

[arjen_1]

Yesterday I got the RME 98/6 PST PCI card and just installed it. It gives me about 25% less CPU load compared to USB audio. Latency is about 5ms. Soundwise it reveals every detail. Basically a VSTi like Imposcar sounds really different compared to internal audio and my usb audio card. It's like the cutoff filter has opened 10% more. Difference is remarkable.
Makes you wonder about what you should use as a reference.   

Drivers are simple and work fine. Just follow the manual and don't forget to install 2 AISO driver files in every AISO Drivers directory. (settings app. & aiso driver file) The one thing I don't understand is the AISO setting of the card. You can only choose between a 16bit or 32 bit audio stream. Before I could choose 8, 16 or 24 bit audio. I guess 32 bit data stream is the same as choosing 24 bit audio?

@Chris....you bought a great card! 

[IIO]

the bit depth settgins is something RME specific and i dont know by heart what exactly it means.

generally audio is always sent as 32 bit float to hardware drivers, at least in OS9 (ASIO EASI Soundmanager ...) – with protools as the only exception.

if you dont like to read manuals, do the following: sum 50 tracks and see if it makes a CPU difference when you switch from 32 to 16. if not, the option simply does nothing.

[arjen_1]

if you dont like to read manuals, do the following: sum 50 tracks and see if it makes a CPU difference when you switch from 32 to 16. if not, the option simply does nothing.

Actually I read the manual over and over. The setting influences buffer/latency. The only thing I can't determine is if the card really plays back on 24 bit or 16 bit. Normally you would choose the playback resolution in the AISO driver. That's not the case here...So does it play 24 bit automatically if I e.g. set record to 24bit in Cubase? Or?

[IIO]

The setting influences buffer/latency. The only thing I can't determine is if the card really plays back on 24 bit or 16 bit.

thats unlikely that a card has a 16 bit AD/DA mode in addition to 24 bit – but i dont know RME enough. it might be that 16 bit is available for ADAT out, for example.

Normally you would choose the playback resolution in the AISO driver.

normally you wouldnt have different resolutions in one converter.

the converters from motu, m-audio, sony, eventide, mykerinos, usbaudio, terratec, avid, apogee, eventide, tc, and digidesign always only support 16 OR 24 bits format.

though i remember that at least one of the nuendo branded 19" units also had different modes (and they are hammerfall based.)

some of the high end sony cards and the weiss converter also have different modes - but of course at different ins and outs.

So does it play 24 bit automatically if I e.g. set record to 24bit in Cubase? Or?

nah, the setting of the software is independant from that. the setting of the host program only affects reading and writing files, everything releated to the card must be available in the ASIO controls.

[arjen_1]

Hi IllO,

Thanks for your responses. I found out that the RME Mac OS9 driver is 100% the same as the windows version, but it has a totally different function. Someone was very lazy.  Basically you can't use the  digi 96/8 manual wich is all over the internet. There's a specific Mac manual on the driver CD which solves all mysteries.

In AISO mode you can choose 2 different buffer sizes for 16 bit playback/recording which is mentioned as a 16 bit data stream in the AISO driver. The 2 - 32 bit buffer options enable 24bit playback/recording. As usual the host program must have the same resolution. Sample rate can only be chosen in the host program. Latency is less on higher samplerates. Don't know why that is.

While working with VSTi's I always used a 44.1 Khz resolution. With this new card I've got the option to work with less latency on 48 Khz or 96 Khz. What would be the most optimal samplerate to use with e.g. Atmosphere, ImpOscar, Vanguard etc?

Greetz,
Arjen

[IIO]

Don't know why that is.

because that part of the overall latency which comes from the card and its driver always comes in samples. shorter samples== shorter latency.

[IIO]

What would be the most optimal samplerate to use with e.g. Atmosphere, ImpOscar, Vanguard etc?

for instruments it does not really matter, but the RME PCI cards offer a very god quality at samplingrtates higher than 44.1, so if you are recording a lot (or of you use this computer to send VSTi´s to an analog mixer) you will have a benefit of using 88.2 or 96 kHz quality wise.

i would not recommend the same for a m-audio interface, cheaper converters sound better at 44.

that, of course, at the cost of CPU (of up to 2 times more CPU when the samplingrate is doubled...)

p.s. this wont apply to you and is not related to the converter questions, but higher samplingrates are also interesting when you sta on the digital layer, i.e. imposcar -> waves ren verb -> render to disk, and even more interesting for processes like pitch shifting or programs like melodyne or metasynth.

[supernova777]

for instruments it does not really matter, but the RME PCI cards offer a very god quality at samplingrtates higher than 44.1

GOD quality;)

"go forth my son, and VST"
amen. lol
 

[DieHard]

Some real world Numbers...

With a M-Audio 2496...

DMA Buffer Size 128.. Latency = 2.9ms (Can't playback 16 Tracks, 32 Bit, Audio has dropouts), VSIs are very playable, 8 tracks of 24 bit audio with VSTis are No problem

DMA Buffer Size 256.. Latency = 5.8ms (Playback 16 Tracks of 32 Bit and NO Audio dropouts), VSTis are somewhat playable

[arjen_1]

These are the choices/numbers of the RME card:

Choice->Buffer size->Resolution->44.1 kHz->48 kHz->88.2 kHz->96 kHz

A-> 46 ms/16 bit  2048 s 16 bit 46.4 ms 42.7 ms 23.2 ms 21.3 ms

B-> 23 ms/32 bit 1024 s 32 bit 23.2 ms 21.3 ms 11.6 ms 10.7 ms

C-> 11 ms/16 bit 512 s 16 bit 11.6 ms 10.7 ms 5.8 ms 5.3 ms

D-> 6 ms/32 bit 256 s 32 bit 5.8 ms 5.3 ms 2.9 ms 2.7 ms

Some results of my tests:
Option D (= 24bit audio playback) gives best results for playing VSTi's. Playback has dropouts while using VSTi's. Without VSTi's no dropouts.
Option B ( = 24bit audio playback) gives best overall result. No audio dropouts. VSTi's are playable etc.

[arjen_1]

for instruments it does not really matter, but the RME PCI cards offer a very god quality at samplingrtates higher than 44.1, so if you are recording a lot (or of you use this computer to send VSTi´s to an analog mixer) you will have a benefit of using 88.2 or 96 kHz quality wise.

p.s. this wont apply to you and is not related to the converter questions, but higher samplingrates are also interesting when you sta on the digital layer, i.e. imposcar -> waves ren verb -> render to disk, and even more interesting for processes like pitch shifting or programs like melodyne or metasynth.

That's interesting. I am going to give it a try. Curious if I can hear an increased quality while using a higher sample rate. I heard no difference with my ESI.  Otherwise the processing benefits are also something to take into account.

[supernova777]

just got the RME DIGI96/8 PST 
$20 .. what a score!!!!!!!
the guy who sold it to me, (the original owner) bought it at a discounted price because he worked for one of the biggest music stores in canada, Long & Mcquade..
he said it cost about 750$-800 retail at the time
best deal ever!

[IIO]

Curious if I can hear an increased quality while using a higher sample rate. I heard no difference with my ESI. 

yeah thats what i would expect. there are exceptions, but the main thing what counts is the stability/quality of the clock.

if you have a MOTU interface and you want to record guitar and voice, a samplingrate of 192 kHz will give you worse result than 44.1 kHz, simply because of the clock beeing "not exact enough" for such a high samplingrate.

[supernova777]

IIO u really gotta learn how to use quote blocks bud

[supernova777]

arjen
does the mac manual for the RME say anything about multi-configurations using more than one of these cards?
the card has connections at the top for sync in, sync out, and cd audio in

from the sound on sound article: http://www.soundonsound.com/sos/feb02/articles/rmedigi0202.asp

Output & Clock Options
It can get confusing dealing with digital clocks manually, but unless you do it correctly you'll end up with skips and jumps in your audio. To make things easier, RME provide an intelligent AutoSync mode that sticks with the low-jitter internal clock unless a valid clock is detected at the active input, when it switches over automatically. This should be a great help during 'on the fly' recordings, since you don't have to rely on the external clock being present before you start, and it also enables multiple DIGI cards to be synchronised from one input signal. However, there are some situations in which this mode can cause digital feedback, when you should switch the card to Master clock mode. A third Word Clock option is available if you've fitted the optional WCM module.

Overview
On the card itself are various additional connectors and jumpers. There's a digital input suitable for connection to a CD-ROM's digital output, or to synchronise another DIGI card, and a Sync output for which a suitable short cable is also supplied.

so it seems these are able to be stacked!!! for more I/O!!

[arjen_1]

Hi Chris,

See for yourself! 

Greetz,
Arjen

PS: Definitely a score! I paid more....

[supernova777]

i just saw last night the asio control panel references up to 4 cards