Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards VST: which audio board?

Hello,

I'm writing this thread because i should make a choice to upgrade my PC audio board for piano sounds (pianoteq).

By now i am playing out from a Terratec DMX 6fire 24/96, a decent audio card of several years ago.

But in the meanwhile i see that a large number of audio cards seems to have improved sound quality.

I' considering E-MU 1616m PCI and MOTU Ultralite Hybrid.
The first one is a PCI board and should have better latency (it has also same converters of digidesign equipments), second one has USB/Firewire connection and seems quite good too.

In your experience are there a sound card better at piano sound?

Thank you for your feedback

Ant

The kind of interface has little to do with latency. Even the old/slow USB 1.0 cable runs at over a million bits per second. Even at those "slow" speeds the time of flight down a cable for a 48 bits per stereo sample is nearly zero.

Latency is determined by the size of an internal buffer. The buffer holds between 64 and 1024 or even more samples. These are clocked out a 44,000 per second. the buffer is needed to avoid clicks and crackles. Others may argue that on Windows PCs it's the audio device driver that controls latency No, it the quality of the driver that determines what size buffer you can use.

Think a bout the future. Firewire and USB audio interfaces can be moved to new computers (which likely will lack the same kind of or even an internal bus) Also the USB/Firewire boxes are, well "boxes" and have real knobs and LEDs and full size jacks and mst of all a forward facing headphone jack with a volume control knob.

For simple playback Lexicon's "Alpha" works and is only $90. It's a good name brand but low in features

High in features and also a good brand and on sale now for $180 is what I just bought, A Presonus "FireBox". It has MIDI, 4 inputs and 6 out. optical and connects via Firewire-400.

Firewire is the best way to connect an interface but is really not needed for driving a pair of stereo speakers. It is required for 16 channel recording but in-between is a grey area. I switched to FW400 when I found the Presonus box on sale for such a good price.

All that said, once you get past a certain quality level with an audio interface you gain almost nothing by further upgrades because very soon the speakers and the room itself become the weak links in the chain. It is very hard to set up speakers for a piano but this matters more than the brand of audio interface

Thank you for your answer.

I have ever believed that PCI cards were better at ASIO latency during midi work, as PCI is "more direct" on the motherboard versus a USB/IEEE interface.

I mean latency between key pressure on a master keyboard and sound coming out.

Yes. That is how you measure it, from key press to sound output. The latency is the sum of several causes. There is the time it takes for the MIDI data to reach the computer, some process time and then the sound sample sits in an output buffer and finally it is clocked out of the buffer and into the interface (PCI, Firewire or USB)

As it turns out all of these causes are so small compared to the delay in the output buffer that they can be ignored.

All these causes are tiny. They are engineered to be small except the output buffer. It alone is specifically designed to introduce a delay in the sound output.

The interface delay is measuerd in microseconds and you never notice compred to the 10 to 100 milliseconds of buffer delay.

Or put it this way. There is no storage in the interface. OK maybe one sample of storage. But all of them, even USB are fast enough that the sample gets out the other end before it's time to put the next sample in. But the buffer can hold either 100 or 1000 samples. So it is more then two or three orders of magnitude slower than even the worst possible interface delay. Or to turn those same numbers around, the most you could hope to gain if you could find a zero-delay interface is a 1% improvement.

OK more numbers. The effect of the delay inside the firewire cable is about the same as if you moved your ear 1/8th of an inch further from the speaker. (speed of sound is about one foot every 1/1000th of a second)

Yes Latency is a big and importance issue with digital music but don't go counting microseconds when latency is measured in milliseconds.

ChrisA, would you still say the same with a DAW that records and manipulates multiple channels of audio, and MIDI on top of it? Is this not a bandwidth question, controlled by the clock speed and width of the bus (and the competition for controller cycles)?

I'm looking at the same PCIe card Anthares was considering. The future is here, for me--- I just got a new PC, USB 2 not firewire. By the time this computer wears out, so will the PCIe card be. (Or I will be...)

Maybe I shouldn't ask--- you seem to have a firm opinion. But the 1616 seemed to be in the mid-range, more than I need now, probably about what I'll grow into, less than what a studio would be buying. And something Cubase might be happy with.

http://connect.creativelabs.com/emu/default.aspx

It doesn't seem there's much interest here in the E-mu 1616m PCIe audio/MIDI interface card... but, who knows, maybe someone will search on it one of these days (the way I did).

It was somewhat frightening to open the box and find that the manual talked about installing the PCI card--- my Windows 7 64-bit system doesn't have PCI slots--- and I wondered if I had gotten the wrong product.

The online PDF manual version was somewhat more up-to-date and reassuring, though it did say (more fright) that versions other than XP and Vista (both including the 64-bit versions) are not supported. However, the software installed, and appears to function. Nothing caught fire.

The link above is buried in the "News" section of the emu.com website. It says that there are some drivers for more recent Mac OSX and beyond, and for Windows 7, available now in beta release. There are some 'don't blame us if" warnings of possible terrible things that can happen with the beta drivers, but it says that the final version's release is anticipated 'soon.' The page is dated June 2010. I think I will wait for the final version.

Now you know.

For playback audio quality once you get to a certain level the card makes little difference. Different story if you were recording using microphones, but for playback the amplifier and speakers matter so much more than the brand of audio interface.

As for latency. Almost ALL of the latency is caused by the audio driver software and the operating system.

If you have a working system and need to upgrade you should upgrade the parts that have the most effect first. Speaker placement and room acoustics, brand and model of speaks, the kind of amplifier. These will all have large noticeable effect. But once you are using an decent audio interface you have very little to gain by swapping it for another.

That said, there may be another good reason to swap out the card. The the best software device driver that you want to use may not suport your card and the card may not work with the OS you want to use. Inother words, choose the software first then buy whatever hardward works best with the software. It may be that you already have that card. Im my case I want an interface that works with Mac OS X and no driver at all. I would not buy a card at all if it required some special driver software.

Also if you intend to record audio then shopping for a new audio interface makes sense as the quality of the preamps built into the interfaces varies.

OK one other reason to replace the card is simply cabling and user interface. Internal sound cards do not place the controls where you need them and force you to make conection in back of a computer. A Firewire interface will have a full size 1/4" headphone jack that you can place right on top of the piano and it wil have seporate physical knobs the volume controls for headphone and monitor speakers. For recording there are even better reasons to use the external box (analog gain controls)

But for playback you may already have one that is as good as the best

Thanks, Chris. Part of the story is, yes, I record using mikes. No doubt the Realtek card that came with the PC is fine for playback (it supports 7.1 surround... with amplified speakers, of course). But, Cubase 5.5 didn't like the integral card, and its audio recording function was disabled. Important functionality. So, that's why looked elsewhere. Also, the Windows 7 machine lacks MIDI ports, and two of the three outboard machines I would like to use with it require them; no, USB-MIDI would not work with them. Besides that, the E-MU card supports S/PDIF (useful) and ADAT (which I don't use), and supports 5.1 surround if I ever get so far. I think I would have to move to a new house to have room for it.

BTW, the E-mu 1616 does include an E-MU ASIO driver for Windows.

I have been using the 1616M for almost 2 years now and I love it. I had it on both Vista x64 and 7 x64 with many less issues in 7 than Vista (sometimes audio would distort and require a reboot). The Patchmix software is extremely useful if a bit confusing. I have mine hooked to a Behringer ADA8000 for more channels.

The main reason I purchased the E-MU was for expandability and balanced outs (which are fantastic!).

Mine isn't the PCIe version but I would have no problems getting another one.

Is it a Realtek ALC888 or similar? If that is the case, audio recording is disabled by drivers that come with Windows. If you install Realtek's own drivers, you can record the Stereo Mix or any of the inputs. With the default Windows drivers you can only record the inputs but no stereo mix recording allowed.

BTW it is one of the best integrated sound chips out there. Of course it cannot substitute a good audio card, but it is quite correct for playback and even some basic recording. I tried Rightmark benchmark (http://audio.rightmark.org/download.shtml) and it performed surprisingly well taking into account what it is. But I would not use it for recording microphones.