You are not logged in. Please login or register.
RME User Forum → FireWire & USB series → Which USB interface?
OK, it's something I know very little about. What I do know is that I've heard differences between clocks and when I upgraded the ULN-2 with the 3D card it sounded much better. It can make quite a big difference.
Yes, better clock may give better sound. Hard to say if DF and BF internal clocks are on par or one of them has better clocking.
AFAIK internal clocks have been on par since the beginning of RME (not the beginning of digital, of RME which is later in time). Improvements have been in the AD and DA sides (filtering, oversampling, surrounding electronics).
Only in the first few generations were there some clock improvements possible. But I guess some were also a myth. Yes some interfaces (early digidesign) sounded different on very good external master clock. But knowing if this was really better is difficult. If the PLL (if any) was bad and it was, external clock no matter how good would give worse jitter then internal and thus distortion, which as we know always sounds better. Our brain prefers slightly distorted as long as it is not to obvious. In a blind test with multiple amps in the hi end range and one very cheap, there was a slight preference (if any) for the very cheap (really cheap like 200$ vs above 2000$). This among pro audio engineers.
Also when using the digiface headphones, the clock is only used for playback, so does not affect your recordings or rendering quality. In theory you should use your digiface in slave mode, with indeed the disadvantage of setting samplerate, If your external device would be slave you are dependent on it to have something similar to steadyclock fs.
Yea, for recording I'd just use the Neve 1073 dpd clock. It's just nice to know really whether it's a different clock to the BF or not. From what I've gathered in forums, including audiosciencereview, I think FS means a different more accurate quartz and circuit. But can't be sure.
I've asked about the ULN-2 having something similar to Steadyclock in the gearspace forum, so might get an answer on that.
I will test external Vs internal with the Digiface and uln-2. In regards to the Digiface quartz used, I'm still interested to know if it's the same one used in the Babyface Pro FS. MC said in this thread:
https://www.forum.rme-audio.de/viewtopic.php?id=26677
"the crystal used in the DAC is referred to as the lowest jitter crystal currently available, outperforming the famous Crystek ones".
Is this also in the Digiface, or is it only in the ones with FS? Quick yes or no and I'll be satisfied 
Is steadyclock used when using internal clock? Does it use 2 clicks, one at 2.4ghz and one at maybe 25 MHz. The 25 is for the pll, which is used even in internal clock mode? But Vinark said "No steadyclock is only active when in external sync." So I'm confused.
I have another question about the madiface driver. I was reading that it used isochronous transfer, whereas the usb one doesn't. If I'm mostly going to be only using around 2 - 4 channel each way, would using the Digiface USB mean it sends all channels even if they're not used, and is there more chance of data errors?
I don't think there is an option for 'iso streaming' for the Digiface USB is there?
Maybe, I am wrong, check it on RME spec pages...
But Digiface USB IMHO uses madiface driver. I do not know what Babyface uses, but rest of my RME devices FF UC, UCX, 802 use IMHO USB driver (for sure different one than DF USB), if USB is used.
Yea it used the madiface one. I hadn't realised that the Digiface uses isochronous transfer. I assume that the Babyface is asynchronous, with having fewer channels? If this is the case, I think I might just cancel the Digiface order and get the Babyface. Especially as I'm using it with a laptop, and I don't want to stream lots of unused channels. Is there no async mode on the Digiface?
The best way for you, IMHO, is what I wrote few days ago.
Get the both, compare them, decide what suits better..... And return the one that does not suit you.... Or the both... Or keep the both...
To save the hassle of getting both, installing both drivers, testing a project that uses lots of CPU, seeing which one starts choking first, sending one back, etc, I thought I'd ask here first.
OK, how much time have you already spent searching, asking and so? And have you found if DF or BF is better for you?
I have a 9-year-old system where the CPU from 2016 is now ranked 754th in terms of performance on Passmark.com:
https://www.cpubenchmark.net/cpu.php?cp … mp;id=2869
Nothing really slow, but also nothing exceptionally fast compared to modern CPUs. But a good performant system with good drivers and thus low "DPC latencies" where drivers do not block the CPU cores to be able to work efficiently on any system load, especially DAW workloads with real-time demands.
I run occasionally an "artificial" load test with Cubase on it, documented here in my blog: https://www.tonstudio-forum.de/blog/Ent … cks-de-en/
Not that I would actually have such big projects / high-performance demands in real-life.
The initial purpose was to find out what high DAW load the system supports without having audio loss.
The next step was to compare my USB3 based UFX+ (now UFX III) 188 (!) channel interface with a PCIe based RayDAT card which has 72 channels I/O. So, in theory the RayDAT might have two advantages here, fewer channel count and PCIe express card (later more to the latter).
The result is, as you can see in that blog article, that both RME products (USB and PCIe based) are both able to handle a high DAW load at single and double speed without issues (audio drops). Playback has been performed using the lowest ASIO buffer sizes possible. For both drivers (MADIface and HDSPe driver) 32 samples at 44.1 and 64 at 96 kHz.
The current purpose of this load test is to validate the performance of my system after changes (OS, driver, HW, settings).
This test was also useful a while ago to demonstrate, that USB based products do not behave worse compared to PCIe-based products under high CPU load. https://forum.rme-audio.de/viewtopic.ph … 92#p210092
Reason: one forum member claimed that USB based products would not work reliably with system loads over 80% CPU, only PCIe-based products could do that. https://forum.rme-audio.de/viewtopic.ph … 63#p210063
My Cubase project alone doesn't deliver such a high CPU load, so I had to run CPU-Z's "system stress test" in parallel, which utilized all CPU cores to a sustained load of 98-99%. The result was that there were no issues playing back audio with the USB based UFX III even under this high load.
@TomVY: therefore, I wonder what problems you expect with the mentioned RME interfaces of the same good quality.
They have all the same good design: FPGA based, no 3rd party communication chips in use, USB communication all done by RME / the FPGA. With the same good drivers and even a much lower channel count.
One thing while we are at it, because you mentioned it a few postings ago, that all audio channels are instantly being transferred even if not in use. Yes, but don't worry so much. This is simply a fact, but doesn't cause any issues on today's systems. My system shows 0% CPU when it has nothing to do.
Two other statements / examples:
- RME supports up to 140 channels over USB2. So why do you worry about much lower channel counts?
- from another area but also I/O related: stress tests over 10Gbit LAN show only 0-1% CPU load on my system.
It seems to me that you are "counting beans" for no real reason and making it only harder for yourself to find the proper choice.
If there is any issue, then it is most likely on your system (bad drivers, high DPCs latencies or other HW/BIOS issues).
That's the reason there are still companies around offering turn key systems for DAW or Video work.
The RAW CPU (usual benchmarks) performance alone doesn't count.
The difference is that audio has real-time demands, where it is important to process data in time.
For you, the advice not to chase shadows. I would simply pick the product which serves your demands best (type/amount of) channel and feature wise.
Since you've already bought it, I wouldn't panic now. You could also gain experience with the product first. If in doubt, RME products can still be sold at good prices on the used market because they are of good quality and there is such long support for drivers and TotalMix FX.
Thanks for the information. I am thinking more and more that Babyface would be better though, as sending 34 channels for playback even though I only need 2 seems like a waste of energy. I'm used to using uln-2 in asynchronous mode, where you can choose how many channels are active over usb. With a really old Mac I use it with I get better performance having less channels active. But the Windows laptop I have is just a couple of years old, so it should be fine. I'll try it and decide then if Babyface might be better really for a bit more money
By the way the Mac I have been using with 10.6 (actually it's hackintosh), the mboard, CPU and 4gb! of ram is from 2008. My laptop has 64 GB ram. And it will have 16tb storage when I've upgraded it
Thanks for the information. I am thinking more and more that Babyface would be better though, as sending 34 channels for playback even though I only need 2 seems like a waste of energy. I'm used to using uln-2 in asynchronous mode, where you can choose how many channels are active over usb. With a really old Mac I use it with I get better performance having less channels active. But the Windows laptop I have is just a couple of years old, so it should be fine. I'll try it and decide then if Babyface might be better really for a bit more money
I don't know if it's only me, but maybe a little criticism of the manoeuvres without wanting to offend you in any way.
I have the impression you're trying to square the circle here.
It is difficult to advise someone who has constantly changing and sometimes contradictory criteria and requirements (technical data, the highest quality, mobility, lowest energy consumption, bus powered, the lowest possible number of channels).
You were also telling money is no issue, excellent, so we can really talk about good solutions.
You would rather not waste money, fair enough, nobody of us wants this. But … Please come up with real and for recording or your business "meaningful" requirements which can be fulfilled in combination.
In other words: I would try to concentrate on the really relevant things; otherwise you'll just make your life/or your choice difficult unnecessarily.
Here again the recommendation is not to buy too small. You're not doing yourself any favours by buying too small, at some point you'll have a new need and something like that will come faster than you think, especially the more flexible the devices are (TotalMix FX routing/features).
And if you just need something for playback, then just get a DAC without the huge possibilities of TotalMix FX:
RME ADI-2 DAC FS or ADI-2 Pro FS R BE or if you want it with a turntable connection and other features: ADI-2/4 Pro SE.
The features I need are: usb, windows support, adm, mobile form factor, 1 headphone, 1 digital in/out (AES/spdif or adat), good clocking and quality. Preferably also usb powered, asynchronous and 2 balanced outputs . I don't need any other features, and if I do later I will sell and upgrade. The device that best fits this seems like the Babyface. I bought the Digiface, as it had all this except lower quality clock and headphone out than bf, isochronous transfer, no XLR out. The lack of analog output just means I use the uln-2 for that for now.
Anyway, I don't want to keep pestering with questions, it was just that I keep finding out information that is not easy to find (isochronous or asynchronous, internal clock accuracy). Df not having FX, adi-2 not having adm, this stuff is in the spec page or manual.
Just curious.... Why have you bought DF when BF seems to meet your criteria better than DF?
Maybe better internal clock, for sure better clocking on external clock, better phone output for sure, XLR out.....
Just curious.... Why have you bought DF when BF seems to meet your criteria better than DF?
Maybe better internal clock, for sure better clocking on external clock, better phone output for sure, XLR out.....
The other thing which comes to my mind is, that he initially told about his two quality preamps, where I can only assume that he might want to be able to operate them in parallel.
But even if he would like to connect only one at a time, he would be in trouble because one of the preamps (Neve) had only analogue I/O, if I googled the right product. This: https://www.thomann.de/de/ams_neve_1073 … stereo.htm
Analogue inputs are not available on the Digiface USB, and I think the plan was not adding one more device to the setup (A/D converter) via one of the 4 ADAT/SPDIF ports.
I am still of the opinion that an UCX II would offer him more options, more powerful headphone outputs, more features and, in total, more "sustainability after purchase" at a still small (mobile) form factor.
Just curious.... Why have you bought DF when BF seems to meet your criteria better than DF?
Maybe better internal clock, for sure better clocking on external clock, better phone output for sure, XLR out.....
Because the DF has my needed things: usb, windows support, adm, mobile form factor, 1 headphone, 1 digital in/out (AES/spdif or adat), good clocking and quality.
It doesn't have 2 of the preferred things (asynchronous, balanced analog out). But as I have the Metric halo uln-2 interface, I can connect it to that. The uln-2 only has Mac drivers, which is the main reason I need an rme interface. MH are working on Windows drivers though, but no ETA yet.
Ramses, the uln-2 has 2 mic pres, but never used them. I use the Neve 1073 DPD which is 2 pres with AES digital out. And I have an AES to optical converter. I also chose df over bf as it is cheaper, and I don't want to lose the uln-2 at the moment, as it has analog phones and monitor controls, and I'm not 100% sure on digital volume yet. I would also need some fixed attenuators with the bf. The final reason is that if I replace my current analog monitors with kh120 II with spdif in, the uln-2 will be redundant as long as the phones on it works well enough for me.
Recommendation for next time. A small uncomplicated ASCII drawing as an overview is very helpful to better recognize how the mentioned devices are to be connected. To be able to deduce from this what type and number of I/O ports are required as a minimum. With a lot of text and information scattered in different places in an increasingly large thread, it is difficult to keep the overview, especially if you do not know the setup and the intended way to work with it.
At the very end of this thread some clarity was added, many thanks for that. I would have liked it more at the beginning.
Yea sorry, I have been a bit all over the place. You obviously made that excel chart so you wouldn't have to deal with people like me I appreciate your time and if I have any more queries, rest assured I will post them here
No problem and many thanks that we can talk about that in such a relaxed way :-)
RME User Forum → FireWire & USB series → Which USB interface?
Powered by PunBB, supported by Informer Technologies, Inc.