...Researched the heck out of it & saw many recommendations for the RME stuff at the high-end. Finally bit the bullet about a year ago and went for a used RME babyface off e-bay. Pricey even used but 100% solid & reliable. Zero issues. Couldn't be happier with the money spent. Hope it helps.
Yes that is why the RME's are pricey - solid, reliable and long term support for them, but couldn't be happier with mine.
]]>Works great for me using a Mac Mini usb2 port. Latency is non-detectable at 128 samples/96khz/1.3ms. I am using these with my HD600 Sennheiser Headphones. The constant current drive on the Apogee really lets the large Senn headphones shine...a highly recommended pairing.
Compared to my previous setup: echo indigo io express card + Little Dot MK V headphone amp, sound and latency is very similar except bass is bit more controlled and top end
sounds very slightly softer on the Apogee. The MK V gave me a little *too much* bass, for example, on the D4 which now sounds more natural. The Bluthner A/B sounds amazing and I am now just realizing that the old C3(?) was perhaps based on the Bluthner Grand (Before the Bluthner was available), since the overall timbre and range are so similar.
Only downsides to the Apogee are (1) Effects sounds like aliquot strings, key-off, tines, etc are much more prominent due to the increased fidelity of the Apogee and so the volume on these needs to be turned down. The increased clarity also means that in some cases the Pianoteq hammer hardness settings need to be reduced and/or the microphone positions need to be set farther from the virtual piano(2) No MIDI in but I use an inexpensive M-audio UNO single cable adapter for that (3) Cost, though if you look hard you can find the Apogee Groove at a substantial discount like I did. (4) Gets slightly warm but really this is pretty minor, especially since I have it hanging down in space and well ventilated.
The Apogee saves me a lot of space compared to the Little Dot ( which has a rather hefty dual mono design) and it's bus powered as well, eliminating cord and cable clutter.
All in All, highly recommended!
]]>...
I used an applicated called LatencyMon to track down the culprit. It was the acpi.sys Battery Monitor drivers.
Once I disabled these in the device manager the problem went away completely. Of course, this means that I don’t get any battery information. So, I’m constantly having to turn them on/off depending on the task I’m performing (music = off, coffee shop work = on). It’s a pretty annoying bug on a device that carries a premium price.http://imgur.com/a/yJp5c
.....
Edit: I have my device set to 64 samples (1.5ms). Works perfectly.
Sweet! Thanks for sharing!
]]>I am using Pianoteq on my i5 dGPU Surface book with a Focusrite Scarlett 6i6. I had intermittant lag that made the device unusable for real-time piano playing within Pianoteq.
You can see by the following screenshot that the CPU usage was only at 26%, so that shouldn’t be an issue.. but those red spikes are when the crackle and pop occur. Those coincided with the apci.sys spiking.
I used an applicated called LatencyMon to track down the culprit. It was the acpi.sys Battery Monitor drivers.
Once I disabled these in the device manager the problem went away completely. Of course, this means that I don’t get any battery information. So, I’m constantly having to turn them on/off depending on the task I’m performing (music = off, coffee shop work = on). It’s a pretty annoying bug on a device that carries a premium price.
Note. I know that there were some bugs with the Surface Pro 4 regarding old battery drivers when the device first shipped. It appears as though my device has the latest drivers (2015-11-13) version 1.2.0.2.
Edit: I have my device set to 64 samples (1.5ms). Works perfectly.
]]>But got distracted by my alternative keyboard, possibly the find of the century: a 1885 Steinway "F" with 20-year-old rebuild for only $2250. Never mind the Casio, the next plan for a digital keyboard is a PNOScan 2 for installation in the Steinway (for those rare times I don't want the sweet Steinway sound, or when I want to run Home Concert Xtreme, etc.):
]]>Went to bring my keyboard and laptop to my local Sam Ash to try different interfaces there. The salesman there also told me to buy a Mac. And a MIDI to Thunderbolt interface. A new keyboard would be cheaper...
Anyway, I was wrong on the USB vs MIDI latency difference - whether I use my WK-3000 with a MIDI port, or a newer non-Casio keyboard with a USB interface, the result is the same - a small but perceptible (and thus annoying) delay. It's easily heard by have both sound from Pianoteq as well as the native keyboard speakers playing. It also was present with two interfaces. So it has to be in the soundcard somewhere.
I saw a youtube video on windows 7 and 8 delays that talked about shutting the record feature "off" on the soundcard, but turning it on from the Realtec utility - I haven';t made this work (only not work so far)
]]>Definitely worth checking out the new Focusrite Clarett 2Pre thunderbolt interface with less than 1ms latency capabilities and extremely high quality
Hope this helps,
Kindest Regards,
Chris
Indeed ! If the Surface Pro 4 had a Thunderbolt port, this would be a good choice ! ...
]]>Hope this helps,
Kindest Regards,
Chris
]]>from what you wrote in the other thread (http://www.forum-pianoteq.com/viewtopic.php?id=4120), there is most likely some kind of problem with your USB midi interface when combined with Windows 10 (given you said this particular problem did not exist on your older win 7 laptop), which you also sort of confirm in your statement that the latency problem does not exist when you click a key in the Pianoteq window directly versus hitting a key on your keyboard.
This is not at all audio-related, then. (The midi signal is NOT being processed by the audio chipset in your computer, it goes straight through USB in your setup.)
Though an external audio interface including midi-in should obviously fix this problem, you might as well try to get a properly working USB-MIDI interface (cable). This will leave you free to find the most suitable audio solution (DAC etc.) later without limiting you to those that include a midi interface. (Since the midi input path is separate from the audio output path, having these in the same device brings no tangible benefits processing-wise, though you may of course prefer to have one less gadget and cable to deal with...)
Others have stated previously that they experienced no problems when switching their Pianoteq setup to win 10, so I doubt there is some general problem with Pianoteq's access to midi interfaces on win 10. But perhaps someone using a plain USB-MIDI cable (i.e. not a USB connector with a vendor specific driver) can confirm?
If this is feasible for you, you could try booting a Linux LiveCD and see if the demo version of Pianoteq exhibits the same behavior then (i.e. significant latency when using your midi keyboard, but not when clicking on a key in the Pianoteq UI). For such a brief test, you do not need to bother with an optimised Linux setup, as a few extra ms will not affect the overall outcome.
Dan
EDIT: link to other thread wasn't clickable
]]>I think I see the difference between what you're running and what I'm running: I have an older keyboard which has only the old fashion 5 – DIN plug MIDI IN and OUT ports. My keyboard does not have a USB output. Therefore, my soundcard has to do a MIDI to USB translation, and I think that's where the slowdown must be occurring. Yesterday I went to a friends house and tried a later model Casio keyboard which had a USB output. I did not experience any latency.
I do not know whether the audio interfaces that are sold that have MIDI inputs and USB outputs do any conversion any faster than what may be happening in the Surface Pro itself.
]]>Audio device type: ASIO
Device: ASIO4ALL
Sample rate: 44100
Audio buffer size: 192 samples (4.4 ms)
Perf:
Internal sample rate: 44100
Maximum polyphony: 96
To connect my keyboard to the Surface Pro I simply use a decent USB cable.
On my main computer I have the settings a bit higher, but the latency is identical on both -- imperceptible. ASIO4ALL is likely all you need.
]]>