Suggestions for cubase box

[flaminghakama]

Hello,

I've realized that it's time to upgrade my recording hardware. I was hoping for some general suggestions.

I mostly record ensembles, recording 12-16 tracks simultaneously. My front end is miscellaneous outboard pres, daisy chained through an Apogee Rosetta 800 into a RME UFX, and then into the computer via USB 3 (everything clocks off an external Antelope clock).

My intention is to migrate from running Logic on a mac to start using Cubase as my DAW. Part of the reason for the upgrade is to switch DAWs.

But the main reason for the upgrade is that I'm trying to fix a problem with blips that seem to be related to the computer IO and/or drive. These blips tend to occur in all tracks at the same time, and occur more frequently the longer things have been running.

After first experiencing them, That was after adjusting things within Logic, like the buffer size, which did not seem to help the issue. What did help was when I switched to recording to an external drive rather than the system drive. However, when I started to record more simultaneous tracks and use a higher sample rate, the blips have come back.

Does this type of problem sound familiar to anyone, and what type of issue could it be? Putting it another way, what would the minimum requirements be for a system to handle recording 16 tracks at 48kHz? I am currently running on a mac mini with quad core and 16GB of ram. My external drive is a Glyph drive connected via FW800.

Second question: does anyone have any experience with whether Cubase performs better on a mac of PC? I would assume that it shouldn't matter much, and that I could get more bang for my buck with a PC than a mac. Are these good assumptions?


Thanks in advance for your suggestions.


David Elaine Alt
composer@davidelainealt.com

[Len911]

into the computer via USB 3, My external drive is a Glyph drive connected via FW800.

FireWire 800 was about half as fast as USB 3.0 in our write tests, turning in scores ranging from 55 MBps and 60 MBps. The read scores were faster, though at 72.3 MBps and 74.5 MBps, they were still considerably slower than USB 3.0.

http://www.macworld.com/article/2039427 ... ally-.html

https://www.glyphtech.com/product/studio
If you have the glyph that has a choice, fw800, esata, or usb 3.0, then try the usb 3.0 connection.

If that was the main reason for the upgrade, and that fixes the problem, maybe you don't need an upgrade. You don't give any info about the mac mini, so it's difficult to know if the current version of Cubase would even run on it? A daw that runs on multiple platforms seems an advantage, if only because you don't have to totally relearn a new program and possible project swaps. I have Cubase, an older version, and like it very much.

My front end is miscellaneous outboard pres, daisy chained through an Apogee Rosetta 800 into a RME UFX, and then into the computer via USB 3 (everything clocks off an external Antelope clock).

A lot seems to be going on there,lol! Daisy chained? You seem to have a lot of preamps, a/d conversion and clocks going on,lol!

[mojobone]

Theoretically your rig should be more than capable of reading/writing 16 tracks at 24-bit, 48kHz, and 16G is plenty of RAM, unless you're 32-bit OS and using lots of virtual instruments. Your interfaces are most likely USB 2.0 and won't get any faster by being plugged into a USB 3.0 socket. Still there shouldn't be a bandwidth problem on the front side via USB 2.0, unless they're maybe all plugged into the same USB bus? Without more info, I can only give hints.

Is your Mac Mini maxed out? Try more RAM; it's usually cheaper than a new system.

When this blipping occurs, check to see if any of your interconnect cables are warm.

Are you using a non-powered USB hub?

Tell us more about your Mini, like is it slower than 2Ghz and please describe your daisy-chaining in detail; it sounds to me as though that's where your bottleneck could be.

[Len911]

Your interfaces are most likely USB 2.0 and won't get any faster by being plugged into a USB 3.0 socket.

That's true, I didn't look up the specs and assumed it was usb 3.0, which it isn't, because the current ufx isn't.

USB 2.0 . The Fireface UFX has been uncompromisingly optimized for highest performance under Windows and Mac OS. It uses a special customized firmware for every operating system. Like other RME USB 2.0* interfaces, the UFX provides revolutionary ultra-low latencies even with multiple channels.

* The Fireface UFX is compatible to USB 3 chipsets.

[flaminghakama]

Thanks everyone for your help and ideas.

Ok, so here is additional detail about my setup, as requested.


1) I plug mics into an XLR snake.

2) The snake connects to preamp ins, which include Manley, True, TL-Audio, Seventh Circle Audio and the pres on the RME Fireface UFX.

3) The pre outs (except for the RME) connect via XLR-TRS snake to a TRS patch bay, a Neutrik NYS-SPP-L1.

4) Likewise, the converter A/D inputs from the Apogee Rosetta and RME connect to another Neutrik patch bay.

5) Then, I patch from the pre outs to the converter ins between the patch bays. (Except for the channels using the RME pres, which are hardwired to RME converters.)

6) The Apogee is connected to the RME via 4 optical cables, 2 for A/D and 2 for D/A, with each cable supporting 4 channels.

7) The RME is connected directly to the computer via USB, no hub--I was incorrect in my OP, in that this is a USB 2.0, not 3.0. Also connected to the RME, but not used during recording, is an L2 connected via AES/EUB. There is also an unused FW400 output on the RME.

8) The Apogee and RME converters (and the L2) are slaved to an a Antelope Isochrome OCX master word clock, via coaxial cables.

9) Here's the info on the mac mini. This is the "late 2012" model with 4 cores, and I am running OSX 10.11.6

Model Name: Mac mini
Model Identifier: Macmini6,2
Processor Name: Intel Core i7
Processor Speed: 2.6 GHz
Number of Processors: 1
Total Number of Cores: 4
L2 Cache (per Core): 256 KB
L3 Cache: 6 MB
Memory: 16 GB
Boot ROM Version: MM61.0106.B0A
SMC Version (system): 2.8f0

10) The drive I record to is a Glyph GT062E, connected to the mac via FW800 directly, not daisy chained through any other device. I have other drives daisy chanied off this drive, but those are not in use during recording. This drive also has USB 2.0 and eSata connections, which are not currently in use.

11) The other connections to the mac include keyboard and mouse, connecting directly via USB (no hub), and a backup drive connected via USB 3.0, a Seagate GoFlex Desk Adapter, which is not used during recording.

12) Within Logic, I choose the RME as the input and input device. The input channels show up as 1-8 the RME A/D converters, channels 9-12 the RME A/D converters for its onboard pres, channels 13-14 MIDI, and channels 15-22 the pass through from the Apogee A/D converters. Logic is set to 24 bit recording, 48kHz sample rate, I/O buffer size 64 samples. I am not using any software monitoring or input monitoring.

The only other programs running on the mac during recording is TotalMix (if I'm using input monitoring, it is through TotalMix, not Logic.)



In terms of the history of this setup, I started with a 2009 mac mini and the Apogee converters, recording 8 simultaneous channels on the system drive at 41.1kHz. That worked fine for a few years. I added the RME in 2011 when I wanted to get a more channels, and that worked fine until I started doing more than 10 channels, last year.

The first attempt to fix this last year was to lower the I/O buffer size in Logic, which had no noticeable effect. Then I tried to upgrade the mac with a 2016 model mini. That changed the sound of the clicks, but did not eliminate them. Once I researched it more, it seemed like recording to a different drive was the likely solution. Once I started recording on the external drive, the clicks went away. During this process I learned that my 2012 mac mini was more powerful than the 2016 mini, so I upgraded it to OSX 10.11.6 and have been using that rather than the 2016 model.

The clicks came back once I started recording 14 channels at 48kHz. The next session I went down to 44.1kHz, and I tried more frequent saving, recording smaller segments and even rebooting after each hour or two. I got fewer clicks, but they were still there.


I appreciate your help. Does this reveal any clues? I'm assuming I'll need to just try a different computer, since I think that each of the connections I'm attempting should be in spec. I used to like mac minis, now that they have jumped the shark and are not a intended as a pro setup. So, I figure I might as well look to the future and figure out what I really need as appropriate hardware.


Thanks.


David Elaine Alt

[andygabrys]

hi David,

Its probably safe to say that you recording front end, convertors, and clocking aren't the problem. Most of us would kill to have what you have.

It sounds like there are 3 possible problems:

1) CPU isn't keeping up. This is unlikely as its a 2.6 gHz i7 4 core (probably 8 virtual cores)
2) RAM isn't enough to keep cycling whats recording - but again unlikely as 16 GB is usually enough.
3) Hard drive scheme isn't keeping up.

My money is on #3) Hard drive scheme. Especially as you noted that as you increased the load (higher sample rate which makes file size much larger) and number of simultaneous tracks then the blips got worse.

There are several aspects to consider here.

1) the drive itself. A 7200 RPM drive that is dedicated to recording and streaming audio only (no samples, no double duty as a backup, no double duty as a system drive) should be sufficiently fast to record 16 tracks at 48 kHz. You can do the research and run the numbers but I would say you should be good especially if there is a cache of 32 MB or greater.

2) the free space on the drive - a mechanical hard drive requires a large amount of free space to swap and write efficiently. So if the drive is any fuller than 50% of capacity, you are going to run into issues. Again, you can do the research, but you will see read / write speed drops off almost exponentially as you run past 50% full. So if this is the case, backup old stuff on this drive and then delete stuff to make it 20% full.

3) the enclosure - You didn't specify how you are running the Glyph (and I don't own one) but it looks like a RAID enclosure that can be run as RAID 0, 1 or as one large volume. Here: http://w.glyphtech.com/products/gt062e/performance.php you can see that the numbers using eSATA look fairly good. But using FW 800 they are pretty poor, and if you are running in RAID 1 (mirror) then you aren't getting any extra performance, you are just getting redundancy.

That's it.

Check free space. If not enough, move stuff off and erase to get it back.
use a faster interface speed than FW 800.

Now....if you want to speed your $$$ - your Mac Mini can take 2 internal drives.

Macsales.com (OWC) has lots of mechanical drives and SSDs to add a second drive. The idea being that the interface speed to the internal drive is way faster than FW800, and you can use the existing Glyph enclosure as a mirror RAID 1 backup drive.

If it were me, I would look at an SSD of at least 1 TB size. A good low cost high performance SSD is the Samsung EVO 850 drive available on Amazon.com.

regarding Logic - if you are needing to monitor real time while recording (like musicians with cans listening and overdubbing, or recording to pre-recorded tracks, or a click track) then you should try lower LATENCY by reducing buffer sample size (max of 64 samples really - otherwise its too much). If you don't need this, and its an acoustic ensemble not needing to monitor, put the recording buffer at maximum (1024 samples). The CPU will be much more lightly loaded.

In Logic > audio preferences - you can also try adjusting the number of processing threads, and processing buffer range - it happens on my machine (12 core mac pro) that Large works better.



Hope that helps.

[Len911]

Have you tried connecting the RME with Fire Wire instead of the usb port? That's what I'd try first.

I'm thinking that usb 3.0 is relatively new. RME

The secret of success: RME is the only manufacturer not to use a third party USB or FireWire audio technology, but an own, self-developed Hammerfall Audio Core with an outstanding performance and uncompromised pro audio features.

, and because

In rare cases, USB 2.0 devices may have problems connecting to USB 3.0 ports on Macs.

https://help.elgato.com/customer/portal ... -3-0-ports, and also because you can google the specs for recording 16 channels of audio with very modest equipment going back many,many years, and also because you had no problems recording 8 channels with your Apogee on the older mac and it wasn't until you used the RME with only 2 channels more before you started having problems...

I can see where using a proprietary usb or firewire instead of a popular third party software might not grab the attention of Apple and compatibility checks. The kind of back and forth firmware upgrades, driver upgrades, to be compatible with newer OS upgrades,etc., may be why the older mac seems more powerful than the newer one with the RME?? The elgato link has some general suggestions about how to deal with the usb 3.0 issue that might be useful.

[mojobone]

regarding Logic - if you are needing to monitor real time while recording (like musicians with cans listening and overdubbing, or recording to pre-recorded tracks, or a click track) then you should try lower sample rates (max of 64 samples really - otherwise its too much). If you don't need this, and its an acoustic ensemble not needing to monitor, put the recording buffer at maximum (1024 samples). The CPU will be much more lightly loaded.

Andy, did you mean 'lower latencies', here'? Otherwise, agreed on all counts. Len's FW suggestion also seems plausible.

[andygabrys]

regarding Logic - if you are needing to monitor real time while recording (like musicians with cans listening and overdubbing, or recording to pre-recorded tracks, or a click track) then you should try lower sample rates (max of 64 samples really - otherwise its too much). If you don't need this, and its an acoustic ensemble not needing to monitor, put the recording buffer at maximum (1024 samples). The CPU will be much more lightly loaded.

Andy, did you mean 'lower latencies', here'? Otherwise, agreed on all counts. Len's FW suggestion also seems plausible.

Yes Mojo - thanks - I corrected my post above.

[andygabrys]

Re: Mac mini power

2012 mini typically had i7 processors.

2016 mini also have i7 processors available but not in the base model (which has i5). Did you buy a base model 2016 mini?

Seems strange that the 2012 would be faster?