Audio data can only be an issue when the power supply of the BF is unstable or insufficient. That can happen when using bus power, with a really bad (or extended) USB cable.

Red warnings in DC indeed only talk about the disk throughput and ASIO performance. But if the BF becomes unstable it will affect anything. That the problem shows only up in stereo is more like a theory, it seems. But easy to check, record some 'empty' additional channels to proove it in the future.

Hi, Mathias,

I have been wracking my brain as to "differences" in setups and one that I may have done is used the USB3 connector on the right side of the computer for the times that failed and on the left side for those that did not. (SO MANY VARIABLES). I THINK the internal cabling is longer to the right side (as it has to go around the DVD drive).

Any thought on whether this MIGHT be a factor?

I'm trying to design tests that will show both failures and success. Ugh. Don't have the time for this. LET ME REPEAT, I DO NOT SEE THIS AS AN RME ISSUE. I am writing hear because RME is smart and thorough as are their users and I hope not to go down this rabbit hole alone. I hope someone else has experienced something like this.

Another thought, do you believe in the rule "always record to an external drive, not the system drive?" If so, would an SDXC card be an option (the laptop has a built-in reader/writer). Is that worth a try. As you may recall, I solved the FFUFX recording challenge to a thumb drive by using SDHC cards in a USB "stick" reader.

Thank you again.

Cheers,

Richard
(FFUFX, BF, 2 MultiFace II)

I finally had two long nights to play and run extended tests.

Machine 1 Dell Vostro 3460 2012 Win 7 Home Premium 64 bit, i5 3230M 2.6 GHz, 8 GB RAM, 500 GB spinning drive, four USB 3.0 ports.

Machine 2 Lenovo Think Pad (2010) T410 reloaded from scratch in 2014 Win 7 pro 64 bit, i5 M520 2.4 GHz, 4 GB RAM, 320 GB spinning drive, four USB 2.0 ports.

Passmark Benchmarks (today) Machine 1: 3933;    Machine 2: 2397 Both are dual core with hyperthreading (four threads).

Both machines have the latest updates and are running Microsoft Security Essentials as their virus protection.

Here is what I found about DigiCheck Global Record and the whole process:

The left time display appears to turn red with buffers lost because of communications errors with the BF.

The right disk loading display appears to turn red with data lost because the disk could not keep up with the BF.

To test this, I connected my pair of DPA 4006 TL mics to the BF through the normal field cabling (using XLR-5 and a direct connection to the 15-pin D connector. There is a breakout at the mics. About 65 feet of quad mic cable.

I set the BF to 192 kHz sampling frequency and recorded 24 bits. I left the optical signals off.

Finding One: The left display errors were mostly related to one apparently junky USB cable. Thanks for pointing this out Matthias. I could not create the problem with a better cable on any of the ports. That seemed to reliably correct the left problem. Sorry for the false alarm on this one and thanks for the troubleshooting assistance.

Finding Two: My old friend Bill Lund who has been doing this with about 50 CDs / year to his credit in Minneapolis said he would always record to an external drive. I realize that I had mounted two drives in my audio tower computers (except I have an Intel RAID-1 with two disks and a hot standby). With an external USB 3.0 1 TB drive, I was able to test over four hours of uninterrupted recording on Machine 1 and 8 hours on Machine 2. In both cases I stopped the tests to get on with other things.

Finding Three: There was nothing I could do to make the internal drive work reliably with 192 KHz audio. Now, I don't use 192 kHz, but I do use 96 kHz. I felt working to 192 kHz gave me the headroom I would normally be content with. The Performance Meter metrics for disk queue on both machines were just under 0.1 on the scale (meaning 10% of capacity.

I could not reliably identify all the pieces of software which tried to run and bogged down the system drive, causing errors. With the external drive, I was able to run a Microsoft Security Essentials quick virus scan while recording and also install the dropbox folder on the Lenovo and browse a few screen shots, all while recording with no glitches.

While I am not thrilled about the external drive, it is a solution. I am thinking my next tests should be to an SD card as both machines have built in SD card readers. That would improve the physical robustness and a CL10 card should have more than enough bandwidth even for 192/24 stereo (or quad). As you may recall, I used a "flag" USB SD card reader on the Fireface UFX with great success.

The key here is to stay away from the system drive. I suspect my previous successes were partially luck with the FFUFX and Global Record. The thing which I think MAY have been happening was that a 10-channel recording at 96 kHz is more of a data load AND the computer was giving that thread a higher priority than the lower bitrate BF threads, but that's a really wild guess.

I am using the full 8K samples buffer in the BF. I note that the FFUFX only goes to 2K samples. Interesting.

I am publishing this here to close the loop and possibly help others who are having challenges. I will test the SD cards when I get the chance and report back. I like that much better than the USB drives on a short leash.

Thanks again for your assistance.

Cheers,

Richard

There was no DVD in the drive. Do I disable that from Device Manager or is there a better way?

As to degrag, I checked the Lenovo machine which is more likely to become my recording machine as its configuration will be more stable as it's used mostly on my workbench with some low-footprint audio software (ARTA and others) to make audio measurements and also check into my home email machine via remote desktop. AND it was just brought up as a fresh machine within the last six months or so.

Machine 2 about 60 % of the HD is full (I have a backup set of the recent Maui trip images on it) It has INDEXING ENABLED (ON) which might be part of the issue, and is 0% fragmented.

I am pretty certain write cache is on, but if you know a quick place to check it, that would be of help.

I did get 7 hours plus of recording to a 32 GB SD card formatted NTFS at 2 channel 192 ks/s However 4 channel had a strange hiccup which might have been an artifact of a previous test, so more testing.

I sometimes really hate computers.

Thanks, Matthias,

OK both machines: Main drive write caching enabled in the Policies (Richtlinien ?) tab.  Machine 1 = 1% fragmented, Machine 2 = 0 % fragmented. Machine 1 500 GB HD with about 100 GB free. Machine 2 320 GB HD with about 100 GB free.

Both have indexing enabled on the main (C:\) drive in each.

I do not know what is arbitrarily / randomly causing the usage of the main drive to spike (and cause it to miss packets) turning the drive section of Global Record red.

It seems as if my testing procedures (with external drive AND SD card) have isolated the issue to the internal HD. I read up on the DPC latency tester and I do not think that this is the latency I'm experiencing. I will give it a try, but I think that's looking at CPU latency issues while I'm fairly certain that I'm seeing HD latency issues (or HD busy issues). But it is worth a look.

I also do not think it practical to stop everything because when updates are installed, version checkers are installed and who knows when they go and look for their associated new versions.

I am running some more tests to the SD card with the BF attached. The FFUFX has its own solution, the SD cards in the USB reader on the front panel.

If I were recording for a living, I think I'd look for a more robust solution, like I have in my tape restoration studio with tower computers, separate data and system drives, and lots of NAS backup. But for the community recording work I'm mostly not doing now (but wish to maintain the capability), I'm thinking this record-to-SD-card-in-PC with the BF will work. There are one or two events a year which suggest I should bring out the FFUFX.

If the quality of the preamps in the Zoom H6 were up to the RMEs, that would solve most of my limited field recording and still let me sell the Sound Devices 722 which, while nice, does not have the transparency in its preamps that the BF and especially the FFUFX have.

In my studio, the BF serves as a backup to the FFUFX should it fail for any reason I have coverage for most of what I do and still have the two Multifaces for 16 channel ingest.

Thanks!

Cheers,

Richard