Let's discuss USB3 stability best practices please

[jlafferty4]

Hi,

I have a personal rMBP 15" 2015 running Yosemite. I proudly stay behind the times because I prize stability over bleeding edge latest/greatest. I tech at times though, so I work on a variety of machines and with whatever setup/software is available. I am having a hell of a time with USB3 - both with my personal camera, Nikon D810, a semi-regular client's D800, as well as an increasing number of clients moving to Canon's 5DS/5DSR.

Before we get into the issue at-hand, please skip over the basics: yes I have the correct and only provider checked in C1 prefs; yes I have multiple cables; yes I work with a powered dock/hub when possible; yes, I make sure I've got a fully charged battery; yes, I try multiple ports; I even trash com.phase prefs at times.

The most consistent stability I get is a single 15ft TetherTools cable into CaptureOne 8.3.4... ever since upgrading to 9.3.x I've had issues, and often the issues are resolved by falling back to 8.3.4 - just changing software they disappear entirely, and sometimes this proves to be not only a good solution, but the *only* solution. But unfortunately... I'm at times in the position where I don't have the option to use 8.3.4, and/or need a longer connection.

Today I'm on a shoot with a MacPro tower running El Capitan 10.11.6. We're using C1 9.3.0.69. All I'm running is C1 and Chronosync, plus TextEdit and Safari. We've got 32gb of RAM here. Yesterday, running a 5DSR to C1 9.3. via a single 15ft TetherTools cable, over an 11 hour period, we had 6-7 dropped connections. Solving the drops was the usual song and dance of restarting Capture, taking the camera battery out, and at worst restarting the computer, and zapping PRAM - some random collection of one or all of these things.

So, in hopes of eliminating problems on day 2 of this job, and also giving the photographer a little more space from the tower, I brought a TetherTools active repeater cable and purchased their TetherBoost Pro unit (the small cable with the additional 5v power line - *not* the CoreController). I had previously purchased the TetherBoost Pro but returned it when I saw no apparent improvement in tethering the D810 to my personal laptop, but after seeing comments here figured I would give it another go.

We were off to a smooth start, but somewhere around 2 hours in, the 5DSR dropped connection, and was *completely unrecoverable*. We did everything I know to do - swap cables, drop the TetherBoost, restart, swap batteries, everything. Nothing brought the camera back at all. Ultimately we swapped out camera bodies *and* dropped the repeater and TetherBoost. We're back up an running but I'm left puzzled as to why a connection could start and work fine, but then deteriorate? Is it possible the power carried over TetherBoost messed with the 5DSR port?

So here I am on lunch, combing the hell out of the forums here for answers and I'm hoping to aggregate the best possible answers. Here's what I've got so far:

1) Always use a powered hub - best solutions seem to be OWC or Belkin Thunderbolt, but I've heard the cheap Amazon 7-port powered hub is good

https://www.amazon.com/Plugable-SuperSp ... B008ZGKWQI

What do you suggest or have good experience with?

2) If you do not have access to location power, use a TetherTools solution that runs on the external battery - but which one? Is the Core Controller more stable than the TetherBoost Pro?

Core Controller - https://www.bhphotovideo.com/c/product/ ... ller.html/

TetherBoost Pro - https://www.bhphotovideo.com/c/product/ ... oller.html

3) A single 15ft cable is best. A booster or power regulator of some sort is advised when using an active repeater. This returns me to the above - Core Controller or Tetherboost Pro?

4) Update your camera's firmware

What else? The seeming randomness of my problems across a set of different cameras, machines and software is driving me nuts.

[jlafferty4]

New session fixed tether issues completely with the above mentioned 5DSR body ¯\_(ツ)_/¯

[OddS]

New session fixed tether issues

I think you should file a report to Phase One support, your observation may be an important piece in a larger puzzle.

[Jim_DK]

Supported cable length for the USB3 spec is 3m/9ft. Any more and you are technically unsupported, but as you have found out you can improve "what works" with powered hubs every three meters.

Alternatively - if you need distance then there are companies making fiber optic USB cables in lengths up to 100m - which would in theory remove all power/timing/distance issues compared to traditional copper cables. I haven't tested it with 5DS though.

[Christian Gruner]

In addition, always use a brace/cable relief, supplied or 3 pary, to completely fixate the USB3 Micro Plug to the camera.
That is actually were most problem lies.

[jlafferty4]

Hi y'all - thanks for writing!

I always work to prevent cable connection issues - seems the USB3 design has a way of undermining my best efforts

Yesterday the setup was gaffed to the capture cart to prevent it from being pulled from the tower, then the repeater gaffed to the main cable, then run through a Tetherblock and into the camera (right-angle USB3 cable).

Even still I do think the sequence of events was cable stress at camera connection -> cable disconnect -> session corrupted.

I cannot send logs as it's not my personal machine. Perhaps I can have their in-house people do it. I will try.

On a related note... I see that C1 9.x and El Capitan no longer writes com.phase files in the Library -> Preferences folder, correct? Historically trashing these had been a consistent go-to in handling issues. Going forward it looks like creating a new session and overwriting the corrupted .cosessiondb file might work?

Also: spoke with Josh at TetherTools - really clear and helpful. He advised that in my setup, optimal placement of the TetherBoost Pro is actually between the repeater and the cable attached to the camera.

[Terence2]

I've been using a Canon 5DSR and have had the most success with this setup:

Anker 10-port powered USB 3.0 hub (I bought this one to handle the various ext HDs and thumb drives, etc) - purchased from Amazon

USB 3.0 cable with 90-degree plug: http://www.usbfirewire.com/parts/rr-asm ... ASML-180GR

I'm running this with a late 2012 rMBP with minimal issues - yes, occasionally get a crash but it's stable enough to not be out of the ordinary. I've also done one shoot with a new iMac 5K (late 2015)...didn't have any issues with tethered connection.

The main drawback is the length of the cable because it's so short (~15 feet). I gaff tape the plug to the camera body and lightly clamp the other end to the camera cart (without pinching the actual cable), just to prevent slight tugs from removing cable. The cart has to live next to camera and it's a total pain shooting in locations like houses or when walking is needed. But, for me, it's stable enough so I go with it.

[jlafferty4]

Thanks Terence. All good now. It was a one-time hardware misshap that corrupted the Session. Totally missed it in going through my usual diagnostic checklist. Works great now.

This was compounded by some issues on my personal machine, which I've now solved as well, thanks to diligent help from the CaptureOne team via a support ticket.

For reference, I will be going forward using a TetherTools 15ft repeater, into their Tetherboost Pro, then into a 6ft or 10ft USB3 cable; passing this through a powered USB3 hub (in my case, the $30 7-port Anker commonly found on Amazon). So far seems super stable and fast, as it should.

Anker hub: https://www.amazon.com/Anker-Charging-A ... B014ZQ07NE

Tetherboost Pro: https://www.tethertools.com/product/tet ... ontroller/

[ben_US]

Hey "jlafferty4"-

On a related note... I see that C1 9.x and El Capitan no longer writes com.phase files in the Library -> Preferences folder, correct? Historically trashing these had been a consistent go-to in handling issues. Going forward it looks like creating a new session and overwriting the corrupted .cosessiondb file might work?

They are written there, though now stored in an OS cache so manually deleting those files from ~/Library/Preferences does not behave the same way. You've got to run this in Terminal (preferably with Capture One closed):

defaults delete com.phaseone.captureone9

also, similarly, running this command will wipe the recent documents list (which is handy if you're having network issues and Capture One keeps calling back to a session on a network)

defaults delete com.phaseone.captureone9.LSSharedFileList

replace captureone9 with captureone8 if you're running that version.

[mark.mrkjhn]

Hi,

A little late to this dance (this forum post, not C1) but have been recently experiencing camera disconnections where as previously it never happened to me. C1 10.x, El Capitan, Macbook Pro late 2013, the usual bag of multiple cables and camera bodies etc., etc., etc....

"It was a one-time hardware misshap that corrupted the Session. Totally missed it in going through my usual diagnostic checklist. Works great now."

What diagnostic do you use for that, and how does one know of a corrupted Session and as being the cause for the bad tethering?

I've followed all the same procedures you've listed, along with different computers at times; I've had good luck previously in the last 6 plus years until now. Not sure if it's an OS thing with El Capitan and the way it's regulating power where as before, never had a problem with tethering.

Thanks,
Mark

[photoGrant]

Hey "jlafferty4"-

On a related note... I see that C1 9.x and El Capitan no longer writes com.phase files in the Library -> Preferences folder, correct? Historically trashing these had been a consistent go-to in handling issues. Going forward it looks like creating a new session and overwriting the corrupted .cosessiondb file might work?

They are written there, though now stored in an OS cache so manually deleting those files from ~/Library/Preferences does not behave the same way. You've got to run this in Terminal (preferably with Capture One closed):

defaults delete com.phaseone.captureone9

also, similarly, running this command will wipe the recent documents list (which is handy if you're having network issues and Capture One keeps calling back to a session on a network)

defaults delete com.phaseone.captureone9.LSSharedFileList

replace captureone9 with captureone8 if you're running that version.

Assuming this can run through all versions of the software, why isn't there a button in preferences to clear the cache? You're trying to get Photographer's in the terminal?

[ben_US]

Assuming this can run through all versions of the software, why isn't there a button in preferences to clear the cache? You're trying to get Photographer's in the terminal?

Running the command while the application is open will cause the application to crash.

[photoGrant]

Assuming this can run through all versions of the software, why isn't there a button in preferences to clear the cache? You're trying to get Photographer's in the terminal?

Running the command while the application is open will cause the application to crash.

Definitely not ideal for someone who doesn't know what they're doing -- could a dedicated button in the app that is programmed to restart the application during the process be easier for users?

[ben_US]

Definitely not ideal for someone who doesn't know what they're doing -- could a dedicated button in the app that is programmed to restart the application during the process be easier for users?

Yes, but it would also make it easier to wipe your workspace/keyboard shortcuts/etc back to vanilla inadvertently.

[JasonRanalli]

So are folks generally seeing a better performance in stability with these boost devices?

I'm fortunate in that I have had no corrupted sessions/catalogs but have had the occasional dropped connection.