[Mel Matsuoka]

I would try the “Refine Soft Matte” filter. It often does wonders on chunky edge detail on stuff like hair…it can be indistinguishable from magic in some instances, or it can make things worse in others.

[Mel Matsuoka]

Just to follow up on my own post, for the benefit of anyone who finds themselves asking this same question:

Pretty much all my research, as well as direct feedback from SNS, led me to the conclusion that for our particular facility, shielded 6A was way overkill. So we decided to bite the bullet on unshielded Cat-6A. And boy, it still hurt!

Studio Network Solutions support told me that while it’s certainly possible to do 10Gig over Cat6, they don’t recommend it for their SNS EVO product, if only because 10GigE certification over copper requires Cat-6A. They implied that tech-support for their product might be problematic without 6A certification, because they wouldn’t be able to support a configuration that is technically not certified for 10Gig over copper.

In our case, it turned out that the cost of labor was actually lower than the cost of the materials, because not only did we have to spec 6A cabling, jacks and a patch-panel (almost all of which needed to be ordered from the mainland, because apparantly, 6A installations are almost never done in my city of Honolulu), the 6A cabling also had to be plenum-rated in order to meet fire/building code, which jacked up the price of materials massively (almost 2x-3x the cost of standard UTP/CMR Cat6 cable). The cost of shielded 6A would have probably caused us to jump out of the window, due to sticker shock and subsequent depression.

If you ask a potential cabling contractor to install 6A in your facility (particularly if they are merely “IT guys”, to paraphrase one of Bob’s favorite pet-peeves), don’t assume that they appreciate the fact that our type business requires every last drop of performance and stability that is possible in a 10GigE copper connection.

My contractor, who clearly knows his stuff, was up-front with me in saying that 6A installations were basically unheard of in our town, surprisingly tried to convince us to go with cheaper Cat6 cabling instead. When I told him that my NAS vendor insisted on 6A, he sympathized with me, and proceeded to order the 6A parts. When he installed the “6A” patch panel in our rack, I noticed that the labeling on it simply said “Cat6”. When I asked him if the panel was actually a 6A-certified panel, he said that the patch-bay vendor informed him that their Cat6 panel was effectively identical in performance to a panel that purports to be “6A”, and it wasn’t worth the added expense for us.

When the RCDD-certified cable-certification people came in later to run 6A certification tests on all the runs, using a horrifically expensive Fluke DTX-1800 cable analyzer, all of the cable runs “barely failed” 6A certification (btw, if you’re gonna spend the money on 6 or 6A, insist on having those runs actually “certified” with a cable analyzer, and not just simply “tested” for continuity). The cabling contractor immediately blamed the patch-panel he originally installed, and within 2 hours of the test results, had all his guys come back in and re-install an honest-to-God 6A patch-panel (a shielded 6A patch panel, to boot) and re-test everything. I just got the general feeling that 6A is such a rare animal in the networking world, that even the product vendors themselves can’t seem to be trusted to offer reliable information!

Now, from a “real world” standpoint, I probably could have gotten away with the “barely failed” 6A runs. But one of the main reasons why I pushed my company (a company consisting of only 4 people, btw) to plunk down a huge chunk of money on this new SNS EVO system is so that I can spend more of my valuable time and mental energy performing the actual roles printed on my business card–“Colorist/Finishing Editor”–instead of being the de facto IT guy/Google-searcher-by-proxy/Techsupport-bitch in the facility, simply because I listen to a lot of tech podcasts and fearlessly know my way around a Bash terminal session. It doesn’t make sense to not be overly pedantic with things like Cat6A vs. Cat6, especially when the NAS vendor itself recommended against the latter.

I almost felt embarrassed presenting the estimated cost for this 6A upgrade to my boss, because it would seem ridiculous to someone who is used to the low cost of garden-variety Cat5e installations. But when I put things in perspective, the amount of money I might have saved up-front by going the cheaper Cat6 route could have ended up as false economy if we ended up having performance issues with the NAS later on. I can understand how SNS would have a hard time supporting their product, if they cant be assured that the physical connections to it are not up to an established spec.

Spending the additional money on 6A hurts like a mofo. But because I did so, I can now just tell my editors to call/email SNS techsupport directly, rather than having them bug me with their NAS problems whenever I’m in the middle of a color-grading session. In that context, trying to save a couple thousand dollars by cheaping out on network cabling is hardly worth it, because if I charged my company for being their Colorist as well as the in-house “IT guy”, then my salary should be rightfully at least twice as much as it is now!

[Mel Matsuoka]

I’m using NeatVideo OFX on 12.5.2 on macOS 10.10.5, and I’m not experiencing any significant instability or slowness that can be attributed to this plugin. I’m assuming that you ran the Neat Video “Optimize Performance” profiler for your OFX version?

The only thing that bugs me about using NeatVideo in Resolve is that even if you cache the node you’ve applied NV to, when you switch a node into Highlight view downstream of that node, Resolve completely ignores the upstream node cache! This makes things like adjusting Qualifiers feel like you’re submerged in a vat of molasses.

Admittedly this is not a NV-specific problem, but it’s a big enough problem that I will often hesitate to use NV unless I really need it.

I’ve submitted this as a bug/feature-request way back when node-caching was first introduced, so it’s a real bummer that this problem still persists in 12.5.2.

[Mel Matsuoka]

If you’re talking about storage boxes for “bare” hard drives, I’m a huge fan of the CRU Dataport Anti-static DriveBoxes (formerly made by WiebeTech). They are compact, and very well made and sturdy for typical storage applications. You can buy 10 of them for $50 USD, and they are well worth it.

For home use I like the Orico PHP-35 storage boxes, as they have nice interlocking grooves, so they are secure when stacked, but I wouldn’t recommend them for frequent usage, since the hinges on those things seem like they would eventually break from repeated fatigue.

[Mel Matsuoka]

Well, after more investigating, I figured out that the possible culprit could be the Noise Industries “FXFactory” plugin package.

Even after trashing my Premiere prefs, uninstalling Premiere, deleting all Adobe Media Caches, etc…then reinstalling via the godawful Adobe Creative Cloud Desktop app (did I mention that I hate the Adobe Creative Cloud Desktop app?), Premiere would still take forever to start up.

I noticed that while it sped through the plugins this time, it would slowly trudge through a seemingly endless series of random, hexadecimal files. I thought at first these were Cache files, but their persistence even after deleting all the Adobe cache files was puzzling. Using the good old Unix “find” command, I determined that these hex filenames were being loaded from the “~/Library/Application Support/Adobe/Common/Plug-ins/7.0/MediaCore/FXFactory” folder.

After deleting all of those files, and then completely disabling FXFactory, Premiere was able to start up again quickly

I created detailed documentation of this issue here, if anyone is curious:

https://bit.ly/2cA7N5n

[Mel Matsuoka]

I of course meant to say Premiere Pro 2015.3, not 2015.4!

[Mel Matsuoka]

[Robert Withers] “Sometimes it’s necessary to keep a machine running under an earlier OS if you want to use an earlier software.”

That’s the harsh reality that I’ve come to assume at this point.

I’m gonna see how well FCP7 runs in a VMWare Fusion virtual machine running Yosemite. I really don’t want to maintain a separate Yosemite partition just so that I can install & run FCP7, since I just need it to load old projects for exporting XMLs out of them.

If that doesn’t work, I guess I’ll create a temporary Yosemite partition just to install FCS3, and then use Migration Assistant to bring the installation over to the El Cap partition.

[Mel Matsuoka]

[Jim Scott] “I can’t remember where I ran across this, but apparently because Rosetta is not supported in Mavericks and later, those OSs will not allow installation of FCS3. If you had upgraded your OS on a drive that already had FCS3 installed on it, it would have been available. Whether or not it would work well, or at all, is another story. But since you did a clean install to a non-Rosetta supporting OS, that new OS is preventing installation.
“

I’m not sure what Rosetta has to do with any of this, since as far as I know, Final Cut Studio (3) is Intel only.

FCP7 installs and works just fine on Yosemite & Mavericks.

[Mel Matsuoka]

The issue is not with Yosemite. The issue is with El Capitan.

[Mel Matsuoka]

[Bob Zelin] ”
You are using a Netgear GS752TX switch. This distributes to your clients via 1GbE, not 10GbE. There is a 10G uplink port that connects to your Netgear ReadyNAS 4220, but your client connection is 1G. “

Actually, this is incorrect. We have 3 Mac Pro clients (each of them with Small Tree PETG2NDA 10g Ethernet adapters) which are connected to the GS752TX via the SFP+ ports.