3 Post Moderation Rule for New Boarders (25 Jan 2021)
New boarders will have their posts moderated - Don't worry if you cannot see your post immediately.
USB3 vs TB1 will I see a huge difference?
16 Dec 2013 21:26 - 16 Dec 2013 21:43 #36075
by Topher5
USB3 vs TB1 will I see a huge difference? was created by Topher5
Will Thunderbolt 1 make a real world difference that is genuinely noticeable.
When editing multi-camera 1080p content in FCPX?
Compared to USB3 drive both spec'd at 7200rpm.
When editing multi-camera 1080p content in FCPX?
Compared to USB3 drive both spec'd at 7200rpm.
Last edit: 16 Dec 2013 21:43 by Topher5.
Please Log in to join the conversation.
- thefluentone
-
- Offline
- Senior Member
-
Less
More
- Posts: 68
- Thank you received: 5
16 Dec 2013 21:54 #36082
by thefluentone
Replied by thefluentone on topic USB3 vs TB1 will I see a huge difference?
Single drive, most likely not. However, going from my 7200 rpm usb externals o a pegasus R6 was a huge improvement. Simple things like opening evets or scrolling through projects. I still have to play with my multicam footage as I just got my drive this weekend. But I can give you some differences later tonight if you like.
Please Log in to join the conversation.
16 Dec 2013 21:57 #36083
by BenB
Replied by BenB on topic USB3 vs TB1 will I see a huge difference?
Eventually you can get enough streams to bog down USB 3. I doubt you'd ever get that bad with Thunderbolt. If I could afford both, I'd go with Thunderbolt. But with Thunderbolt, you'd want a RAID. A single HDD won't make much difference.
Please Log in to join the conversation.
16 Dec 2013 22:00 - 16 Dec 2013 22:03 #36085
by Topher5
Replied by Topher5 on topic USB3 vs TB1 will I see a huge difference?
I see so the bottle neck is the hard drives, at this rate.
So Raid O or Raid11
Is it possible to take two USB3 ext HD and Partion them as one HD?
A poor mans Raid 0.
In disk utility.
Or stripe (Raid 0) two SDD's maybe?
So Raid O or Raid11
Is it possible to take two USB3 ext HD and Partion them as one HD?
A poor mans Raid 0.
In disk utility.
Or stripe (Raid 0) two SDD's maybe?
Last edit: 16 Dec 2013 22:03 by Topher5.
Please Log in to join the conversation.
16 Dec 2013 22:06 #36086
by Topher5
Replied by Topher5 on topic USB3 vs TB1 will I see a huge difference?
Also would the drives benefit formatting to HFS versus exFAT32?
Please Log in to join the conversation.
- thefluentone
-
- Offline
- Senior Member
-
Less
More
- Posts: 68
- Thank you received: 5
16 Dec 2013 23:38 #36093
by thefluentone
Replied by thefluentone on topic USB3 vs TB1 will I see a huge difference?
I'm not sure if Disk Utility will let you do a software RAID across different External interfaces. It'd be interesting to see. Geerally though with other drive systems, there is a hardware controller for the RAID or as I found out Drobo does theirs via software. Hardware RAIDS are always faster. RAID 1 is a mirror, ad RAID is a stripe. Mirror are slower than regular dives, Stripes are usuallay faster to a degree depending on the implementation and interfaces. If you need something cheaper and fast, you could do somehting like the J4 with SSDS, or a G Raid. Both are still more than a pair of USB 3 externals, but less than a full size array. TW, other useful levels of RAID used are 5 &6, which use a concept of striping to a set of drives, while reserving one drive to calulate parity for RAID 5, and 2 drives to calulate parity for RAID 6. hese setups usually require hardware RAID controllers and at least a few drives. The reason I say that, I'm not sure if OSX can handle more than 2 drives in a software RAID configuration.
Please Log in to join the conversation.
- Darren Roark
-
- Offline
- Platinum Member
-
17 Dec 2013 01:04 #36096
by Darren Roark
Replied by Darren Roark on topic USB3 vs TB1 will I see a huge difference?
You can create a RAID from externals using Disk Util, however I haven't tested it for speed or reliability.
About the T-Bolt vs USB 3.0, I bought a Seagate T-bolt adapter for the cheap seagate Backup Plus drives you can get at Costco. I did some speed tests on my retina MBP 15" using the blackmagic speed tests. (This is from memory so forgive me) They were pretty similar with the thunderbolt doing slightly better with sustained speeds on large files. They both got around 110mb/s when the drive was new and never written to. Now, using the drive full of footage, that's another story...
My wife is producing a feature length doc that I brought in the 3000+ clips and timelines from FCP7 to FCPX. In order for her to have easy access to all the footage so she can do stringouts for the editor, I transcoded everything to proxy. (took days) On her 13" retina MBP it beachballs too often using the USB. The Thunderbolt adapter seems much more stable as the project seems much more responsive. I switched back to the USB adapter and the frequent beachballing was back. I even tried a different USB adapter bottom, same thing. My guess is FCP X is checking and fussing all over the drive platter so often in the background that on a slow single green drive it really maxes out it's capabilities. Maybe Thunderbolt is better at handling that, I mean having microchips in the cables and being expensive should give you something right?
I just remember in the firewire days, enclosures were not created equal. If I were on a better quality USB 3.0 enclosure, I wouldn't be surprised to get better results.
About the T-Bolt vs USB 3.0, I bought a Seagate T-bolt adapter for the cheap seagate Backup Plus drives you can get at Costco. I did some speed tests on my retina MBP 15" using the blackmagic speed tests. (This is from memory so forgive me) They were pretty similar with the thunderbolt doing slightly better with sustained speeds on large files. They both got around 110mb/s when the drive was new and never written to. Now, using the drive full of footage, that's another story...
My wife is producing a feature length doc that I brought in the 3000+ clips and timelines from FCP7 to FCPX. In order for her to have easy access to all the footage so she can do stringouts for the editor, I transcoded everything to proxy. (took days) On her 13" retina MBP it beachballs too often using the USB. The Thunderbolt adapter seems much more stable as the project seems much more responsive. I switched back to the USB adapter and the frequent beachballing was back. I even tried a different USB adapter bottom, same thing. My guess is FCP X is checking and fussing all over the drive platter so often in the background that on a slow single green drive it really maxes out it's capabilities. Maybe Thunderbolt is better at handling that, I mean having microchips in the cables and being expensive should give you something right?
I just remember in the firewire days, enclosures were not created equal. If I were on a better quality USB 3.0 enclosure, I wouldn't be surprised to get better results.
Please Log in to join the conversation.
- thefluentone
-
- Offline
- Senior Member
-
Less
More
- Posts: 68
- Thank you received: 5
17 Dec 2013 02:29 #36101
by thefluentone
Replied by thefluentone on topic USB3 vs TB1 will I see a huge difference?
Darren, USB uses the CPU to process traffic. More frequent traffic means the CPU has to spin a thread to process the bus. Thunderbolt has its own controllers like firewire did.
Please Log in to join the conversation.
- Darren Roark
-
- Offline
- Platinum Member
-
17 Dec 2013 04:02 #36107
by Darren Roark
Replied by Darren Roark on topic USB3 vs TB1 will I see a huge difference?
That deserves a hearty click of the thank you button.
That makes a lot of sense. I had activity monitor open while I was swapping the adapters back and forth. FCP X went into "Not Responding" mode fairly often with USB.
I'd really like to find a way to explain the benefits of the thunderbolt up-charge to clients. Benchmarks don't tell the whole story.
That makes a lot of sense. I had activity monitor open while I was swapping the adapters back and forth. FCP X went into "Not Responding" mode fairly often with USB.
I'd really like to find a way to explain the benefits of the thunderbolt up-charge to clients. Benchmarks don't tell the whole story.
Please Log in to join the conversation.


