I guess you should try to find the max vertical height. Create a 3840x2400 mode for the MateView. Then try different heights up to 2560. You can also try lower refresh rates. The display might not accept a signal less than 30Hz, but you want to see what the macOS will allow before you find out what the display will allow. In that case you might want a second Mac for screen sharing in case the iMac accepts a mode that the display does not work with. If there's a maximum number of pixels limit, then you might want to go with a smaller width, like 1000, while trying to find the max height. But then you might run into an issue where macOS doesn't like a mode because the aspect ratio is too different...
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
What have you done with an early Intel recently?
- Thread starter Project Alice
- Start date
- Sort by reaction score
Well that was quick 'n' easy - 2400 pixels it is.
I used the Asus MG24UQ since that accepts widths of up to 4096 and heights of more than 4000 (plus refresh rates as low as 10 Hz).
3840×2400 works but shows glitches I'm inclined to attribute to the display since that resolution is fine on Bertha.
Heights of 2401, 2404, 2416, 2496 and 2560 are all rejected (using lower refresh rates to keep pixel clock below 540 MHz).
2560×2880 doesn't work either so it's clear now why the UP2715K doesn't talk to the Iris.
I wonder how that max height of 2400 came about, since only 2304 (for 4096×2304) "has to work".
Curiously, the HD 3000 can do 3840×2560 but I only tested it on Mavericks whereas the 6200 was tested only on Mojave - and I've come to expect different versions of macOS to behave differently.
The MateView accepts 3840×2560 at 25Hz.
Last edited:
Say hello to 2011
2011 Macs gained SATA III along with Sandy Bridge.
Wow. I didn’t realize it was as late as Sandy Bridge. I would have expected it happened prior to then by a year or two — like Arrandale. This is good to know.
That said, I did learn pretty quickly how the SuperDrive SATA bus on an early 2011 is SATA II and not SATA III.
A week after buying my early 2011 MacBook Pro as a factory refurb, I added a drive caddy in lieu of the SuperDrive to give the system two drives — one SSD for the system and one HDD for /Users. My original plan was to put the SATA III SSD in the caddy and the HDD in the usual drive bay, but this was one of those SSDs whose controller didn’t gracefully downshift to slower SATA standards on slower SATA buses. Consequently, I had to swap them before I could get the system to boot properly.
I'd always give the SSD the faster bus, thus moving the HDD to the caddy.
Yeah. I considered putting two SSDs into mine for a RAID 0 ('cause... whatever!) but quickly abandoned plans due to that.
It was, at the time, a situation where I was planning to have multiple data HDDs which I could reasonably swap physical drives easily without touching the system install. In retrospect, it was an inefficient idea. But to this day, that MBP still runs with two drives in that original /System on SSD and /Users on the HDD configuration.
Yup.
I find macOS' approach of keeping the
/Users directory on the same partition as the system... weird. All my Linux installs have /home on a different partition or drive so I can nuke the system partition without worrying about my data.
Same - but that's not easy-peasy i.e. Mac-like
Out of curiosity, what SSD was this? For the past several years I've been accustomed to just buying whatever SSD was on sale/the best value at my local computer shop. But after the experience I had with the WD Green SSD I recently bought, it's a little disappointing to now be wary about what might/might not work in my Macs.
Moved my Unibody upto to Mojave in order for it to be a better bridge between my iPhone 7, and older iOS devices that I collect and still use quite frequently in this house. While I do have a 12 core Mac Pro 5,1, the MacBook Pro 5,1 is my primary legacy iOS device sync / backup device for devices ranging from an iPod touch 2nd gen, iPhone 4, all the way up to this 7.
One side benefit is my audio sync to my office HomePod Mini is far better when watching videos. I actually keep it outputting audio to that device as my primary casual listening speaker.
Had this thing since new and it is still kicking and does well for what I need it for.
I mean, you can, even if Apple doesn’t like to set it up that way. That’s why I used that above blog link (which I’ve had bookmarked all these years) to set up my 2011 MBP with /Users in a different volume. Apple relies vigorously on those volume UUIDs, which is definitely a quirk of OS X.
Oh my, this was back in 2011. If memory serves, the model was a Patriot Pyro III 60GB unit. I’ve long since given away that SSD. That MBP is now using a Recdata 256GB SSD for the /System.
With my early 2008 MBP, I’m now using a 500GB WD Blue m.2 SATA inside an m.2-to-2.5 adapter.
Yeah. For Thunderbolt 3 or 4 systems only. These adapters have been around for a while actually. Not compatible with Thunderbolt 1 or 2 systems sadly.
Last edited:
Ignore the fact that the below screen grab is from my work MBP (2015 MBP). The point here is, two Minis doing work.
Backup going off on the A1283 (Late 2009) and iTunes streaming Drone Zone to my BT sound bar on the A1176. This is from Screen Sharing on the MBP.
You might also note the Geektool display of my console.log and LittleSnitch showing my incoming and outgoing connections.
PS. I used Nightingale on the A1176 a bit earlier to stream music from my NAS to my BT sound bar - all with the Mini being controlled headless through Screen Sharing. I may have decided to keep the Mini on Snow Leopard, but as you can see it's far from useless.
Backup going off on the A1283 (Late 2009) and iTunes streaming Drone Zone to my BT sound bar on the A1176. This is from Screen Sharing on the MBP.
You might also note the Geektool display of my console.log and LittleSnitch showing my incoming and outgoing connections.
PS. I used Nightingale on the A1176 a bit earlier to stream music from my NAS to my BT sound bar - all with the Mini being controlled headless through Screen Sharing. I may have decided to keep the Mini on Snow Leopard, but as you can see it's far from useless.
LOL!
Streaming Drone Zone on the MacPro at the same time as the Mini, so the sound bar is effectively a 'center' speaker.
Photo does not show the right speaker.
Streaming Drone Zone on the MacPro at the same time as the Mini, so the sound bar is effectively a 'center' speaker.
Photo does not show the right speaker.
Why wouldn't it be compatible with Thunderbolt 1 or Thunderbolt 2 systems? Do they have an issue with Titan Ridge?
For Thunderbolt 2 Macs, you should be able to connect one 4K60 display or two 4K30/1440p60 displays.
Same should be true for Thunderbolt 1 Macs (except not 4K60).
The interesting thing with the OWC adapter is that the Thunderbolt cable is detachable. I wonder if it could be powered by an Apple Thunderbolt 3 to Thunderbolt 2 adapter? I know the Apple adapter has enough power for a Thunderbolt device (Apple Thunderbolt to FireWire adapter or Gigabit Ethernet adapter) but probably not enough power for a Thunderbolt 3 device. In that case, you would need to connect the OWC adapter to a Thunderbolt 3 dock which would make the OWC adapter redundant since you would just connect the two displays to the dock instead (unless your dock has HDMI instead of DisplayPort).
I thought that a Thunderbolt 3 to dual-head adapter wouldn't be able to extract the two discrete DisplayPort signals when connected to a TB 1 or 2 system, so the displays would be mirroring each other. If dual-head does work, all the better.
Thunderbolt 3 to dual head adapter is just a Thunderbolt dock without any USB or PCIe. The Thunderbolt 3 controller in the adapter is the same as the Thunderbolt 3 controller in a dock.I thought that a Thunderbolt 3 to dual-head adapter wouldn't be able to extract the two discrete DisplayPort signals when connected to a TB 1 or 2 system, so the displays would be mirroring each other. If dual-head does work, all the better.
The Thunderbolt 3 controller has a DisplayPort output for one DisplayPort output of the adapter. This is the DisplayPort output of a dock (or any Thunderbolt 3 device with a DisplayPort output - for example, the OWC Mercury Helios S3).
The Thunderbolt 3 controller has a downstream Thunderbolt port that is used for the second DisplayPort output. This is the same as a Thunderbolt 3 dock or device that has a USB-C to DisplayPort adapter connected to the downstream Thunderbolt 3 port.
The Thunderbolt 3 controller has an upstream Thunderbolt port.
Unused in the dual head adapter is the Thunderbolt 3 controller's PCIe and USB ports.
Since the OWC dual head adapter uses Titan Ridge, one of the DisplayPort outputs might be usable with a USB-C (non-Thunderbolt) host.
There's probably a USB billboard device in case you connect the adapter to a USB-C port that doesn't support Thunderbolt or DisplayPort Alt Mode.
@joevt: This reddit post says dual monitors via an OWC TB3 dock connected to a (TB2) 2015 MBP aren't working (but were using two daisy-chained TB2 docks).
[If TB3 docks weren't so expensive I'd give it a try myself.]
[If TB3 docks weren't so expensive I'd give it a try myself.]
Last edited:
Two 4K60 displays to Thunderbolt 2 is difficult. You need a third display that is limited to HBR link rate (1440p60). It's not enough for the display to be using 1440p60 timing - it has to have a max link rate of HBR (not HBR2) - it cannot have a 4K60 capability (or a 1440p120 mode, etc). Connect that first, then when you connect the first 4K60 display, it will also be limited to HBR link rate (because that's the max allowed by the remaining bandwidth of Thunderbolt 2). Then disconnect the 1440p display. The 4K60 display will still be using HBR link rate. Connect the second 4K60 display and it will also be limited to HBR link rate. Each 4K60 display will then be limited to HBR link rate (4K30 or 1440p60).
His dual Elgato setup worked because one of the displays was limited to 4K30.
https://help.elgato.com/hc/en-us/ar...at-ports-does-Elgato-Thunderbolt-2-Dock-have-
For this situation, it would be nice if macOS had a button to alter the DisplayPort link rate manually. I think you probably just need to poke some Thunderbolt registers which are documented in the Alpine Ridge document that appeared in the forums. Maybe you could enable two HBR2 connections over Thunderbolt 2. It should work as long as the combined pixel clock of both displays does not exceed 833 MHz @ 8bpc or 667 MHz @ 10bpc (multiply pixel clock and bpp and make sure the result is less than 20 Gbps).
For Titan Ridge, maybe you could enable two HBR3 connections like Apple does for the XDR in dual tile mode. Dual HBR3 should work as long as you don't exceed a combined pixel clock of 1667MHz @ 8bpc or 1333MHz for 10bpc. The XDR is 648.91MHz x2 = 1297.82MHz total. Apple doesn't enable dual HBR3 if there's a Thunderbolt device or optical cable between the XDR and the host. If we could change the link rate ourselves instead of relying on Apple's trick then we could add daisy chaining to XDR which some people need to separate the XDR from the Mac at a greater distance. Putting the Mac in a server room and connecting the XDR with a super long optical Thunderbolt cable currently only works for single tile mode up to 4K or dual tile up to 5K or requires a GPU that supports DSC for single tile 6K.
With the ability to change the link rate in code, you wouldn't have to do the connection gymnastics I described above to connect two high bandwidth displays (two displays that could exceed the max bandwidth of Thunderbolt).
It seems to me that all Thunderbolt DisplayPort connections should have the max link rate, then the OS should be able to monitor the bandwidth pool and disable display modes that the monitor supports but cannot be used because the bandwidth pool would be exceeded. This is similar to what is necessary to handle an MST Hub (why doesn't Apple support MST for multiple displays?). A method is needed to map GPU DisplayPort connections to Thunderbolt controllers so you can see the path of the DisplayPort signal and know where the bottlenecks are. I think this should be possible if you can match info from the DisplayPort input of the Thunderbolt controller to the output from the GPU.
For calculating the bandwidth pool: in the Thunderbolt chain, you have to consider the connections between all the Thunderbolt controllers in the chain (10 Gbps, 20 Gbps, 40 Gbps). Before and after the Thunderbolt chain could exist a chain of MST Hubs so you have to consider those as well (various DisplayPort link rates and link widths for each MST Hub in the chain). Another consideration for bandwidth pool is the max capabilities of the GPU. That one is a little more difficult to deal with.
Bonus points if the UI can show the modes that would be possible if the bandwidth used by another display was reduced. It should have options that can reduce the bandwidth of a display, other than reducing the resolution - this means pixel depth (8bpc, 10bpc, 12bpc, 16bpc; Apple, please add 6 bpc for fun!), chroma sub-sampling (4:2:0 and 4:2:2) and DSC.
We had agreed awhile back Sandy Bridge could sneak into the Early Intel forums. I used OCLP to get 10.15.7 on my iMac 12,2. It works perfectly, and while I was never the biggest fan of Catalina, there is some features I had been missing in High Sierra. Non Metal acceleration is working perfectly on the Radeon 6970. Not bad for a 10 year old iMac.
I just used a late-2007 17" MacBook Pro to post this very message!
