Mac Pro Late 2013 and Samsung 4k monitor problem

Why the image is showing up upside down is beyond me.

But here they are. Be in touch if you'd like them. No profit here. Just helping out a fellow community member with the same 4K issue. 🙂

<Email Edited by Host>

Thanks pacioccone for the offer, but I think based on the information I received today from Accell (below for those interested), I'm going to hang tight here and hope to be seeing it in stock soon.

Thank you for contacting Accell. The demand for B143B-007J has far exceeded our expectation and it currently backordered. We have rushed shipments from our factory and should arrive in about 2 weeks. Once received it will be available on our website http://www.accellcables.com/products/ultraav-mini-displayport-to-displayport-1-2 -cable. We apologize for the wait.

Poyda wrote:

You need a good Mini DisplayPort to DisplayPort 1.2 cable AND you need to use the custom EDID / timings that have been discussed from page 6 onwards in this thread.

Thanks for confirming that as I saw the responses about cables and got excited 😉

Alas you still need to ***** around like you're on Linux or some such OS. 😉

It seems that this isn't even solved in Yosemite Beta.

<Edited by Host>

All good, mate! I hope you enjoy the monitors with the new cables.

I spoke with Apple Care who had a response from engineering regarding this thread and the bugs referenced in the thread. Engineering is putting this back on Samsung and stating this has nothing to do with Apple. Why would this be Samsung's issue? 😐

I guess maybe they thing samsung's EDID data is bad? But blaming on samsung seemsa bit implausible given it works fine on a nvidia GTX 680 or under windows with ATI cards.

I thought I saw a macrumors post that the latest 10.10 DP finally fixed this, but it was in a long thread and not sure I could find again.

Most people on this forum seem to be using the Samsung U28D590D or the Asus PB287Q, but since their release, a few other 4K monitors are on the market. Was wondering if anyone here has any experience with using the following monitors with the new Mac Pro?

Viewsonic VX2880ml

Philips Brilliance 288P6LJEB

AOC U2868PQU

Lenovo Thinkvision Pro 2840m

pacioccone wrote:

I spoke with Apple Care who had a response from engineering regarding this thread and the bugs referenced in the thread. Engineering is putting this back on Samsung and stating this has nothing to do with Apple. Why would this be Samsung's issue? 😐

It's not a Samsung issue if it works in Windows. Has Engineering or anyone in this thread tried one of these 4K 60 Hz SST monitors in Boot Camp on the same Mac? If it works in Windows but not in Mac OS X, then it's a Mac OS X driver problem.

Terscion wrote:

Most people on this forum seem to be using the Samsung U28D590D or the Asus PB287Q, but since their release, a few other 4K monitors are on the market. Was wondering if anyone here has any experience with using the following monitors with the new Mac Pro?

Viewsonic VX2880ml

Philips Brilliance 288P6LJEB

AOC U2868PQU

Lenovo Thinkvision Pro 2840m

I can't tell if those monitors are SST or MST. The Samsung U28D590D or the Asus PB287Q are SST. Any SST monitor should just work (except for this AMD issue) since they are like normal monitors except with higher bandwidth. I don't know how MST monitors work in Mac OS X - whether or not Apple has to add support for each MST monitor, or if the driver can detect that the monitor requires MST and work with that. If Apple needs to add support for each MST monitor, then maybe we can figure out how that's done and add support ourselves, or maybe the EDID can be edited to make it appear as a supported MST monitor.

The engineer clearly doesn't have all the information. I'm using my U28D590 right now on my Late 2013 Mac Book Pro and it works perfectly even using an older mini display port to display port cable. Its an AMD driver / signal issue. The MBP puts out a clean 3840x2160 @ 60Hz horz. 133.3khz vert. and it was like a quarter the price of the MP.. 😠

II'm using samsung monitor with new Mac Pro with D300 graphics and it also works fine. No artifacts on the screen in 60hz - no need to modify edid file. It could be actually AMD problem with the drivers. Of course for us is all Apple who is to blame. But for sure I wouldn't say it's samsung fault.

chriskuta wrote:

The engineer clearly doesn't have all the information. I'm using my U28D590 right now on my Late 2013 Mac Book Pro and it works perfectly even using an older mini display port to display port cable. Its an AMD driver / signal issue. The MBP puts out a clean 3840x2160 @ 60Hz horz. 133.3khz vert. and it was like a quarter the price of the MP.. 😠

Right, the MBP uses either Intel integrated or NVIDIA graphics (isn't external video always using NVIDIA graphics?). Since the monitor works with that, then there's a problem with the AMD graphics or with the AMD drivers.

romanheym wrote:

II'm using samsung monitor with new Mac Pro with D300 graphics and it also works fine. No artifacts on the screen in 60hz - no need to modify edid file. It could be actually AMD problem with the drivers. Of course for us is all Apple who is to blame. But for sure I wouldn't say it's samsung fault.

Another user in this thread also reported no issue with the D300. Noone has reported an issue with the D300. So it's either a problem with the D500/D700 or the AMD driver for the D500/D700. We need to see some tests with Boot Camp.

There's a possibility that the NVIDIA graphics is working better than standards to allow working with a monitor that doesn't meet standards - but that's probably crazy talk and the engineer shouldn't be so lazy trying to track down this problem.

Hello all,

I've just managed to get Windows 8.1 installed on my nMP with D700 (which wasn't easy btw), and tried to check the resolution/timings on the Windows platform, I get stuck at 2560x1440@60Hz, no other option above that in the Adjust Resolution section in Control Panel. Anyone managed to get 3840x2160 on Windows? Or any ideas on how to change it?

I have googled around for an hour with no luck....

dmolrub wrote:

Hello all,

I've just managed to get Windows 8.1 installed on my nMP with D700 (which wasn't easy btw), and tried to check the resolution/timings on the Windows platform, I get stuck at 2560x1440@60Hz, no other option above that in the Adjust Resolution section in Control Panel. Anyone managed to get 3840x2160 on Windows? Or any ideas on how to change it?

I have googled around for an hour with no luck....

Did you install the latest AMD Catalyst drivers (14.4)?

http://support.amd.com/en-us/download/desktop?os=Windows+7+-+64#amd-catalyst-pac kages

On the same page is listed a beta driver (14.7 Beta). The beta driver mentions the following possibly relevant issue (AOC is different from the ASUS and Samsung displays discussed in this thread):

http://support.amd.com/en-us/kb-articles/Pages/latest-catalyst-windows-beta.aspx

Collaboration with AOC has identified non-standard display timings as the root cause of 60Hz SST flickering exhibited by the AOC U2868PQU panel on certain AMD Radeon™ graphics cards. A software workaround has been implemented in AMD Catalyst 14.7 RC driver to resolve the display timing issues with this display.

• Users are further encouraged to obtain newer display firmware from AOC that will resolve flickering at its origin.
• Users are additionally advised to utilize DisplayPort-certified cables to ensure the integrity of the DisplayPort data connection.​​​

This suggests that there might be a problem with the displays (at least for the AOC display). So try the 14.4 driver first and if you see flickering, go to the 14.7 driver. If the flickering goes away, then it would be nice to know the workaround they used to see if it's different than the work around we're using in Mac OS X.

Hi joevt,

Is it possible to make a data line for the override file that is actually 60 Hz. for the Samsung 4k monitor? I have been using the 50 Hz file for a while and it has been much better than the 56 Hz which causes the picture overlap on the right hand side but still a little lagging. 57 Hz also causes the same overlap but it appears to be thinner to the edge. The monitor seems to be okay going past 57 Hz and so does the nMP (D500 model). I have just switched over to the 54 Hz. now and the picture is getting better all the time.

Thanks,

PB

PB_NB wrote:

Hi joevt,

Is it possible to make a data line for the override file that is actually 60 Hz. for the Samsung 4k monitor? I have been using the 50 Hz file for a while and it has been much better than the 56 Hz which causes the picture overlap on the right hand side but still a little lagging. 57 Hz also causes the same overlap but it appears to be thinner to the edge. The monitor seems to be okay going past 57 Hz and so does the nMP (D500 model). I have just switched over to the 54 Hz. now and the picture is getting better all the time.

Thanks,

PB

There's no way to know unless you find a mode that works or you find that all possible (at least reasonable) modes don't work. You can use SwitchResX to try different modes. I guess you would want to play with the horizontal blanking values. The horizontal blanking period consists of 3 parts: the front porch, the sync pulse, and the back porch. For CVT-RB timings, it looks like this (16 pixels per character):

–––__––––– (48 front porch, 32 sync width, 80 back porch)

You can move the sync pulse left or right (make a change in the front porch and the opposite change in the back porch) which would have the affect of moving the display left or right on a CRT (not sure what the result would be on an LCD).

You can change the width of the sync pulse (make a change in the sync pulse and the opposite change in the front or back porch).

You can make the total blanking period shorter or longer, but then you would need to adjust the pixel clock to keep the vertical refresh rate at 60 Hz (which means a line will take the same amount of time, but the blanking will be a different fraction of the time).

If the display works in Boot Camp on the same Mac, then you could get the data line from that and use it in Mac OS X. If the data line is the same then it means the Mac OS X driver has some other problem.