Pairing ProMotion Mac with Studio Display?

[singularity0993]

I'm planning to pair a 2021 MacBook Pro with a Studio Display. But it looks like the Studio Display does not have ProMotion 120hz. So the MBP screen is supposed to be a lot smoother in scrolling and animations. I use the screen mainly for coding so I guess 120hz isn't that important, but I surely can feel the difference between 120hz and 60hz.

So what does it feel like to use the two screens side by side? Will Studio Display be noticeably more laggy? I can still cancel the order if that really is a problem.

On the other hand, I am aware of the 27-inch MiniLED 120hz Display that is supposed to arrive in October. But since I don't really need the MiniLED, I don't think 120hz alone can justify the ($3000+?) price tag. What do you think?

 

[astorre]

I got MBP 120hz, next to it 100hz 1440p ultrawide and next to it 60hz 1080p and I dont notice much during coding, I noticed laggy behaviour when I put one display to 30hz, otherwise its fine

 

[unrigestered]

why don't you just disable ProMotion on your MBP and see for yourself?

 

[singularity0993]

why don't you just disable ProMotion on your MBP and see for yourself?

I did. The difference was noticeable but not unacceptable. But side-by-side comparison is a different thing and since my Studio Display hasn't arrived yet I couldn't do side-by-side comparison. That's why I'm asking.

 

[unrigestered]

in the end, only you can decide though
- there are people who don't notice any difference
- there are people who notice the difference, but don't care all that much about it
- there are people who notice the difference and for them it's the most important feature ever

and last but not least
- there are people who don't notice a difference but it's still the most important feature ever to them too

so everyone will give you different advice.
but is it the right one tailored for you?

 

[Dealmans]

It’s jarring going from a 14” MacBook to a 24” iMac.

 

[singularity0993]

From my observation ProMotion on iPhone and iPad is far more noticeable than it is on Mac (because Mac has less animation?) . I guess I should wait until the ASD arrives and see it in action.

 

[Sheepish-Lord]

It’s jarring going from a 14” MacBook to a 24” iMac.

Jarring? Seems a little extreme as I have the 16” and 24” and sometimes its hard to tell especially if scrolling at slower speeds. Now the XDR vs non-XDR is a game changer especially HDR content but that’s to be expected?

 

[unrigestered]

my view on it:
ProMotion is nice, but I’m not realizing it all that much after a while. Mostly straight after going from non-PM to PM.

Scrolling is still not completely fluid and still has its hick ups. It’s just that they are not as frequent ie going from several times a second to just once every couple of seconds with PM.
PM should also be better for video footage shot at 24 fps and preserve battery life better when using more or less “ still images“ while doing word processing, etc. as the frame rate can automatically adopt to something like 10 fps when nothing is happening that would need higher refresh rates.

Apple is also using pretty slow panels and IMO the bluriness during scrolling is way more pronounced with ProMotion on vs off.

If i had to choose, i‘d get ProMotion, but I’m also fine without it

 

[BeatCrazy]

On the other hand, I am aware of the 27-inch MiniLED 120hz Display that is supposed to arrive in October. But since I don't really need the MiniLED, I don't think 120hz alone can justify the ($3000+?) price tag. What do you think?

There won't be a 27" mini-LED ProMotion display in October, so that's one less thing to consider

 

[Alex Cai]

I'm planning to pair a 2021 MacBook Pro with a Studio Display. But it looks like the Studio Display does not have ProMotion 120hz. So the MBP screen is supposed to be a lot smoother in scrolling and animations. I use the screen mainly for coding so I guess 120hz isn't that important, but I surely can feel the difference between 120hz and 60hz.

So what does it feel like to use the two screens side by side? Will Studio Display be noticeably more laggy? I can still cancel the order if that really is a problem.

On the other hand, I am aware of the 27-inch MiniLED 120hz Display that is supposed to arrive in October. But since I don't really need the MiniLED, I don't think 120hz alone can justify the ($3000+?) price tag. What do you think?

why don't get a high fps display from Benq/Aoc? It's cheaper

 

[singularity0993]

why don't get a high fps display from Benq/Aoc? It's cheaper

For me, higher resolution is more important than higher refresh rate. And there is no other Retina display in the market other than Apple and LG's (both using even the same panel).

Actually I don't think higher refresh rate really makes things smoother... well maybe a bit BUT it does make you feel that lower refresh rate is laggy. I shouldn't have turned on ProMotion in the first place 🤣

 

[singularity0993]

There won't be a 27" mini-LED ProMotion display in October, so that's one less thing to consider

But Ross Young accurately predicted every display-related change in Apple hardware before...?

 

[BeatCrazy]

But Ross Young accurately predicted every display-related change in Apple hardware before...?

Ask Ross Young how you're going to connect a Mac to a 5K/120Hz display. That kind of bandwidth isn't supported by Thunderbolt 3/4. You need DisplayPort Alt Mode 2.0 via USB4, which isn't supported by any M1/M2 hardware.

And Apple has never released a display that didn't work with at least a few old/existing Macs.

 

[singularity0993]

Ask Ross Young how you're going to connect a Mac to a 5K/120Hz display. That kind of bandwidth isn't supported by Thunderbolt 3/4. You need DisplayPort Alt Mode 2.0 via USB4, which isn't supported by any M1/M2 hardware.

And Apple has never released a display that didn't work with at least a few old/existing Macs.

I guess you could use DSC? We already know Apple used DSC in Pro Display XDR so it wouldn’t surprise me if they use the same tech to support 5K 120hz. And AFAIK, TB4 supports up to 8K@60hz, which is more data transmitted per second than 5K@120hz, why would 5K@120hz be limited by bandwidth?

After all, a display analyst couldn’t possibly be unaware of such bandwidth limitation, if it really exists.

 

[BeatCrazy]

I guess you could use DSC? We already know Apple used DSC in Pro Display XDR so it wouldn’t surprise me if they use the same tech to support 5K 120hz. And AFAIK, TB4 supports up to 8K@60hz, which is more data transmitted per second than 5K@120hz, why would 5K@120hz be limited by bandwidth?

After all, a display analyst couldn’t possibly be unaware of such bandwidth limitation, if it really exists.

The scenario I described assumes use of DSC. 8K@60Hz is for 8bpc via Thunderbolt, and macOS uses 12bpc.

Ross Young can crown himself with whatever designation he'd like.. but until he shows us the math behind how a 5K120 12bpc external display can connect to any current or near-future Mac, the math just doesn't support his prediction.

Calling @joevt for any additional insight.

 

[Dealmans]

Jarring? Seems a little extreme as I have the 16” and 24” and sometimes its hard to tell especially if scrolling at slower speeds. Now the XDR vs non-XDR is a game changer especially HDR content but that’s to be expected?

Everyone's tolerance to pro motion seems different

 

[singularity0993]

The scenario I described assumes use of DSC. 8K@60Hz is for 8bpc via Thunderbolt, and macOS uses 12bpc.

Ross Young can crown himself with whatever designation he'd like.. but until he shows us the math behind how a 5K120 12bpc external display can connect to any current or near-future Mac, the math just doesn't support his prediction.

Calling @joevt for any additional insight.

Ok… But whether it is possible or not, I’m not going to buy a $3000+ monitor. Studio Display will still be the best Retina display deal on the market even if Apple somehow manages to make 5K XDR 120hz happen.

 

[joevt]

5K120 is 1903 MHz = 22.8 Gbps using DSC @ 12bpp. That would require HBR3 link rate (25.92 Gbps). I don't think I've seen DSC work at HBR3 link rate yet on a Mac? Or no one has tried. Apple Pro Display XDR is HBR2 link rate when it's using DSC.

8K60 is 2090 MHz (using CVT-RB2 timing) which should be doable with HBR3 link rate using DSC @ 12bpp which can do 10bpc.
HDMI is 2376MHz which would require DSC @ 10bpp when using HBR3 link rate. Even DSC @ 8bpp should be able to do 10bpc.

I have yet to see anyone prove that the dscTargetBPP preference has an effect on macOS DSC bpp choice. In theory it should make macOS add modes with DSC using 8bpp or 10bpp (or whatever you chose for the preference) for any display that macOS allows to use DSC but I haven't tested it yet.

I still need to test if the DisplayPort/DisableDSC field in the display override file can be used to enable DSC. This field is read by displaypolicyd which only reads override files from /System but files in /System can't be easily modified so I wrote a WhateverGreen patch to change the file path that displaypolicyd uses.
https://forums.macrumors.com/threads/mac-mini-4k-120hz.2267035/post-31207285

EDIT (Oct 3, 2024):
Turns out that the dscTargetBPP preference can only be changed if the Apple Internal SIP bit is set, and that bit cannot be set by csrutil. I don't know how to change that bit.

My WhateverGreen fork (Intel Macs only) has a new patch to change dscTargetBPP (add a boot-arg dscbpp=8 - for Catalina and later). If you're using Open Core or OCLP, then you can replace Lilu and WhateverGreen with my versions. If you use any other Lilu based kexts then they need to be recompiled using the headers from my Lilu.kext. That patch, in conjunction with the CheckTimingWithRange patch (add a boot-arg -cdfon - For Tiger and later), can enable 4K240 and beyond on Intel Macs with a GPU that supports DSC (tested with 6800XT and Sequoia). DP to HDMI adapters are not tested.

I think Apple Silicon Macs automatically support DSC with values ≤ 12 (4K240 tested on M3 with Sequoia).

 

[Tyler O'Bannon]

5K120 is 1903 MHz = 22.8 Gbps using DSC @ 12bpp. That would require HBR3 link rate (25.92 Gbps). I don't think I've seen DSC work at HBR3 link rate yet on a Mac? Or no one has tried. Apple Pro Display XDR is HBR2 link rate when it's using DSC.

8K60 is 2090 MHz (using CVT-RB2 timing) which should be doable with HBR3 link rate using DSC @ 12bpp which can do 10bpc.
HDMI is 2376MHz which would require DSC @ 10bpp when using HBR3 link rate. Even DSC @ 8bpp should be able to do 10bpc.

I have yet to see anyone prove that the dscTargetBPP preference has an effect on macOS DSC bpp choice. In theory it should make macOS add modes with DSC using 8bpp or 10bpp (or whatever you chose for the preference) for any display that macOS allows to use DSC but I haven't tested it yet.

I still need to test if the DisplayPort/DisableDSC field in the display override file can be used to enable DSC. This field is read by displaypolicyd which only reads override files from /System but files in /System can't be easily modified so I wrote a WhateverGreen patch to change the file path that displaypolicyd uses.
https://forums.macrumors.com/threads/mac-mini-4k-120hz.2267035/post-31207285

Is there any way that apple could make a 5K and 6K 120Hz panel that used 2 Thunderbolt busses and each ran half the image?

 

[joevt]

Is there any way that apple could make a 5K and 6K 120Hz panel that used 2 Thunderbolt busses and each ran half the image?

You are basically describing how the Dell UP3218K works. It uses two HBR3 x4 connections, one for each half of the display. Using Thunderbolt to transfer the two HBR3 signals doesn't really add anything except the ability to include PCIe tunnelling for USB 3.1 gen 2 or to chain other Thunderbolt devices.

Note that the Apple Studio Display has a mode for GPUs that don't support DSC where the Thunderbolt cable carries two HBR2 x4 connections, one for each half of the display.
Same with the Apple Pro Display XDR except in that case the single Thunderbolt cable carries two HBR3 x4 connections. Those connections each have a pixel clock of 648.91MHz which is just about the max that Thunderbolt can carry for 10bpc RGB with no DSC (648.91MHz * 2 halves * 30bpp = 38.9 Gbps).

For 5K120, each half would be 966 MHz. HBR3 can do that only at 8bpc (The Dell UP3218K is also limited to 8bpc at 8K60). So for best results without DSC, Apple would need to create a display with 4 DisplayPort inputs (which can be carried by two Thunderbolt cables using HBR2 x4 for each). The display can be divided into four using one of these timings for each quarter:
1280x2880@120Hz = 498MHz (all the tiled displays I've seen always split the display horizontally like this)
2560x1440@120Hz = 483MHz (the display is split once horizontally and once vertically)
5120x720@120Hz = 476MHz (arranging the tiles vertically is the most efficient use of bandwidth)

I've never seen a tiled display with four tiles work in macOS or Windows but I think it might be possible. I think displaypolicyd does allow for 4 tiles when it's verifying an mtdd file. I would like to create a patch that makes macOS believe that four separate displays are each a tile of a single display and see what happens. Then maybe @Amethyst1 can use his IBM T221 as a single display.
https://forums.macrumors.com/thread...supported-external-gpus.2175397/post-27557716

6K120 might be doable using HBR3 x4 with DSC@9bpp.
Without DSC, then four tiles can be done with one of these timings for each tile:
1504x3384@120Hz = 681MHz. x4 = 81.6 Gbps (definitely too much for two Thunderbolt connections)
3008x1692@120Hz = 664MHz. x4 = 79.7 Gbps (probably too much for two Thunderbolt connections)
6016x846@120Hz = 655MHz. x4 = 78.6 Gbps (might be doable by two Thunderbolt connections, 39.3 Gbps each; remember that 60Hz only uses 38.9 Gbps)

Maybe Apple will make a 5K+ 120Hz display when Intel Macs are no longer supported - when all Macs support DSC or when they support DisplayPort 2.0.

 

[fwmireault]

On the other hand, I am aware of the 27-inch MiniLED 120hz Display that is supposed to arrive in October. But since I don't really need the MiniLED, I don't think 120hz alone can justify the ($3000+?) price tag. What do you think?

Slightly unrelated to the main topic, but regarding this rumor, Ross Young said he has not confirmed it will have promotion (Source). He said it will have mini-LED, but 120hz is not guaranteed, probably for technical reasons as some have said in previous answers (although it might be possible with DSC on TB4, but I am no expert on this matter). We might have to wait for TB5 to get this level of bandwidth

 

[theorist9]

Is there any way that apple could make a 5K and 6K 120Hz panel that used 2 Thunderbolt busses and each ran half the image?

....6K120 might be doable using HBR3 x4 with DSC@9bpp.
...

Maybe Apple will make a 5K+ 120Hz display when Intel Macs are no longer supported - when all Macs support DSC or when they support DisplayPort 2.0.

I calculated Apple should be able to do 6.75k@120 (giving a 35.56" Retina display @218 ppi), with much less compression (12 bpp), using HBR3x2 and a single TB 3/4 port. Did I get something wrong here?:

You've explained that, including the active pixels, and horizontal and vertical front porch and sync width, the XDR is 6096*3516.

Thus, without DSC, at 30 bpp, it requires a bandwidth of 6096*3516 *60*30/10^9 = 38.58 GHz, which it can send over TB 3/4 using dual HBR3 tiling (whose max is 38.9346 GHz).

You also said Apple's default DSC is 12 bpp:

12bpp is the default DSC target bpp for macOS. I don't know if macOS can do DSC @ less or more than that.
4:2:0 8bpc is also 12bpp.

With that, TB could accommodate 1.123 times as many pixels horizontally and vertically while also doubling the frequency to 120 Hz (here I've assumed the needed porch and sync scale linearly with the increase in the no. of pixels in each direction; even if that's not the case, the values should be close). This yields a 35.56" 120 Hz Retina display that can be driven using a single TB cable:

6096*3516 *1.123^2*120*12/10^9 = 38.92 GHz
6016*1.123 = 6756 (active horiz pixels)
3384*1.123 = 3800 (active vert pixels)
sqrt[(6756/218)^2 + (3800/218)^2] = 35.56"

For comparison, the XDR's active pixels are 6016*3384. At 218 ppi, that gives a diagonal of:
sqrt[(6016 /218)^2 + (3384/218)^2] = 31.66"

 

[joevt]

I calculated Apple should be able to do 6.75k@120 (giving a 35.56" Retina display @218 ppi), with much less compression (12 bpp), using HBR3x2 and a single TB 3/4 port. Did I get something wrong here?:

By HBR3x2, you mean dual HBR3 x4? I think I make that mistake sometimes when discussing dual tile displays.

When I said HBR3 x4 with DSC@9bpp, I meant single tile.

Here's a bash script (any other scripting language would probably be faster) to calculate the max for single tile and dual tile:

for ((bpp = 8; bpp <=12; bpp++)); do
	fps=120
	maxgbpstile=20 # half of Thunderbolt 3 
	maxgbps=25.92 # HBR3 x4
	tilew=0
	thew=0
	gotit=1
	for ((w = 1024; gotit; w += 16)); do
		gotit=0
		mhztiled="$(edid-decode  -S --cvt w=$((w/2)),h=$((w*9/16)),fps=$fps,rb=2 | perl -pE 's/.* (\d+.\d+) MHz.*/$1/')"
		if (($(bc <<< "$mhztiled * $bpp < $maxgbpstile * 1000"))); then
			tilew=$w; gotit=1
		fi
		mhz="$(edid-decode  -S --cvt w=$((w)),h=$((w*9/16)),fps=$fps,rb=2 | perl -pE 's/.* (\d+.\d+) MHz.*/$1/')"
		if (($(bc <<< "$mhz * $bpp < $maxgbps * 1000"))); then
			thew=$w; gotit=1
		fi
	done
	printf "Max   dual tile 16:9 %d Hz display @%dbpp for %.3f Gbps x 2 is %d x %d\n" $fps $bpp $maxgbpstile $tilew $((tilew*9/16))
	printf "Max single tile 16:9 %d Hz display @%dbpp for %.3f Gbps     is %d x %d\n" $fps $bpp $maxgbps $thew $((thew*9/16))
done

Result:

Max   dual tile 16:9 120 Hz display @8bpp for 20.000 Gbps x 2 is 8272 x 4653
Max single tile 16:9 120 Hz display @8bpp for 25.920 Gbps     is 6688 x 3762
Max   dual tile 16:9 120 Hz display @9bpp for 20.000 Gbps x 2 is 7792 x 4383
Max single tile 16:9 120 Hz display @9bpp for 25.920 Gbps     is 6304 x 3546
Max   dual tile 16:9 120 Hz display @10bpp for 20.000 Gbps x 2 is 7392 x 4158
Max single tile 16:9 120 Hz display @10bpp for 25.920 Gbps     is 5968 x 3357
Max   dual tile 16:9 120 Hz display @11bpp for 20.000 Gbps x 2 is 7040 x 3960
Max single tile 16:9 120 Hz display @11bpp for 25.920 Gbps     is 5696 x 3204
Max   dual tile 16:9 120 Hz display @12bpp for 20.000 Gbps x 2 is 6736 x 3789
Max single tile 16:9 120 Hz display @12bpp for 25.920 Gbps     is 5456 x 3069

6736 is pretty close to 6.75K.

With that, TB could accommodate 1.123 times as many pixels horizontally and vertically while also doubling the frequency to 120 Hz (here I've assumed the needed porch and sync scale linearly with the increase in the no. of pixels in each direction; even if that's not the case, the values should be close).

I used the CVT-RB2 method to calculate porch and sync. In the above script, it's calculated by the edid-decode command using the --cvt ....,rb=2 option.

There might be reasons for not making a dual tile display that uses DSC.

 

[theorist9]

By HBR3x2, you mean dual HBR3 x4? I think I make that mistake sometimes when discussing dual tile displays.

When I said HBR3 x4 with DSC@9bpp, I meant single tile.

Yes. I was using the "x2" in HBR3x2 to mean two HBR3 connections => dual-tile HBR3. Or, more completely: dual-HBR3, where each HBR3 uses x4 DP. Sorry for not making that clearer.

There might be reasons for not making a dual tile display that uses DSC.

Any idea on what those reasons might be?

Even if they could build a dual-tile display that requires DSC to be used with TB3/4, might Apple not want to build it for visual-quality reasons?

For instance, they could have driven the XDR without having to bother with dual-tile HBR3 if they had used DSC, yet they chose not to. Could that have been because they thought DSC might interfere with critical color-grading work? [Not that the monitor is quite capable of that—it doesn't meet Dolby Vision certification requirements for use as an HDR mastering monitor—but that may have been Apple's aspiration when they designed it.]